Trench-Forearc Geology:

Sedimentation and Tectonics on Modern

and Ancient Active Plate Margins
 Trench-Forearc Geology:
Sedimentation and Tectonics on
Modern and Ancient Active
Plate Margins

edited by
Jeremy K. Leggett
Department of Geology, Royal School of Mines,
Imperial College of Science and Technology, London

1982

Published for
The Geological Society of London,
by Blackwell Scientific Publications
Oxford London Edinburgh
Boston Melbourne
Published by DISTRIBUTORS
Blackwell Scientific Publications USA
Osney Mead, Oxford O X 2 0 E L Blackwell-Mosby Book Distributors
8 John Street, London WCIN 2ES 11830 Westline Industrial Drive
9 Forrest Road, Edinburgh EH1 2QH St. Louis, Missouri 63141
52 Beacon Street, Boston, Massachusetts 02108, USA Canada
99 Barry Street, Carlton, Victoria 3053, Australia Blackwell-Mosby Book Distributors
120 Melford Drive, Scarborough
Ontario M1B 2X4
First published 1982 Australia
Blackwell Scientific Book Distributors
© Copyright 1982 The Geological Society. 99 Barry Street, Carlton
Victoria 3053
All rights reserved. No part of this publication
may be reproduced, stored in a retrieval system, British Library Cataloguing in Publication Data
or transmitted, in any form or by any means, Trench-forearc geology.
electronic, mechanical, photocopying, recording 1. Geology, Structural
or otherwise without the prior permission of I. Leggett, Jeremy K.
the copyright owner. 551.1'36 QE601

ISBN 0-632-00708-7

Text set in 9/10 pt Linotron 202 Times, printed and

bound in Great Britain at The Pitman Press, Bath
 Contents
Page
Preface: LEGGETr, J. K . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

JAPAN
TAIRA,A., OKADA,H., WHITAKER,J. H. McD. & SMITH,A. J. The Shimanto Belt of Japan:
Cretaceous-lower Miocene active margin sedimentation ........................................ 5
VON HUENE, R. & ARTHUR, M. A. Sedimentation across the Japan Trench off northern
Honshu Island . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
OGAWA,Y. Tectonics of some forearc fold belts in and around the arc-arc crossing area in
central Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
SHIKI, T. & MISAWAY. Forearc geological structure of the Japanese Islands ..................... 63

CENTRAL A M E R I C A
MOORE, J. C., WATKINS, J. S., McMILLEN, K. J., BACHMAN, S. B., LEGGETr, J. K.,
LUNDBERG, N., SHIPLEY, T. H., STEPHAN, J.-F., BEGHTEL, F. W., BUTT, A., DIDYK,
B. M., NIITSUMA, M., SHEPHARD, L. E. t~ STRADNER, H. Facies belts of the Middle
America Trench and forearc region, southern Mexico: results from Leg 66 DSDP ......... 77
SHIPLEY, T. H., LADD, J. W., BUFFLER, R. Z. & WATKINS, J. S. Tectonic processes along the
Middle America Trench inner slope .................................................................. 95
MCMILLEN, K. J., ENKEBOLL, R. H., MOORE, J. C., SHIPLEY, T. H. t~z LADD, J. W.
Sedimentation in different tectonic environments of the Middle America Trench,
southern Mexico and Guatemala ...................................................................... 107
VON HUENE, R., AUBOUIN, J., AZEMA, J., BLACKINTON, G., CARTER, J. A., COULBOURN,
W. T., COWAN,D. S., CURIALE,J. A., DENGO, C. A., FASS, R. W., HARRISON, W., HESSE,
R., HUSSONG, D. M., LADD, J. W., MUZYLOV, N., SHIKI, T., THOMPSON, P. R. &
WESTBERG, J. A summary of Deep Sea Drilling Project Leg 67 shipboard results from
the Mid-America Trench transect off Guatemala ................................................. 121
LUNDBERG, N. Evolution of the slope landward of the Middle America Trench, Nicoya
Peninsula, Costa Rica ................................................................................... 131

SOUTH A M E R I C A
KULM, L. D., RESIG, J. M., THORNBURG, T. M. & SCHRADER,H.-J. Cenozoic structure,
stratigraphy and tectonics of the central Peru forearc ............................................ 151
MOBERLY, R., SHEPHERD, G. L. & COULBOURN, W. T. Forearc and other basins, continental
margin of northern and southern Peru and adjacent Ecuador and Chile ......... ............. 171
EVANS, C. D. R. & WHITTAKER,J. E. The geology of the western part of the Borb6n Basin,
North-west Ecuador ..................................................................................... 191

ALEUTIANS
MARLOW, M. S., COOPER, A. K., SCHOLL, D. W. & MCLEAN, H. Ancient plate boundaries in
the Bering Sea region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
NtLSEN, T. H. & ZUFFA, G. G. The Chugach Terrane, a Cretaceous trench-fill deposit,
southern Alaska . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
BVRNE, T. Structural evolution of coherent terranes in the Ghost Rocks Formation l Kodiak
Island, Alaska . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
vi Contents

ASIA AND A U S T R A L A S I A
MOORE, G. F., CURRAY, J. R. & EMMEL, F. J. Sedimentation in the Sunda Trench and
forearc region ............................................................................................. 245
VAN DER LINGEN, G. J. Development of the North Island Subduction System, New Zealand 259

ATLANTIC
WESTBROOK,G. K. The Barbados Ridge Complex: tectonics of a mature forearc system ...... 275
PUDSEY, C. J. & READING, U. G. Sedimentology and structure of the Scotland Group,
Barbados ................................................................................................... 291
MALOD, J.-A., BOILLOT, G., CAPDEVILA, R., DUPEUBLE, P.-A., LEPVRIER, C., MASCLE, G.,
MOLLER, C. • TAUGOURDEAU-LANTZ,J. Subduction and tectonics on the continental
margin off northern Spain: observations with the submersible C y a n a . . . . . . . . . . . . . . . . . . . . . . . . 309

MEDITERRANEAN
LE PICHON, X., HUCHON, P., ANGELIER, J., LYBI~RIS, N., BOULIN, J., BUREAU, D., CADET,
J. P., DERCOURT, J., GLA~ON, G., GOT, H., KARIG, D., MASCLE, J., RICOU, L. E. &
TnlEBAULT, F. Subduction in the Hellenic Trench: probable role of a thick evaporitic
layer based on Seabeam and submersible studies .................................................. 319
KENYON, N. H., BELDERSON, R. H. & STRIDE, A. H. Detailed tectonic trends on the central
part of the Hellenic Outer Ridge and in the Hellenic Trench System ......................... 335
WEZEL, F. C. The structure of the Calabro-Sicilian Arc: result of a post-orogenic intra-plate
deformation ................................................................................................ 345

MAKRAN OF IRAN AND PAKISTAN

WHITE, R. S. Deformation of the Makran accretionary sediment prism in the Gulf of Oman
(north-west Indian Ocean) ............................................................................. 357
ARTHURTON, R. S., FARAH, A. & AnMED, W. The Late Cretaceous-Cenozoic history of
western Baluchistan, Pakistan--the northern margin of the Makran subduction complex 373
MCCALL, G. J. H. & KIDD, R. G. W. The Makran, Southeastern Iran: the anatomy of a
convergent plate margin active from Cretaceous to Present ..................................... 387

CALIFORNIA
BACHMAN,S. B. The Coastal Belt of the Franciscan: youngest phase of northern California
subduction ................................................................................................. 401
AALTO, K. R. The Franciscan Complex of northernmost California: sedimentation and
tectonics .................................................................................................... 419
BLAKE, M. C. JR, JAYKO, A. S. & HOWELL, D. G. Sedimentation, metamorphism and
tectonic accretion of the Franciscan assemblage of northern California ...................... 433
COWAN, D. S. Deformation of partly dewatered and consolidated Franciscan sediments near
Piedras Blancas Point, California ..................................................................... 439
INGERSOLL, R. V. Initiation and evolution of the Great Valley forearc basin of northern and
central California, U.S.A ............................................................................... 459

F O R E A R C T E R R A N E S IN O R O G E N I C BELTS
HESSE, R. Cretaceous-Palaeogene Flysch Zone of East Alps and Carpathians: identification
and plate-tectonic significance of 'dormant' and 'active' deep-sea trenches in the
Alpine-Carpathian Arc ................................................................................. 471
LEGGETr, J. K., MCKERROW, W. S. & CASEY, D. M. The anatomy of a Lower Palaeozoic
accretionary forearc: the Southern Uplands of Scotland ........................................ 495
HEPWORTH, B. C., OLIVER, G. J. H. & MCMURTRY, M. J. Sedimentology, volcanism
structure and metamorphism of the northern margin of a Lower Palaeozoic accretionary
complex; Bail Hill-Abington area of the Southern Uplands of Scotland .................... 521
 Contents vii

FACIES, P E T R O L O G Y AND MODELS

UNDERWOOD, M. B. & BACHMAN,S. B. Sedimentary facies associations within subduction
complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
MAYNARD, J. B., VALLONI, R. & Yu, H.-S. Composition of modern deep-sea sands from
arc-related basins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
KARIG, D. E. Initiation of subduction zones: implications for arc evolution and ophiolite
development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
 Preface

This book is a collection of papers on an aspect the Publications Committee of the Geological
of plate tectonics of which our understanding is Society decided to allow a relatively high
at present limited. In the mid-1970s, prior to amount of review material so that the final
the recent phase of IPOD active margin drill- volume would be a useful comprehensive and
ing, few geologists would have anticipated that up-to-date reference for the geologist interested
at the start of the 1980s so many new questions in this field.
concerning the nature of tectonic and The papers fall into 11 more or less natural
sedimentary processes in forearc regions would groups, the first eight of which are defined on
have come to light. the basis of geography. These concern Japan,
1980 seemed to be a good time to synthesize Central America, South America, the Aleu-
current research and future problems. Several tians, Asia and Australasia, the Atlantic, the
fascinating transects of active margins had been Mediterranean, and the Makran of Iran and
drilled in the preceding years, with largely Pakistan. All these sections include data from
enigmatic results. With the exception of drilling both land-based and marine studies. Papers on
on the Barbados Ridge in early 1981, IPOD the geology of the Franciscan and Great Valley
active margin activities were about to enter a terranes of California, in section 9, have been
period of abeyance. Several academic institu- separated from papers on other possible ancient
tions had completed informative cruises on forearc terranes, in section 10, whose setting in
active margins, and numerous studies of emer- orogenic belts makes their reconstruction more
gent areas in modern forearc regions and of equivocal. In the final section are three papers
proposed ancient forearc terranes had come on facies models, petrology of forearc sedi-
to fruition. For these reasons the British ments, and the origin of forearc-related
Sedimentological Research Group, a Specialist ophiolites.
Group of the Geological Society of London, In editing this collection of papers I have
decided in 1979 to convene a three-day interna- tried to provide the interested geologist with a
tional conference on the theme of 'Trench and clear impression of the current state of research
Forearc Sedimentation and Tectonics in Mod- on forearc regions of active plate margins. If
ern and Ancient Subduction Zones'. The meet- this goal has been achieved, he or she will
ing was held on 23-25 June 1980, at the Geolo- hopefully be impressed with the enormous
gical Society, Burlington House, Piccadilly, amount of data which has accrued over recent
London. It was attended by more than 180 years, though should be equally aware that
earth scientists from 17 countries. there are many puzzling variations in the
Most of the contributions in this book were geology of forearcs and that there are many
presented orally or as poster sessions. Many of disagreements in the interpretation of indi-
the presentations were subjected to lengthy and vidual forearc areas. Variations in forearc tec-
often heated discussions. I consider it unfortun- tonics will become clear in comparing, for
ate that it has not proved possible to incorpor- example, the Late Cretaceous and Palaeogene
ate selections from those useful discussions in forearc geology of land areas of SW Japan in
the final volume. the Shimanto Belt (Taira et al.) with the
The ambitious title of the book is a conces- Neogene behaviour of the Honshu offshore
sion to brevity rather than an exact description forearc (von Huene & Arthur). Equally im-
of the contents. The emphasis is on tectonics pressive are the differences in tectonic evolu-
and sedimentation. However, in studies of fore- tion seen along the same active margin in South
arc geology, perhaps more so than in any other America (Kulm et al.; Moberly et al.) and
topic in geology, it becomes difficult to consider highlighted in two drilled transects of the Cen-
tectonic and sedimentary phenomena indepen- tral American margin (Moore et al. ;von Huene
dently. For this reason, these topics are discus- et al.). Differences in interpretation of the same
sed side by side in many of the papers. Meta- area include the Hellenic Trench, where
morphic and igneous phenomena are also consi- Kenyon et al. and Le Pichon et al. arrive at
dered in several contributions. differing conclusions using different geophysical
The papers fall into three approximately equal techniques, and the Franciscan Complex of
categories: contributions with entirely new California, where Blake et al. and Bachman/
data, review papers on areas of special interest, Aalto arrive at differing palaeogeograptiic inter-
and papers which combine reviews with new pretations using detailed mapping of differ-
data and ideas. As at the original meeting, ent parts of the same terrane. The paper by
 Preface
Wezel gives a salutory r e m i n d e r that p e r h a p s seas authors regarding American spellings in some
we are all barking up the w r o n g tree in a d h e r i n g figure captions where I was so requested.
too closely to the tenets of plate tectonics.
I have p r e f e r r e d not to a t t e m p t the tradition- ACKNOWLEDGMENTS:
al editor's s u m m a r y paper. My overall impress- I am greatly indebted to many people. First, to the
ion is that the r e c e n t flood of data on the Council of the Geological Society: without their
forearc geology of active margins has c r e a t e d agreement to fund the conference, coverage of the
just as m a n y , if not m o r e , n e w p r o b l e m s than it subject in this book would not have been so com-
has solved. This is g o o d for o u r science, and plete, and the conference participants would not have
enjoyed the meeting as much as they did. Second, to
augurs well for a stimulating future. If this the staff at the Geological Society: they tolerated
collection of papers achieves o n e thing, I h o p e endless impossible deadlines in the administration of
it will be to e n c o u r a g e the f u n d i n g which will be the conference. Third, to the contributors: all very
necessary to tackle these p r o b l e m s , both at sea busy men, who nonetheless found time to produce
and on land, over the next few years. their manuscripts on time. Fourth, to referees and
advisors too numerous to mention: they put in hours
of their time to assist--Gwyn Thomas in particular.
Note on editorial procedure: Fifth, to Mac (W. S. McKerrow) for all his support
As editor I have followed wherever possible the and encouragement. Finally, to my parents, Jessie
editorial proceedures and format used in the Journal and Sally: they know what for.
of the Geological Society. The papers are written by
geologists whose specialities cover a variety of fields, J. K. LEGGET1~, Department of Geology,
and who hail from many different countries. I have Royal School of Mines, Imperial College,
tried to standardize scientific terminology, nomencla- London SW7 2BP.
ture, and spelling. I have made concessions to over- June 1981.
 The Shimanto Belt of Japan: Cretaceous-lower Miocene
active-margin sedimentation

A. Taira, H. Okada, J. H. McD. Whitaker & A. J. Smith

SUMMARY: South-east of the two paired metamorphic belts of Kyushu, Shikoku and
Honshu, and separated from them by the Chichibu and Sambosan belts and the Butsuzo
Tectonic Line, lies a belt of mildly metamorphosed sedimentary rocks called the Shimanto
Supergroup. Their area, extending from the Nansei Islands through Kyushu, Shikoku, Kii
Peninsula, Akaishi and Kanto Mountains to the Boso Peninsula, is comparable in size with
the combined Franciscan Formation and the Great Valley sequence in California. A
Cretaceous lower group is flanked on the Pacific side by an upper group of Palaeogene to
early Miocene age. The Shimanto sediments, mostly sandstones and mudstones, were
deformed by cyclic subduction into open and isoclinal folds which close south-eastwards,
and thrust slices which in places incorporate basaltic pillow lavas and radiolarian cherts.
Sandstones from Kyushu and Shikoku show marked changes of composition with
stratigraphic position and the feldspar content can be used to distinguish Cretaceous from
Tertiary sandstones. Data on sandstone petrography and palaeocurrents indicate that
sediments of the Shimanto Supergroup were probably derived from the NW. Precambrian
gneisses and older Mesozoic granitic rocks of the Korean Peninsula may have contributed
sediments in Early Cretaceous times, when the Shimanto Terrane was located to the east
of southern Korea. The unroofing of granites in the Inner Zone of SW Japan may also
have contributed to the Cretaceous and later sedimentation. Quartz-rich, well-sorted
sands in younger Shimanto sediments may have been recycled from older Shimanto
formations.
The thick Shimanto sedimentary sequences were laid down in a variety of environments
within a forearc basin on accretionary complex, trench-slope break, trench inner slope and
trench settings; first in the Cretaceous and again, in a more southerly position, in the
Palaeogene and early Miocene. In the shallower waters on the inner (arc) side of the
Palaeogene forearc basin, coarsening- and thickening-upward deltaic sequences were
deposited. In deeper water further offshore, submarine channel and fan complexes,
base-of-slope slump deposits and red shales with cherts accumulated. M~langes with
basalts and slump olistostromes occupied the trench-slope break, while inner trench wall
perched basins on accretionary basement were filled with coarsening-upward flysch,
slump-olistostrome facies and m61anges with metabasalts: similar lithologies probably
occupied a trench fill.
Sedimentation and deformation were controlled by intermittent cyclic subduction of the
Kula Plate towards the north. Phases of tension and down-faulting, forming long, narrow
intra-arc basins, alternated with those of compression. Comparable sedimentary environ-
ments are found today off SE Japan on the submarine terraces, trench-slope break, inner
slope (including perched basins) and fill of the Nankai Trough.

To Western geologists, Japan is best known for tion plus the Great Valley sequence of California.
its two pairs of metamorphic belts and for its The purpose of this paper is to summarize the
igneous rocks, well summarized by Miyashiro structural and sedimentological features ot the
(1973), and m o r e recently by various authors in Shimanto Belt and to reconstruct the conditions
Tanaka & Nozawa (1977). Less well known, of deposition of Shimanto rocks. We believe
because the literature is mainly in Japanese, is they accumulated through Cretaceous and
the mildly m e t a m o r p h o s e d Shimanto Belt lower Tertiary times in an active margin setting
lying to the SE of the Sambagawa metamor- consisting of forearc basin on accretionary
phic and Chichibu and Sambosan Belts and c o m p l e x - - t r e n c h - s l o p e b r e a k - - t r e n c h inner
separated from them by the Butsuzo Tectonic s l o p e - - t r e n c h environments (terminology of
Line (Fig. 1). Extending some 1800 km along Dickinson & Seely 1979). Thrusting during
strike from the South Nansei Islands to the accretion, with the production of m61anges
Boso Peninsula and up to 70 km wide across the bearing blocks of ocean-floor pillow basalts,
strike in the Kii Peninsula, the Shimanto Belt is hyaloclastites, cherts and sandstones was a fea-
comparable in area with the Franciscan Forma- ture of this active margin.
6 A . Taira et al.

cut" ~Lower ShimantoGroup (N.Belt)

2~~°'
. ~'~ ~wA#-OSHIM~~.~ .rb
~ "S. ~ -i~
~- Upper
AXIALShimanto
LATERAL
CURRENT
CURRENT
Group (S.BeIt)

.~ / % o _~o ~oo ~oo~

FIG. 1. Map of Japan to show the distribution of the Lower and Upper Shimanto Groups, with
generalized palaeocurrents. K = Kochi; N = Nichinan; S -= Shizuoka: U = Uwajima. BTL =
Butsuzo Tectonic Line. MTL = Median Tectonic Line.

General geology of the Shimanto Belt the Southern Belt this facies contains gabbro,
diabase and serpentinite. Recently, 'umber'
The Shimanto Supergroup is now exposed in deposits comparable to modern oceanic sedi-
the Nansei Islands, South Kyushu, South Shi- ments have been found between pillows of
koku, Kii Peninsula, Akaishi and Kanto Moun- basalt in the Mineoka Group in the Boso
tains and on the Boso Peninsula (Fig. 1). This Peninsula (Tazaki et al. 1980). Large-scale
long, narrow belt is divisible along its length olistostromes are common in both belts, but are
into two major tectonostratigraphic units, a better developed in the Southern Belt.
Northern Belt (Lower Shimanto Group) and The overlying flysch facies in both belts is
Southern Belt (Upper Shimanto Group) characterized by terrigenous turbidites alternat-
(Teraoka 1979). The former is characterized by ing with mudstones, infrequently intercalated
uppermost Jurassic to Cretaceous strata, the with acidic tuff layers. In the Northern Belt,
latter by Palaeogene and Lower Miocene rocks. ammonites and inoceramids have been found
Both these groups are characterized by meta- sporadically (Matsumoto & Okada 1978), while
basalt and mudstone facies generally overlain bivalves and gastropods occur rather commonly
by flysch facies, forming successions many in some limited sequences in the Southern Belt.
kilometres in thickness (Kanmera 1976). These Trace fossils are prolific locally, e.g. in Shikoku
rocks are highly deformed, often into isoclinal (Katto 1960, 1964). Generally, however, fossils
folds, and there is important repetition by are rare and subdivision of the Shimanto Super-
imbricate thrusting (Sakai 1978). There is group must use sandstone petrography.
general low-temperature regional metamorph- Sandstones of the Northern Belt (Lower Shi-
ism of the prehnite-pumpellyite and greenschist manto Group) tend to be highly feldspathic in
facies. composition, less sorted and coarser grained
In the cross-sections of the Shimanto Belt, than the quartz-rich, better sorted Upper Shi-
sediments tend to become younger oceanwards manto sandstones of the Southern Belt. In the
(Fig. 2) as does the unconformity between the mature arenites of the latter group, extremely
highly deformed Shimanto strata and the over- well-rounded quartz grains abound. These
lying mildly deformed post-Shimanto sedi- sandstones are further characterized by the
ments. A similar trend continues out of the stable heavy mineral suite of zircon-tourmaline-
Shimanto Belt into the present trench inner rutile, and by a higher muscovite content. In
slope (Fig. 2, columns 5 & 6). both groups, microcline is a major constituent,
The lower metabasalts in both the Northern but it decreases upwards. The total amounts of
and Southern Belts are characterized by pil- feldspars also decrease upwards: the Cre-
lowed basalt and hyaloclastite (Kanmera 1976; taceous sandstones of Kyushu, for example,
Suzuki & Hada 1979; Tsuchiya et al. 1979). In have more than 35% feldspar, Tertiary ones
 The S h i m a n t o Belt o f Japan 7

Shimanto Belt in Kyushu Off SW J a p a n

B eB el l t S .. , "
N, Tanega- Upper Lower Shikoku gii Akaishi
Ma I N o r t h l S o u t h ] shima Inner Trench Slope Peninsula Mtns,
O.
Tanabe &
,-. :Op.'.'.'-lnag.a G p : K ~" Ace ~ [ Kumano
Misaki ," G r o u p s , ' , .
• .Mioc. ......
" " • '''" [ . . ' . . . . ,,'..,:
-- . ~ _ f ~ --
,'"
• .. G r o u p
, .' ' ' -- ~ ' " " "
F __
20. . . . , iNichinan
;_GroupF / f- S
? """ " " : &uro/ ,~ . . . .
Group gawa5
°~
~ Oligoc c /H y u g /
a -- ~-r- r 7 t 7M u rfo t o - / i f p Gr°u~r
GrouPs Kumage
~~ ~ 40: f Group
Group # shigawa
~ e c Group £ f
~ J Mikado $ )~ f
D
Pa]aeoc,
F
ef
Fm
F
#? i
Owa3ima .idaka- ~7'Shi- y
Late ¢J- ; f if' J & Nano- gawa 3 manto

Cret.
"~ s ~o,o-
tsuka
ix/ i//i
gawa
Fms. etc
f Groupf Group"

P ,// / ?
o ~ o roupJ
~ Acc = Accretionary prism
~ F

0 ~ ~100-r j~/; P = Piei . . . . . . . . . . . . . . .

aro & / /
Hayama f /
early Pleistocene
Fms. etc

Cret.
120- T = Early t o
= Unconformity
Mid-Miocene
;/ //
/ i -- F -- = Fault

]40 ~ = Deformed strata

_
2-/f _ .,>/ i I:/
= Post-orogenic s t r a t a

Fro. 2. Simplified stratigraphy of the Shimanto Supergroup and immediately overlying strata in four
areas. Many formation names are omitted. The four Kyushu columns show the southward younging
of the Shimanto rocks, which is continued offshore (next two columns, from Okuda et al. 1979). Acc,
P, K and T are based on acoustic stratigraphy.

less than 35%. Moreover, the ratios between the Southern Uplands of Scotland by Floyd
orthoclase and plagioclase and those between (1975): these zones can be traced along strike
orthoclase and microcline alter with stratigra- for up to 200 km (J. D. Floyd, pers. comm.
phy. Teraoka (1979), for example, found that 1980).
the content of feldspar and the microcline/
orthoclase ratios tend to decrease upwards from
the Upper Cretaceous. Okada (1977) disting- Regional geology of the Shimanto Belt
uished three petrographic zones in the Shiman- The best-studied areas (Shikoku and Kii) are
to of Kyushu. Lower Shimanto Group (Cre- described first, then the Akaishi, Kanto and
taceous) sandstones (zone I) are characterized Boso areas to the NE, and finally Kyushu and
by ill-sorted, angular-grained feldspathic are- the Nansei Islands to the SW.
nite containing more than 35% feldspar clasts.
Sandstones of zones n and III are richer in
Southern Shikoku
quartz and the grains are better sorted and
more rounded. These sandstones of zones II In the southern half of the Island of Shikoku,
(lower sequence of the Upper Shimanto Group, the Shimanto Belt extends about 230 km along
lower Palaeogene), and m (upper sequence of strike (Fig. 3): as in other regions, it is much
the same group, upper Palaeogene to lower folded and faulted. In eastern Shikoku, the
Miocene) contain 20-35% and 10-20% of feld- triangle-shaped Muroto Peninsula provides ex-
spars respectively. Thus the Cretaceous, lower cellent continuous exposures along the coast-
Palaeogene and upper Palaeogene in Kyushu line while in the west, the type sections of the
are distinguishable from one another on the Shimanto Supergroup, here 70 km wide, are
basis of sandstone petrography (Okada 1977; exposed along the Shimanto River.
Teraoka 1977, 1979). Interestingly, similar
strike-parallel petrographic zones have been (a) The Lower Shimanto (Cretaceous) Belt
delineated in a supposed ancient accretionary The Cretaceous rocks of this belt are sepa-
complex in the Ordovician and Silurian rocks of rated from a Jurassic subduction complex (the
 O0

"~. o ~,o 2? ...... ~oK~ " "

FIG. 3. Geological sketch map of the Shikoku Shimanto Belt.

(1) Cretaceous shallow marine facies; (2) turbidite facies (wave mark for slumping); (3) slump and olistostrome facies; (4) m61ange facies; (5) basic
volcanics; (6) Eocene conglomerate and sandstone; (7) Eocene and Oligocene shallow marine facies; (8) Miocene shallow marine facies; (9) Miocene
granitic intrusives; (10) faults. (A) Neocomian to Turonian shallow marine facies (Doganaro Formation); (B) Albian turbidite facies (Hayama
Formation); (C, E, G, I) Upper Cretaceous turbidite facies; (D, F, H) m61ange facies; (J) Campanian to Maastrichtian shallow marine facies
(Nakamura and Arioka Formations); (K) Turonian to lower Campanian shallow marine facies (Uwajima Group); (L) Ohyama-misaki Formation;
(M) Naharigawa Formation; (N) Muroto Formation; (O) Palaeogene shallow marine facies (Hirata Formation); (P) Palaeocene? m61ange facies; (Q)
Eocene slump and olistostrome facies; (R) Eocene turbidite facies; (S) Miocene shallow marine facies (Misaki Formation). (Data from Katto et al.
1977; Katto & Taira 1978; Suzuki & Hada 1979; Sano et al. 1979; Katto et al. 1980; Taira et al. 1980.)
 The S h i m a n t o Belt o f J a p a n 9

AB CDEF G H I

~ : ~ MR . . . .
", 1 / i" "" " < ~ . . - 7 S ; ~ :

FiG. 4. Schematic reconstruction of the Shimanto cross-section during the time of Palaeogene
subduction. Letters correspond with Fig. 3. (1) Extensional fracture of oceanic crust in the trench
outer slope; (2) olistostrome of basalts and pelagic sediments mixed with trench sediments;
(3) accreted m61ange; (4) turbidite and slump deposits of perched basin on the trench inner slope;
(5) turbidites in the accretionary forearc basin. BTL = Butsuzo Tectonic Line. (Based on Taira et al.
1980.)

Sambosan Group) by a major thrust fault, the While tight, closed folding occurs locally, the
Butsuzo Tectonic Line (Figs 3 &4) (Taira et al. succession is repeated by imbricate faults which
1979b). They can be divided into: (1) a shallow control the overall structural framework. Slaty
marine facies, (2) a turbidite facies and (3) a cleavage sub-parallel to the bedding is ubi-
m61ange facies (Figs 3 & 4) (Taira et al. 1979a; quitous.
Suzuki & Hada 1979; Taira et al. 1980). Each Ammonites and inocerami have been found
tacies is usually fault bounded. at only a few localities (Katto et al. 1980), but
(1) The shallow marine facies mainly occurs in microfossils, especially radiolarians, have been
three small areas: along the Butsuzo Tectonic found more recently. Although the structural
Line (A of Figs 3 & 4); in the Uwajima area of framework cannot be dated in detail, an impor-
western Shikoku (Figs 3K & 4K); and along the tant ~esult has been obtained (Katto et al. 1980;
southern margin of the Cretaceous belt near the Taira et al. 1980): the oldest age found in this
boundary with the Palaeogene (Nakasuji area) facies is Albian for the Hayama Formation
(Figs 3J & 4J). which occupies the northernmost turbidite belt
Near the Butsuzo Tectonic Line in Kochi (B of Figs 3 & 4). Southwards, several belts of
Prefecture the Doganaro Formation, composed turbidite facies (C, E, G, I of Figs 3 & 4) show
of interbedded sandstone and mudstone, yields overall southward younging. However, in each
Neocomian brackish molluscan faunas, and individual turbidite belt (generally between
Turonian fully marine types (Katto & Tashiro 2000 and 5000 m in thickness) the age, based
1978). In the Uwajima area, the Supergroup on radiolaria and the stratigraphic succession
consists mainly of the Lower Cretaceous indicates a northward-younging trend. This
(Albian) Kitanada Group and the Upper Cre- compares with the model for modern accretion-
taceous (Coniacian to Santonian) Uwajima ary wedges (Seely et al. 1974; Moore et al.
Group (Teraoka & Obata 1975). The former is 1979), supposed ancient analogues such as the
mainly pelitic, the latter is sandy with abundant Ordovician-Silurian Scottish Southern Uplands
molluscs. In the Nakasuji area, several locali- (Leggett et al. 1981) and the Franciscan of
ties yield upper Campanian to Maastrichtian California (e.g. Bachman 1981).
molluscan faunas, serpulid limestones and In western Shikoku, the Coniacian to
Zoophycus trace fossils and there is intensive Campanian turbidite facies (Nonogawa Forma-
bioturbation (Tanaka 1977; Katto & Tashiro tion, G of Figs 3 & 4) is widely distributed.
1979b). This facies, of shallow marine origin, (3) A zone of m~lange facies, which separates
roughly defines the outer margin of the frontal the belts of turbidite facies in some places (D,
arc area, which apparently migrated south- F, H of Figs 3 & 4), consists of a chaotic mixture
wards through Cretaceous times. of blocks of sandstone, chert and basic volca-
(2) The turbidite facies forms a major part of nics in pervasively sheared shale matrix (Fig.
the Cretaceous Shimanto Belt in Shikoku. 5d). The size and shape of the blocks is varied,
Sandstone layers of various thicknesses in- but phacoidal and lenticular shapes with exten-
terbed with black shales and occasional grey sive foliation structures abound. The composi-
and red shales. Slumps and olistostromes are tion of the sandstone blocks is similar to the
common. This facies generally shows an east- turbidite beds, though some differences are
west structural trend with steeply dipping beds. reported (Suzuki & Hada 1979). The m61ange
 e,,,.

Fl6. 5. (a) Huge block of basalt, Muroto Formation (Palaeogene) m61ange complex. East side of Muroto Peninsula, Shikoku. The pillow structures
show the block to be inverted, Height of cliff in view is about 12 m.
(b) Coherently bedded flysch sequence, west of Tatsukushi, western Shikoku. Arrow indicates figure for scale.
(c) Slump facies, Muroto Formation, west side of Muroto Peninsula.
(d) Close-up view of Cretaceous m61ange facies, with phacoidal sandstone lenses. Okitsu coast SW of Kochi, Shikoku.
 The Shimanto Belt o f Japan 11

facies contains less K-feldspar and more are part of an upper-fan facies (Katto & Taira
siliceous matrix than the other facies. Blocks of 1978). It also contains slump and olistostrome
basic volcanics, generally metamorphosed to deposits which may represent base-of-slope en-
the pumpellyite-prehnite facies, range in size vironments.
from gravel to units of mappable size. Pillow The Naharigawa Formation (M of Figs 3 &
structures are commonly observed. In some 4), containing Eocene nannofossils, radiola-
larger blocks, pillow lava is overlain by metal- rians and molluscs, is a flysch sequence of
liferous black mudstone and red radiolarian interbedded turbidite sandstones (up to 10 m
chert. Radiolaria in such cherts are older than thick) and shales, with minor amounts of tufts
those of the argillaceous 'matrix' of the m61ange (Fig. 5b). Trace fossils include Helminthoidea,
and surrounding turbidite facies (Taira 1979; Nereites, Paleodictyon and Spirorhaphe sug-
Nakaseko et al. 1979; Katto et al. 1980). For gesting deep-water deposition, a deduction sup-
example, in belt D (see Figs 3 & 4), the facies ported by recent finds of traces of the last two
contains Valanginian red radiolarian chert rest- genera on the surface of modern ocean sedi-
ing on basaltic pillow lava, but is embedded in a ments at depths ranging from 1436 to 3895 m
Coniacian to Santonian 'matrix'. In the south- for Paleodictyon and from 3358 to 5119 m for
ern m61ange belt (H in Figs 3 & 4 and other Spirorhaphe (Ekdale 1980). Palaeocurrent
localities), Albian red radiolarian chert blocks analysis indicates east to west dominated axial
are found with basaltic lava blocks and this current directions (Fig. 1) (Katto & Arita
m61ange facies is overlain by Coniacian to 1966).
Campanian turbidites. These data indicate that To the south the Muroto Formation (Figs 3N
the lavas and cherts within the m61ange facies & 4N) is a m61ange complex containing slump-
are emplaced older exotic blocks from an olistostromes (Fig. 5a,c) and flysch. The
oceanic plate mixed with presumed arc- or m61ange consists of blocks of ocean-floor
continental margin-derived granitic clastic de- hyaloclastite, basalt (Sugisaki et al. 1979),
tritus. chert, tuff and sandstone in pervasively sheared
shale and tuff matrix. The m61ange material is
(b) The Upper Shimanto (Palaeogene) Belt not yet well dated, but radiolarians from the
The lithofacies of the Palaeogene rocks in sheared grey shale indicate a Palaeocene and
Shikoku resemble the Cretaceous rocks, but the early Eocene age, the oldest age so far obtained
overall facies relationships are better estab- in the Muroto-hanto Group.
lished and their depositional setting can be Large slump-olistostrome deposits are com-
deduced in more detail. mon in the Muroto Formation, especially in a
The representative Muroto-hanto Group of major zone ranging up to 700 m in thickness.
south-eastern Shikoku illustrates the rela- The flysch facies consists of a coarsening up-
tionships. Nannofossils, radiolarians and mol- ward sequence: shale, some of it red, interca-
luscs suggest an age from early Eocene to lated with tuff passing up into numerous turbi-
Oligocene (Katto & Tashiro 1979a). The group dite sandstone units. Radiolarians suggest an
consists of three units (from north to south) : Eocene age for the post-m61ange flysch facies.
the Ohyama-misaki, Naharigawa and Muroto Recent work on the Tsuro Formation at the
Formations (L, M, N of Figs 3 & 4) (Katto et al. south tip of the Muroto Peninsula has revealed
1961; Katto & Taira 1978). Foraminifera and Radiolaria of lower Miocene
The Eocene Ohyama-misaki Formation (Figs age (Taira et al. 1980).
3L & 4L) consists of conglomerates, sandstones
and shales formed within a fining-upward de- (c) Sedimentation and tectonics
positional cycle. The conglomerate beds, which The lithofacies and structural pattern of the
show channelling, scour-and-fill, large-scale Shikoku Shimanto Belt resembles other exam-
grading and trough cross-bedding, contain large ples of forearc basin deposits resting on an
boulders of various lithologies interpreted as accrenonary complex (cf. Dickinson & Seely
derived from the frontal arc areas including the 1979, fig. 3, lower left). The overall southward
Cretaceous Shimanto, Chichibu and Sambaga- younging trend of the Shimanto Supergroup
wa Belts. Sandstones show graded bedding, shows that the accretionary process was domi-
parallel- and trough cross-bedding and abun- nant through Cretaceous and Palaeogene times.
dant rip-up clasts typical of some turbidite The conditions of deposition are best
sequences. This formation is interpreted as documented in the Muroto-hanto Group (see
sumbarine channel- or canyon-fill conglomer- Fig. 4), with the Ohyama-misaki and Nahariga-
ates and sandstones, and over-bank, thin- wa Formations interpreted as Eocene accre-
bedded shales and sandstones, which possibly tionary forearc basin sediments (5 of Fig. 4),
12 A . Taira et al.
the former representing submarine channel (or facies, but the m61ange facies contains blocks of
canyon) fill and slope sediments and the latter older age.
representing a possible submarine fan complex These features of the m61ange facies indicate
that filled a major part of the accretionary original formation as olistostrome deposits,
forearc basin (Katto & Taira 1978; Taira 1979). possibly in a trench (Taira et al. 1980). The
The Muroto Formation can be interpreted as oceanic plate materials such as basaltic pillow
trench-slope break to trench inner slope de- lava and radiolarian chert may have been de-
posits and accretionary basement (subduction rived from the faulted scarps of the oceanic
complex) (Taira 1979; cf. Bachman 1981). The crust or seamounts on the outer slope of the
broader m61ange zone which contains the large trench (see 1 and 2 of Fig. 4) as has been
metabasalt complex possibly makes up a base- observed in the Peru-Chile Trench (e.g. Kulm
ment high at the trench-slope break. The flysch et al. 1981). Such trench-filling olistostrome
and slump facies separated by the m61ange zone deposits may have been accreted to the arc
with imbricated thrust faults are interpreted as front together with arc- or continental margin-
small perched accretionary basins on the trench derived clastic sediments (3 of Fig. 4). Most of
inner slope (4 of Fig. 4) similar to those de- the pelagic sediments and the oceanic lithos-
scribed by Karig & Sharman (1975), Moore & phere itself have probably been subducted.
Karig (1976) and Lewis (1980).
Similar arrangements of lithofacies can, with Kii Peninsula
some variations, be deduced throughout the
Palaeogene Shimanto Belt. In Western Shiko- Exposures of the Shimanto Supergroup con-
ku, there is an extensive development of slump tinue eastwards from Shikoku into the Kii
deposits containing blocks of shallow marine Peninsula (Figs 1 & 6). There, between the
origin such as limestones containing large Fora- Butsuzo Tectonic Line and the Pacific Ocean,
minifera, and sandstones rich in molluscan fos- thick accumulations of Shimanto Supergroup
sils (Hiromi m61ange) occurring in an upper sediments are characterized by dark mud-
slope setting (Figs 3Q & 4Q). stones, turbidites and conglomerates, and the
Similar features occur in Cretaceous sedi- whole sequence is deformed (Fig. 6, sections).
ments (see Fig. 4A-K); shallow marine facies Here the Shimanto has been studied by the
suggest deposition south-eastwards of an ex- Kishu Shimanto Research Group led by T.
posed frontal arc area. Contemporaneous turbi- Harata of Wakayama University and T.
dites (e.g. the Nonogawa Formation) were Tokuoka of Shimane University (Kishu Shi-
probably filling an accretionary forearc basin manto Research Group 1970, 1975; Harata et
(cf. the Naharigawa Formation during al. 1978). Central to much of their discussions
Palaeogene times). The other turbidite facies of the Shimanto geology of the Kii Peninsula is
were either trench fill deposits or filled small their belief in the existence of a pre-Miocene
perched basins on the trench inner slope. The ancient Kuroshio continent once lying to the
abundant slump and olistostrome deposits south. This account owes much to the work of
within this facies indicate that there was much the Research Group but the conclusions here
active slumping of the turbidite cover as well do not necessarily reflect the views of the
as gravity failure of accreted m61ange base- group.
ment. As in other areas, the geology of the Kii
The origin of the m61ange facies (Fig. 5d) is Shimanto Supergroup reflects a migration of a
not clear, but the following observations set depo-centre towards the SE. The oldest Shi-
some constraints: (1) Oceanic plate materials manto sediments are assumed to be Late Cre-
incorporated into the Shimanto terrane are a taceous on the basis of inoceramid fossils (I.
very small proportion of the whole, occurring (Mytiloides) aff~ labiatus and I. (Platyceramus)
only as blocks. There is no evidence of amakusensis), though these may have been
wholesale incorporation of oceanic crust itself transported into the basin, whilst the youngest
into the Shimanto Supergroup. (2) The blocks are lower Miocene in age on the basis of their
in the m61ange facies show variety of size and molluscan fauna (see Kishu Shimanto Research
shape and are interpreted as olistostrome de- Group 1970).
posits. Bedded chert and sandstone blocks The Kii Shimanto Supergroup has been di-
show slump folding. (3) The shale 'matrix' is vided into three groups: the Hidakagawa, Oto-
often very sheared, with abundant quartz veins. nashigawa and Muro Groups (see Fig. 2). Each
In some places, it shows a schistose appearance. group is separated by major reverse faults or
(4) The age of the m61ange material indicates a tectonic lines with steep dips to the north (Fig.
southward younging trend, as for the turbidite 6).
 The Shimanto Belt of Japan 13

,~ ~.o:~.~.-'g" W A GROUP ~ '

--34°N r ~ ...... " y..lxG ~' ____ .~____L

07" 9 . ~ -, -,-.-. g" ~ \ ~

-'4 / ;,- ~e...,~ ..,, Go-,o - H ~ .......... . ..'L-"'~'~.vJ,';~;~-.~

,,-,, "i., o~O~>7~.,.'..~

.............
"~Yn'.",::~ I
,,',-,',, ,-, "...~~ . . . . ',~'_...,-:~:W
. . . . - ......

o ~o ~OK~ tc~'.. <u°'~-. t~il'~+~' .-" >-"

, ' ' "-" --,Ooo,~ "----:.i~'.-- "/
/

(a) M2 (b) ii: .:. ~. ~,l

M&U LOWERMIDDLE&UPPER PARTSOFTHE MUROGROUPRESPFCTIVELY

HIDAKAGAWA GROUP OTONASHIGAWAGROUP MUROGROUP KUMANO 0 ] KM

tH 1 - 3 ) (01-3) (M 1 - 3 ) GROUP(KI I
0 20KIM

FIG. 6. Map and sections of the Kii Peninsula. The Hidakagawa, Otonashigawa and Muro Groups of
the Shimanto are overlain unconformably by the Miocene Tanabe and Kumano Groups. Also shown
are offshore exposures of the Tanabe Group (T) in submarine canyons in the outer Kii Strait. TL
= Tectonic Line. Section (a) is along AA t (A.J.S.), (b) is a detailed E-W section through part of the
Muro Group (after Suzuki 1975).

(a) Hidakagawa Group ing Ryujin Formation. More than 7 km thick, it

Exposed between the Butsuzo Tectonic Line is composed of massive, usually sandy, turbi-
in the north (see Fig. 6) and the Gobo-Hagi dites, which may be associated with basalts and
Tectonic Line, it is composed of three forma- radiolarian cherts, and subordinate shales.
tions, with complex structural relationships. Roughly contemporaneous with Hidakagawa
The oldest, the Nyunokawa Formation, is sedimentation, turbidite deposition of the Izu-
more than 2 km thick: it and contemporaneous mi Group was taking place in a narrow basin
formations are found near the Butsuzo and the north of the Median Tectonic Line (Fig. 1).
Gobo-Hagi Tectonic Lines. The lower part is
generally finer grained, composed of shales and (b) Otonashigawa Group
muddy turbidites, while higher the turbidites Only differentiated from the succeeding
become sandier and massive. Conglomerates, Muro Group relatively recently, it is about
often coarse, are intercalated in the upper part 1.5 km thick (Hatenashi Research Group 1975)
and contain a wide variety of rock types: and is exposed between the Gobo-Hagi Tecto-
rhyolites are abundant, with granites, sand- nic Line and the Hongu fault. Though no
stones, cherts and some limestones. The suc- reliable palaeontological evidence has been dis-
ceeding Ryujin Formation is about 3 km thick covered, the Otonashigawa Group has been
and finer grained, being composed of shales ascribed mainly to the Eocene. Again the
and muddy turbidite sequences with only occa- lowest formation (the Uridani Formation) is
sional more sandy sequences. Basalts with pil- predominantly shaly and again there are basalts
low lavas and rhyolitic tufts occur. and acid tufts. The middle formation, the Low-
The youngest formation is the Miyama er Haroku Formation, is the most extensive
Formation and it lies to the north of the preced- formation and is somewhat sandier than the
14 A. Taira et al.

Uridani Formation. The Upper Haroku Forma- tion to the synsedimentary tectonic structures,
tion completes a coarsening upward sequence with major folds being regarded as slump struc-
and in the southern outcrops includes the tures, and the occurrence of clastic dykes. Much
Kizekkyo conglomerates which contain clasts remains to be done in the Muro Group areas to
up to boulder size. Sandstone clasts predomin- determine the relationship between slump de-
ate, with chert, granite, acid volcanics and posits and the early 'folding'. Here, much of the
limestones less abundant. folding must have been penecontemporaneous
(c) Muro Group with sedimentation for all the folding and much
The most southerly and youngest of the Kii of the faulting was completed before the uncon-
Shimanto Supergroup is estimated to be 7.5- formable deposition of the sediments of the
9 km thick, and fossils, though possibly trans- Tanabe and Kumano Groups of middle
ported, indicate an Oligocene to early Miocene Miocene age. Palaeontological and radiometric
age. Again the lowest formation, some 1200 m dating clearly show that only a very short time
thick, is mainly fine grained whilst the middle separated these groups from the Muro Group.
formation is thicker and sandier with some The last events in the Shimanto succession of
intercalated conglomerates. The uppermost Kii reflect proximal sedimentary conditions in-
part is between 3 and 4 km thick and is of mixed cluding slumps, locally derived olistoliths,
composition: shaly horizons are associated with rounded quartzite pebbles, washouts, clean or
conglomerates, which have muddy and sandy relatively clean conglomerates and even ripple
matrices, and breccia-bearing mudstones occur. marks, all affected by penecontemporaneous
Large olistoliths, up to several metres across, folding. The Kishu Shimanto Research Group
occur near the southern limit of the peninsula (1970) sees much of this as evidence of the close
(Fig. 7b). The clasts in the conglomerates are proximity of a now-lost southern 'Kuroshio
composed predominantly of acid volcanic rocks continent'. However, recent work by Tokuoka
and also include cherts, orthoquartzites, sand- & Okami (1979), although they still insist on a
stones, granites, and limestones. Spectacular supply from the south, suggests to us that in the
slumps occur (Fig. 7a) and clastic dykes are Palaeogene the orthoquartzite pebbles were
common. The character of this formation sug- recycled from the Lower Cretaceous Tetori
gests that it was deposited near its source, Formation, lying to the north. A boulder of
though many of the pebbles, including the granitic gneiss from the Muro Group gave K-Ar
orthoquartzites, are extremely well rounded: ages of 63.3 Ma on biotite and 70.4 Ma on
their possible origin is discussed later. muscovite, suggesting that the provenance of
the gneiss may be the Ryoke metamorphic belt
(d) Sedimentation and tectonics (Shibata & Nozawa 1973).
Essentially, these Shimanto Supergroup suc-
cessions show three sequences represented by Akaishi and Kanto Mountains
the three groups, each coarsening upward and
implying either a simple shallowing or an in- The Shimanto rocks of this deeply dissected
creased input from the source area. Each group mountain region have been described by Oga-
is separated by some sort of hiatus, though it is wa & Horiuchi (1978) and Ogawa (1981). The
not clear if this is solely due to tectonic events. Akaishi Belt trends NNE-SSW and is separated
The tectonic lines themselves must represent from the Ryoke, Sambagawa, Chichibu and
major planes of movement since they now Sambosan Belts on the west by the Butsuzo
effectively separate the three Shimanto groups Tectonic Line and from Miocene groups on the
(Fig. 6). east by the Itoigawa-Shizuoka Line. Deforma-
The palaeocurrent directions for the different tion is represented by large-scale recumbent
groups (for details see Fig. 8) reveal a predomi- folds in the west and a north-south trending
nantly east to west transport, but both southerly anticlinorium in the east. Metamorphism is of
and northerly directions are important. In the the usual pumpellyite or greenschist facies. The
Muro Group, there are some west to east Shimanto rocks, several kilometres thick, con-
directions. sist of turbidites, sandstones, mudstones and
The Supergroup is much folded and faulted slump deposits, with local limestones and chert,
(Fig. 6, sections), with steeply dipping, vertical, together with basalts and rhyolitic tufts in the
and overturned strata common and much strike lower and middle parts. The lower group, 'Shi-
faulting. Harata et al. (1978) comment that manto s.s.' is Jurassic (in part) and Cretaceous
there has been a 30-40% shortening in the Kii in age (Fig. 2). The upper, Setogawa Group, is
Peninsula south of the Butsuzo Tectonic Line. of Palaeogene age, probably passing up into the
In the Muro Group, Suzuki (1975) draws atten- Miocene.
 °J

.....

d
FIG. 7. (a) Large slump structure, Muro Group, S Kii. Arrow indicates figure for scale.
(b) Olistoliths in mudstone, Sarashikubi Formation, Muro Group, S Kii. The block to the left of Dr Shiki is approximately 4 m in diameter. The
matrix forms the wave-cut platform in the foreground.
(c) Thinner bedded sandstones and shales within the thick cross-bedded sandstone facies, Kumage Group (Palaeogene), N of Nishino-omote,
Tanegashima. Mainly shale (right) and mainly sandstone (left) forming a coarsening- and thickening-upward deltaic sequence.
(d) Slumping of plastically deformed packets of thin sandstones and mudstones, Kumage Group (Palaeogene), Minato, N Tanegashima. Compare the
deformation of the Aberystwyth Grits, Wales (Davies & Cave 1976). Hammer scale arrowed.
16 A. Taira et al.

IV. MIDDLE MIOCENE

~,5--

.......... ~..~-"-" ,,
KUMANO I

TANABE I)P.
t

(
III. OLIGOCENE TO EARLY MIOCENE \
\

II. ?EOCENE

GOf50- H/kGI T.L. .--'" .-" \

. ~" IO','ON,",SH,GAWAO,'.I
II
I. LATE CRETACEOUS il m • • m J• gl~lll
! ( I z u M , Gp.)L -/

_. IIII II ...... .
~l~ CURRENT DIRECTIONS

T,L, TECTONIC LINE

FIG. 8. Changing palaeocurrent patterns through the Kii Shimanto and up into the middle Miocene.
 The Shimanto Belt of Japan 17

In the east-west-trending Kanto Belt, the ment of Shimanto rocks. This relatively thin
northern Ogochi Group, Lower to Upper Cre- group (1000 m + ) is made up of hemipelagic
taceous in age, consists of quartz/feldspar sand- siliceous shale, flysch-type sandstones, shales
stones bearing Inoceramus, and mud-rich and olistostromes, with basalts and serpenti-
slumps. In the south, the Upper Cretaceous nite. Previously thought to be of Palaeogene
Kobotoke Group consists of sandstones, mud- age (Imai 1977), Miocene-type radiolaria have
stones and turbidites forming sandy deltaic recently been found in the lower part of the
aprons marked by slumping: basalts, dacitic Mineoka Group (K. Nakaseko, pers. comm.
tufts and chert are also present. There are no 1980).
Palaeogene strata in the Kanto Mountains. We now return to the Shimanto develop-
These Shimanto rocks are interpreted by ments SW of Shikoku.
Ogawa & Horiuchi (1978) as arc-trench gap and
trench slope sediments that were later de- Southern Kyushu
formed by collision with the northward-moving The Shimanto of Kyushu (Fig.9) is compli-
Izu microcontinent, lying at the northern end of cated in structure as shown in a map by Hashi-
an aseismic ridge (Matsuda 1978): transcurrent moto (1962). In the Northern Belt, the Lower
movements were also important (Ogawa 1981). and Upper Cretaceous Morotsuka Group
(Fig. 2) consists of phyllites, sandstones and
Boso Peninsula mafic volcanic rocks with minor cherts in its
The Mineoka Group in the southern Boso lower part and mainly sandstones in its upper
Peninsula represents the easternmost develop- part'(Sakai 1978). In the Southern Belt, the
I • ,>/ ~^~ SHIKOKU
131°

33 o /.f
KYUSHU

17 11 I"
~, 132 °
,.~
Shimizu Belt

/
133 °

uga Belt

/ \~ # - _
t I \\~ ~32 o _~ Su4 Upper 8
I ~ ~TrmTrm~

i ._J _ o_

'-" \ f "~, rlTllTn Su2 Lower o

~,

.~ SI3 Upper ~

"~r .~ F 6_ SI2 Middle

o-~'~ ~ O 5 0 km L0w.Cref.l~
"-,~': I L i a I
131 ° 132 °
I I

FIG. 9. Map of the Shimanto of Kyushu and West Shikoku. MTL = Median Tectonic Line; UYTL =
Usuki-Yatsushiro Tectonic Line; BTL = Butsuzo Tectonic Line; NT = Nobeoka Thrust; SI = Lower
Shimanto; Su = Upper Shimanto. (Teraoka 1979, fig. 10, reproduced by permission of the author
and the Geological Society of Japan.)
18 A . Taira et al.

highly deformed Palaeogene Mikado Forma- claystones with abundant radiolarians (inten-
tion, of limited extent, is rich in pillowed sively studied by Nakaseko et al. 1979) are
basalts: it is separated from the Cretaceous Belt closely associated with metabasalts.
by a northward-dipping thrust. Other lower
Tertiary rocks are the Hyuga Group (Eocene to Northern Nansei Islands
early Miocene) in the north and the Nichinan
Group, shaly in the lower part and sandy The best studied Shimanto rocks, the Eocene
above, in the south: this group is late Eocene to to Oliogocene Kumage Group, are well ex-
early Miocene in age. posed on Tanegashima (Okada & Whitaker
The petrographic work on the sandstones of 1974; Suzuki et al. 1979, Hayasaka et al. 1980).
Kyushu has already been described (p. 6). Rocks of similar age occur on Yakushima
Further details on the stratigraphic division of (Hashimoto 1956) and other smaller islands.
the Shimanto of Kyushu and the distribution of The Kumage Group occurs in tight folds,
sandstone compositions are given by Imai et al. including isoclines, with parallel axial planes
(1975), Okada (1977), Teraoka (1977, 1979) trending NNE-SSW, which are separated by
and Teraoka et al. (1979). In the last three many thrust planes. The group consists of three
papers, the sandstone petrography of western tectonostratigraphic units, separated from each
Shikoku is also examined. other by thrusts. They comprise: (1) a flysch
At an early stage of sedimentation of the facies with variegated claystones, (2) a chaotic
Lower Shimanto Group, Precambrian gneisses facies and (3) a thick cross-bedded sandstone
and older Mesozoic granites in the Korean facies.
Peninsula and their equivalents on the Asian (1) The flysch facies, dominant in southern
continent may have been important as the Tanegashima and more than 2900 m thick,
provenance of the Shimanto sediments. Yaska- consists of 1(~50 cm thick turbidities, normally
wa (1975) deduced from palaeomagnetic stu- interbedded with mudstones of similar thick-
dies that in the Cretaceous, southern Japan was ness, with some thin tufts and accompanied by
to the east of today's Korea. The Ryoke gra- radiolarian-bearing black, grey, green and red
nites and acidic igneous rocks in the Inner Zone claystones near the base. The radiolarian
of SW Japan may have played increasingly assemblage is assigned to the Thyrsocyrtis
important roles as source rocks for the Shiman- triacantha Zone (Riedel & Sanfilippo 1974), i.e.
to sediments at the later stages of development. latest early Eocene to earliest middle Eocene
This is supported by palaeocurrent data, (Suzuki et al. 1979), but attempts to date the
summarized in Fig. 1: i.e. current directions are lutites palynologically have not been successful
mostly axial, but lateral currents from the north due to heat degradation (C. Downie, pers.
were also consistently present and, sporadical- comm. 1978). Within the flysch facies, some
ly, currents came from the south. Some geolog- slumping occurs.
ists (e.g. Harata et al. 1978), underlining the (2) The chaotic facies, more than 700 m thick in
importance of the southerly currents have, as the SW of the island, is characterized by peb-
noted earlier, advocated a 'Kuroshio bles and blocks set in scaly claystones. The
Palaeocontinent' composed of Pre-Cambrian sedimentary clasts range from a few millimetres
rocks to the SE of the Shimanto basin, which to tens of metres in diameter. A large block of
contributed sediments that were deposited on pillowed basalt is also incorporated in claystone
Precambrian basement. We prefer that the matrix surrounded by sandstone.
Shimanto sediments of Kyushu, as in Shikoku (3) The thick cross-bedded sandstone facies,
and Kii, were deposited in basins analogous to dominant in northern Tanegashima and around
modern accretionary forearc basins and partly Shimama in the SW, is more than 1300 m thick.
to trench inner slope basins and trenches. This facies is characterized by well-sorted aren-
Therefore, tectonic highs like the trench-slope ires in beds up to tens of metres thick. They are
break (Dickinson 1973), active during sedi- generally of medium sand grain-size and seem
mentation, could have controlled the north- to be lacking in pebbles. Large-scale cross-
ward-flowing currents and may sometines have bedding (Fig. 10a), ripple-marks, ripple-drift
provided a supply of new or recycled clasts (cf. lamination, mega-ripples (Fig. 10b), parting
the Oyashio 'landmass' in the DSDP Japan step lineation and dewatering structures are well
Trench transect (von Huene, Nasu et al. 1978)). developed. The only sole markings recorded
As in Shikoku and Kii, the greenstone facies are very large flute casts and load casts. When
associated with the flysch of Kyushu is of slump structures occur, they are on a spectacu-
oceanic origin (Kanmera 1976; Suzuki & Hada lar scale (Fig. 10c).
1979). Again the siliceous red and variegated Between the sandstones are dark shales and
 The Shimanto Belt of Japan 19

FIG. 10. Thick cross-bedded sandstone facies, Kumage Group (Palaeogene), Tanegashima.
(a) Normal bedding at base of photograph, foreset bedding behind hammer; (b) Mega-ripple;
(c) Large slump fold in massive sandstone.
20 A . Taira et al.

mudstones bearing siderite nodules and trace phyllites and mafic volcanic rocks, is in thrust
fossils: thin coaly beds and lenses are occa- contact with the Eocene Kayo Formation
sionally found above seat-earths. Hayasaka et (850 m) of the Southern Belt, which consists of
al. (1980) recently reported molluscan fossils flysch-type sandstone-shale alternations. On
suggesting an early Miocene age. The thin the basis of Nereites-assemblage trace-fossils,
siltstones and intermediate sandstones (up to Fukuda & Hayasaka (1978) consider their
1 m) which are also present in this facies show depth of formation to be 3500-5500 m. On
convolute lamination, slump structures, ripple- Amami-oshima, the Northern Belt contains
drift lamination, small-scale cross-bedding, thin phyltites and pillowed basalts of the Naze
clastic dykes and tubular trace fossils. Top Formation (1300 m) overlain by Upper Cre-
surfaces sometimes show ripple marks and, at taceous flysch sandstones and shales of the
two localities, sand volcanoes (Okada & Whi- Ogachi (2400 m) and Tatsugo (1000 m) Forma-
taker 1979). Flute, groove, striation and gutter tions, that show coarsening- and thickening-
casts (Whitaker 1973) occur and, less commonly, upward cyclic sedimentation. These in turn are
longitudinal ridge-and-furrow, prod, bounce, covered by Eocene sandstones and shales of the
brush and crescentic scour casts. From these Wano Formation (800 m) (Ishida 1969; Sakai et
sole markings and the foreset beds of the al. 1977). This slightly metamorphosed sequ-
sandstones, a complex palaeocurrent pattern ence is severely deformed into isoclinal folds
emerges (Okada & Whitaker 1974). and SE-verging thrust faults (Kizaki 1978).
This facies shows thickening- and coarsening-
upwards cycles (Okada 1971, 1973), (Fig. 7c)
and well-developed slumps (Fig. 7d). C r e t a c e o u s b a s i n s n o r t h o f the
S h i m a n t o Belt
Environments o f sedimentation
These three facies are interpreted as having Although outside the Shimanto Belt, two types
been laid down in the following environments: of basin lying to the north received Cretaceous
(1) The flysch deposits, with variegated radiola- sediments that may be related to the Cre-
rian shale suggesting deep water, may have taceous Shimanto (Teraoka 1977: Taira 1979;
been deposited in upper to mid-slope perched Taira et al. 1979a).
basins on an inner trench wail.
(2) The chaotic deposits may be part of an Forearc shelf basins
accretionary complex beneath the flysch, form-
ing a trench-slope break ridge, the pillowed These occur within the Chichibu Belt and
basalt block having been incorporated into this show three correlatable fining-upward deposi-
sequence from the ocean floor during mdlange tional megacycles in the Lower Cretaceous and
formation: slumping off this ridge could one in the Upper Cretaceous. Each cycle lasted
account for disturbances in the other two facies. 10-20 Ma, similar to cycles in the accretionary
(3) The thick cross-bedded sandstone sequence forearc basin, and is related to synchronous
may be forearc basin deposits on the arc side of vertical movements of SW Japan.
the trench-slope break ridge. The coals and
sideritic nodules suggest a delta top environ- Intra-arc basins
ment, with the thick sandstones as shallow
marine sandy delta deposits. The siltstones and (i.e. between the volcanic and frontal arcs).
mudstones, sometimes slumped, could have These formed during the Late Cretaceous. The
formed on an unstable delta-front slope. The best-known are the Mifune, Goshonoura, Ono-
variable palaeocurrents would support this in- gawa, Himenoura and the 300 km long Izumi
terpretation. Basins (Fig. 1) which, like the forearc shelf
If our conclusions are correct, it follows that basins, had a life-span of about 10 Ma. Coarse
facies (1) and (3) may be synchronous in their turbidites accumulating by longitudinal accre-
times of sedimentation and younger than facies tion at 1-4 m/1000 yr gave enormous thicknes-
(2). Unfortunately, fossils are too sparse or not ses of sediment up to 40 km thick in the Onoga-
sufficiently diagnostic in the Tanegashima Shi- wa Basin (Teraoka 1970) and 50 km thick in the
manto Rocks for confirming this suggestion. Izumi Basin (Suyari 1973) in fault-bounded
troughs (compare the 25 km thick Devonian
Southern and Central Nansei Islands sediments of the Hornelen Basin, Norway
(Steel & Gloppen 1980)). Synclinical in struc-
On Okinawa, the Early Cretaceous Nago ture, these basins were filled from east to west
Formation of the Northern Belt, consisting of but closed from west to east (Suyari 1973; Taira
 The S h i m a n t o Belt o f Japan 21
+
*,
St_

MTL

BTL

FIG. 11. Schematic block diagram showing inferred environments of deposition of the Lower
Shimanto Group during the Cretaceous (not to scale) (after Niitsuma et al. 1979; Taira 1979; Taira et
al. 1979a). (1) Seaward limit of Ryoke high T/P belt; (2) Intra-arc basins (Izumi, Onogawa, etc.),
mainly Upper Cretaceous, with very thick longitudinally and rapidly deposited clastics; (3)
Sambagawa low T/P and Chicbibu Belts, with (4) small forearc shelf basins with thin fluvial and
shallow marine sands and clays, mainly Lower Cretaceous; (5) to (8), Lower Shimanto accretionary
forearc basin; (5) deltaic sandstones and clays; (6) flysch in deeper water with variable current
directions; (7) submarine canyon, fan channel and fan system; (8) slumps and olistostromes from
trench-slope break; (9) trench-slope break (partly submergent, partly emergent); (10) slumps and
olistostromes on upper inner trench wall; (11) turbidites and (12) canyon-fan system contributing
sediments to perched basin (13); (14) slumps and olistostromes on lower inner trench wall; (15)
canyon-fan system and (16) deep-water turbidites feeding trench (17). MTL = Median Tectonic
Line; BTL = Butsuzo Tectonic Line; Acc = Accretionary prism wedges separated by thrusts with
m61ange (M); OP = intermittently subducting oceanic plate. Conditions during the Palaeogene for
the Upper Shimanto Group are visualized as being broadly similar, but with the Cretaceous Lower
Shimanto deformed and exposed to erosion and new accretionary forearc basin, trench-slope break
and trench developing further towards the right (SE).

1979). Filling was cyclical: the Himenoura the Chichibu and Sambosan sediments on to the
Basin, for example, accumulating five cycles earlier paired metamorphic belts, a long, nar-
each of 1-2 Ma duration (Tashiro, Taira & row, NE-SW trending forearc basin (or series of
Matsumoto, in preparation). The alternating basins) developed above an accretionary com-
tensional and compressional phases affecting plex to the SE of the Chichibu and Sambosan
these basins are attributed by one of us (Taira belts. Environments cited in the following
1979) to subduction proceeding in cyclical man- account are numbered in Fig. 11. In these
ner. We shall return to this idea later, after accretionary forearc basins (see Fig. 11-6) great
considering the tectonic and sedimentary set- thicknesses of Cretaceous sandstones and turbi-
ting of the Shimanto Supergroup. dites accumulated, fed by rivers draining the
Eo-Japanese landmass to the north and the
Korean region which then lay directly west of
Kyushu (Yaskawa 1975). The shallow water
Evolution of the Shimanto Supergroup
sandstones that accumulated on the shelf and in
Towards the end of Jurassic times, after north- deltas (5) were relatively angular and coarse
ward subduction of the Kula Plate had accreted grained, ill-sorted and rich in feldspars (espe-
22 A . Taira et al.

cially microcline) but with K-feldspar content trench slope. On the trench inner slope,
decreasing through Cretaceous times. Lateral perched basins trapped flysch and slump de-
and axial turbidity currents (6) and canyons and posits. Blocks in the slumps of western Shikoku
fans (7) were active. As subduction proceeded, and the later slumps of Kii are shallow-water
accretionary wedges at the outer margins of the sandstones and limestones that came to rest in
forearc basins continued to form a positive and upper trench slope basins.
active trench-slope break (9). This brought During the early to middle Miocene, the
older m~langes and olistostromes (with large Palaeogene Shimanto suffered the same fate as
basaltic blocks and red chert derived from an the Cretaceous Shimanto, becoming tightly
oceanic trench or its faulted outer wall) into folded, sliced into fault blocks, thrust into
shallow waters or above sea-level, and slump accretionary wedges and, in places, mildly
masses (8) and turbidity currents were shed
northwards into the forearc basins and south- /x
wards down the inner wall of a deep-sea trench
(10), (11), (12). Terraced basins (13), perched
on the inner trench wall, accumulated some of
these slump deposits and turbidites while the
rest of the clastic detritus remained on the
I
slopes of the trench wall or reached the bottom
of the trench (17) via turbidity currents (16),
slumps (14) or through canyons feeding fans
(15).
North of the Butsuzo Tectonic Line, Lower
Cretaceous forearc shelf basins (Fig. 11-4) with
thin fluvial and shallow marine sands and
offshore muds rest on the Chichibu sequence
and further north, beyond the Median Tectonic
Line, Upper Cretaceous sediments of excep-
tional thickness occupy down-faulted intra-arc B
basins.
Towards the end of the Cretaceous, con-
tinued subduction had deformed, mildly meta-
morphosed and exposed some of the accretion-
ary forearc basin deposits and created younger
accretionary wedges towards the SE. Thus, new
forearc--trench-slope break--trench inner
slope--trench environments evolved, ocean-
ward of the Cretaceous ones, and these lasted
through the Palaeocene, Eocene and Oligocene
into the early Miocene. The Palaeogene accre-
tionary forearc basins were now fed mainly
from the newly exposed Lower Shimanto (Cre-
taceous) rocks. In consequence, the Palaeogene
forearc sandstones tend to be less angular, finer
grained, better sorted and more quartz-rich
than their Cretaceous equivalents, with less FIG. 12. (A) At the back of the diagram,
feldspars and, in particular, diminishing compression results in accretionary prism
amounts of microcline upwards. Sandy deltas developing on inner trench wall and keep-
occupied the arc side of the forearc basin in the ing the intra-arc basin closed. At the front,
SW (Tanegashima area) while submarine fan the steepening dip of the downgoing slab
complexes with channels built out to fill the gives relative tension, resulting in the
opening of an intra-arc rift-type basin.
major part of the central forearc basins (Shiko-
(B) At the back of the diagram, there is
ku region). relative tension, as in (A), front. At the
The Palaeogene trench-slope break consisted front, partial detachment of the slab begins
of a broad m61ange zone containing a large a new cycle of subduction with compress-
metabasalt complex forming a basement high. ion closing the intra-arc basin. (From Taira
This zone, like its predecessor, shed detritus 1979, fig. 6, with permission of The Earth
north into the forearc basin and south on to the Monthly, Kaiyo-shuppan Co., Tokyo.)
 The Shimanto Belt o f Japan 23

metamorphosed. Their uplift and erosion arc basins, as described by Lewis (1980) in New
caused clastic sediments to be shed into new Zealand (Whitaker 1982).
basins, such as those of the Miyazaki Group in The northward movement of the Kula Plate
southern Kyushu and the Tanabe and Kumano in Shimanto times changed to a westward
Groups of western and eastern Kii respectively movement of the Pacific Plate with transform
(Fig. 2). Many of these Miocene deposits may motion parallel to the Shimanto trend in its
be traced offshore (the Tanabe Formation eastern part. This, followed by the opening of
being exposed in the walls of present-day sub- the Shikoku backarc basin, the Pliocene-Recent
marine canyons, Fig. 9) and, with Pliocene and renewed northward subduction under Shikoku
Pleistocene deposits, make up the fill of today's and H o n s h u and the northward m o v e m e n t of
terraced forearc basins. The modern trench- the Izu Block to collide with the Akaishai &
slope break is flanked on the Pacific side by Kanto Shimanto Belts (Matsuda 1978) m a k e a
the inner wall of the Nankai Trough (Fig. 2, complicated but intriguing sequel to Shimanto
columns 5, 6) with basins perched on accretion- events (Smith 1982).
ary wedges (Inoue et al. 1977).
This shift SE of Shimanto forearc basin to
trench environments from their Cretaceous ACKNOWLEDGMENTS:A.T. wishes to express his appre-
positions, through Palaeogene positions, to ciation to Professor J. Katto, Drs M. Tashiro and M.
their present sites, seems to have been effected Okamura and other faculty members of Kochi Uni-
at intervals by cyclic subduction (Niitsuma et al. versity for their encouragement. H.O. wishes to
1979; Taira 1979; Taira et al. 1979a). This gave thank Professor S. Hayasaka and Dr Y. Teraoka for
rise to alternating periods of compression (max- constructive comments and Messrs T. Arakawa and
imum subduction with accretionary prism de- K. Suzuki for their help in the field. H.O. and A.T.
velopment) and tension, which produced fault- are grateful for a Grant-in-Aid for Scientific Resear-
ches from the Ministry of Education, Japan (Grant
ing and new intra-arc rift-type basins such as the No. 434041). J.H.McD.W. is grateful to the Royal
Izumi, Onogawa and Mifune Basins on the arc Society for grants for two long visits to Japan, and to
side of the forearc basins ('retroduction' of the many Japanese geologists, especially his co-
Taira (1979), 'roll-back effect' of Dewey 1980). authors Professor H. Okada and Dr A. Taira, who
This idea, developed by one of us (A.T.) in made these studies not only possible but enjoyable.
several papers (Taira 1979; Katto & Taira A.J.S. is grateful to the Royal Society and the Japan
1978), could also account for simultaneous Society for the Promotion of Science for grants for
compression at one place and tension at several visits to Japan, and to many Japanese geolog-
another, as illustrated in Fig. 12 (from Taira ists, particularly members of the Kishu Shimanto
Research Group, for their guidance and compan-
1979, fig. 6). Should subduction have been ionship in the field. Professor Charles Downie kindly
oblique, a lateral component of movement examined many specimens for palynomorphs. Con-
could have resulted perhaps in considerable structive comments on the manuscript were made by
strike-slip along the faults (such as the Median Dr J. K. Leggett, Professor J. Tarney, Dr R. yon
and Butsuzo Tectonic lines) bounding the intra- Huene and a Japanese reviewer.

References
(J signifies that the paper is in Japanese: J,E indicates EKDALE, A. A. 1980. Graphoglyptid burrows in mod-
a summary in English). ern deep-sea sediment. Science, 207, 304-6.
BACHMAN, S. B. 1981. The Coastal Belt of the FLOYD, J. D. 1975. The Ordovician Rocks of West
Franciscan: youngest phase of northern Califor- Nithsdale. Thesis, PhD, Univ. St Andrews (un-
nia subduction (this volume). publ.).
DAVIES, W. & CAVE, R. 1976. Folding and cleavage FUKUDA, Y. & HAYASAKA,S. 1978. Trace fossils from
determined during sedimentation. Sediment. the Eocene Kayo Formation in Okinawa-shima,
Geol. 15, 8%133. Ryukyu Islands, Japan. Rep. Fac. Sci., Kagoshi-
DEWEY, J. F. 1980. Abstract from GSL Meeting, ma Univ. (Earth Sci. Biol.) l l , 13-25,
London, June 23-25. HARATA, T., HISATOMI, K., KUMON, F., NAKAZAWA,
DICKINSON,W. R. 1973. Widths of modern arc-trench K., TATEISHI, M., SUSUKI, H. & TOKUOKA, T.
gaps proportional to past duration of igneous 1978. Shimanto geosyncline and Kuroshio PaleD-
activity in associated magmatic arcs. J. geophys. land. Suppl. J. Phys. Earth, 26, 357-66.
Res. 78, 3376--89. HASmMOTO, I. 1956. Geological succession and struc-
- - & SEELY, D. R. 1979. Structure and stratigraphy ture of the undated strata in Yaku-shima, and
of forearc regions. Bull. Am. Assoc. Petrol. some facts concerning the Kumage Group in the
Geol. 63, 2-31. western part of Tanega-shima, Kagoshima Pre-
24 A . Taira e t al.

fecture. Rep. Earth Sci. Dep. Gen. Educ. Formation in Susaki area, Kochi Prefecture. Res.
Kyushu Univ. 2, 23-34. Rep. Kochi Univ. Nat. Sci. 27, 143-50 (J,E).
1962. The sedimentary complex of uncertain -- & TASmRO, M. 1979a. A study on the molluscan
ages in South Kyushu. Rep. Earth Sci. Dep. Gen fauna of the Shimanto terrain, southwest Japan,
Educ. Kyushu Univ. 9, 13-69 (J,E). part 2: bivalve fauna from the Muroto-hanto
HATENASm RESEARCH GROUP 1975. Stratigraphy and Group in Kochi Prefecture, Shikoku. Res. Rep.
geologic structure of the Otonashigawa-Muro Kochi Univ. Nat. Sci. 28, 1-11.
Group in the Kii Peninsula, southwest Japan. -- & TASmRO, M. 1979b. A study on the molluscan
Monogr. Assoc. Geol. Collab. Jpn 19, 157-66 fauna of the Shimanto terrain, southwest Japan,
(J,E). part 3: on the bivalve fauna from the Arioka,
HAYASAKA, S., FUKUDA, Y. & HAYAMA, A. Nakamura and Susaki Formations in Shimanto
1980. Discovery of molluscan fossils and the (Northern) terrain, Kochi Prefecture. Res. Rep.
paleoenvironmental aspects of the Kumage Kochi Univ. Nat. Sci. 28, 4%58.
Group in Tanegashima, South Kyushu, Japan. - - T A S H I R O , M., TAIRA, A. & OKAMURA, M. 1980.
Prof. Saburo Kanno Mem. vol. Tsukuba Univ. Biostratigraphy of the Cretaceous Shimanto Belt
59-71). in the Susaki area, Shikoku. Geol. News, Tokyo,
IMm, I. 1977. Cretaceous and lower Tertiary Systems 310, in press.
in thc Shimanto Belt. In: TANAKA, K. & Klsltu SttIMANTO RESEARCH GROUP 1970. Sedimento-
NOZAWA, T. (eds). Geology and Mineral Re- logical and paleontological studies of the Muro
sources of Japan 1, 3rd edn, 207-13. Geol Surv. Group of the Kii Peninsula. Mem. Fac. Educ.
Japan, Hisamoto. Wakayama Univ. Nat Sci. 20, 75-102 (J,E).
- - T E R A O K A , Y. & OKUMURA, K. 1975. Structural -- 1975. Development of the Shimanto geosync-
division of the Shimanto Terranc of Kyushu, line. Monogr. Assoc. Geol. Collab. Jpn 19, 143-
southwest Japan. Monogr. Assoc. Geol. Collab. 56. (J,E).
Jpn 19, 179-89 (J,E). KIZAKI, K. 1978. Tectonics of the Ryukyu Island arc.
INOUE, E., ISHIBASHI, K., ISttlitARA, T., KINOSHHA, Suppl. J. Phys. Earth, 26, 30 I-7.
Y., JosmMA, M. & TAMAKI, K. 1977. Geological KULM, L. D., RESIG, J. M., TilORNBUR(;, T. M. &
map off outer zone of southwest Japan, 1 : 1 SCHRADER, H. J. 1981. Cenozoic structure, stra-
million. Mar. Geol. Map Ser. 8. Geol. Surv. tigraphy and tectonics of the central Peru forearc
Japan. (this volume).
lSHIDA, S. 1969. Wano Formation (Eocene) in Ama- L~.:l-, S. M. 1974. The tectonic setting of Korea with
mi-Oshima. Ryukyu Islands, Japan. J. geol. Soc. relation to plate tectonics. Tech. Bull. U.N.
Japan. 75, 141-56. ESCA P, CCOP, 8, 3%53.
KANMERA, K. 1976. Comparison between ancient and LEC;GE'n, J. K., MCKERROW, W. S. & CASVY, D. M.
modern geosynclinal sedimentary bodies. I, II, 1981. The anatomy of a Lower Palaeozoic
Kagaku (Sci.) 46, 284-91,371-8 (J). accretionary forearc: the Southern Uplands of
KARIG, D. E. & SHARMANIII, G. F. 1975. Subduction Scotland (this volume).
and accretion in trenches. Bull. geol. Soc. Am. - - MCKERROW,W. S., MORRIS, J. H., OLIVER, G. J.
89, 377-89. H. & PmLLIPS W. E. A. 1979. The north-western
KArro, J. 1960. Some problematica from the so- margin of the Iapetus Ocean. In: HARRIS, A. L.,
called unknown Mesozoic Strata of the southern HOI,I,AND, C. H. & LEAKE, B. E. (eds). The
part of Shikoku, Japan. Sci. Rep. Tohoku Univ. Caledonides of the British Isles Reviewed. Spec.
Sendai, 2nd Ser. (Geol.) 4. 323-34. Publ. geol. Soc. London, 8, 499-512.
1964. Some sedimentary structure and proble- LEwis, K. B. 1980. Quaternary sedimentation on the
matica from the Shimanto Terrain of Kochi Prefec- Hikurangi oblique-subduction and transform
ture, Japan. Res. Rep. Kochi Univ. 13, Nat. Sci. margin, New Zealand. In: BALLANCE, P. F. &
1, 6, 14 pp. READING, H. G. (eds). Sedimentation in Oblique-
- - & ARITA, M. 1966. Geology of the Muroto slip Mobile Zones. Spec. Publ. Int. Assoc.
Peninsula, Shikoku, Japan. Res. Rep., Kochi Sedimentol. 4, 171-89. Blackwell Scientific Pub-
Univ. 15, Nat. Sci. 1, 8, 59-63 (J,E). lications, Oxford, 232 pp.
- - KOJIMA, J., SAWAMURA,T. & SUYARL K. 1961. MATSUDA, T. 1978. Collision of the Izu-Bonin arc
The Geology and Mineral Resources of Kochi with central Honshu: Cenozoic tectonics of the
Prefecure, with Geologic Map. Kochi Prefecture Fossa Magna, Japan. Suppl. J. Phys. Earth, 26,
(J). 409-21.
- - S U Y A R I , T., KASHIMA, Y., HASHIMOTO, I., MATSUMOTO, T. & OKADA, H. 1978. Evaluation of
HADA, S., MITSUL S. & AKOJ1MA, I. 1977. Sur- molluscan fossils from the Mesozoic of the Shi-
face Geologic Map of Shikoku. Kochi Regional manto Belt. Proc. Japan. Acad. 54B, 235-330.
Forestry Office. MIYASHIRO, A. 1973. Metamorphism and Metamor-
- - & TAIRA, A. 1978. Lithofacies and depositional phic Belts. Allen & Unwin, London.
environments of the Muroto-hanto Group. Geol. MOORE, G. F. & KARIG, D. E. 1976. Development of
News, Tokyo, 287, 21-31 (J). sedimentary basins on the lower trench slope.
-- & TASHmO, M. 1978. A study on the molluscan Geology, 4, 693-7.
fauna of the Shimanto terrain, southwest Japan, MOORE, J. C., WATKINS, J. S., SHIPLEY, T. H.,
part 1: on the bivalve fauna of the Doganaro BACHMAN, S. B., BEGHTEL, F. W., BuTt, A.,
 The Shimanto Belt of Japan 25

D1DYK, B. M., LEGGEIT, J. K., LUNDBERG, N., 93 (in press).

MCMILLEN, K. J., NIlTSUMA, N., SHEPHARD, L. STEEL, R. & GLOPPEN, T. G. 1980. Late Caledonian
E., STEPHAN, J.-F. & STRADNER, H. 1979. Prog- (Devonian) basin formation, western Norway:
ressive accretion in the Middle America Trench, signs of strike-slip tectonics during infilling. In:
Southern Mexico. Nature. London, 281,638-42. BALLANCE, P. F. & READING, H. G. (eds).
NAKASEKO, K., NISHIMURA, A. & KANNO, K. 1979. Sedimentation in Oblique-slip Mobile Zones.
Study of radiolarian fossils from the Shimanto Spec. Publ. Int. Assoc. Sedimentol. 4, 79-103.
Belt. Spec. Issue News Osaka Micropaleont. 2, Blackwell Scientific Publications, Oxford, 232
1-49. PP.
NIITSUMA, N., TAIRA, A. & KITAZATO, H. 1979. SUGISAKI, R., SUZUKI,T., KANMERA,K., SAKAI,T.,
History of the plate boundaries in the Japanese SANO, H. 1979. Chemical compositions of green
Islands. Earth Monthly, Tokyo, 1, 193-205 (J). rocks in the Shimanto Belt, southwest Japan. J.
OGAWA,Y. 1981. Tectonics of some forearc fold belts geol. Soc. Japan, 85, 455-66.
in and around the arc-arc crossing area in central SUYARI, K. 1973. On the lithofacies and the correla-
Japan (this volume). tion of the Izumi Group of the Asan Mountain
-- & HORIUCHI, K. 1978. Two types of accretionary Range, Shikoku. Sci. Rep., Tohoku Univ. Hatai
fold belts in central Japan. Suppl. J. Phys. Earth, Mere. Vol, 6, 489-95 (J,E).
26, 321-36. SUZUKI, H. 1975. Deformed structures and deforma-
OKADA, H. 1971. A pattern of sedimentation in tional history of the Palaeogene Muro Flysch
clastic sediments in geosynclines. Mere. geol. in southwest Japan. Monogr. Assoc. Geol. Col-
Soc. Japan, 6, 75-82 (J,E). lab. Jpn, 19, 167-77 (J,E).
-- 1973. Coarsening-upwards cycle. Tohoku Univ. SUZUKI, K., NAKASEKO, K. & OKADA, H. 1979.
Sci. Rep., 2nd Set. (Geol.) 6, 429-37 (J,E). Stratigraphy, structure of the Kumage Group in
1977. Preliminary study of sandstones of the southern Tanegashima (Abs.) 86 Ann. meet.
Shimanto Supergroup in Kyushu, with special geol. Soc. Japan, 134 (J).
reference to "Petrographic Zone". Sci. Rep., SUZUKI, T. & HADA, S. 1979. Cretaceous tectonic
Dept Geol., Kyushu Univ. 12, 203-14 (J,E). m61ange of the Shimanto belt in Shikoku, Japan.
& WHITAKER, J. H. McD. 1974. Shimanto J. geol. Soc. Japan, 85, 467-79.
Group in Tanegashima off South Kyushu (Abs.). TAIRA, A. 1979. Formation of sediment body in the
81 Ann. meet. Geol. Soc Japan, 201(J). arc-trench system and cyclic subduction model.
-- & WHITAKER,J. H. McD. 1979. Sand volcanoes Earth Monthly, Tokyo, 1,860-8 (J).
of the Palaeogene Kumage Group, Tanegashi- -- KATTO, J, & TASHIRO, M. 1979a. The Cretaceous
ma, southwest Japan. J. geol. Soc. Japan. 85, and Cenozoic geologic development of south-
187-96. west Japan and the tectonism of arc trench
OKUDA, Y., KUMAGAI,M. & TAMAKI, K. 1979. Tecto- system. Geol. News, Tokyo, 296, 27-40 (J).
nic development of the continental slope and its , - - & OKAMURA,M. 1980. Geology and
peripheral area off southwest Japan in relation to origin of the Shimanto Belt in Kochi Prefecture,
sedimentary sequences in sedimentary basins. J. Japan. In: TAIRA, A. & TASHIRO, M. (eds).
Japan Assoc. Petrol. Technol. 44, 279-90. Geology and Paleontology of the Shimanto Belt.
RIEDEL, W. R. & SANFILIPPO, A. 1974. Radiolaria Rinyakosaikai Press, Kochi, Japan (in press)
from the southern Indian Ocean, DSDP Leg. 26. (J,E).
Initial Rep. Deep Sea drill. Proj, 26, 771-813. - - NAKASEKO,K., KATTO, J., TASHIRO, M. & SAITO,
SAKAI, T. 1978. Geologic structure and stratigraphy Y. 1979b. New observations on the Sambosan
of the Shimantogawa Group in the middle Group in the western Kochi Prefecture. Geol.
reaches of the Gokase River, Miyazaki Prefec- News, Tokyo, 302, 22-35. (J).
ture. Sci. Rep., Dept Geol., Kyushu Univ. 13, TANAKA, K. 1977. Shimanto Supergroup in the Suku-
23-38 (J). mo area, southwestern Shikoku. Bull. geol.
ONO, K., MOMOKI, Y., OTSUKA, H. & Surv. Japan, 28, 461-76 (J,E).
HAYASAKA, S. 1977. Geology of the northern -- & NOZAWA,T. (eds) 1977. Geology and Mineral
part of Amami-Oshima. Geol. Studies Ryukyu Resources of Japan, 3rd edn, 1, 430 pp. Geol.
Islands, 2, 11-23 (J,E). Surv. Japan.
SANO, H., KANMERA,K., & SAKAI, T. 1979. Sedi- TAZAKI, K., INOMATA, M. & TAZAKI, K. 1980. Um-
ments associated with greenstones of the Shi- bers in pillow lava from the Mineoka tectonic
manto terrain. J. geol. Soe. Japan, 85, 435-44 Belt, Boso Peninsula. J. geol. Soc. Japan, 86,
(J,E). 413-6.
SEELY, D. R., VAIL, P. R. & WALTON,G. G. 1974. TERAOKA, Y. 1970. Cretaceous formations in the
Trench slope model. In: BURK, C. A. & DRAKE, Onogawa Basin and its vicinity, Kyushu, South-
C. L. (eds). The Geology of Continental Mar- west Japan. Rep. geol. Surv. Japan, 237, 87 pp.
gins, 249-60. Springer-Verlag, New York. (J,E).
SHIBATA, K. & NOZAWA, T. 1973. K-Ar ages of -- 1977. Comparison of the Cretaceous sandstones
gravels of orthoquartzite and gneiss from the between the Shimanto Terrane and the Median
Muro Group. Bull. geol. Surv. Japan, 24, 551-3. Zone of Southwest Japan, with reference to the
SMITH, A. J. 1982. The Neogene to Recent geology of provenance of the Shimanto geosynclinal sedi-
Japan and its surrounding seas. Proc. Geol. Ass. ments. J. geol. Soc. Japan, 83, 795-810 (J,E).
 26 A. Taira et al.

1979. Provenance of the Shimanto geosynclinal TSUCHIYA, N., SAKA1,T. & KANMERA,K. 1979. Mode
sediments inferred from sandstone compositions. of occurrence and petrological characteristics of
J. geol. Soc. Japan, 85, 753-69 (J,E). greenstones of the Shimanto Terrain in the Mimi
• OBATA,~I. 1975. Stratigraphy of the Upper River area, Kyushu. J. geol. Soc. Japan, 85,
Cretaceous Uwajima Group. Mem. Nat. Sci. 445-54 (J,E).
Museum, Tokyo, 8, 5-20 (J,E). VON HUENE, R., NASU, N. et al. 1978. On Leg 57
- - OKUMURA,K. & IMAI, I. 1979. Sandstones of the Japan Trench transected. Geotimes, 23, 16-20.
Shimanto Supergroup in the Mimi-kawa area, WHITAKER, J. H. McD. 1973. 'Gutter casts', a new
Kyushu, southwest Japan--with reference to the name for scour-and-fill structures: with examples
structural division of the Shimanto terrane. In: from the Llandoverian of Ringerike and Mal-
HARA, I. et al. (eds). The Basement of Japanese m6ya, Southern Norway. Nor. geol. Tidsskr, 53,
Islands, 133-51. Prof. Hiroshi Kano Mem. Vol., 403-17.
Akita Univ. (J,E). 1982. Cretaceous-Palaeogene geology of south-
TOKUOKA, T. & OKAMI, K. 1979. Orthoquartzite west Japan. Proc. Geol. Ass. 93 (in press).
clasts and the problem on the basement rocks of YASKAWA, K. 1975. Palaeolatitude and relative posi-
the Japanese Islands. In: HARA, I. et al. (eds). tion of south-west Japan and Korea in the Cre-
The Basement of Japanese Islands, 601-23. Prof. taceous. Geophys. J. R. astron. Soc. 43, 835-46.
Hiroshi Kano Mere. Vol., Akita Univ. (J,E).

A. TAIRA, Department of Geology, Kochi University, Akebono-cho 2-5-1, Kochi,

780, Japan.
H. OKADA, Institute of Geosciences, Shizuoka University, Shizuoka, 422, Japan.
J. H. McD. WHITAKER, Department of Geology, The University, Leicester LEI
7RH, England.
A. J. SMITH, Department of Geology, Bedford College, Regent's Park, London
NW1 4NS, England.
 Sedimentation across the Japan Trench off northern Honshu
Island

Roland von Huene & Michael A. Arthur

SUMMARY: The convergent margin along the International Program of Ocean Drilling
(IPOD) Japan Trench transect was sampled by dredging, piston coring, and drilling by the
Deep Sea Drilling Project (DSDP). This sampling was within an extensive network of
single-and multichannel seismic reflection records. The Quaternary uplift of northern
Honshu Island, climatic change, and lowered sea-levels, provided coarser grained
sediment during the Quaternary than during the late Miocene and Pliocene. Within the
Quaternary sediments, those in the forearc basin, in a slope basin, and on the trench floor
are coarser grained and accumulated more rapidly than the clayey sediment from the
trench slope. But overall, the sediment sampled from all environments is mainly clay and
silt. Rates of accumulation varied greatly and prior to 3 Ma ago they were higher in trench
slope basins than in the forearc basin. Tectonism along the Japan Trench transect must
have changed the Miocene-Pliocene morphology by tilting some of the former trench
slope basins; this may have actually triggered slumping and erosion rather than allowing
sediment deposition.
The lower slope has a relatively large number of channels and a greater amount of
slumped material then is found elsewhere; thus it may contain vertical and lateral changes
of facies in sediments deposited over relatively short periods of time. Morphological
features and sedimentation patterns on the lower slope are greatly affected by local
tectonic structure which has changed significantly in the past 2-3 Ma. Slope basin and
trench-floor deposits are distinguished from those of the main forearc basin by the relative
abundance of material deposited by mass movement. Channels commonly cross the
forearc basin and trench slope and provide the transport paths for coarse sediment to any
part of the margin. No specific characteristics appear exclusively in a given environment,
but the relative abundance of features can help identify broad environments of deposition.

Introduction veloped during regional subsidence of the outer

forearc area which was once subaerially ex-
The Japan Trench transect, a 100 km wide posed and is now buried beneath 1-3 km of
corridor across the convergent margin off sediment (yon H u e n e , Nasu et al. 1978). This
northern Honshu Island, was studied using a contrasts greatly with certain other outer fore-
variety of geophysical methods, conventional arc areas which appear dominated by uplift
sampling, and drilling by the G l o m a r Challen- (Moore, Watkins et al. 1979, 1981). A relatively
ger during Legs 56 and 57 (Fig. 1). This effort fiat terrace occupies the forearc which has
was concentrated in the forearc area, where a subsided below the normal 200 m maximum
network of reconnaissance single channel seis- depth of continental shelves; this is called the
mic reflection records (Tamaki et al. 1977) was Sanriku deep-sea terrace (Nasu 1964). Seaward
obtained prior to the acquisition of five 24- of the deep-sea terrace along the trench land-
channel C D P seismic reflection records (Nasu ward slope, is a 4000 m deep narrow t e r r a c e - -
et al. 1980). M a n y dredge samples, several the mid-slope terrace--which is structurally a
piston cores, and seven D S D P drill sites pro- slope basin and we use the morphological and
vided samples from the forearc basin, and structural terms interchangeably.
across the Japan Trench. The distribution of In this paper we summarize sample and seis-
sediment types recovered in samples combined mic information to define the Neogene and
with the structure shown in seismic records help Ouaternary sediment facies in each morpho-
outline a late Miocene to Quaternary history of tectonic environment. We describe the gross
sedimentation in the Japan Trench. structure of the margin and examine the dis-
Geological structure across the Japan Trench tribution of Quaternary sediment types and
transect consists of landward dipping reflections morphological features. Then we examine the
beneath strata paralleling the sea-floor. This succession of older sediments and relate this
pattern is similar to the structure seen in seismic succession to tectonic history. Our observations
records of other convergent margins. However, of sediment facies and the sedimentary features
the Neogene sediment section off Honshu de- associated with the late Neogene environments

27
 28 R. von H u e n e & M . A . A r t h u r

142 ° 143 ° 144 ° 145 °

4:3 °
I
o 50 IO0 KILOMETERS ,..S)--:".~2:,.
t i I " ";'"=~
...~:..!,:.::,
CONTOUR INTERVAL 500 METERS
.,...~
-::..= KUSHIRO
EXPLANATION CHANNEL
v Core

~!i:i ~7464
t:.
•
o
Dredge
DSDP site
4600

42 °

,000,\,, V I P~ Erimo

\\ I 500 " ~ \..~ HIDxl

~ o466

41 °

HACHINOHE

JNOCl

~
470 ~ • o
Hochinohe
RENCH
4~)8 ORI
AXIS

47~1

40 °

"!"..:..: .434
JNOC 2
Mi, 475

HONSHU I.

FIG. 1. Pacific side of northern Japan showing the coast of northern part of Honshu and southern
Hokkaido Islands. Lines indicate 24-fold CDP seismic reflection records; circled dot indicates DSDP
drill site; dot indicates location of dredge hauls; and triangle indicates location of piston core.
Channels indicated by dashed lines.
 S e d i m e n t a t i o n across the J a p a n T r e n c h 29

lead to some predictions of features that may be ture of the lower slope is badly obscured by
found at convergent margins like the Japan diffractions (Fig. 2). On record JNOC-2, which
Trench. shows the part of the slope where most of the
drilling was done, the deep poorly defined
reflectors dip landward. Overlying these reflec-
Structure of the Japan margin tors are strata sub-parallel to the seafloor which
are the slope deposits that were drilled at
The Japan margin off Honshu consists of the DSDP sites 435, 440, 441, and 434. The land-
Neogene Ishikari-Hidaka forearc basin, a ward-dipping strata may have been penetrated
topographically simple trench landward slope, a only at site 434.
trench floor with generally little ponded fill, The trench floor is too deep to be drilled by
and a block-faulted slope seaward of the trench Glomar Challenger. It has thin ponded sedi-
(Fig. 2) (Mogi 1978; Ishiwada & Ogawa 1976). ment covering about 2 km of the trench floor
The configuration of the Ishikari-Hidaka basin that was sampled by piston coring. In record

0 50 K ILOMETERS I00 150

J- I i i 1 I i 1 i i I 1 i i i I

DEEP SEA TERRACE 438 439 ~ SLOPE

z
5 o

t~
to

- --... _ - f ~ ~ c- ~ A _ ~ . ~ j
JNOC I --------- - ~ --i

0 50 KILOMETERS I00

..... ._ 5 --. 440 z 5

"'TO~-, NFOO;~--OCEANCROST
FIG. 2. Line drawings of 24-fold CDP seismic reflection records JNOC-1 (A) and JNOC-2 (B);
locations of DSDP drill sites indicated by site numbers. Vertical axis is 2-way time (seconds)
horizontal axis is distance (kilometres). Double line indicates angular unconformity.

is like that of most forearc basins, although it is JNOC-1 (Fig. 2A) a large toe projects from the
in 1500-3000 m of water. In some places sedi- foot of the trench slope. This toe is interpreted
ment of the basin makes a gradual transition to as sediment transported by mass movement
the trench slope (Fig. 2A); in other places it is from the slope (Arthur et al. 1980; Nasu et al.
marked by a sharp break in slope (Fig. 2B). 1980).
Neogene sediment filling the Ishikari-Hidaka The slope seaward of the trench consists of
basin rests on an angular unconformity that can igneous oceanic basement overlain by a uni-
be followed throughout the seaward part of the form layer of sediment that follows the topogra-
basin. phy (Fig. 2). Ocean crust and overlying sedi-
The slope landward of the trench is divided ment are broken by normal faults as they are
into an upper and lower part by a narrow but flexed downward into the trench. The upper
continuous mid-slope terrace (Fig. 2). The up- part of this sediment is composed of hemipela-
per slope is underlain by a continuation of the gic material from a terrigenous source.
Ishikari-Hidaka basin sediment section and the
unconformity at its base. The mid-slope ter-
race, formed by a narrow slope basin in which Sampling of Quaternary sediment
sediment has ponded, marks the seaward extent Early studies
of the unconformity and a basic change in the
structure of the margin (von Huene et al. 1980). Descriptions of sediment samples from the
Seismic records that show the internal struc- northern Honshu and Hokkaido margins have
30 R. v o n H u e n e & M . A . A r t h u r

TABLE 1. Dredged and cored samples after Yuasa et al. (1977)," Inouchi et al. (1977); Hasegawa et
al. (1977); Nasu et al. (1960); Nasu & Sato (1962). Samples in the vicinity of the Japan Trench
transect are as shown in Fig. 1, and the lithologies penetrated at D S D P sites are shown in Fig. 3

Station Depth Lithology

no. (m)

Deep sea terrace (lshikari-Hidaka Basin)

443 1710- Numerous cobbles and pebbles, and sandy silt. Clasts include sandstone, tuff, limestone,
1700 rhyolite, andesite, biotite granite, diorite, biotite hornfels, metavolcanic rock, and
partly calcareous metamorphic rock. Age 1.65 to 2.1 Ma
444 1080 Olive-brown silt with pebbles of pumice, scoria, and siltstone
467 1150 377 cm core of silt, sandy silt, and clay broken by thin sand and volcanic sand beds
468 1740- Medium volcanic sand, and cobbles of volcanic rock, weakly metamorphosed rock, and
1730 clinopyroxene-olivine basalt, pebbles of chert, pumice, shale, and volcanics. Dredge site
in or near a canyon
469 1420- Fine sand with pumice, and shale pebbles, dredged from side of Hachinohe Knoll
1415
470 1650 376 cm core of clay with very thin sand layers from base of Hachinohe Knoll
479 1850 462 cm core of silt with six thin sand beds
E1T 1690 A dredge yielding sand, mud, and a few pebbles
E1P 1800 A core of mud with sand

Trench upper slope

435 2800 Silty clay or silt, and pumice
437 3050- Micritic limestone, sandy limestone, and tuffaceous sand clasts
2850 Limestone; pale bluish green in colour, with grains of plagioclase, hypersthene, augite,
brownish glass, and rock fragments.
Arenite; pale greyish brown with grains of quartz, plagioclase, clinopyroxine, and
siliceous rock fragments. Dredged across a canyon. Age 0-0.65Ma
439 2520- Sandy silt, pebbles of chert, slate, and sandstone, sedimentary rock fragments, and scoria.
2500 Age 0-0.65 Ma
445 3120- Sandy olive-brown silt with pebbles and gravels of chert, siltstone and pumice
3050
448 2520- Sandy silt with pebbles of granite, chert and siltstone. Age 0.4-2.1 Ma
2500
473 3110- Semiconsolidated silt, siltstone, and pebbles of chert, volcanic and metamorphic
3050 rocks. Age 0.9-1.4 Ma
475 2510 Clay and pumice dredged from top of slope
477 3330- Silt, siltstone, and pumice dredged from outer lip of a small terrace
3290
478 3400- Silt, cobbles of siltstone and andesitic rock, pebbles of pumice, siltstone, igneous
3380 rock, and shale
EDS-2- 2300 Gravel of various rock types
S 1 and S2
EDS-2-B 1 430 Sandy mud from side of Kushiro Canyon
476 4770 556 cm core of clay with silt beds and thin layers of sand, volcanic sand, and plant
fragments. Core from mid-slope terrace basin
440 4980 Silt, sedimentary rock fragments, gravels, pumice, and scoria dredged from mid-slope
terrace. Age 1.65-2.1 Ma

Trench lower slope

472 4870- Clay including pebbles and semiconsolidated silt. Pumice and tuffaceous siltstone (?)
4560 fragment. Age 0.4-0.7 Ma
474 5280- Silt, semi-consolidated siltstone, and siltstone
5150
480 7000- Silty clay, siltstone and a dolerite clast. Age 2.9-4.4 Ma
6850
E2P 6935 324 cm core of blue green mud, and some red-brown mud
 Sedimentation a c r o s s the J a p a n T r e n c h 31

Trench
434 7900 Pebbles of olivine basalt, chert, and mudstone. Corer was stopped by gravel
438 7300 475 cm core of clay with many thin sand and tuff beds and a 50 cm sand bed at base of core
447 7400 442 cm core of clay with thin beds of sand
471 7330 541 cm core of clay with four very thin tuffaceous sand and sand layers
JEDS-2-C 2 8005 85 cm core of mud with sand

Trench seaward slope

446 5790- Reddish-brown clay with manganese nodules and manganese oxide coated pebbles, and pale
5740 greenish grey clay
442 5400 464 cm core of silt, sand beds, clay and tuff layers
E2T 6700- Sand and mud with numerous gravels were dredged along with contaminated red clay.
7340 Wood fragments and bamboo
E3P 5430 258 cm core of blue green mud mixed with red brown clay and some sand

been reported since the rebuilding of Japan's interpreted as turbidites. One sample from a
scientific capability after World War II (Table channel contains significant amounts of sand.
1). Terrigenous sand, gravel, and mud were Samples from the slope are mostly clay and silt
sampled from the Pacific side of northern Hon- (Yuasa et al. 1977; Inouchi et al. 1977).
shu and Hokkaido in the late 1950s by Nasu Quaternary sediment samples were obtained
(Nasu et al. 1960; Nasu & Sato 1962). Of great by DSDP drilling at the edge of the deep-sea
interest to Nasu and his colleagues were the terrace (Sites 438, 439), on the upper slope
large amounts of gravel and boulders recovered (Site 435), on the mid-slope terrace (Site 440),
by dredging (Nasu 1964), which, since the Leg on the lower slope (Sites 434, 441), and at the
56-57 DSDP drilling, are thought to be Pleis- top of the seaward slope of the trench (Site 436)
tocene ice-rafted materials. Interest in the about 20 km east of the trench floor (Figs 1 & 2,
Japan Trench increased with the seismic studies Table 2). The sediment across the transect,
of Ludwig et al. (1966) and the area was one of regardless of location, is predominantly green-
the first recommended for drilling by the IPOD ish-grey diatomaceous clay and silt which con-
Active Margins Panel. tains varying amounts of volcanic ash and scat-
tered sand generally consisting of volcanic rock
fragments, plagioclase, and quartz.
Sampling programme
In preparation for future IPOD drilling, the Forearc basin
Geological Survey of Japan carried out a sam-
pling programme within their network of recon- Of the seven surface samples from the fore-
naissance geophysical tracklines (Honza 1977). arc basin, two dredge samples (468, 469) con-
Off northern Honshu, 28 sites were sampled tain mainly sand (Yuasa et al. 1977; Inouchi et
and an additional 19 sites were sampled along al. 1977). Both samples are near a topographic
the adjacent southern Kuril Trench. Sampling high, Hachinohe Knoll (Fig. 1), and one is in a
was by short dredge hauls, all of which reco- canyon area. Three piston cores consist of clay
vered muddy sediment, and by piston coring and silt with some sand, and the remaining two
(Table 1). Most dredge samples were taken dredge samples are silt and sandy silt.
from steep slopes, and most piston cores from Site 438, near the top of the trench landward
flat areas. Almost all samples contain ice-rafted slope, yielded about 30% muddy sand of Pleis-
material, and the remaining material is pre- tocene age in the first 50 m penetrated. Muddy
dominantly silt and clay. The samples were fine- to coarse-grained sand beds range in thick-
examined for foraminifers, radiolarians, and ness from thin laminae to perhaps 10m.
nannofossils, and eight contained assemblages The beds cannot be correlated between two
suitable for dating (Hasegawa et al. 1977). holes that are about 800 m apart and so they
Sediment samples from the southernmost appear to have a restricted lateral extent.
Kuril Trench resemble material from the Japan However, sand is confined mainly to the Pleis-
Trench. Despite the narrow shelf without large tocene and is a minor constituent in the under-
sediment traps, and several well-defined sub- lying sediment. The section from 50 to 916 m
marine channels that would allow sediment to consists of uniform diatomaceous clay and
bypass the shelf and slope (Fig. 1), the two claystone of Pliocene to late early Miocene
cores from the Kuril Trench are composed of age and contains occasional pumice pebbles and
clay interbedded with thin layers of fine sand ash layers (Fig. 3, Table 2).
32 R. v a n H u e n e & M. A . A r t h u r

438 438B 435 440 441 434 436

15475m 1558.0m 3401m 4.507 m 5655m 5985.8m 5240m

.... :c--:---~--~ L- --- 2;22

i
" ~ IJ, Pliocene

k.>---~
--5_
.,,,v

i7:¢
_- - 4 /
z,._
--_# L. Pliocene
___
~ - U . Miocene
200

=v / / , y,.c 2
-~UZ
¢ / / ~ M. Miocene
4oo~
L , , L . . Miocene
- - ~ 439 ". ~ ¢ U. Cretaceous
l
438B 1656m ~-~, ~b "" "~i? z
1564,5m -- O--~//be " . . ~:'x'P" ----~ £V:
Z'--:
N 600 ~

/ o
dw
..J
.__ g
~.-~ --:-~

Boo g
~--- MtOCene "
a

"~Cene I000

~_U. Oli.ggocene

U Cretaceous
FXPLANATION
1200

D~otomoceous
c~oys'o~e ~ Calcareous
r~oc,ofoss,~s [~ Vo'con,c~ow'n
(mcludln~ pumice)

Coal ~ B. . . . ia { Graded bed,,ng

FIG. 3. Lithology of DSDP cores from Legs 56 and 57. Site 436 is not within the area of Figs 1 or 2.

The vertical distribution of sand is consistent Groups of small channels incised into the
with changes in sediment source inferred from upper sediment layers are seen in seismic re-
Pleistocene sea-level changes and vertical tec- cords from the area off Hachinohe and just
tonism. The shoreward flank of the forearc south of Miyako (Figs 1 & 4) (Nakajima 1973;
basin is a 20 km wide eroded platform (Magi Sakurai et al. 1975; Magi 1978). Most of these
1978) which rises 1000 m or more above the channels are prominent on the 2 ° slopes be-
deep sea terrace. The adjacent coast is cut into a tween the eroded coastal platform and the main
massif of Palaeozoic and Mesozoic rock that part of the deep sea terrace. At station 468 (Fig.
was uplifted during the Quaternary; short 1) one such channel may have been dredged,
streams drain the massif area. During the because the sample had thin layers of sand
Miocene and Pliocene the area of the massif (Yuasa et al. 1977). In some places seismic
was a lowland and much of the surrounding records indicate broad overbank deposits
area was flooded (Chinzei 1966). Thus, during associated with the channels (Fig. 4). Signifi-
the pre-Quaternary period, local terranes were cant transport of sediment in channels is indi-
probably only of minor importance as sediment cated by the coarse sediment that bypasses the
sources. During times of lowered sea-level the shelf and is deposited in the mid-slope terrace
coast was further seaward and the landward basin and the trench. Certainly, the very high
flank of the Ishikari-Hidaka forearc basin was Quaternary rate of accumulation in the mid-
eroded (Fig. 4). This erosion may have been a slope basin--about 10 times that on the deep-
major source of sediment. sea terrace--suggests bypassing. Other evi-
 TABLE 2. Lithology
Lithological
unit and
sub-unit/
thickness Depth
Site (m) Core (m) Age Lithology
434 1/101.5 434-1-434-11 0-101.5 Lower Pliocene Clayey diatomaceous ooze, largely uniform, pebbly (pumice),
434A-1 to Pleistocene in part; greyish olive-green (5GY 4/2); sediment semilithified
434B-1 (lower Pleistocene below Core 3
missing)
2/199.5 434-12-434-33 101.5-301.0 Lower Pliocene Spicular diatomaceous mudstone and diatomaceous spicular
434A-2 mudstone, vitric in part, some thin (< 1 cm) ash layers. ~.~°

Scattered pebbles present (sedimentary and volcanic rocks).

Sediment highly fractured in some intervals; greyish olive-
green (5GY4/2), dusky yellow-green (5GY5/2), greyish-
green (10G4/2), and light olive grey (5Y4/2) with abundant
mottles
3/160.5 434B-2-434B-18 295.5-456.0 Lower Pliocene (pos- Vitric diatomaceous mudstone and diatomaceous vitric
sible repetition of mudstone; volcanic glass more abundant with diatoms and
diatom zones) sponge spicules, minor terrigenous grains; some carbonate r~
layers and pebbles; mostly olive-grey (5Y3/2) with slight
colour variations to greenish-black (5GY2/1) and greyish
olive-green (5GY3/2)
4/153.0 434B-19-434B-34 456.0-609.0 Upper Miocene Tuffite (30%-60% ash) and clayey tuffite generally without
siliceous microfossils; no ash layers recovered; carbonate
pebbles common; variable colour, mostly olive-grey (5Y3/2),
dark greenish-grey (5GY4/1, 5G4/1) greenish-black
(5GY2/1), and greyish olive-green (5GY3/2)
5/>28.5 434B-35-434B-37 609-637.5 Upper Miocene Vitric diatomaceous mudstone and diatomaceous vitric
mudstone; vitric component less than 25%, diatoms as much as
25%; marlstone pebbles common; dark green-grey (5GY4/1)
to greyish olive-green (5GY3/2)
435 1/244.5 435-1-435-11 0-149.5 Lower Pliocene Spicule-diatom ooze to diatom mud (diatoms: 10-40%);
435A-1-435A-16 149.5-244.5 to Pleistocene pebbles common (pumice, igneous rocks, argillite); (ash layers
present; sediment commonly burrowed; lithified below
about 225 m; greyish olive (10Y4/2) to dusky yellow-green
(5GY4/2)
436 1/103 436-1-436-11 0-103 Upper Pliocene to Vitric diatomaceous mud and diatomaceous mud (20% diatoms,
Holocene 15% ash); olive-grey to greyish olive-green; ash layers and
pockets
 TABLE 2 (continued) ~t~
Lithological
unit and
sub-unit/
thickness Depth
Site (m) Core (m) Age Lithology

2/62 436-12-436-18 103-165 Pliocene Vitric muddy diatomaceous ooze; higher biogenic silica than
Unit 1; thin ash beds or patches; dusky yellow-green to
greyish-olive (5GY4/2), commonly mottled with brownish-
grey (5YR5/1)
3A/80 436-10-436-26 165-245 Pliocene Diatomaceous vitric mud to stiff vitric diatomaceous mud
and minor muddy diatomaceous ooze; irregular ash content
3B/67 436-27-436-33 245-312 Upper Miocene Diatomaceous mudstone and vitric mudstone; lithified at
about 245 m; similar to 3A but coiour changes downward
from greenish-grey (5G6/1-5/1) to moderate yellow-brown
(10YR5/2)
4/48 436-34-436-38 312-360 Miocene Radiolarian diatomaceous mudstone; bright moderate
yellow-brown (10YR5/2, 6/4, 5/4). Transition to pelagic
clay; diatoms and ash decrease downward
5/19 436-39-436-40 360-379 Eocene and Pelagic clay: very dark brown (10YR2/2-2/1), stiff,
Oligocene (?) scattered burrows and lenses
6/>18 436-41-436-42 379-? Cretaceous Pelagic clay with chert (poor .recovery); brownish-black fragments
of chert in dusky yellow-brown pelagic clay (10YR 2/2)
437 No sediment recovered
438 1/107.0 438-1-438-12 0.0-107.0 Upper Pliocene Diatomaceous sandy, silty clay, and silt, pebbly (pumice and
to Pleistocene black greywacke [?]), dark grey, olive-grey, and dark olive-
grey (5Y4/2-3/2, 3/1); numerous ash layers
1/51.5 438A-1-438A-4 0.0-51.5 Pleistocene Diatomaceous sandy, silty, pebbly clay, olive-grey to dark
olive-grey (5Y3/2-4/2) in part; numerous ash layers
2 438A-5-438A-80 51.5-817.0 Lower Pliocene Homogeneous to burrow-mottled (lower part) clayey diatom
438 2A/312.5 438A-5-438A-32 51.5-363.0 through upper ooze, diatom ooze, and diatomaceous clay (soft to
(cont.) Pliocene firm), mostly olive-grey (5Y4/2) with dark and very dark
grey intervals (5Y5/3); numerous ash horizons (green-grey
to grey), occasional pebbles (pumice), and siliceous sponge
remains
2B/230.0 438A-33-438A-56 363.0-593.0 Upper Miocene Diatomaceous claystone and clayey diatomite (silicified),
to lower burrow-mottled throughout, minor ash layers in upper
Pliocene portion. Most of sub-unit is olive-grey, dark olive-grey, grey to
(lowermost) very dark grey. Sponge remains occur only in upper part.
Lower part of sub-unit silicified. Rare pebbles (pumice)
 TABLE 2. (continued)
Lithological
unit and
sub-unit/
thickness Depth
Site (m) Core (m) Age Lithology

2C/224.0 438A-57-438A-80 593.0-817.0 Middle Miocene Mostly burrow-mottled claystone and diatomaceous claystone
to lower showing dewatering structures and some microfaults.
upper Miocene Thin calcareous concretions and thin limestone beds common.
Pebbles (pumice) and ash horizons rare. Sub-unit 2C
is predominantly dark grey, grey, and grey-green (5Y4/2-
5/1, and 5GY5/1-4/1)
3/61.0 438A-80-438A-86 817.0-878.0 Lower Miocene to Sandy claystone and diatomaceous claystone, vitric in part.
lower middle Clayey sandstone and siltstone alternates with claystone.
Miocene Burrow-mottling common. Colours mostly grey to dark grey
and grey-green (5YG5/1-4/1 and 5GY4/1)
2B 438A-1,438A-2
Overlap with Hole 438
m
2C 438A-3,438A-4 --
Continuous coring begins
3/766.5 438A-5-438A-11 849.5-916.0 Lower and [to] Greyish olive-green (5GY3/2) and olive-grey to dark olive-
middle Miocene grey (5Y4/2-3/1) claystone; intensely burrow-mottled (5Y
6/4 and 5Y4/3) concretionary limestone layers, uncommon
thin tuff layers, pumice clasts and fragments, microfaults
and dewatering veins, tension fractures, sharp transition to
next unit below
Hiatus
4/77.5 438A- 12-438A-22 916.0-993.5 Lower Miocene Interbedded turbidites and silty claystone, greenish-black
(Section 1) (5G2/1, 5Y4/1) to olive-grey, Bouma turbidite sequences
are mostly 'ae' with a few 'ade' and 'acde'; medium-
grain to fine- sand, silt grading upward to clay, sharp
basal contacts, scoured; lithic wacke composition (micro-
crystalline quartz lithic fragments), some mollusc fragments.
A few tuff layers, mottled clayey siltstone at base-sharp
transition to next unit below
5/104.5 438A-22 (Section 993.5-1098.0 Oligocene Gray to olive-grey (N5 to 5Y4/1 and 2.5Y4/0) massive
2)-438A-32 sandstone and siltstone moderately well sorted, friable, no
(Section 1) sedimentary structures, abundant molluscan debris and
articulated pelecypods; sandstone calcite-cemented in lowest
part of unit. Wood chips, quartz, lithic fragments (sands
are lithic arenites) common
 TABLE 2. (continued)
Lithological
unit and
sub-unit/
thickness Depth
Site (m) Core (m) Age Lithology

6/47.5 438A-32 (Section 1098.0-1145.5 Oligocene (?) Boulder to pebble conglomerate and breccia, mostly mud-
2)-438A-37 supported, angular to sub-rounded clasts in clay-size matrix;
(Seclion 1) boulder and smaller clasts of dacite (blue grey to greenish-
grey) and pebbles and granules of very dark grey silicified
silty claystone (similar to Unit 7). Boulders up to 65 cm in
diameter. Several individual conglomerate units. Basal 60 cm
is olive-grey (5Y4/1) plastic clay (firm)
Hiatus
438A-37 (Section 1145.5-1156.5 Cretaceous Dark to very dark grey (n4 to N2), well-indurated, silicified
2)-438A-39 claystone and clayey siltstone with thin (<1 cm) calcareous
graded silt beds; contorted into slump folds but some
horizontal bedding t~
440 1 440-1-440-5 0.0-38.0 Upper Pleistocene Greenish-black to olive-grey (5G2/1 to 5Y4/2) clayey sand
(cont.) and Holocene and gravel, poorly sorted, from dusky yellow-green to olive-
grey (5GY5/2 to 5Y3/2) diatomaceous clay, sandy diatom
ooze and greyish-olive (10Y4/2) diatomaceous silty and
sandy clay; some graded bedding in coarser-grained intervals
Rounded pebbles of various igneous and sedimentary
rocks common. Some thin ash layers
440-5-440-26 38.0-380.0 Lower Pliocene Mostly dark olive-grey to olive-grey (5Y3/2-4/2) diatomaceous
and upper clay and (below 168 m) diatomaceous claystone.
Pliocene Randomly dispersed rounded pebbles common; burrow
mottling increases downhole. Minor thin ash layers.
440-27--440-71 380.0-808.0 Lower Pliocene Generally olive-grey to dark olive-grey (5Y4/2-3/2) diatomaceous
and upper clay and (below 168 m) diatomaceous claystone.
Pliocene Zones of folded beds (due to mass movement), veins and
microfaults. Thin (1 cm) ash layers common throughout
441 1 441-1-441-6 0.0-132.0 Upper Pleistocene Dark greenish-grey (5GY4/1) silty diatomaceous clay at
441A-1-441A-2 and Pleistocene top, predominantly olive-grey (5Y4/2) diatomaceous clay
with minor ash layers; rounded pebbles up to 2 cm in diameter
consisting of basalt, dacite, pumice, and some calcareous
pebbles or concretions. Minor clayey silt or sand layers.
Slightly mottled throughout. One graded silty tuffite noted
at base of Unit 1
 TABLE 2. (continued)
2 441-7--441-9 132.0-380.0 Lower Pleocene Dark green-grey (5GY4/1), dark olive-grey (5Y3/2), and
441A-3-441A-4 and upper olive grey (5Y4/1) diatomaceous claystone, minor clayey
HI, H2 a Pliocene diatomite, vitric diatomaceous claystone, and a few thin
tuff beds. Some calcareous mottles and concretions. Entire t~
unit is characterized by fractures, parting, and some black
veins (rehealed). Recovery is mostly in the form of < l c m
chips (fissility) and very poor
3 441A-5 380.0-410.0 Lower Pliocene Breccia of dark grey (N3) siltstone: rounded to angular siltstone
fragments in more clay-rich matrix or gouge. Dewatering
veins at top of unit. Several graded and ungraded
very fine-grained to coarse-grained sandstone beds
interbedded with claystone (5Y6/1). Bottom of unit
chosen on basis of signature on geophysical log (FTC). Fair
recovery
4 441A-6, 441A-7 410.0-504.0 Lower Pliocene Mostly olive-grey (5Y3/2) diatomaceous claystone and t~
claystone, some siltstone, possible redeposited claystone,
minor tuff layers, and calcareous concretions; poor recovery
and highly fractured material
5 441A-8-441A-15 504.0-687.0 Upper Miocene Heterogeneous lithology consisting of olive-grey (5Y3/2) clay-
H4 and lower stone, greenish-black to greenish-grey (5G2/6--6/1) silty claystone,
Pliocene calcareous claystone and carbonate-cemented claystone breccia, to
dark greenish grey (5GY4/1) resedimented claystone, 45° dip. Lower
part of unit composed of greenish-black claystone (5G2/1-3/1) and
olive-grey silty claystone (5Y4/1) with some light olive-grey (5Y5/1)
silty tuffite and (5Y3/1) vitric silty claystone. Entire unit exhibits high
fissility; intensely fractured except for breccia units.

aH signifies 'wash cores,' sediment cored through an interval longer than the core barrel. These are archived and sampled as a normal cores.
38 R. v a n H u e n e & M . A . A r t h u r
0 50 100 Km
o I ~ m i ~ I, l 1 ~ ~ .......I ...... t. !

I A
sediment blanket
migra t i ng
channel
ove rban k
deposit

&-'~ ~ ,~',~~', 7 1 r , ~ i ~ , ~ .......... I / / •

"~'~" ) ;.', ;~" ~ -I ~ ~ _~,~ .... ,~ i~m" /

,.~,~,~.;,~ ........... ~ ~ii~:~' ~ZI~..

/
channels overbank sediment

ships track p a r a l l e l to s t r i k e II track parallel to dip

turn

(A)

FIG. 4. (A) Examples of seismic reflection records along ships tracks paralleling the trend of the trench
and coast off northern Honshu. Record A shows a migrating channel which is being filled and the
associated overbank deposits which merge with draped deposits filling a trough. Part of this record
parallels the dip of the slope and shows ponding in small tectonic features at the top of the trench
slope.

dence for sediment bypassing is the displace- which are interpreted as deep-sea fans (Sakurai
ment of Pleistocene microfauna from shelf and et al. 1974, in Magi 1978). The third type of
upper slope environments to depths of 4000 m sediment deposit seen in seismic records con-
and more at Sites 440 and 434 and the occur- sists of layers conforming roughly to the
rence of calcareous microfauna below the car- topography, similar to the sediment blanket
bonate compensation depth (CCD) in dredged described by Inouchi & Tamaki (1977). The
samples from the trench lower slope (Hasagawa published seismic data are insufficient to map
etal. 1977). the distribution of each type of deposit but such
Three types of sediment deposits inferred a map might be made if all of the original
from single-channel seismic records occur on published and unpublished records in this area
the deep-sea terrace. First, sediment is ponded were studied.
in lows behind a ridge (Fig. 4). Secondly, some A change in sedimentation caused by tecton-
ponded sediment has a convex-upward surface ism is suggested by the prominent sediment
(Fig. 4) resembling the sediment lobes associ- pond in seismic record JNOC-2 (Fig. 2). Arthur
ated with channels just north of Hachinohe, et al. (1980) suggest that a former trough, along
 S e d i m e n t a t i o n a c r o s s the J a p a n T r e n c h 39

~
0

ent blanket
i -. Hachinohe Knoll

2 eroded coastal ~.,'~'~ / / / ~ ~ . ~ % . ~

._1

\1 !

3 ~ _ "~_./~,~k trench slope

o oo
VE = 20:t \ ~"llI
\,,
From Honza et ol, 1978 \/\\jj
"t
(B)

Flo. 4. (B) Line drawings of seismic records showing sediment lobe, filled trough, and sediment
blanket from Honza et al. (1978).

which sediment was transported parallel to the which is commonly associated with sandy sedi-
slope, was disturbed by tectonism in the late ment, hemipelagic silt and clay appear to pre-
Pliocene and Pleistocene. This trough extended dominate on the deep-sea terrace from the
for perhaps 200 km along the deep sea terrace sample information. The fine grain size may be
from Hidaka Channel and emptied on to the exaggerated by a sampling bias because samples
trench slope just south of the location of seismic are located far from shore and mainly in the
record JNOC-2. Arthur et al. (1980) suggest inter-channel areas. In the central Gulf of Alas-
that continued uplift of the anticline that forms ka, for example, sand and gravel predominate
the seaward bank of the trough caused it to fill within 25 km of the shore (Carlson et al. 1977);
by lessening or reversing its gradient (Fig. 5). the sample closest to shore on the deep-sea
The trough probably passed over an outcrop of terrace is about 90 km off Hachinohe. South of
basement material that previously formed a the transect area, near the Kuji River, Nasu
high point in the unconformity surface. Im- (1964) reports abundant sand and gravel in the
mediately below the unconformity and beneath near shore area.
the trough, seismic refraction velocities of DSDP drill holes on the deep-sea terrace
6.3 km s -1 are reported (Nagumo et al. 1980), have been sited on structural highs where sub-
values suggesting that the high on the uncon- surface targets are at shallow depths. Thus, in
formity surface (Fig. 5A) is igneous rock. Sedi- the forearc basin, interchannel areas were sam-
ment first transgressed over the high point and pled preferentially. Perhaps sampling closer to
buried it (Fig. 5A). The region then began to the eroded coastal platform or in channels
tilt seaward and local uplift formed a 'shelf edge would have produced more coarse material
high' (deep-sea terrace edge), the relief of because the coarse sediment found in the slope
which was diminished by sedimentation (Fig. basin and trench must have been transported
5B). Further local uplift and regional tilting from shoreward regions in channels. We sus-
surpassed the rate of hemipelagic sedimenta- pect a bias toward fine-grained sediment in the
tion required to maintain a level surface. sample data and have taken this into account in
Despite abundant channelling and ponding our conclusions.
40 R. yon H u e n e & M. A . A r t h u r
KILOMETERS
15 20 25 50 35

C.

Ple=stocene?
4

Pliocene
......~ - ~ ~!~-~_.~~ - ~: , : : ~ ~:~,~,!~'~ ~

FIG. 5. Development and filling of trough in seismic record JNOC-2, Fig. 2. This sequence was made
by flattening successively deeper horizons on a depth section of the seismic record. The dashed line
over basement is a permissible alternate interpretation. We interpret a local uplift at the edge of the
deep sea terrace, indicated by a simple arrow, superimposed on a regional tilt of the margin,
indicated by the heavier arrow.

The information regarding the distribution of Hesse (1977) examined the sedimentary
Quaternary sediment facies on the shelf and structure in piston cores from the slope with x-
morphological features is summarized as fol- radiographs. Cores from the upper slope
lows---silt and clay dominate samples from in- appear enriched in sand; the sand grains are
terchannel areas well offshore but one sample, scattered throughout the muddy matrix. An
possibly from a channel, is mostly sand. Chan- undetermined amount of the sand and granule-
nels are observed in bathymetric and seismic size material is likely to be of ice-rafted origin.
records. Seismic data show sediment bulges, Site 435 on the upper slope yielded diato-
some of which have been interpreted as deep- maceous silty clay with sparse sandy silt, one
sea fan lobes; they show some sediment pond- bed of sand, and volcanic ash from the top 40 m
ing behind tectonic ridges, and a blanket of (Fig. 3 and Table 2). Site 435 has the same
sediment in many other areas. Tectonism has hemipelagic character as Site 438 except that
changed local morphology, and the mode of the proportion of sand in the upper Pleistocene
sedimentation appears to have changed corres- interval of 435 is 10% or less, far less than in
pondingly. samples from Site 438.
Quaternary sediment seen in seismic records
Trench upper slope down the upper slope is generally in a layer that
follows the topography (Fig. 4). In the multi-
The trench upper slope was sampled at nine channel seismic records, older beds are trun-
dredge sites. As a whole the slope samples are cated at the seafloor in some places and the
finer grained than those from the forearc basin. Quaternary deposits may have been lost by
Four samples consist of clay with varying slumping (Fig. 2) (Nasu et al. 1979). A well-
amounts of silt, five samples are silt with sparse defined slump is recorded on the upper slope by
sand layers, and one consists of micritic lime- Nakajima (1973). It has a clear headwall scarp
stone (Table 1). and a jumbled footwall. This slump block mea-
 S e d i m e n t a t i o n across the J a p a n T r e n c h 41

sures about 35 km in the downslope direction essentially continuous for about 250 km along
and is as much as 500 m (0.6 s) thick. Slumps in the Japan Trench. Seismic records across the
the Geological Survey of Japan seismic records mid-slope terrace show ponded sediment in
are difficult to outline because of the high many places (Nakajima 1973, record VII; Nasu
vertical exaggeration of the records (37:1). The et al. 1979, record 78-4; Tamaki et al. 1977,
progressive seaward loss of continuity in reflec- records 18, 19, 22, 23, 24) and records with low
tions indicates tectonic deformation, some of vertical exaggeration show channels (Nakajima
which can be interpreted as listric faulting 1973, record VI). At D S D P Site 440 the pond-
(Nasu et al. 1980). Channels across the upper ing in the present slope basin began in the
slope are seen in topographic contours and Pleistocene and was succeeded by sedimenta-
particularly in seismic and bathymetric records tion at a relatively rapid rate of accumulation
paralleling the strike of the slope (Fig. 4). (160m Ma -~) until the present.
Seismic records made on transits to and from
the area drilled by Glomar Challenger show Trench lower slope
large (10-20 km wide) canyons at 50-60 km
spacings all along the Japan Trench upper slope The five samples dredged from the trench
(Fig. 4). Some are filled with sediment includ- lower slope are dominantly silt and clay with
ing slump debris, whereas others contain some semiconsolidated siltstone. Most samples
eroded areas. Filled depressions observed in from the trench slope are no older than 2 Ma.
the records may once have been major canyons. Calcareous microfossils were present in all sam-
Inter-canyon areas commonly have a uniform ples from the trench lower slope, and two
blanket of sediment. From the scanty seismic samples contain abundant planktonic foramini-
records made parallel to the strike of the slope fers and nannofossils. Their recovery from
it is clear that seismic records parallel to the dip deep below the CCD probably indicates con-
of the slope give an oversimplified impression siderable down-slope transport. One lower
of sediment processes on the trench slope. slope sample is between 2.8 and 4.4Ma old,
More data parallel as well as transverse to the which is much older than the other samples and
slope are needed to quantify the relative im- suggests that it is a displaced fragment from an
portance of individual morphological features older upslope outcrop. At Sites 434 and 441 on
off Japan. the lower slope most of the Pleistocene section
is missing owing to slumping and non-
Slope basin (mid-slope terrace) deposition. The beds sampled below the first
10 m are Pliocene or older. Some are re-
A single piston core from the slope basin sedimented claystone breccia which indicate
consists of turbidite mud and massive mud even more down-slope mass movement than at
broken by thinly laminated sand layers which Site 440 (Scientific Party, Legs 56 and 57,
are not megascopically graded (Inouchi et al. 1980).
1977). Rapid sedimentation in this environment Multichannel records across the trench lower
means that deep penetration of cores is neces- slope show truncation of reflections at the
sary to get a vertical succession of representa- seafloor (Nasu et al. 1979). The steep slope
tive facies, such as was obtained at DSDP Site (4°-8 °) and rugged terrain indicate little Quat-
440. ernary sediment deposition; sediment erosion
Site 440, on the trench mid-slope terrace, and slumping are typical, a conclusion con-
yielded muddy sand and gravel interbedded firmed in D S D P drill holes. Some channels
with silty clay in the first 40 m penetrated. continue from the upper slope across the lower
Below this depth, there are very few sand beds slope, the most obvious one being Hidaka
in the diatomaceous clay and claystone section Channel. In the adjoining Kuril Trench, the
(Fig. 3 and Table 2). The composition of the well-developed Kushiro channel is also con-
sediment is much like that recovered at Site tinuous across the entire slope (Mogi 1978).
438, but several zones of syndepositionally
folded mud below 300 m probably originated Trench floor
from mass movement (Scientific Party, Leg 57,
1980). Despite its proximity to the trench, the Piston cores from the trench contain
total section is in a normal stratigraphic order hemipelagic mud, mud turbidites, and some
and records slope sediment deposition in water turbidite deposits composed of fine sand. Fine
more than 2000 m deep. volcanogenic sand is apparently redeposited by
In the network of bathymetric and seismic bottom traction processes rather than turbidites
records, the mid-slope terrace appears to be (Hesse 1977).
42 R. y o n H u e n e & M . A . A r t h u r

The trench floor yielded somewhat coarser area are described only in general terms. Sur-
material than the trench slope. The three piston face and near surface samples in conjunction
cores consist of clay with thin beds of fine sand, with seismic and bathymetric data do, however,
volcanic ash, and pumice. One core barrel allow us to infer a very general distribution of
apparently bounced off a gravel bed and Quaternary sediment facies in the Japan Trench
another bottomed in a sand bed. Transect area. DSDP samples, which are from
Except at its northernmost end, the trench greater depths, show the vertical change in
floor generally has only a small amount of facies between the late Miocene and Quater-
ponded sediment (Fig. 2). At the north end of nary.
the Japan Trench a large sediment mass associ- The distribution of Quaternary sandy facies
ated with the Hidaka Channel fills the trench along the Japan Trench margin is similar to that
and displaces the trench axis seaward (Figs 1 & in the areas discussed by Underwood & Karig
2A) (Nasu et al. 1980; Arthur et al. 1980; (1980), Moore et al. (1981), Underwood et al.
Odonera & Honza 1977). The sediment mass (1980), and Underwood & Bachman (1981).
may have been deflected southward by Erimo Sediment of the forearc basin, slope basin, and
Seamount which is now on the trench floor. The trench is characterized by coarser-grained
sediment moving down the Hidaka Channel, material than is found in sediment of the slope.
perhaps augmented by slumps from the adja- In the forearc basin only one of the eleven
cent landward slope, where truncated beds crop samples has no sand; in the mid-slope basin
out along steep slopes (Nasu et al. 1979), must much sand and gravel was sampled; in the
have accumulated in the trench more rapidly trench mostly mud was cored, but two of the
than it could be subducted. three cores bottomed in sand or gravel. On the
other hand, only four of nine samples from the
Trench seaward slope trench upper slope had any trace of sand and no
sand was reported from the trench lower slope.
Two piston cores from the trench seaward Overall however, the great bulk of sediment
slope yielded red-brown clay, greenish clay, sampled is clay and silt and discrete layers of
and silt with sand and tuff. The seamounts that sand are rare. Facies characteristics are not
were dredged appear to be covered with a layer distinct enough to identify the original environ-
of silt. Further examination of the sediment on ment without knowledge of sample location.
seamounts is necessary in order to assess how The morphological features and gross
much sediment on the margin falls through the sedimentary structures are best seen in seismic
water column (e.g. as from aeolian transport) lines parallel to the regional trend of the mar-
as opposed to that transported by mass-flow or gin. Such records show sediment draping,
other means which presumably would not be ponding, lobate sediment masses (probably
able to climb the flank of a sediment. fans), channels with overbank deposits and
Site 436, on the outer swell of the seaward slumps. Certain combinations of these features
slope, yielded about 300 m of diatomaceous silt are more commonly observed in one environ-
and clay with much volcanic ash, underlain by a ment than in others although most features can
pelagic sequence. This Neogene section is be found in all environments. Despite the large
largely hemipelagic material, similar to that amount of seismic data off northern Honshu, it
sampled on the landward slope. is difficult to quantify the relative occurrence of
The seaward slope is covered by hemipelagic each type of morphological feature or sediment
and pelagic sediment that conforms to the structure. From our study of the geophysical
topography. It is possible that the dredge sam- data, we conclude that the forearc basin consists
ples containing red clay were taken along fault mainly of ponded and draped sediment with
scarps that exposed the pelagic lower part of the common lobate masses, channels, and over-
ocean section. Quaternary sediment at Site 436 bank deposits. Similar features are seen on the
is about 85 m thick, and its lower boundary trench upper slope, but these channels and
cannot be resolved well in multichannel seismic slumps are more common there. The trench
records, probably because of the poor resolu- slope basin and trench floor also show features
tion of this technique at great water depths. seen in the forearc basin, but most of the
sediment is ponded, slumps have been arrested,
and channels are few. Details of the trench
Summary of sediment sampled and lower slope are less clearly resolved in seismic
rates of deposition and bathymetric records. A single short line
parallel to the margin shows a canyon with
Many samples from the Japan Trench transect truncated beds and possibly some overbank
 Sedimentation across the Japan Trench 43

deposits (Arthur et al. 1980). Large slump between 25 and l l 0 m M a -~ at Sites 438 and
scars are inferred on the lower slope from mor- 439 in the forearc basin, whereas higher rates of
phology. as much as 2 7 0 m M a -1 are indicated at sites
The Glomar Challenger made two seismic on the slope (Fig. 6). In contrast, rates of
records parallel to the orientation of northern sediment accumulation seaward of the trench
Honshu during transits to the Japan Trench are between 14 and 5 0 m M a -1. There are
transect. Features in the upper sediment layers major differences in the water contents of sedi-
of the forearc basin and the trench upper slope ments at each of these sites, so various sediment
can be compared in the two records (Table 3), components have been separated (Fig. 7).
Rates of accumulation of biogenic silica are
TABLE 3. Comparative amounts of various generally the same at all sites but are slightly
sedimentary structures along two Glomar Chal- higher in the Pliocene than at other times. The
lenger seismic lines parallel to the regional terrigenous component differs greatly between
topographic or tectonic trend the forearc basin and the lower slope. Surpris-
ingly, the lower slope sites (440,441,434) have
Sedimentary Forearc basin Trench upper slope Miocene and Pliocene accumulation rates
structure (%) (%) roughly 1.5 to 3.5 times greater than sites in the
Ponded sediment 45 10 forearc basin (438) and upper slope. Rates on
Draped sediment 25 10 the lower slope are about as high as the very
Channels and high Pleistocene rate measured at Site 440
overbank 30 60 which probably corresponds to formation of the
Slump 0 10 slope basin and sediment ponding. At Site 436
Unknown 10 the accumulation rates of biogenic silica, volca-
nic ash, and total sediment increased toward
but the comparison is tentative because three- the top of the core because the area of the site
dimensional structure can only be inferred from moved closer to the Japan margin.
the two-dimensional seismic record. The re- The high rates of accumulation on the trench
cords suggest, however, a trend from ponded lower slope are probably due to a tectonic
and draped sediment bodies in the forearc basin process, such as creation and disruption of
toward more canyons and slumps on the upper basins on the lower slope, because it is difficult
slope. to otherwise explain higher rates of accumula-
Across the Japan Trench margin, Quaternary tion on the 5° trench slope than in the nearly fiat
sediment has more coarse grained material than forearc basin. The change from anomalously
the underlying Pliocene and late Miocene sedi- high rates, to the present lower rates of accu-
ment. The increased grain size which in Quater- mulation on the slope coincides with regional
nary sediment includes gravel in many samples, tectonic events on land, such as a change in the
may be attributed to ice rafting, to a lowering of stress field (Nakamura & Uyeda 1981), and
the sea-level which caused the erosional explosive volcanism (Cadet & Fujioka 1980). A
platform along the coast (Honza et al. 1978; compilation of the number of ash layers per
Nakajima 1973; Sakurai et al. 1975; Mogi 1978), million years indicate that and the middle
and to Quaternary uplift of the coastal massif Pliocene was a time of great airfall ash deposi-
(Chinzei 1966). In contrast, Pliocene and tion. The pattern of accumulation of dissemi-
Miocene sediment from the deep sea terrace nated volcanic glass entrained in the sediment
(three holes at Sites 438 and 439) and the slope (Fig. 7) only partially reflects the trends infer-
(four sites, of 15 holes) contains no discrete red from discrete ash layers. The ash accumula-
sand beds. Apparently relatively little coarse tion pattern of Site 436 also reflects the in-
sediment was being supplied to the margin prior creased proximity to the volcanic source with
to the Quaternary when the adjacent part of time as shown by accumulation rates of ter-
Honshu was a lowland and partly submerged. rigenous material. At the sites on the lower
The grain size differences in pre-Quaternary slope (440,441,434), the accumulation rates of
samples from across the margin are small; ash follow the same trends as those of ter-
however, the number of sites are few for such a rigenous sediment. The similarity suggests sub-
large area. marine transport of much of the volcanic glass,
Despite the uniform lithostratigraphy from probably in dilute suspension turbidity cur-
one DSDP drill site to another (Fig. 3), rates of rents. Further evidence of a change in rate or
pre-Quaternary sediment accumulation vary pattern of tectonism in the Pliocene is seen in
widely between the environments sampled. the sequence of benthic foraminifers recovered
Simple plots of age versus depth indicate rates from Site 438 and 439 (Keller 1980).
44 R. y o n H u e n e & M . A . A r t h u r

...... o. E ..... ELO. RI ....... OCENE I ........ 0 ....

[2 4 I 6 8 I0 12 14 16 18

\ %.,-.,',
~ 300 \ \\ \ \ \ \ \ k ~" ''*"*** **,, ........ SiTE
....... SITE 435-UPPER
436-SEAWARDSLOPE
SLOPE
', ***',,,~,, - - SITE 438-FOREARE BASIN

-- -- -- SITE 440- MID-SLOPE BASIN

\ \ \\ Seoword -- -- SITE 441 LOWER SLOPE
\ \ ', ~ slope
450 \\ \ \',
\ \ ",.

600 k \
E / \ \
] \\ \
Z50 \\ \

Mid-slope ~
bosin
Foreorc bo$in
900 I I I I l I _2 . . . . i

FiG. 6. Plots of sediment age versus depth in DSDP cores, uncorrected for compaction.

When analysed in a manner developed by also indicate subsidence (von Huene et al.
van Hinte (1978), the benthic assemblages show 1980).
rapid subsidence of the margin until about Soon after the period of change from subsi-
3 M a ago and then a rapid uplift (Fig. 8) dence to uplift at Sites 438 and 439, and the
(Arthur et al. 1980). The lithologies of rocks Pliocene increase in explosive volcanism, basins
from which the foraminifers were recovered on the lower slope were rearranged, sediment

300 - Total sediment

8iogenic silica 434

441
250
[~] Volcanic ash

Hiatus
200

150 .o
4381439 ._ .
E

I00

435
: iii'i?~::i:ii,.~.., !:.::.)i i 456
< 50- ,.=,!J~i ~ i!i!:::!:!:!:
. . . . . . . •

I ! I I I ' ' ' I'aq I "~ I ' ' ' ' ' '
Ig

0 4 8 12 16 20 0 4 0 4 8 0 4 0 4 8 0 4 8 12
AGE (MY)

FIG. 7. Rates of sediment accumulation in DSDP cores, after Arthur et al. (1980). These are the same
data as in Fig. 6 except that the water has been subtracted and components have been separated.
 Sedimentation across the J a p a n Trench 45
0 mentation paths and depocentres must have
SUBAERIAL
EXPOSURE
shifted through time and caused changes in
deposition and in facies associations. In the
500 mid-slope area these changes are significant
WAT E R
over 1Ma periods or less whereas in the
forearc basin area significant local changes took
OOO place over 2-3 Ma periods or longer. Thus it
appears that the morphology which also influ-
ences sedimentation, changes at very rapid
z 1500 rates near the trench compared to areas near
£
the arc.
o One interesting feature in the Japan Trench
2000
is the large amount of material filling the trench
floor opposite Hidaka Channel The widest part
D I M E NT
of this sediment mass, more than 20 km wide,
almost bridges the trench floor. Perhaps this
body represents a large deep-sea fan. The up-
per surface of the toe is marked by diffractions
on seismic record JNOC 1 (Fig. 2), perhaps
5OOO
AGE(MYBP) from slump blocks or fan channels. If a sedi-
20 ment mass bridges the trench, it could provide a
LATE EARLY I MIODLE LATE transport path for coarse material seaward of
OLIG. MIOCENE ] MIOCENE MIOCENE PLIO. \
the trench floor.
PLEISTOCENE

FIG. 8. Vertical movement at Sites 438 and

439 indicated by benthic foraminiferal
assemblages and plotted in the manner
described by Van Hinte (1978) (after Conclusions
Arthur et al. 1980 and Keller 1980).
Although no sediment, sedimentary structure,
or morphological feature is unique to any
particular environment along the Japan Trench,
certain features appear more common to one
was ponded in the mid-slope basin (Fig. 2), and area relative to another. The samples are
the bank at the edge of the forearc basin in insufficient to establish sediment facies associa-
JNOC-2 was uplifted (Fig. 5). The stress field tions typical of each morphological element
on Honshu also changed during this period defined by the bathymetry or observed in seis-
(Nakamura & Uyeda 1981) and slow uplift mic records. The high resolution seismic re-
began in the area of the coastal massifs (Chinzei cords however, are sufficient to make an esti-
1966). We propose that during the Pliocene mate of the areas covered by each type of
the lower slope sediment depocentres were re- morphological feature or sediment deposit,
arranged in their present form. As the slope as we have done in Fig. 9. Additionally, re-
was tilted, failure resulting in mass movement arrangement of sediment paths and depocen-
caused the Pleistocene hiatus observed at Sites tres can be said with confidence to be more
434 and 441. rapid near the trench than in the forearc basin.
This Neogene tectonic history of the conver- Our major conclusions are as follows:
gent margin off northern Honshu illustrates (1) Neogene and Quaternary sediment on
how the forearc basin area is tectonically more the Japan Trench convergent margin is mainly
stable than the slope area. This relative stability fine grained with rare discrete layers of sand.
is also shown by the depths of the erosional (2) For specific morphologies and sediment
angular unconformity at the base of the structures seen in seismic records the facies
Neogene section as seen in seismic records (Fig. associations cannot be determined without a
2). At its seaward end this unconformity has great number of additional samples tied to the
subsided as much as 6 km in Neogene time, a seismic and bathymetric data.
depth indicating very rapid rates of tectonism, (3) Thick (200 m) hemipelagic sediment of
whereas near the present coast of Honshu the terrigenous origin blankets the deep-ocean
unconformity is indistinct and there was very pelagic sediment of the seaward slope of the
little subsidence and perhaps minor uplift. Sedi- trench.
46 R. y o n H u e n e & M. A . A r t h u r

MORPHOLOGIC &
SHELF SLOPE TRENCH TRENCH TRENCH
MORPHOTECTONIC AXIS LANDWARD SEAWARD
FEATURE BASIN BASIN BASIN SLOPE SLOPE

upper Iowe,

FANS ~ ~ ~ " - - --

LONGITUDINAL
CHANNELS

TRANSVERSE
m . . . . . . . . . . m
CHANNELS

OVERBANK DEPOSITS m m ~,,,- - -

PONDED DEPOSITS . . . . . ~

DRAPED DEPOStTS

MASS M O V E M E N T
m m m . . . . .

RAPID M O R P H O - . . . . . ~ ~
TECTONIC CHANGE

FIG. 9. Estimate of the frequency with which morphological features are found in various parts of the
Japan Trench convergent margin, from seismic records and bathymetry.

(4) Tectonic disruption changed trench slope of sediment facies shows no distinct difference
basins to unstable slopes, and unstable slopes to from margins that were uplifted.
basins, in a 1-2 Ma period on the lower slope. (7) The coarse terrigenous material sampled
The rearrangement of channels and shifts in may have been associated with transverse chan-
depocentres were less abrupt (2-3 Ma) in the nels. Most of the coarse-grained sediment prob-
forearc basin. The relative rates of change in ably moves across the margin in such channels
sedimentation corresponding to change of mor- which deposit coarse-grained material in mor-
phology may help to distinguish between broad phologically diverse areas of the margin. This
environments in ancient convergent margins. opinion is shared by a number of other authors
This depends, however, on the ability to isolate (Underwood et al. 1980; Underwood 8,: Bach-
tectonic and eustatic influences on sedimenta- man 1981; Moore et al. 1981; Underwood &
tion. Karig 1980).
(5) Local sedimentation patterns at the lead- (8) Most morphological features and
ing edge of the convergent margin may at times sedimentary structures are common to all parts
be superimposed on sedimentary response to a of the margin, but ponded or draped sediment
regional tectonic event as is perhaps shown by without many slumps is most characteristic of
increased volcanism. Such an event was associ- the forearc basin, slumps are most prevalent in
ated with uplift of northern Honshu about slope basins and on trench floor, and canyons
3.5Ma ago and together with a change in are most common on trench landward slopes.
climate and lowered sea-level, these factors No environment appears to be distinguished by
increased the grain size of terrigeneous sedi- a single morphological feature, sedimentary
ment supplied to the margin. Thus, the Pleis- structure, or facies association. The relative
tocene high-energy environments are easier to abundance of an assemblage of features,
recognize by their sandier facies. During pre- however, is suggestive of the bathymetric en-
vious times the same sedimentary processe s left vironment along the Japan convergent margin,
a less diagnostic record in deposits of clay and especially when features in the different en-
silt. This change in sedimentation reinforces the vironments are compared. Given a single sam-
well-known importance o f tectonism in con- ple or even a small group of samples from one
trolling sedimentary facies along convergent locality, it would be very difficult to establish
margins; at sea tectonism influences sediment depositional environment except in the case of
depocentres and distribution systems, whereas deep-ocean pelagic sediment.
on land it influences the character of the sedi-
ment source. ACKNOWLEDGMENTS: We greatly appreciate the re-
(6) The succession of litho- and biD-facies views of our early manuscripts by David Howell,
found off northern Honshu indicate massive Mike Fields, Arnold Bouma, Steve Bachman and
subsidence of the margin; however, the pattern Jerry Leggett.
 Sedimentation across the Japan Trench 47

References
ARTHUR, M. A., VON HUENE, R. & ADELSECK, C. 1966. Sediments and structure of the Japan
1980. Sedimentary evolution of the Japan forearc Trench. J. geophys. Res. 71, 2121-37.
region off northern Honshu (DSDP Legs 56 and MOGI, m. 1978. An Atlas of the Seafloor Around
57). In: Initial Rep. Deep Sea drill. Proj. 56 & 57, Japan, Aspects of Submarine Geomorphology.
pt 1. U.S. Govt Printing Office, Washington. University of Tokyo Press. 96 pp.
CADET, J. & FUJIOKA, K. 1980. Neogene volcanic MOORE, G. F. & KARIG, D. E. 1976. Development of
ashes and explosive volcanism: Japan Trench sedimentary basins on the lower trench slope.
Transect, leg 57. Deep Sea Drilling Project. In: Geology, 4, 693-7.
Scientific Party, Init. Rep. Deep Sea drill. Proj. MOORE, J. C., WATK1NS, J. S., BACHMAN, S. B.,
56 & 57, pt 2. 1027-42. U.S. Govt Printing BEGHTEL, F. W., BUTT, A., DIDYK, B. M.,
Office, Washington. LEGGErr, J., LUNDBERG, N., MCMILLEN, K. J.,
CARLSON, P. R., MOLNIA, B. F., KI1TELSON, S. C. & NUTSUMA, N., SHEPHARD, L. E., SH1PLEY,T. H.,
HAMPSON, JR, J. C. 1977. Map of distribution of STEPHAN, J. F. & STRADNER, H. 1979. The Mid-
bottom sediments on the continental shelf, dle America Trench off Mexico. Geotimes, 24,
northern Gulf of Alaska. Two sheets, scale, 20-2.
1:500 000. U.S. Geol. Surv. miscell. Field Studies , MCM1LLEN, K. J., BACHMAN, S. B.,
Map MF 876. LEGGETY, J. K., LUNDBERG, N., SHIPLEY, T. H.,
CHINZEI, K. 1966. Younger Tertiary geology of the STEPHAN, J.-F., BEGHTEL, F. W., BuTr, A.,
Mabechi River Valley, northeast Honshu, DIDYK, B. M., NnTSUMA, N., SHEPttARD, L. E. &
Japan. J. Fac. Sci. Univ. Tokyo, Sec. II, 16, STRADNER, H. 1981. Facies belts of the Middle
161-208. America Trench and forearc region, northern
HASEGAWA, S., SAKAI, T., OKAMURA, M. & Mexico---results from Leg 66 DSDP (this
TAKAYAMA,T. 1977. Age assignment of the silt- volume).
stone fragments dredged. In." HONZA, E. (ed.). NAKAJIMA, T. 1973. Submarine topography off the
Geological Investigation of Japan and Kurile southern part of Sanriku district. Mag. Earth Sci.
Trench and Slope Areas GH 76-2 Cruise, pp. 82, 136-47.
80-5. April-June 1976. Geol. Survey of Japan. NAKAMURA, K. & UVEDA, S. 1981. Stress gradiant in
HESSE, R. 1977. Softex-radiographs of sliced piston arc-back arc regions and plate subduction. J.
cores from the Japan and southern Kurile Trench geophys. Res. 85, 6419-28.
and slope areas. In: HONZA, E. (ed.). Geological NAGUMO, S., KASAHARA, J, & KORESAWA, S. 1980.
Investigation of Japan and Southern Kurile OBS airgun seismic refraction survey near sites
Trench and Slope Areas GH 76-2 Cruise, pp. 441 and 434 (J-1A), 438 and 439 (J-12), and
86-108, April-June 1976. Geol. Survey of Japan. proposed site J-2B: legs 56 and 67, Deep Sea
HONZA, E. 1977. Outline of the cruise. In: HONZA, E. Drilling Project. In: Scientific Party, Init. Rep.
(ed.). Geological Investigation of Japan and Deep Sea drill. Proj. 56 & 57, pt 1,459-62. U.S.
Southern Kurile Trench and Slope Areas GH Govt Printing Office, Washington.
76-2 Cruise, pp. 1-9, April-June 1976. Geol. NASU, N. 1964. The provenance of the coarse sedi-
Survey of Japan. ments on the continental shelves and the trench
-- , TAMAKI, K. & MURAKAMI,F. 1978. Geological slopes off the Japanese Pacific coast. In: MILLER,
map of the Japan and Kuril Trenches and the R. L. (ed.). Papers in Marine Geology: Shepard
adjacent areas. Geol. Survey of Japan. One Commemorative Volume. New York, 531 pp.
sheet, 1:1 000 000. , hJIMA, A. & KAGAM1, H. 1960. Geological
INOUCHI, Y. & TAMAKI, K. 1977. 3.5 kHz echo results in the Japanese Deep Sea Expedition in
sounder profiling survey. In: HONZA, E. (ed.). 1959. Oceanogr. Mag. 2, 201-14.
Geological Investigation of Japan and Southern , TOMODA, Y., KOBAYASHI, K., KAGAMI, H.,
Kurile Trench and Slope Areas GH 76-2 Cruise, UYEDA, S., NAGUMO, S., KUSHIRO, I., OZIMA,
pp. 17-20, April-June 1976. Geol. Survey of M., NAKAZAWA, K., TAKAYANAGI, Y., OKADA,
Japan. H., MURAUCHI, S., ISHIWADA, Y. & ISHI1, Y.
, YUASA, M. • ODONERA, K. 1977. Cored 1979. Multichannel Seismic Reflection Data
materials. In: HONZA, E. (ed.). Geological Inves- Across the Japan Trench. IPOD-Japan Basic
tigation of Japan and Southern Kurile Trench and Data Series, No 3, Ocean Research Institute,
Slope Areas GH 76-2 Cruise, pp. 78-9, April- University of Tokyo, 22 pp.
June 1976. Geol. Survey of Japan. .... , VON HUENE, R., ISHIWADA,Y., LANGSETH,M.,
ISHIWADA,Y. ~,~ORGAWA,K. 1976. Petroleum geology BURNS, T. & HONZA, E. 1981). Interpretation
of offshore areas around the Japanese islands. of multichannel seismic reflection data, Legs
UN-ESCAP. Bull. CCOP Tech. 10, 23-4. 56 and 57, Japan Trench transect, Deep Sea
KELLER, G. 1980. Benthic Foraminifera. In: Init. Drilling Project. In: Init. Rep. Deep Sea drill.
Rep. Deep Sea drill. Proj. 57. U.S. Govt Printing Proj. 56 & 57, pt 1. U.S. Govt Printing Office,
Office, Washington. Washington.
LUDWIG, W. J., EWING, J. I., EWING, M., MURAUCHI, - - , & SATO, T. 1962. Geological results in the
S., DEN, N., ASANO, S., HO'FFA, H., HAYAKAWA, Japanese Deep Sea Expedition in 1961 (JEDS-
M., ASANUMA, T., ICHIKAWA,K. & NAGUCHI, I. 4). Oceanogr. Mag. 13, 155-66.
48 R. y o n H u e n e & M. A . A r t h u r

ODONERA, K. & HONZA, E. 1977. Bathymetric sur- nary Depositional Environments on the Pacific
vey. In: HONZA, E. (ed.). Geological Investiga- Continental Margin, 211-29. Pacif. Sec. Soc
tion of Japan and Southern Kurile Trench and econ. Paleontol. Mineral.
Slope Areas GH 76-2 Cruise, pp. 10-60, April- UNDERWOOD, M. B. & Khgl6, D. E. 1980. Role of
June 1976, Geol. Survey of Japan. submarine canyons in trench and trench-slope
SAKURAI, M., NAGANO, M., NAGAI, T., KATSURA,T., sedimentation. Geology, 8, 432-6.
TOZAWA, M. ~. IKEDA, K. 1975. Submarine geol- VAN HINTE, I. E. 1978. Geohistory analysis--applica-
ogy off the south coast of Hokkaido. Hydrogr. tion of micropaleontology and exploration geolo-
Res. 10, 1-37. Hydrographic Office of Japan. gy. Bull. Am. Assoc. Petrol. Geol. 62, 201-29.
SCIENTIFIC PARTY, 1980. Init. Rep. Deep Sea drill. VON HUENE, R., LANGSETH, M., NASU, N. & OKADA,
Prof. 56 & 57. U.S. Govt Printing Office, S. 1980. Summary, Japan Trench Transect. In:
Washington. lnit. Rep. Deep Sea drill. Prof. 56 & 57, pt 1.
TAMAKI, K.,, INOUCHI, Y., MURAKAMI, F. & HONZA, U.S. Govt Printing Office, Washington.
E. 1977. Continuous seismic reflection profiling , NASU, N., ARTHUR, M., CADET, J. P., CARSON,
survey. In: HONZA, E. (ed.) Geological Investi- B., MOORE, G. W., HONZA, E., FUJIOKA, K.,
gation of Japan and Southern Kurile Trench and BARRON, J. A., KF.I.I.ER, G., REYNOLDS, R.,
Slope Areas GH 76-2 Cruise, pp. 50-71, April- SHAFFER, B. L., SATO, S. & BELL, G. 1978. Japan
June 1976. Geol. Survey of Japan. Trench transected .on Leg 57. Geotimes, 23,
UNDERWOOD, M. B. • BACHMAN, S. B. 1981. 16-21.
Sedimentary facies associations within subduc- YUASA, M., IONUCHI, Y., ODONERA, K. & KIMURA,
tion complexes (this volume). M. 1977. Rocks and sediments. In: HONZA, E.
, -- t~ SCHWELLER,W. J. 1980. Sedimentary (ed.) Geological Investigation of Japan and
processes and facies associations within trench Kurile Trench and Slope Areas GH 76-2 Cruise,
and trench-slope settings. In: FIELD, M. E., pp. 72-7, April-June 1976, Geol. Survey of
BOUMA, A. H. & COLBOURN, I. (eds). Quater- Japan.

ROI.AND YON HUENE, U.S. Geological Survey, Menlo Park, California, U.S.A.
MICHAEL A. ARTHUR, U.S. Geological Survey, Denver, Colorado, U.S.A.
 Tectonics of some forearc fold belts in and around the arc-arc
crossing area in central Japan

Yujiro Ogawa
SUMMARY: The zonal structure of the old Honshu arc in central Japan is deflected and
cut by the colliding younger Izu-Bonin arc causing peculiar types of accretionary fold
belts. The Shimanto Fold Belt (s.s.), made up of Cretaceous flysch and pre-flysch
sediments, shows a collisional fold pattern: arcward vergence in the inner side and
oceanward vergence on the outer side. The minor structures are also different, with shear
folds on the inner side of the belt and lens folds on the outer side. Structures in the inner
side grew by flattening and those in the outer side by simple shearing. The divergent
development of structural characteristics is explained by the differences in geothermal
gradient and stress fields across the belt. The Izu-Bonin arc has a forearc fold belt situated
between the Honshu and Izu-Bonin arcs. The stratigraphical, palaeogeographical and
structural history of the area shows that it developed in a strike-slip regime associated with
the transform fault between the Eurasian and Philippine Sea plates.

In the south-western part of the Mesozoic- Kimura 1974). Each orogeny has a paroxism
Cenozoic H o n s h u arc of Japan, several accre- around 240 Ma (Permo-Triassic) and 140-
tionary fold belts have been tectonically juxta- 110 Ma (Jurasso-Cretaceous) respectively (Ono
posing since late Mesozoic times, making the 1980). The Sakawa Orogen forms the backbone
area most appropriate for tectonic analysis of of the Eo-Nippon (Eo-Japan) Cordillera
trench and forearc geology (Kimura 1974; Oga- (Kobayashi 1941). Several inliers expose older
wa 1978; Kanmera et al. 1980). The accretion- tectonic zones, made of remnant arcs or micro-
ary fold belts are made of arc- or continent- continental fragments and dismembered
derived flysch sediments which are underlain ophiolite complexes which include high press-
with both concordant and tectonic contacts by ure metamorphic rocks dated at 380-240 Ma
pre-flysch (pelagic) sediments and oceanic or (Maruyama et al. 1978) beside 390 Ma granitic
ophiolitic rocks (Kanmera 1976; Sakai 1978; and gneissose rocks (Ishizaka 1972, the Rb-Sr
Sakai & Kanmera 1981). The old Honshu arc is age is recalculated by Miss R. H a m a m o t o 1980,
cut and bent in its central part by collision of the pers. comm.).
Izu-Bonin arc which causes several peculiar The new Izu-Bonin arc, on the other hand, is
types of accretionary fold belts (Ogawa & a juvenile arc mostly comprising volcanogenic
Horiuchi 1978) (Fig. 1). materials of Tertiary and Q u a t e r n a r y ages.
In this paper ! describe these belts in and Collision between the Izu-Bonin and H o n s h u
around the arc-arc crossing area, especially the arcs might have originated as long ago as the
Cretaceous Shimanto Fold Belt (s.s.) in the Cretaceous (Matsuda 1978; Ichikawa .1980).
Kanto Mountains and the Neogene Miura Fold Voluminous basaltic to rhyolitic lavas and pyro-
Belt in the Miura and Boso Peninsulas (Fig. 1), clastics filled the intra-arc basins, and volcani-
I compare their tectonic features with those of clastics filled the forearc basins. In the backarc
the equivalent fold belts in other areas, and basins volcanogenic materials are interbedded
assess the tectonic significance of collisional and with terrigenous clastics supplied from the old
strike-slip tectonics in their evolution. arc. The Izu-Bonin arc runs through the old
Honshu arc in the great tectonic graben called
the 'Fossa Magna'.
The tectonic status of the central part of the
Tectonic setting old H o n s h u arc and the surrounding areas can
The old H o n s h u arc is composed chiefly of be explained historically using the theory of
orogenic belts comprising middle or upper plate tectonics as follows:
Palaeozoic to Mesozoic 'geosynclinal' sedi- Before about 40 Ma, prior to the establish-
ments and their metamorphic equivalents. ment of the present plate system around Japan,
Paired metamorphic belts constitute the cores the eastern margin of the Eurasian Continent
of the Akiyoshi and Sakawa Orogenic Belts was underthrust by the NNW-subducting Kula
(Fig. 1) (Kobayashi 1941; Miyashiro 1961; or Pacific plate (Emperor Seamount direction)

49
50 Y. O g a w a

(Uyeda & Miyashiro 1974). The parallel tains (the eastern side of the arc-arc crossing
arrangement of the orogenic and accretionary area) by Kosaka (1979) indicate that the fold
fold belts of late Palaeozoic to early Cenozoic belts rotated rather rigidly in the Palaeogene to
age were produced by welding and accretion of Neogene without significant lateral converg-
oceanic, continental and island arc materials ence or dragging of the fold belts. Vertical
during this long subduction phase (Uyeda & faults with minor right-lateral slippage occurred
Miyashiro 1974; Kimura 1974; Ogawa 1978; in the Neogene.
Ono 1980; Kanmera et al. 1980). At present, the Philippine Sea plate borders
When the spreading direction in the Pacific on the Nankai Trough south of Southwest
region changed to WNW (Hawaiian direction) Japan. The plate boundary enters into the
in the late Palaeogene, about 40 Ma ago, the Paci- Suruga Trough or runs along the eastern edge
fic plate began to subduct along the Japan, Izu- of the Izu Peninsula, and then it turns sharply to
Bonin and Mariana trenches, and consequently the SE along the Sagami Trough, joining the
the Neogene to Quaternary Izu-Bonin volcanic Japan and Izu-Bonin trenches at the triple
arc was formed. The Philippine Sea plate was junction called the Japan Triple Junction in this
trapped to the west of the arc and backarc paper (Fig. 1). Present tectonism around the
spreading occurred from 30 to 15 Ma ago, i.e. north-eastern edge of the Philippine Sea plate is
late Oligocene to middle Miocene (Kobayashi said to result from NW or NNW compression
& Nakada 1978). In the Miocene, the Sea of (Seno 1977; Matsuda 1978; Matsubara 1980),
Japan also began to open, associated with large- and lateral compression of the Honshu arc by
scale lateral faulting in the eastern and western the colliding Izu-Bonin arc has been effective
parts of Japan (Otsuki & Ehiro 1978). The since the Cretaceous.
Eo-Nippon Cordillera was detached from the
Eurasian Continent at this stage to form the
detached type of island arc of Dickinson &
Seely (1979). The Philippine Sea plate may Subduction and collision tectonics of
have been separated from the Honshu arc by a the S h i m a n t o Fold Belt in central Japan
strike-slip fault without any significant effects of General structure and stratigraphy
subduction at this stage (Kobayashi & Nakada
1978). M. Takahashi (1980, pers. comm.) con- The structural character of the accretionary
siders that a pulse of subduction of the hot fold belts in the outer part of the Honshu arc
Philippine Sea plate occurred in the middle is essentially imbricate (Kobayashi 1941;
Miocene, since considerable rhyolitic magmat- Kimura 1974; Ogawa 1978). The strata involved
ism occurred on the forearc area of Southwest are broadly 'geosynclinal', comprising alterna-
Japan around 14 Ma (Nakada & Takahashi tions of continent-derived flysch and pre-flysch
1979). argillaceous sediments, associated with tholeii-
Throughout these subduction phases, from tic and alkalic basalt, chert and limestone. The
the Cretaceous to the present, a remarkable sediments in places are intercalated and inter-
lateral bending of fold belts took place in mingled by both sedimentary and tectonic de-
central Japan. The west-east trending zonal formation to form what may be termed
structure in Southwest Japan became sharply sedimentary or tectonic m61anges. They are
bent and cut in the shape of the letter 'W' (Fig. metamorphosed into greenschist or lower grade
1). The en echelon arrangement of the internal facies. Thrusts in the imbricate structure are
fold axes in the Sambagawa Metamorphic Belt mostly inclined inwards, but in the arc-arc
and the rotation of the Cretaceous Ryoke crossing area folds on a variety of scales show
Metamorphic Belt and of the myionite along both inward and outward inclination as will be
the Median Tectonic Line in the western and described later.
eastern parts of the arc-arc crossing area indi- A typical group of accretionary fold belts is
cate that northward inhomogenous compress- well exposed in the Sambosan-Shimanto Ter-
ion effected ductile bending both on the west- rane on the outer side of the Sakawa Fold Belt
ern and eastern sides of the arc-arc crossing in in Southwest Japan (Figs 1 & 4 bottom). In this
the Cretaceous (Hara et al. 1980). The lateral terrane, several Mesozoic to Cenozoic fold
drag of fold belts with northward convergence belts, in which the strata become progressively
in the Akaishi Mountains (Fig. 1) indicates that younger outwards, are arranged parallel to the
plastic left-lateral dragging occurred due to major zonal structure. The late Tertiary to
large-scale lateral slip during the Cretaceous to Quaternary fold belts are restricted to the arc-
Neogene (Kimura 1959, 1961). On the other arc crossing area.
hand, structural analyses in the Kanto Moun- The Cretaceous to early Miocene fold belts in
 Tectonics o f s o m e forearc belts in central Japan 51

ISAKAWA,.,. /7 20Okra~7 y

-,_-,-~
"''-
~.~ I. ~ ,,~-L,/'I ~ ",- ~
~ v ',
I ~q~r
: ~ 4~.,
-...,
1"~: :'-'--'--'-~F':B "~- ~_~
. K A:..Ei.:.T.ON ...:..~Q ~ ~,~ " ff~"~.:.~':~-~f /

0 t SHIMANTO(,~).~t
-A "-~ ~ (-FOL D_BEL T~"~ T0 K Y.O~ S.~

~-~ . _ " _ $ V T A N Z A W A
;
, ~ ~
Mf. FUJI
~,'0
E .~
~ J •
EURASIAN PLATE

] ~ ~ ~PENINSULA

:f --qr,,

v~ T ~ _
o H-.y---

PHILIPPINE SEA PLATE

I)
-- 34"1
I

FIG. 1. Index maps of the forearc fold belts in the arc-arc crossing area in central Japan. From inner
to outer: Sakawa Fold Belt (Ry, Ryoke Metamorphic Belt; Sb-Ch, Sambagawa-Chichibu Belt);
Sambosan F. B.--coarsely dotted; Shimanto F. B.--finely dotted; Setogawa F. B.---dashed;
Oigawa-Tanzawa F. B.--large v symbol; Miura F. B.--small v symbol. M.T.L.--Median Tectonic
Line. A and B indicate the sites of cross-sections in Fig. 2.

the Shimanto Terrane are generally made up of as radiolarian chert, umber-like sediments and
flysch and pre-flysch sediments which have radiolarite) usually within a pebbly mudstone
been explained as forearc and ocean floor de- (Tsuchiya et al. 1979). The pre-flysch sediments
posits (Kanmera 1976; Sano et al. 1979; Sakai & may merit interpretation as accretionary prism
Kanmera 1981; Taira et al. 1981). The flysch basement, mixed by sedimentary and/or tec-
sediments are composed of thick sandy or mud- tonic means, and the overlying terrigenous
dy alternations, often turbiditic. Some large- flysch as mantling lower trench slope sediments
scale slump or slide deposits are mappable as (Suzuki & Hada 1979; Sakai & Kanmera 1981;
olistostrome beds (Kanmera 1977). The pre- Taira et al. 1981).
flysch sediments are largely composed of argil- Some limestone and chert blocks contained
laceous materials including fragments or strips within the Cretaceous argillaceous pre-flysch
of basalt, chert, limestone, as well as of sand- sediments yield Triassic or Jurassic fossils, in-
stone. I formerly considered that the pre-flysch cluding conodonts, corals and radiolarians (Fu-
sediments were conformably overlain by, or jimoto & Suzuki 1969; Nishimiya & Yamagiwa
intercalated to some extent with, the flysch 1973; Yamato-Ohmine Research Group 1979).
sediments and that some basaltic volcanism Permian fusulinids are occasionally found (Fu-
occurred within the flysch basin (Ogawa & jimoto & Suzuki 1969). These blocks of older
Horiuchi 1978). This I now consider doubtful sediments are probably recycled materials de-
since there is no strong evidence that the vol- rived as olistoliths from older (inboard)
canism was in situ. Rather the basaltic rocks accreted strata of the Chichibu or Sambosan
commonly accompany pelagic sediments (such Terranes.
52 Y. Ogawa
INNER SIDE @

~' b A ' ' " OUTER SIDE @

f ~/~\\\W~ ', .O~N~Z~W~

e ~,~,~ ~) g ->~

, - .,-.

Fff;. 2. Fold patterns in the Kobotoke Group of the southern part of the Shimanto Fold Belt in the
Kanto Mountains around A and B in Fig. 1. Small arrows indicate the facing of beds.

Structural characteristics of the Cretaceous

Kobotoke Group in the Kanto Mountains
account is summarized in Fig. 3. Flow folds
The Cretaceous sediments of the north characterized by schistosity are developed at
Ogochi and south Kobotoke Group in the the lower levels in the core of the fold belt.
Kanto Mountains are attributed to the products Metamorphic grade is commonly as high as
of a subduction regime. Large-scale olisto- greenschist facies, decreasing laterally or verti-
stromes characterize the m61ange zones, and cally into pumpellyite facies. In upper levels a
are overlain by terrigenous turbidites. Green- series of shear folds and a series of lens folds are
stones within the pre-flysch facies are mostly developed in the inner and outer zones respec-
large slabs or sheets of metabasalt and those tively (Fig.3) Other structural characteristics
within the flysch facies are volcaniclastic sedi- are also different in the two sides (Fig. 4).
ments such as volcanic conglomerates or sand- In the mid-depths of the inner side, W- or
stones. M-shaped symmetric folds are common and the
Large-scale closed and overturned folds are fold style corresponds to the so-called shear
developed in the flysch sediments of the Upper folds, characterized by development of well-
Cretaceous Kobotoke Group. Folds of various developed axial planar slaty cleavage (e.g. Fig.
styles verge northward in the inner (north) side 2a). Typical folds in this region belong to Class
and verge southward in the outer (south) side 1C or Class 2 of Ramsay (1963). Associated
(Figs 2 & 4). Structures with opposed vergence bedding slip is occasionally recognizable in the
are well developed in the central part of the interturbidite mudstones. In such cases the
group. Similar fold patterns are also recognized bedding slip plane is shown as a segregated vein
in the Shimanto Fold Belt in the Akaishi Moun- of albite and quartz, and the vein is refolded on
tains (Matsushima 1978; Ogawa & Horiuchi a microscale in the axial part of folds. This
1978) and in the Chichibu Terrane in the Kanto indicates that the folds were developed by
Mountains (Takizawa 1979). bedding slip in the early stages of deformation
Differing development of fold styles (Ramsay and by flattening in the later stages.
1963; Kimura 1968) is also distinguishable be- Sedimentary structures such as graded bedding
tween the inner and outer sides of the Kobo- or lamination are preserved, but sole marks are
toke Group outcrop. Terminology used in this not.
 Tectonics of some forearc belts in central Japan 53

In upper levels of the inner side, flexure folds Tectonic significance of the differing fold styles
are developed. Weakly developed slaty cleav- on the inner and outer sides of the Shimanto
age is observed in the muddy part, and bedding Fold Belt
slip is only rarely observed (e.g. Fig. 2b).
In the mid-depths of the outer side, S- or Similar development of differing fold styles
Z-shaped asymmetric folds are common and on the inner and outer sides of the Shimanto
the fold style corresponds to the so-called Fold Belt is also well recognized in the Akaishi
lens folds of Kimura (1968) (e.g. Fig. 2e,f,g). Mountains region (Ogawa & Horiuchi 1978).
Brittle beds such as sandstone are fractured or In both the Akaishi and Kanto study areas
faulted into lensoidal blocks and sometimes fold styles can be compared in terms of brittle
flow along the bedding slip plane. The blocks and ductile responses to deformation. The
were originally deformed by extension or shear alternating sandstone and mudstone layers
fracturing and then stretching and flowage or show different reactions according to two fac-
even rotation to form tectonic lenses in the later tors: mean ductility and ductility contrast
stages of folding. The axial parts of the folds are (Uemura 1981). The mean ductility depends
often recognized only by critical determinations both on temperature and confining pressure
of the facing of beds. Bedding slip and shearing (Handin & Hager 1957, 1958), i.e. it is a
are usually strong and some sedimentary struc- product of tectonic level. The ductility contrast
tures are deformed and destroyed. The types of depends chiefly on temperature when confining
folds are difficult to determine but generally pressure is constant. Under relatively high
correspond to Class 1B or 1C. Vergence of the temperature, the ductility contrast is low, while
folds is mostly outward. Slaty cleavage is some- at lower temperature it is high, as shown by the
times developed, but the texture is quite differ- experimental work of Handin & Hager (1958).
ent from that in the shear folds of the inner Therefore, deformational mode in a horizon
side. where the confining pressure is constant is
In upper levels of the outer side, flexural slip determined by geothermal gradients (Ogawa &
folds accompanied by remarkable bedding slip Horiuchi 1978; Kimura 1979) (Fig. 3).
are commonly developed (e.g. Fig. 2f). No slaty When the ductility contrast is high, deforma-
cleavage occurs. tion during folding proceeds by pronounced

LOW -- GEOTHERMAL GRADIENT -- HIGH

HIGH m DUCTILITY CONTRAST -- LOW

s e r i e s of s e r i e s of
I e n s folds shear folds

v~ .~.. flexure

flexural-slip ./~,

%' , ,

,//~ w ( type ~1 "~- (type A )

FIG. 3. Fold styles and their arrangement according to tectonic level in the Kobotoke Group,
showing series of lens folds and shear folds. For further explanation, see text.
54 Y. Ogawa
bedding slip, and grain boundary slip is con- (Moore 1979), suffering deformation in a lower
fined within individual sandstone or mudstone geothermal gradient. Faulting and fracturing
beds. On the other hand, when the ductility occur and a series of lens folds is likely to result.
contrast is weak, grain boundary slip occurs On the other hand, sediments in the inner side
along planes which can penetrate through the of the belt are far from the subduction zone and
bedding boundaries, and passive folding occurs. much nearer to the volcanic front, and suffer
If bedding slip proceeds, with development of from deformation by flattening in a higher
fractures or faults of the brittle layers, lensoidal geothermal gradient. Thus, steady-state flow
blocks form giving rise to the characteristic lens occurs and a series of shear folds is likely to
fold style (Kimura 1968). Thus the series of lens result.
folds is constituted by flow folds in the lowest In conclusion, sediments deposited in the
tectonic level and lens folds and flexural slip trench and forearc area were progressively de-
folds at higher levels. The series of shear folds formed in a subduction regime (Fig. 4). The
comprises flow folds, shear folds and flexure vergence pattern across the belt around the
folds in upward succession (Fig. 3). These arc-arc crossing area probably results from col-
observations indicate that the lens fold series is lision of the Izu-Bonin and Honshu arcs, giving
the characteristic structural mode in areas of inner vergence in the inner side of the belt and
low geothermal gradient, and that the shear outer vergence in the outer.
fold series is the characteristic mode in areas of
high geothermal gradient.
In the Kobotoke Group, as described pre- Strike-slip tectonics of the Miura Fold
viously, the inner and outer sides of the outcrop
Belt in the Miura and Boso Peninsulas
belt are characterized respectively by series of
shear folds and lens folds (Fig. 2). However, The stratigraphy and structure in the Miura and
the assemblages of metamorphic minerals in the Boso Peninsulas is quite different from that in
two sides do not show obvious critical differ- the other Shimanto terranes. This is because of
ences indicative of different geothermal gra- the singular tectonic position of the area, which
dients. Vertical and lateral changes of meta- is situated on the sandwiched forearc basin
morphic grade should be further studied. between the Honshu and Izu-Bonin arcs (Fig.
In Japan, a similar divergent development of 1). Sediments were supplied both from the Izu
shear and lens folds is observed in the Sakawa volcanic arc to the west and from the Honshu
Orogenic Belt. In the Ryoke high temperature arc to the north. The structural framework
belt, a shear fold zone is underlain by gneiss, consists of gentle soft-sediment folds, forming
while in the Sambagawa-Chichibu high pressure curving outcrop patterns and arranged in N W -
belt, lens folds are underlain by schist with SE trending zones (Fig. 5). These indicate that,
sporadic development of shear folds (Ogawa unlike other Shimanto forearc fold belts, the
1978). This may indicate that the divergent sediments did not suffer significant deformation
development of fold-styles is possibly related to or metamorphism, but rather were subjected to
the paired metamorphism in the arc-trench strike-slip tectonism throughout sedimentation.
system (Ogawa & Horiuchi 1978). This tectonism was probably the result of move-
The tectonically divergent development of ment on the transform fault between the Eura-
fold styles or series of folds and other deforma- sian and Philippine Sea plates.
tional modes is as follows. Sediments of the
Shimanto Fold Belt were deposited in trench
Stratigraphy and palaeogeography
and forearc areas--flysch sediments in the fore-
arc trough or basin and pre-flysch sediments on The Palaeogene to Neogene rocks and sedi-
the inner trench slope or in a trench. The ments are divided from bottom to top into the
pre-flysch sediments contain oceanic and pela- Mineoka Group, Hota Group, Yabe Formation
gic materials, are covered by the flysch sedi- and Miura Group. The Miura Group forms the
ments and occupy the lower stratigraphic and majority of the outcrop area, while the others
structural position. They were accreted to the are restricted to the vicinity of the Hayama-
inner trench slope during subduction, giving Mineoka Uplift Belt and the offshore (sub-
rise to inward-dipping imbricate faults (Karig merged) Okino-yama Uplift Belt (Geological
1974; J. C. Moore & Karig 1976; G. F. Moore Survey of Japan 1976; Kimura et al. 1976)
& Karig 1976). Sediments in the outer side of (Fig. 5).
the Shimanto Fold Belt are much nearer to the
subduction zone, and are pushed into and Mineoka Group
under the flysch sediments by simple shearing Outcropping within the structural horst in the
 Tectonics of some forearc belts in central Japan 55

VOLCANIC ASEiSMIC TRENCH

FRONT FRONT

FORE - A R C AREA

FL YSCH BASIN PRE-FLYSCH

BASIN
~-~~-_=- 'L'~--~';~.~- .....
~_--~ ~~"-J~_.-~.~ 7--

,/~ ! , ,
/ ' 7~_/-',,
- " 10km
/

FORE-ARC FOLD BELT

INNER OUTER
fold style /
,,A~C-,FLEXURE F LE XUR~L._LS~LIP /
/ --" " Y~...~.
" " SHEAR ~ - _ ~ - ~ , ~ ~A~' I ~ . ~ . @ / ~ ' /

D~COkLEME NT

KANTO MOUNTAINS
SAMBOSAN i SHIMANTO FOLD BELT
, ~ ~ ~: ~ ~.. /
",.,~. ,~ 'z~:-~.~.+~ .;~----7, .'~,, ~,.,~ ~,,. . . ~j&coo~. ~ "........~>..~"./5;. '-

OGOCH I G R O U P / '

FIG. 4. Tectonic regime of the Shimanto Fold Belt in the Kanto Mountinas. Notice that pre-flysch
sediments of A n age are accreted under the flysch sediments of Bn + 1 age. Tsunami F. corresponds
to the Tsunamigenic fault of Fukao (1979).

central part of the Boso Peninsula, the Mineoka The amphibolite schist and hornblende gab-
Group is predominantly composed of ophiolitic bro or diorite give K-Ar ages of 38 Ma (late
rocks which comprise tholeiitic pillow lavas, Eocene) and 14 Ma (middle Miocene) re-
dolerite intrusives, and hyaloclastites (Uchida spectively (Yoshida 1974). Mudstone inter-
& Arai 1978). Detrital or sheared serpentinites calated between the pillow lavas yields radio-
and pyroxene gabbro intrusives occur in the larians probably of early Miocene age.
eastern part, with alternating siliceous and cal- Mudstone, alternating with serpentinite, yields
careous mudstones and turbidites in the west- early Miocene foraminifera (Globigerinita
ern part. The sheared serpentinites are com- unicava and Globigerinoides trilobus) (Yoshida
monly intruded by hornblende gabbro and also 1974).
contain blocks of hornblende gabbro, harzbur- Some rocks of the Mineoka Group, especial-
gite, picrite and amphibolite schist (Kanehira et ly the amphibolite schist, pyroxene and horn-
al. 1968; Kanehira 1976). Some ophiolitic rocks blende gabbros, chert and claystone show weak
which crop out as knockers within serpentinite, but significant cataclastic or stilolitic textures.
mudstone or sandstone are emplaced as ser- Only the amphibolite schist has undergone sig-
pentinite m61ange and olistostromes. nificant metamorphism and shearing. The
56 Y. O g a w a

,:., M i l J R A P, EN. :°

.__I" SH I I ~ K U BASIN
RIFTING
.t HAWAI I A N STAGE
-~'EMPEROR SEAMOUNT STAGE

C'-. . . . -.------ -------- -" "-_-__

-, ..,'"
O~ ~ ........../
.....

-_ ~,., ?..._.--,, ~ ~.~.~ ..... .. |

~ .. ", ~ , . -,,.~, [

,/f
0
.~" "-~."-- (":'""" " // _35o N

_~ -- ..-~-.

% " ~ ,'aE

Fw,. 5. General structural pattern in the Miura Fold Belt in the Miura-Boso Peninsulas. White
arrows in the map indicate the palaeocurrents in the middle part of the Miura Group. D, dome or
anticline; B, basin or syncline. For further explanation, see text.

others are of zeolite grade, the metamorphic }'abe Formation

minerals occurring in veins. These differing This formation is composed of basaltic to
indicators of metamorphism indicate that the rhyolitic tuffaceous or pumiceous sediments
rocks were emplaced as tectonic blocks, in- which were derived from the west, i.e. from the
corporated in a serpentinite matrix. Subse- Izu volcanic arc. These volcaniclastic rocks
quently gabbroic dykes were intruded into the overlie the Mineoka and H o t a Groups uncon-
breccias. The assemblage of rocks in the formably. In some areas blocks of pillow basalt
Mineoka Group seems to indicate trapped are intercalated in the basal part of this forma-
oceanic materials sandwiched in the forearc tion (Kimura et al. 1976).
area in the early to middle Miocene (Fig. 6).

Hota G r o u p Miura Group

This group is composed of rhyolitic tuf- This group is widely distributed, and the
faceous sediments and radiolarian-bearing stratigraphy, palaeogeography and structure
siliceous or calcareous siltstone or claystone; it are well known. The group is divided, in
also contains some large blocks of altered tuff ascending order, into the Sakuma, Misaki,
and chert probably from the Mineoka Group. Hatsuse, Zushi and Ikego Formations. Fora-
The strata form gentle dome-basin structures. minifera are of N8 to N21 (middle Miocene
The upper part of the group yields lower- to early Pleistocene) age (Ikebe 1978). The
middle Miocene radiolarians such as Cyrtocap- group unconformably overlies the previously
isella cornuta, C. tetrapera and Stichocorvs del- described units.
montensis. The Sakuma Formation only outcrops in the
 Tectonics o f some forearc belts in central Japan 57
vicinity of the Mineoka Group and has a char- Group, an assemblage very similar to typical
acteristic basal conglomerate composed largely lithofacies of other parts of the Shimanto Ter-
of Mineoka-derived volcanic or gabbroic rocks. rane. The basin was probably dammed up tem-
In the Misaki and later formations, the lower porarily by uplift of the Mineoka Group,
part is generally composed of turbiditic alterna- caused by rotation of the Honshu arc during the
tions of scoriaceous sandstone or conglomerate Miocene in the transform zone between the
and rhyolitic tuffaceous or pumiceous siltstone; Eurasian and Philippine Sea plates (Fig. 6).
the upper part is composed of rather coarse
scoriaceous and pumiceous unsorted sediments Structures of the Miura Fold Belt and their
which often show cross-bedding. Tuffaceous tectonic significance
limestone or calcareous tuff occurs in the lower
part. Scoriaceous or pumiceous materials are Structurally the Neogene strata in the Miura
coarse and thick in the north-western and Fold Belt are characterized by gentle undula-
south-western coastal areas of the Miura-Boso tions giving rise to a dome-basin pattern.
Peninsulas and in these areas large-scale sub- N W - S E trending faults bound the stratigraphic
marine slide and slump deposits are common zones of the fold belt, while folds occur be-
(Ogawa & Horiuchi 1978; Kozima 1980). tween each fault (Fig. 5). Sediments of the
In the middle horizon of the Miura Group, Miura Group are all poorly consolidated. The
around the key tuff bed (called Ok by Mitsu- mechanical properties of fine siltstones of the
nashi & Yazaki 1958), remarkable conglomer- Miura Group are characterized by 'visco-
ates of various kinds are intercalated and trace- ductile' behaviour beyond 100 or 300 bars con-
able over a wide region (Fig. 5). In the south- fining pressure in triaxial compression tests
ernmost part of the Boso Peninsula, greenish (Hoshino et al. 1972). This indicates that the
volcanic sandstone, containing pebbles of fresh sediments have never been subjected to a con-
terrestrially erupted basalt, is widely distri- fining pressure over those values. The specific
buted. Cross-beds in the conglomerate indicate density ranges from 1.44 to 1.91 (Hoshino et al.
northward transport. Restricted distribution of 1972). No significant alteration nor meta-
the peculiar sediments and slump deposits as morphism is recognized except in the rocks of
well as the northerly movement, indicate that the Mineoka Group. Volcanic glass and cristo-
the material was derived from a volcanic island balite or trydimite are common in the tuff beds,
sited to the south or SW. In the middle part and montmorillonite or smectite is present in
of the peninsula, a greenish conglomerate the fine sediments.
contains basalt and gabbro pebbles, and it is Although for the most part poorly consoli-
interbedded within the scoriaceous and dated and unmetamorphosed, the strata of the
pumiceous beds. Cross-beds over a wide region Miura and Boso Peninsulas show very compli-
indicate eastward transport, and grain size and cated fault and fold systems. Generally the
thickness decrease eastward. In the western small-scale folds are slump structures. Strata
part of the Miura Peninsula, a conglomerate from the Hota to Miura Groups are usually
chiefly composed of angular blocks of basalt, displaced laterally or vertically on the de-
scoria and pyroxene gabbro was deposited from cimetre to metre scale. Small-scale faults usual-
various directions including NW and SW (Fig. ly cut these slump structures, but are commonly
5). syn-sedimentary.
These data indicate that much igneous mat- On the western coast of the Miura Peninsula,
erial was supplied from the Izu volcanic arc in minor faults in the Miura Group show the
the west; sediment transport in general was following sequence of development (Ogawa &
from the south or SW in the southern out- Horiuchi 1978; Ogawa 1980). The first-formed
crop area and from the NW or SW in the faults are normal and syn-sedimentary, being
central and northern outcrop area (Fig. 5). At associated with slump beds and debris flows
this time, in the late Miocene, basaltic and which accumulated along the fault scarps. De-
dacitic or rhyolitic volcanism occurred over a watering structures, called beard-like structure
wide area of the Izu Peninsula, with associated (Ogawa 1980) also formed at this stage. Next,
large olistostromes of scoriaceous and thrust faults were formed by lateral compress-
pumiceous rocks. ion, coincident with the present regional strike
In summary, the Neogene, Hota, Yabe and (Ogawa & Horiuchi 1978). These faults may
Miura Groups in the Miura Fold Belt are represent the beginning of large-scale folding in
composed chiefly of volcanic materials depo- a strike-slip regime. The first two types of faults
sited in the forearc basin of the Izu volcanic arc. usually show drag of strata along fault planes
They also include fragments of the Mineoka and show welding of fault planes. However, the
58 Y. Ogawa

last set of faults, vertical or oblique-slip faults, Discussion and conclusions

do not show such drag. This type of fault was
developed at the same time as the large-scale As I have shown, the area of the Miura Fold
folds. Belt has a different sedimentary and tectonic
These three fault systems in the Miura Group history to the Shimanto and Izu Terrane. In the
indicate that faults were g e n e r a t e d during the Shimanto Terrane of the arc-arc crossing area,
formation of the sedimentary basin and con- collisional tectonics occurred coincident with
tinued during lateral compression at the same sedimentation of the Miura Terrane. Sediments
time as large-scale folds were developed. Sedi- in the latter area were deposited during strike-
mentation began in the south at first and gra- slip movements, and olistostromes, slumps and
dually shifted northward (Mitsunashi 1973). slide deposits are common. Moreover the fold
The upper part of the group in each area does axes are not parallel but en echelon or curving.
not show such strong deformation as the lower Such evidence, as well as the changeable stress
part. This indicates that deformation was occur- fields deduced from the fault patterns, indicates
ring during sedimentation. that the area was situated in the strike-slip

CRETACEOUS MIDDLE MIOCENE

-100Ma EURASIAN PLATE 15Ma
-..... I ~, o~: ~N-,-. -
/
--. J
// ~-ARC (SHIMANTO)
,~x,' p¢.?/~/. GROUP
¢
I v V ~---~'~TRIPLE
./7 KULA o, I v " " 11 JUNCTION
"~PALEO- ~ PACIFICPLATE ~'1 v YABE FORMATION
IZU RIDGE
t
PALEOGENE f/ LATE MIOCENE~
40 Ma- ¢ EARLY PLIOCENE
14Ma-
,t ~ / PACIFIC
-',J', =L~" # PLATE
-t~ v ~
/,, ,,
/ - ,sJI///~ ~I~INEOKA
Y ~" I GROUP MIURA GROUP
PHILIPPINE SEA /
PLATE TRAPPED

EARLY MIOCENE LATE PLIOCENE ~ RECENT

19Ma ~ . , i( 3Ma~ ~'
v MIURA-50.50
V v.:.~ --,TRAPPED
_..~. v~V' '~I HOTA GROUP v ".L
II /IZU-BONIN V~,,~ PACIFIC
X/ ~ v ~ PLATE
4,-11,,,~ ARC
v v t
II v ~PHILIPPINE ~
SEA PLATE ~ " ~"

[----.j,. IC-C3 I -qf

FIG. 6. Plate tectonic model of the arc-arc crossing area in central Japan from Cretaceous to Present.
a, subduction zone; b, transform fault; c, direction of plate spreading; d, ophiolite; e, arc volcanism;
f, volcaniclastics. Folded small arrows indicate sediment transport. Present radiolarian data indicate
some Mineoka sediments are of lower Miocene and some Hota sediments are of middle Miocene
age (see text).
 Tectonics of some forearc belts in central Japan 59

regime between two different arcs. Miura G r o u p were also trapped in this trans-
Interestingly, the current stress field in the form zone. These sediments became welded to
Sagami Trough, which is the boundary between the Shimanto Terrane to form the Miura Fold
the Eurasian and Philippine Sea plates, is Belt in the eastern side of the arc-arc crossing
generally of lateral-slip type, with a changeable area (Fig. 6).
component for thrust or normal faults between This trapping and welding may have been
the co-seismic and inter-seismic times (Ando associated with the southward shifting of the
1974; Matsuda et al. 1978; SenD 1980). Judging Japan Triple Junction during Miocene and
from the sedimentological analysis of the strata Pliocene times. The transform fault probably
in the Miura-Boso Peninsulas, the Sagami switched to its present position in the late
Trough was not situated in its present position, Pliocene, after the last sediments of the Miura
but adjacent to the central part of the peninsu- Group were deposited. Some tectonic erosion
las from N W - S E during the Miocene and may have occurred during the triple junction
Pliocene. In the trough at that time numerous migration (Matsubara 1980).
olistostrome, turbidity or debris flow units indi-
cate near-continuous crustal movement along ACKNOWLEDGMENTS: I am grateful to Dr J. K. Leggett
the transform fault. for inviting me to the Trench-Forearc Conference. Dr
The transform seems to have been formed J. K. Leggett and Professor A. J. Smith are gratefully
when the system of plate spreading changed acknowledged for improvement of the manuscript.
Professor K. Kanmera, Messrs S. Tonouchi, K.
around 40 Ma ago. At the same time the
Fujioka and T. Sakai are thanked for their discussion.
crossing and colliding of the Izu-Bonin arc was Professor K. Nakaseko determined the radiolarian
taking place (Fig. 6). Some part of the oceanic names. Messrs K. Horiuchi, H. Taniguchi and K.
crust (the ophiolitic Mineoka Group), and sub- Koga helped me in the field survey. This work was
sequently sediments of the middle Miocene supported by grants from the Ministry of Education
Hota Group, Yabe Formation and a part of the of Japan and the Takeda Scientific Foundation.

References
(J), in Japanese. ICHIKAWA,K. 1980. Geohistory of the Median Tecto-
ANDO, M. 1974. Seismo-tectonics of the 1923 Kanto nic Line of Southwest Japan. Mem. geol. Soc.
Earthquake. J. Phys. Earth, 22, 263-77. Japan, 18, 182-212.
D~CK1NSON,W. R. & Seely, D. R. 1979. Structure and IKEBE, N. 1978. BiD- and chronostratigraphy of
stratigraphy of fore-arc regions. Bull. Am. Japanese Neogene, with remarks on paleD-
Assoc. Petrol. Geol. 63, 2-31. geography. In: HUZITA, K. (ed). Cenozoic
FUJIMOTO, H. & SUZUKI, M. 1969. Geology of the Geology of Japan, pp. 13-34 (J).
basin of the River Obora-gawa, a tributary of the ISmZAKA, K. 1972. Rb-Sr dating on the igneous and
River Arakawa. Bull. Chichibu Mus. nat. Hist. metamorphic rocks of the Kurosegawa Tectonic
15, 1-18 (J). Zone. J. geol. Soc. Japan, 78, 569-75 (J).
FUKAO, Y. 1979. Tsunami earthquakes and subduc- KANEHIRA, K. 1976. Mode of occurrence of serpenti-
tion processes near deep-sea trenches. J. geo- nite and basalt in the Mineoka district, southern
phys. Res. 84, 2303-13. Boso Peninsula. Mere. geol. Soc. Japan, 13,
Geological Survey of Japan (ed). 1976. Geological 43-50 (J).
Map of Tokyo Bay and adjacent areas -- , OKI, Y., SANADA,S., YABE, M. & ISHIKAWA,
(1:100,000). Geol. Survey of Japan. F. 1968. Tectonic blocks of metamorphic rocks at
HANDIN, J. • HAGER, R. V., JR 1957. Experimental Kamogawa, southern Boso Peninsula. J. geol.
deformation of sedimentary rocks under confin- Soc. Japan, 74, 529-34 (J).
ing pressure: tests at room temperature. Bull. KANMERA, K. 1976. Correlation between the geosyn-
Am. Assoc. Petrol. Geol. 41, 1-50. clinal sediments of past and present. I & II.
-- & ~ 1958. Experimental deformation of Kagaku (Sci), 46, 284--91,371-6 (J).
sedimentary rocks under confining pressure: --1977. General aspects and recognition of
tests at high temperature. Bull. Am. Assoc. olistostromes in the geosynclinal sequence.
Petrol. Geol. 42, 2892-934. Monogr. Assoc. geol. Collab. Japan, 20, 145-59
HARA, I., SnoJI, K., SAKURAI,Y., YOKOYAMA,S. & (J).
HIDE, K. 1980. Origin of the Median Tectonic -- , HASHIMOTO,M. & MATSUDA, T. (eds). 1980.
Line and its initial shape. Mem. geol. Soc. Japan, Geology of Japan. Iwanami-shoten, Tokyo. 387
18, 27-49. pp. (J).
HOSHINO, K., KOIDE, H., [NAMI, K., IWAMURA,S. & KARIG, D. E. 1974. Evolution of arc systems in the
MITSUI, S. 1972. Mechanical properties of western Pacific. Ann. Rev. Earth planet. Sci. 2,
Japanese sedimentary rocks under high confining 51-75.
pressures. Rep. geol. Soc. Japan, 244, 200 pp. KIMURA, M., YUASA, M., MASAI, Y. & KANIE, Y.
60 Y. Ogawa
1976. Finding of pillow lava from the Oligo- MOORE, G. F. & KARIG, D. E. 1976. Development of
Miocene formation of the Miura Peninsula. Bull. sedimentary basins on the lower trench slope.
geol. Soc. Japan, 27, 451-7 (J). Geology, 4, 693-7.
KIMURA, T. 1959. A sharp bend of the Median Tecto- MOORE, J. C. 1979. Variation in strain and strain rate
nic Line and its relation to the Akaishi Tectonic during underthrusting of t~ench deposits. Geo-
Line--tectonic significances yielded by lateral logy, 7, 185-8.
faulting. Jpn. J. Geol. Geogr. 30, 215-32. & KARl6, D. E. 1976. Sedimentology, structural
1961. The Akaishi Tectonic Line, in the eastern geology, and tectonics of the Shikoku subduction
part of Southwest Japan. Jpn. J. Geol. Geogr. zone, Southwest Japan. Bull. geol. Soc. Am. 87,
32, 119-36. 1259-68.
- - 1968. Some folded structures and their distribu- NAKADA, S. & TAKAHASHI,M. 1979. Regional varia-
tion in Japan. Jpn. J. Geol. Geogr. 39, 1-26. tion in chemistry of the Miocene intermediate to
1974. The ancient continental margin of Japan. felsic magmas in the Outer Zone and the
In: BURK, C. A. & DRAKE, D. L. (eds). The Setouchi province Of Southwest Japan. J. geol.
Geology of Continental Margin, 817-29. Soc. Japan, 85, 571-82 (J).
Springer-Verlag, New York. NISHIMIYA, K. & YAMAGIWA, N. 1973. Coral fossils
-- 1979. The Japanese Islands: the Geologic His- from the Kosuge Formation, Yamanashi Prefec-
tory (II-A), 245-578. Kokinshoin, Tokyo ture. Trans. Proc. Palaeontol. Soc. Japan, 89,
(J). 15-23.
KOBAYASHI, K. & NAKADA, M. 1978. Magnetic ano- OGAWA, Y. 1978. Structural characteristics and tec-
malies and tectonic evolution of the Shikoku tonisms around the microcontinent in the outer
inter-arc basin. J. Phys. Earth Suppl., 26, $391- margin of the Paleozoic-Mesozoic geosyncline of
402. Japan. Tectonophysics, 47, 295-310.
KOBAYASHI,T. 1941. The Sakawa orogenic cycles and 1980. Beard-like veinlet structure as fracture
its bearing on the origin of the Japanese Islands. cleavage in the Neogene siltstone in the Miura
J. Fac. Sci. Univ. Tokyo. H, 5, 219-578. and Boso Peninsulas, central Japan. Sci. Rep.
KOSAKA, K. 1979. Cretaceous and post-Cretaceous Dep. Geol. Kyushu Univ. 13, 321-7 (J).
faulting in the northeastern Kanto Mountains. & HORIUCHI, K. 1978. Two types of accretion-
J. geol. Soc. Japan, 85, 157-76 (J). ary fold belts in central Japan. J. Phys. Earth
KOZlMA, N. 1980. On the layers of disordered sedi- Suppl., 26, $321-36.
mentation in the Misaki Formation in the south- ONO, A. 1980. A model of the formation of the
western part of the Miura Peninsula, south Ryoke-Sambagawa paired metamorphic belt. J.
Kanto, Japan (Part I). J. geol. Soc. Japan, 86, Assoc. Jpn. Min. Petrol. econ. Geol. 75, 31-7
313-26 (J). (J).
MARUYAMA,S., UEDA, Y. & BANNO, S. 1978. 208-240 OTSUKi, K. & EHIRO, M. 1978. Major strike-slip faults
M.Y. old jadeite-glaucophane schists in the and their bearing on spreading in the Japan Sea.
Kurosegawa Tectonic Zone near Kochi City, J. Phys. Earth Suppl., 26, $537-55.
Shikoku. J. Jpn. Assoc. Mineral. Petrol. Econ. RAMSAY,J. G. 1963. Folding and Fracturing of Rocks.
Geol. 73, 300-10(J). McGraw-Hill, New York. 568 pp.
MATSUBARA, Y. 1980. Izu Peninsula and Philippine SAKAt, T. 1978. Geologic structure and stratigraphy
Sea plate. Chikyu (Earth Monthly), 2, 157-63 of the Shimantogawa Group in the middle
(J). reaches of the Gokase River, Miyazaki Prefec-
MATSUDA, T. 1978. Collision of the Izu-Bonin arc ture. Sci. Rep. Dep. Geol. Kyushu Univ. 13,
with the central Honshu--Cenozoic tectonics of 23-38 (J).
the Fossa Magna, Japan. J. Phys. Earth Suppl., & KANMERA, K. 1981. Stratigraphy of the Shi-
26, $409-21. manto Terrane and the tectono-stratigraphic set-
, OTA, Y., ANDO, M. & YONEKURA, N. 1978. ting of greenstones in the northern part of
Fault mechanism and recurrence time of major Miyazaki Prefecture, Kyushu. Sci. Rep. Dep.
earthquakes in southern Kanto district, Japan, as Geol. Kyushu Univ. 14, 31-48 (J).
deduced from coastal terrace data. Bull. geol. SANO, H., KANMERA,K. & SAKAI,T. 1979. Sediments
Soc. Am. 89, 1610-8. associated with greenstone of the Shimanto Ter-
MAXSUSmMA,N. 1978. Geology of the Southern Alps rane. J. geol. Soc. Japan, 85, 435-44 (J).
of Japan--geologic structure of the Shimanto SENO, T. 1977. The instantaneous rotation vector of
belt in the Akaishi Mountains. J. nat. Sci. Shi- the Philippine Sea plate relation to the Eurasian
moina Educ. Soc. (Geol.) 1, 119-34 (J). plate. Tectonophysics, 42, 209-26.
MITSUNASrn, K. 1973. Geologic development of south -- 1980. Changeable regional stress field. Chikyu
Kanto and Niigata sedimentary basins from (Earth Monthly), 2, 146-54 (J).
Miocene to Pleistocene. J. Assoc. geol. Collab. SuzuKI, T. & HADA, S. 1979. Cretaceous tectonic
Japan, 27, 48-65 (J). melange of the Shimanto belt in Shikoku, Japan.
-- & YAZAKI, Y. 1958. The correlation between J. geol. Soc. Japan. 85, 467-79.
the Cenozoic strata in Boso and Miura Peninsu- TAmA, A., OKADA, H., WmTAKER, J. H. McD. &
las by the pyroclastic key bed (No. 1). J. Jpn. SMma, A. J. 1981. The Shimanto Belt of Japan:
Assoc. Petrol. Technol. 23, 16-20 (J). Cretaceous-Lower Miocene sedimentation in
MIYASHIRO,A. 1961. Evolution of metamorphic belts. forearc basin to deep-sea trench environments
J. Petrol. 2, 277-311. (this volume).
 Tectonics o f s o m e f o r e a r c belts in central J a p a n 61

TAKIZAWA, S. 1979. Stratigraphy of the Chichibu Peninsula. J. geol. Soc. Japan, 84, 561-70.
Terrane in North Kanto Mountains, In: UEMURA, T. 1981. Deformation facies, series and
INAMORI, J. (ed.). Biostratigraphy of Permian grades. J. geol. Soc. Japan, 87, 297-305.
and Triassic Conodonts and Holotinrian Sclerites UYEDA, S. • MIYASHIRO, A. 1974. Plate tectonic
in Japan, 89-102 (J). model and Japanese Islands: a synthesis. Bull.
TSUCHIYA, N., SAKAI,T. & KANMERA,K. 1979. Mode geol. Soc. Am. 85, 1159-70.
of occurrence and petrological characteristics of YAMATO-OHMINE RESEARCH GROUP 1979. Triassic
greenstones of the Shimanto Terrane in the Mimi limestone blocks within the Shimanto Group in
River area, Kyushu. J. geol. Soc. Japan, 85, Shikoku. J. geol. Assoc. Collab. Japan, 33, 59-61 (J).
445-54 (J). YOSHIDA, A. 1974. Findings of foraminifera from
UCHIDA, T. • ARA1, S. 1978. Petrology of ultramafic Mineoka Mountains, Chiba Prefecture. Chishit-
rocks from the Boso Peninsula and the Miura su News, Geol. Surv. Japan, 233, 30-6 (J).

YUJIRO OGAWA, Department of Geology, Faculty of Science, Kyushu University,

Hakozaki, Fukuoka 812, Japan.
 Forearc geological structure of the Japanese Islands

Tsunemasa Shiki & Yoshibumi Misawa

SUMMARY: Under the forearc region off the Japanese Islands a conspicuous basal
acoustic reflection plane can be traced at the surface of the acoustic basement from the
oceanic plate into the bottom of the accretionary prism. Recent studies indicate this plane
canno~ be the upper boundary plane of the subducting oceanic plate, but must be the
upper layer of the oceanic basaltic basement (or chert layer).
The tectonic activity of the Wadati-Benioff zone appears to be a thrust movement which
cuts the boundary between the oceanic basement and the overlying sediments. Thrusts cut
the basal reflector and extend into the accretionary prism. Large earthquakes which occur
at intermediate and shallow depths under the continental shelf and slope are a direct
manifestation of these thrust movements.
Thrust faulting also occurred in past geological time. Acidic and basic magma, possibly
generated along the earthquake zone at intermediate depths, ascended along the thrust
faults or associated normal faults. Intrusions occurred 15-16 Ma ago in the forearc region of
Southwest Japan and 22 Ma ago in that of Northeast Japan; subsequently basins subsided
landward of the intrusions. In some cases the forearc igneous rocks carried continental and
oceanic crustal xenoliths. Good examples of such forearc igneous rocks, including
xenoliths, occur in the Permian and Triassic structural belts of the Japanese Islands.
The model described here is specific to the forearc regions of the Japanese Islands. We
envisage that different tectonic parameters such as convergence rate and sediment supply
may engender different structural processes in other forearcs.

develop and expand our previous ideas and to

Introduction
present a modified model of the forearc struc-
The Japanese (or Honshu) Island Arc may be ture of the Japanese Islands.
divided into two parts: the Northeast (or Toho-
ku) Arc and the Southwest Arc. The former is a
typical island arc displaying the characteristic The accretionary prism and basal
arrangement of a deep trench (in this case the acoustic reflection plane
Japan Trench), a gravity minimum arcward of
the trench, earthquake hypocentres along a Although a large number of studies have been
dipping Wadati-Benioff zone, with an inner made in the forearc region of the H o n s h u Arc,
volcanic arc and an outer sedimentary arc it is only in recent years that the precise struc-
above the Wadati-Benioff zone. The Southwest ture of the region has been discussed. The
Japan Arc, however, is not a typical arc: it lacks discovery of the unexpectedly small size of the
recent volcanism and a deep trench (though the accretionary prism off Tohoku is one of the
Nankai Trough may be regarded as a trench) most important results of recent multi-channel
(Hoshino 1963) (Fig. 1). seismic reflection surveys and D S D P - I P O D
The Japan Trench-Island Arc system is the legs 56 and 57 (Nasu et al. 1978, 1979; yon
best studied Pacific convergent margin. Careful H u e n e & A r t h u r 1981). A n o t h e r discovery is
re-examination of existing data and new and that no true pelagic ocean floor sediments were
more detailed investigations of the forearc re- drilled in the continental slope of the Japan
gion of this system could lead to a better Trench forearc region ( O k a d a & Sakai 1979;
understanding of forearc sedimentation and Langseth, O k a d a et al. 1978). In the case of the
tectonics in general. Shiki (1978) and Shiki & Southwest Japan forearc region, Leg 31 drilling
Misawa (1979, 1980) have attempted such an and air-gun surveys indicate that most of the
investigation, and presented a new model of the accretionary sediments under the lower con-
forearc structure of the Japanese Islands, espe- tinental slope (inner trench slope) comprise
cially that under the inner slopes of the Nankai terrigenous and hemi-terrigenous materials
Trough and the Japan Trench. Several major (Karig, Ingle et al. 1975). In such examples,
publications (e.g. Murauchi 1979; Nagumo where trench turbidites are apparently the only
1980) have appeared since the Shiki and Mis- materials accreted, the prism may be called a
awa papers. In this contribution, we wish to 're-accretionary prism' (Shiki & Misawa 1980).

63
64 T. S h i k i & Y. M i s a w a

;£W <-, /, J+'%+.

FIG. 1. Map showing the area discussed and Permian-Triassic structural belts in Southwest Japan.
MSB: Maizuru structural belt.
SSL: Su-etakegawa structural line.
KSB: Kurosegawa structural belt.

Another problem concerns the structure of acoustic basement, which can be regarded as
the forearc region. A conspicuous acoustic re- basaltic basement there, is not flat under the
flection plane can be observed in multi-channel Shikoku Basin and the Nankai Trough (South-
profiles, at the bottom of the accretionary prism west Japan Trench) and uneven basement fea-
under the lower continental slope of the Hon- tures can be traced under the inner slope of the
shu Arc. In the profile across the Japan Trench Nankai Trough (Shiki & Misawa 1979). Other
shown by Beck et al. (1975) this major basal air-gun records by the Geological Survey of
reflector is apparently flat. It is generally re- Japan (Inoue 1978; Okuda et al. 1979). show
garded as the slip plane between the oceanic similar features more clearly (Fig. 2C). Such an
plate and the continental plate. However, the uneven surface cannot be regarded as a plane of
reflection plane is traceable from the surface of thrust slip occurring simply at the top of the
the acoustic basement under the ocean floor ocean crust and below pelagic sediment of
arcwards under the inner trench slope. It there- Layer 1.
fore seems most likely that the relationship More recent investigations (Langseth & Oka-
between the lower part of the accretionary da et al. 1978; Nasu et al. 1979) reveal that the
prism above the reflection plane and the layer basal reflection plane in the Japan Trench
beneath the plane must be similar to that region is cut by thrust faults (Figs 3 & 4). This
between the ocean bottom sediments above the further supports interpretation of the basal
reflection plane and the layer beneath the reflection plane as a non-conformity or uncon-
plane. Hence, the surface must be a non- formity above the chert layer rather than a
conformity or unconformity within a strati- major d6collement surface.
graphic section. The thrusts which cut the basal reflection
This non-conformable or unconformable extend into the accretionary prism, some of
stratigraphic relationship is more apparent in them approaching the seafloor, forming an
Southwest Japan. As is shown in Fig. 2(A), the imbricate structure in the prism• Some of these
 Forearc geological structure of the Japanese Islands 65
0 sec--
A

• o .....

O ec--
Tosa terrace ' ~.~.. -"-;~

4 298 DSDP 297

0 sec-- \ I. Nankai
_ T'\ trough

-,~ Hyugo terrace ;, f ~ -~ _ j . _ / - : - . . ~

2" ~"~X"~ ^ Nankoi

- . %---..x~k trough

0 50 km 6 - - - - - ]- :~'f """~
6 ' L ~

8-

Fl6.2. Single-channel seismic reflection profile across Nankai Trough.

(A) Across central part of Nankai Trough (Misawa 1976).
(B) Off Kochi through Tosa Terrace and DSDP sites 297 and 298.
(C) Across western corner of Nankai Trough and Hyuga Terrace (Inoue 1978; Okuda et al. 1979).

LUMP

• " ~ " " 1 - . "~ ."T-" , ~ S L U M P

z~_f/ /..._. 434

•, o 1 1 7. . . . . . . y . . Z - ~ . ~ ' ~ 10
ACOUSTIC ~~rrh')U rm''"'
BASEMENT
0 I0 20 km
I- I q

FIG. 3. Seismic reflection profile across the Japan trench and inner trench slope, off Miyako,
Northeast Japan.
Note that the reflection plane under the accretionary prism is cut by thrust faults (after Leg 56
report; Langseth, Okada et al. 1978).

thrusts may unite at depth, as was suggested by surface of the oceanic lithosphere and appear at
Van Huene (1978) for the Aleutian forearc the trench bottom. However, recent detailed
region. studies have yielded results somewhat different
from this general conception.
For instance, Suzuki & Okada (1977) relate
Thrust faults and earthquakes seismicity around Northeast Japan, to a fault
plane which does not continue to the trench axis
It has long been postulated that the Wadati- but extends into the continental crust, appearing
Benioff zone must extend along the upper at the sea-floor in mid-slope. A similar distribu-
 MIGRATED DEPTH SECTION
Line 78-4 Shot Number O - 1 4 6 0

coo ~ . . . . .
• .7~ 1 _ ~ . . . . . . . . . . . . .
1o;o . I',o0..
~ ;= ~i¸ v

TRENCH SLOPE OUTER TRENCH SLOPE

7 T - %~:: 5 ....... T
..... :
= ~JAPAN TRENCH

F .- .... .

211
m

.10

.15kin

FIG. 4. Multi-channel seismic reflection record across Japan Trench (line 78-4) BM: acoustic basement (from Nasu et al. 1979)
 Forearc geological structure o f the Japanese Islands 67

tion of earthquake foci is seen in the figures of several speculative explanations were suggested
Hasegawa et al. (1978). Earthquake foci also for the occurrence of these rocks. However, in
occur, of course, along the upper surface of the Southwest Japan, various igneous rocks such as
lithosphere just under the trench. However, basalt, diabase, gabbro, granophyre, and
these earthquakes are different in character to quartz-porphyry, occur in a forearc setting on
the intermediate depth (30-60 kin) earthquakes the east coast of Honshu Island at Cape Shiono-
under the continental slope, as shown by Yoshii Misaki and Cape Muroto-Misaki (Yoshizawa
(1979). 1954a,b; Imoto et al. 1970; Yajima 1972a,b;
Tectonic deformation associated with large Yajima & Fukuda 1976; Miyake 1981) and on
earthquakes along trenches has been described Tosabae Bank, off Shikiku Island (Niino 1935).
and interpreted in terms of low angle thrust Hisatomi & Miyake (1981) clarified that the
faulting between the continental block and igneous activity at Shiono-Misaki was intimate-
ocean floor. Sawamura (1953) noticed such ly associated with uplift and formation of a
deformation along the Nankai Trough about tectonic high which dammed sediments in the
30 years ago. Plafker (1972) studied the thrusting forearc basin.
of listric surfaces related to the 1964 Alaska The ages of the forearc ijzneous rocks from
earthquake. The main shock of the Tokachi- Northeast and Southwest Japan are different.
oki earthquake of 16 May 1968 (M = 7.9; depth The 4°Ar-39Ar age of the Northeast Japan rocks
'0 km') seems to have been caused by thrust is around 22 Ma (Ojima 1979), while the rocks
motion along the upper plane of the dual deep of Southwest Japan are inferred to be 15-16 Ma
seismic zones under the Japan Trench forearc old based on biostratigraphy (Hisatomi &
region (.Ichikawa & Kawamura 1971; Watanabe Miyake 1981). The geological setting of both
1971). A thrust fault was also assumed from sets of rocks is similar; in both cases the intru-
analyses of the data of the Miyagi-oki earth- sion or eruption of magma seems to be con-
quake (12 June 1978; M = 7.4; depth 40 km), nected with the formation of a tectonic high and
and was regarded as a branch of the Wadati- a forearc basin landward of the high.
Benioff zone. Its extension seems to appear on Studies on the rocks of Shiono-Misaki and
the slope on the outer side of the structural high Muroto-Misaki (Yoshizawa 1954a,b; Yajima
(Nagumo 1980). 1972a,b; Yijima & Fukuda 1976; Miyake 1981;
It seems possible, therefore, that some of the Hisatomi & Miyake, in press) show a compli-
large earthquakes are caused by motion on cated occurrence of various rocks (acidic, and
thrust faults which reach into the accretionary basic to ultrabasic), and poor K content. The
prism and appear at the mid-slope seafloor, linear outcrop distribution suggests the pre-
linking at depth with the Wadati-Benioff zone. sence of a tectonic line; this has been called the
'Shiono-Misaki-Osumi line' (Nozawa 1975) and
the 'Shiono-Misaki zone' (Hisatomi & Miyake
1981). It seems possible that the igneous mater-
ial gained egress along thrusts or normal faults.
Forearc igneous rocks Shiki & Misawa (1979) called these igneous
Present occurrence rocks 'continental slope intrusive rocks' and
Shiki & Misawa (1980) called them 'forearc
One of the most important results of Leg 57 igneous rocks'.
of the DSDP was the discovery of conglomer- Shiki & Misawa (1979, 1980), have suggested
ates and upper Oligocene course sandstones that the occurrence of such igneous rocks on the
resting unconformably on Upper Cretaceous continental slope may not just be a feature of
siltstones (Nasu et al. 1978; von Huene & the Japanese Islands. For example, gabbroic to
Arthur 1981). It was assumed that the con- granitic rocks from the structural high in the
glomerate pebbles (of acidic to intermediate Kuril-Kamchatka forearc region, which in-
volcanic rocks) originated from a palaeo-island truded pre-Neogene deformed volcanogenic
(Oyashio Ancient Landmass) which was situ- sediments (Gnibidenko et al. 1978) may be of
ated off Northeast Japan. Geochemical studies similar origin. Boninite and other igneous rocks
by Fujioka & Nasu (1978) indicate that the found in the Mariana Islands forearc on DSDP
rocks might belong to a calcalkalic and thole- Leg 60 (Hussong & Uyeda et al. 1978) could be
itiitic rock series of island arc-type. Since the forearc igneous rocks, as might the boninites of
location is so far east (i.e. oceanwards) of the the Bonin Islands (Ogasawara Islands) them-
present volcanic front of the Northeast Japan selves.
arc the finding surprised some geologists, and
68 T. S h i k i & Y. M i s a w a

vinces. Rocks in the Permian and Triassic

Problems of generation Kurosegawa and Maizuru structural belts (Fig.
Although the rocks appear on the seafloor of 1) may be examples. The rocks of these belts
the present forearc region, their intrusion or are more complex than those of the younger
eruption occurred some time ago in all cases. forearc igneous rocks mentioned above. They
The geological setting of the forearc igneous include igneous rocks (such as trondhjemite
rocks of the Nankai Trough and Japan Trench adamellite, quartz diorite, granophyre, gabbro,
suggest that the generation of the rocks possibly diabase), metamorphic rocks (such as amphibo-
occurred in response to Tertiary tectonic move- lite, amphibolite schist, and serpentinite) and
ments such as the rise of the structural high and Silurian sedimentary rocks (Hada et al. 1979;
the depression of a forearc basin. The presence Igi 1973; Igi et al. 1979; Yamashita 1979). The
of the trench at the time of the generation of the acidic igneous rocks of the belts have semi-
rocks is not absolutely proven. oceanic chemical characters, i.e. high Na/K
Clearly the genesis of such magma presents a ratios, high K/Rb ratios and low initial 87Sr/
serious problem. The heat flow flux is generally 86Sr ratios (Ishizaka & Yanagi 1975). A proto-
small at recent trenches and in adjacent zones; clastic texture is another feature of many of
neither high temperatures nor magmatism can these acidic rocks: such a texture can be caused
be expected. At 30-50 km beneath the sedi- by shear stress suffered at the time of intrusion.
ments and continental crust, however, the Although many of these rocks were dated as
temperature must be somewhat higher and Permian to Triassic, others are more than
magmatism may exceptionally occur. The study 400 Ma old (Ishizaka 1972; Nishimura et al.
of Hayakawa & Iizuka (1979) on the relation 1976; Nishimura 1979; Nohda 1973; Shibata et
between earthquakes and the process of partial al. 1977). Anatexis or remobilization of pre-
melting is interesting; they note that a compara- Silurian or Precambrian rocks has been claimed
tively high heat flow zone is observed off South- in these cases (Tomita 1954; Noda 1961; Kano
western Japan, probably occurring in response et al 1961; Matsumoto et al. 1962). The most
to partial melting at the hypocentral depths of significant feature of the geological setting of
great earthquakes. Certainly the thermal gra- the Maizuru belt is its occurrence along the
dient in these high heat flow zones shows that boundary between an outer uplifting province
the temperature at hypocentral depths (30- and subsiding 'geosynclinal' province of late
40 km) can approach the melting point of wet Permian and Triassic age. Maruyama (1978)
peridotite, supporting the possibility of magma has interpreted the Kurosegawa belt as having
genesis in forearc regions under certain special between located in a palaeo-forearc, and the
geothermal and geological conditions. Such Maizuru belt appears to have a similar origin.
magma genesis may have occurred in the past, We do not maintain that the geological set-
though no instances of contemporary forearc ting of the rocks of the Permian and Triassic
volcanism have been reported. Various kinds of structural belts is identical with that of the rocks
lower crustal rocks including both oceanic and of younger forearc igneous rocks. However, the
continental ones may melt, or be caught as resemblance in character (including complexity
xenoliths and rise to shallower crustal levels. It and occurrence along faults or tectonic lines)
is natural that the presence of some zone of suggests some similarity in setting and geologi-
weakness makes the ascent of such magma and cal significance.
xenoliths easier. The imbricate thrusts men-
tioned above or other fractures associated with
the thrusts could provide such planes of weak- Discussion and conclusions
ness. We believe that such a model explains the
occurrence of Tertiary igneous rocks in the The conspicuous basal acoustic reflection plane
Japanese forearc region. which can be traced under the accretion~,ry
prism from the oceanic plate must be the upper
surface of oceanic Layer 2 (basaltic basement,
or the chert layer). It is not a tectonic slip plane.
The Permian and Triassic structural D6collement resulting from subduction occurs
b e l t o f Japan and associated forearc along thrust faults which cut the basal reflector
(Figs 5 & 6). Large earthquakes at intermediate
igneous rocks and shallower depths under the continental
Shiki & Misawa (1979) pointed out that some shelves and slopes are the direct result of these
examples of forearc igneous rocks may be fault movements (Figs 6 & 7). This model does
found in much older Japanese geological pro- not deny relative movement between the con-
 Forearc geological structure of the Japanese Islands 69

NW TOSABAE SE
TOKI BASIN NANKAI TROUGH
-O

7< H /4 /ly / /'/~,~,¢,7 "B~

, ,
~ ,
u .
I / ~' ¢ , / ~ ~ , , / ' -~..-
/-/- ;~i --- -7 ~ . o.

.I,/ , . k~;; /_ ~ '~, ~-_./

:./'/:7 /-
fk,- I /i'~. "/.'/14~.~C././d,f .i

+,.<$7;
•

-"
ifP/'"Y'l / I _ _ J 2 km
z z / / I0 km

FIG. 5. Schematic structural profile across Nankai Trough off Central Shikoku, showing basal
reflection plane (BM), thrust faults, forearc igneous rocks, and illustrative seismic velocities.
Modified from Shiki & Misawa (1979), and based on references quoted therein.

W E

STRUCTURAL INNER TRENCH SLOPE

- HIGH OUTER TRENCH S L O P E - O krn
STRUCTURAL
HIGH
---:"~-~ _1 , 7 " ~ . . . .
_ ~ TRENCH
:-:L" N~".~_-_~-_2.3- ~ N~• ..-za.or.~--'-::-:~
/:~:': ~ . " - 7 , ~ - " ~ ; 7:-~'~. ".7.;'~. _ ~

~, i /.,-, ~ c r . ~ "I":-4,- , . . .., .

- ' ~ ~ ~ -20

8-Z
I _ _ I
I0 k m

-30

Fro. 6. Schematic structural profile across Japan Trench and inner trench slope off Hachinoe,
Northeast Japan, showing basal reflection plane (= acoustic basement, BM), thrust faults,
accretionary prism, and illustrative seismic velocities and earthquake loci. Revised from Shiki &
Misawa (1979, 1980), and based on references quoted therein and Honza et al. (1978).
 --4

Japan Trench

NEOGENE
W AF PALEOGENE DSDP
U. CRET. 438 439 435 440 441 4 3 4 4:36
0 km E
I 1 I I I I
_ _ 1.7" 1.6 Okm
MESOZOIC , ------- 2.4-2.8- 2-0
-- 4 " 3 ~ -- ---- --
6"2 •, - PALEOZOIC .
5q -- 5q - / 1.8 1.8 2-3
~ "---'---. ~ ....... ' , , 5-8 BM
I0-- + .5-9 1/ g
+ ~ 4.7 -
--I0
+ 6.6 j
ISLAND ARC B A S E M E N ~ + • " 6.7 7,2 ".,.
/ +
6,6 + . ;,_V "--,,ry/~-~..- K ~ S / d ~ . - - , ->~'-- "~,- ' 7.9
~7,'2-"-
20-- + + 7.6
+ x, ,A ''~ ~" ~ ~ - ~ y S - ~ ' ~ '- ' -- + + -- 20
,/~.~,,,~.(,~,j~_~.>.-~ -,-, :__.__-- +
i- ?j +
, -~ i" 8.0 " 4 "''~ . ~-6"81 " - "
._ ~ ~,7 / ~ . . ~ - , _----- + +
+ +
MOHO k~
+
30~ " ..,
--30
8.0
~..~.,~" \ + + + o 5 0 km
* ~ + L , , I I I
Y.~,I ' ~ Lj ~ / *-~- * + +
40-- /

fy~_

+
f

FIG. 7. Schematic structural profile through DSDP sites across Japan Trench and forearc region off Miyako, northeast Japan, showing acoustic
basement (BM), thrust faults, accretionary prism, forearc igneous rocks, and idealized distribution of earthquake foci. Based on numerous
sources quoted in Shiki & Misawa (1979, 1980). Seismic velocities are based mainly on collective study of seismic refraction information by
Murauchi (1979) and on multi-channel reflection data in Nasu et al. (1979). Values of the velocity may not be accurate at deep horizons in each
vertical sequence.
 Forearc geological structure of the Japanese Islands 71

tinental lithosphere and the oceanic litho- The model given here is based upon the
sphere. However, the mechanism of the move- results of many investigations and studies by
ment in our model differs from that of current many scientists, both at sea and on the Japanese
interpretions. Our mechanism, i.e. of thrust mainland. Further studies, especially deeper
faulting which cuts the boundary between the drilling, the more precise determination of
continental lithosphere and the oceanic litho- earthquake hypocentres and stress patterns,
sphere, does not lead to such a rapid subsidence and closer observation of the heat flow flux are
of the oceanic lithosphere as is held in current required to obtain a more accurate knowledge
views on plate tectonics. of the structure and tectonics of forearc regions
Although currently small, the size of the of the world including those of the Japanese
accretionary prism may have been variable in Islands.
the past, depending not only upon speed of
subduction of the oceanic plate, but also on
factors such as the supply of sediment from land
as well as the amount of sediment trapped in
forearc basins. ACKNOWLEDGMENTS:The writers wish to express their
Magmatism has occurred in the past at hearty thanks to Professors N. Nasu, A. Asano, S.
greater depths in the forearc. Acidic and Nagumo, Drs J. Kasahara, H. Kagami, and Mr K.
basic magmas ascended from depth and in- Fujioka of the University of Tokyo, Dr A. Mizuno of
the Geological Survey of Japan, Professors H. Aoki
truded into some of the fault crush zones,
of the Tokai University and K. Nakazawa, Dr D.
sometimes carrying up continental and oceanic Shimizu, Messrs Y. Miyake and K. Hisatomi of the
crustal xenoliths. Although currently-exposed Kyoto University, and many others for their stimu-
forearc igneous rocks are likely to be old, lating discussions and help. The writers are also in-
magma genesis may still be taking place in cer- debted to Miss T. Imai for the preparation of the
tain circumstances under the forearc. manuscript.

References
BECK, R. H., LEHNER, P., DIEBOLD, P., BAKKER,G. HONZA, E., TAMAKI, K. & MURAKAMI, F. 1978.
& DOUST, H. 1975. New geophysical data on key Geological map of the Japan and Kuril trenches
problems of global tectonics. Proc. 9th World and the adjacent areas. Marine Geology Map
Petroleum Congress, Tokyo, 2, 3-32. Series, 1:I 000 000. Geol. Surv. Japan.
FUJIOKA, K. & NASU, N. 1978. Chemical composition HOSHINO, M. 1963. Southwest Japan Trench. Mar.
of the volcanic rocks from the Oyashio Ancient Geol. Tokyo, 10, 10-15 (in Japanese with
Landmass. Tectonophysics EOS Trans. Am. geo- English abstract).
phys. Union, 59, 1183. HUSSONG, D. & UYEDA,S. et al. 1978. Leg 60 ends in
GNIBIDENKO, H. S., KRASNY,M. L. & PoPov, A. A. Guam. Geotimes, 23, 10, 19-22.
1978. Tectonics of the Kuril-Kamchatka deep- ICHIKAWA, M. & KAWAMURA,Y. 1971. Mechanism.
sea trench. Tectonophysics EOS. Trans. Am. In: General Report on the Tokachi-oki Earth-
geophys. Union, 59, 1184. quake of 1968. Eigaku Pub. Co., 33-8.
HADA, S., SUZUKI, T., YOSHIKURA,S. d~: TSUCHIYA, IGI, S. 1973. The metagabbros and related rocks
N. 1979. The Kurosegawa Tectonic Zone in of the 'Yakuno complex' in the inner, zone,
Shikoku and tectonic environment of the outer Southwest Japan. Rep. geol. surv. Japan, 248,
zone of Southwest Japan. In: The Basement of 1-39.
the Japanese Islands. Prof. Hiroshi Kano Memo- IGl, S., NAKAZAWA, K. & KURODA, Y. 1979. The
rial Volume, 341-68 (in Japanese with English basement complex in the Maizuru structural belt.
abstract). In: The Basement of the Japanese Islands. Prof.
HASEGAWA, A., UMINO, N. & TAKAGI, A. 1978. Hiroshi Kano Memorial Volume, 143-52 (in
Double-planed structure of the deep seismic Japanese with English abstract).
zone in the northeastern Japan area. Tectono- IMOTO, N., KANEKO, K., NAKAYAMA,I., YOSHITAN1,
physics, 47, 43-58. A., TEMPAKU, T. & SUZUKI, H. 1970. The
HAYAKAWA,M. & IIZUKA, S. 1979. A mechanism to occurrence of so-called acidic igneous rocks in
explain the earthquakes around Japan by a pro- the north-eastern part of the Shionomisaki
cess of partial melting. In: UYEDA,S., MURPHY, Peninsula, Wakayama Prefecture. Bull. Kyoto
R. W. & KOBAYASHI K. (eds). Geodynamics Coll. Educ. 37, 91-7 (in Japanese with English
of the Western Pacific. Suppl. Issue, J. Phys. abstract).
Earth, 43-58. INOUE, E. (ed). 1978. Investigation of the Continental
HISATOMI, K. & MIYAKE, Y. 1981. Upheaval move- Margin of Southwest Japan June-July 1975
ment and igneous activity in the Shionomisaki (GH75-4 Cruise). Geol. Surv. Japan, 88 pp.
area, Kii Peninsula, Southwest Japan. J. geol. ISHIZAKA, K. 1972. Rb-Sr dating on the igneous and
Soc. Japan (in Japanese with English abstract). metamorphic rocks of the Kurosegawa tectonic
72 T. Shiki & Y. Misawa
zone. J. geol. Soc. Japan, 78, 569-75 (in west Japan. In: The Basement of the Japanese
Japanese with English abstract). Islands. Prof. Hiroshi Kano Memorial Volume,
ISmZAr,A, K. & YANAO, T. 1975. Occurrence of 201-16 (in Japanese with English abstract).
oceanic plagiogranites in the outer tectonic zone, NODA, M. 1961. Pre-Cambrian (?) rocks in the outer
Southwest Japan. Earth planet. Sci. Lett. 27, zone of Southwest Japan. J. geol. Soc. Japan, 67,
371-7. 346-9 (in Japanese with English abstract).
KANO, H., NAKAZAWA,K. & SHIKI, T. 1961. Consid- NOHDA, S. 1973. Rb-Sr Dating of the Yatsushiro
eration on the Permian back-grounds of the granite and genesis, Kyushu, Japan. Earth
Maizuru district. J. geol. Soc. Japan, 67, 463-75 planet. Sci. Lett. 20, 140-4.
(in Japanese with English abstract). Nozawx, T. 1975. A preliminary note on the distribu-
KARIG, D. E., INGLE, J. C., JR, el al. 1975. Initial Rep. tion of granitic rocks on outer side of Southwest
Deep Sea drill. Proj, 31, U.S. Govt Printing Japan, in Kii-Shikoku-Kyushu Area. Assoc.
Office, Washington. Geol. Collaboration Japan Monogr. no. 19,
LANGSETH, H. & OKADA, H. et al. 1978. Near the Problems on geosynclines, 209-12 (in Japanese
Japan Trench, transects begun. Geotimes, 23, 3, with English abstract).
22-7. OJIMA, M. 1979. History of the Earth. Iwanami Pub.
MARUYAMA, S. 1978. The dismembered ophiolite Co. (in Japanese).
belt. Earth Sci. (Chikyu Kagaku), 32, 317-20 OKADA, H. & SAKA1,T. 1979. Leg 56 deep sea drilling
(in Japanese with English abstract). in the Japan Trench--1. Mar. Sci./monthly, 11,
MATSUMOTO, T., NODA, M. & MIYAHISA, M. 1962. 756-62 (in Japanese).
Kyushu Region. Asakura Pub. Co. (in OKUDA, Y., KUMAGAI,M. & TAMAKI,K. 1979. Tecto-
Japanese). nic development of the continental slope and its
MATSUZAWA,A., TAMANO,T., AOKI, Y. & IKAWA,T. peripheral areas off Southwest Japan in relation
1980. Structure of the Japan Trench subduction to sedimentary sequences in sedimentary basins.
zone, from multi-channel seismic-reflection re- J. Japan Assoc. Petrol. Technol. 44, 279-90 (in
cords. Mar. Geol. 35, 171-82. Japanese with English abstract).
MISAWA, Y. 1976. Submarine geology off the south- PLAFKER, G. 1972. Alaskan earthquake of 1964 and
western part of the Fossa Magna. Mar. Sci./ Chiliean earthquake of 1960, Implications for arc
monthly, 8, 611-6 (in Japanese with English tectonics. J. geophys. Res. 77, 901-25.
abstract). SAWAMURA,T. 1953. Relation between the activities
MIYAKE, Y. 1981. Geology and petrology of the of the outer earthquake zone in Southwestern
Shionomisaki igneous complex, Wakayama Pre- Japan and the geologic structure and crustal
fecture, Japan. J. geol. Soc. Japan. 87, 383-403 movements of Shikoku and its vicinity. Res. Rep.
(in Japanese with English abstract). Kochi Univ. 2, 2, 1-46 (in Japanese with English
MURAUCm, S. 1979. Mechanism of subduction in abstract).
Japan Trench. Mar. Sci./monthly, 11, 799-806 SHIBATA,K., IGI, S. & UCHIUMi,S. 1977. K-At ages of
(in Japanese). Hornblendes from gabbroic rocks in Southwest
NAGUMO, S. 1980. Seismic activity and geological Japan. Geoch. J. 11, 57-64.
structure across the Japan Trench and forearc. SHIKI, T. 1978. Structure and tectonics under the
In: SUGIYAMA,R., HAYAKAWA,M. & HOSHINO, continental slopes of the Honshu Arc. Tecto-
M. (eds). Earthquake. Tokai Univ. Press, nophysics EOS Trans. Am. geophys. Union, 59,
Tokyo, 25-40 (in Japanese). 1183.
NASU, N., HONZA, H., FUJIOKA, K. & SATO,T. 1978. SHIKl, T. & MISAWA, Y. 1979. A few problems
Outline of drilling in Japan Trench by Glomer concerning the structure under the continental
Challenger, Leg 57. Bull. Japan Soc. Sedimentol- slope and its origin. Earth Sci. ( Chikyu Kagaku),
ogy, 15, 12-15 (in Japanese). 33, 208-24 (in Japanese with English abstract).
NASO, N., TOMODA,Y., KOBAYASHI,K., KAGAMI,H., SHIKl, T. & MISAWA,Y. 1980. Geologic structure and
UYEDA, S., NAGUMO, S., KUSHIRO, I., OZIMA, tectonics under the inner slopes of the Nankai
M., NAKAZAWA, K., TAKAYANAGI,Y., OKADA, Trough and the Japan Trench. Geotectonics, 6,
H., MURAUCHI, S., ISHlWADA, Y. & ISHII, Y. 98-109 (in Russian).
1979. Multi-channel seismic reflection data SUZUKI, S. & OKADA, H. 1977. Regional variation in
across the Japan Trench. IPOD-Japan Basic Pn velocity and seismic activity in and around
Data Series, no. 3. Ocean Res. Inst., Univ. Northeastern Japan. In: Subsurface Structure of
Tokyo. Hokkaido and its Surroundings, and its Meanings
NHNO, H. 1935. On sounding at Tosabae. Geogr. in Geosciences, a Symposium, 1-12 (in
Rev. Japan, l l , 679-87 (in Japanese with Japanese).
English abstract). TOMITA, T. 1954. Geologic significance of the color of
NISHIMURA, S., KATSURA, K. • SASAJIMA, S. 1976. granite zircon, and the discovery of the Pre-
Fission track ages of zircon in conglomerates Cambrian in Japan. Mem. Fac. Sci. Kyushu
occuring in the Maizuru District. J. geol. Soc. Univ. 40, 135-61.
Japan, 82, 413 (in Japanese with English ab- VON HUENE, R. 1978. A section across the Aleutian
stract). arc based on shallow and deep geophysical data.
NISHIMURA, Y. 1979. Metamorphic rocks of 300- Abstr. Int. Geodynamics Conference 'Western
400 My ages found in the inner zone of South- Pacific' and 'Magma Genesis', Tokyo, 172.
 F o r e a r c g e o l o g i c a l structure o f the J a p a n e s e I s l a n d s 73

VON HUENE, R. & ARTHUR, M. A. 1981. Sedimen- YAJIMA,T. & FUKUDA,H. 1976. On the Shionomisaki
tation across the Japan Trench off northern igneous complex, Wakayama Prefecture. Mere.
Honshu Island (this volume). Fac. Educ. Saitama Univ. 25, 27-44.
VON HUENE, R., NASU, N. et al. 1978. On Leg 57, YAMASHITA,N. 1979. A history of the researches on
Japan Trench transected. Geotimes, 23, 4, 16-21. the Kurosegawa Tectonic Zone. In: The Base-
WATANABE,H. 1971. Tsunami sources, aftershocks, ment of the Japanese Islands. Prof. Hiroshi Kano
and earthquake mechanism on the Tokachi-oki Memorial Volume, 299-318 (in Japanese with
earthquake of 1968. In: General Report on the English abstract).
Tokachi-oki Earthquake of 1968. Eigaku Pub. YOSHtl, T. 1979. A detailed cross-section of the deep
Co., 137-51. seismic zone beneath Northeastern Honshu,
YA.IIMA, T. 1972a. Petrology of the Murotomisaki Japan. Tectonophysics, 55, 349-60.
gabbroic complex. J. Japan Assoc. Min. Pet. YOSHIZAWA,M. 1954a, b. On the gabbros of the Cape
econ. Geol. 67, 218-41. of Muroto, Shikoku Island, Japan. Mere. Fac. Sci.
YAJ1MA, T. 1972b. Petrochemistry of the Murotomi- Univ. Kyoto, 20B, 291-84, 21B, 193-212.
saki gabbroic complex. J. Japan Assoc. Min. Pet.
econ. Geol. 67, 247-61.

TSUNEMASA SHIKI, Department of Geology and Mineralogy, Kyoto University,

Kyoto 606, Japan.
YOSHmUMIMISAWA,Faculty of Marine Science and Technology, Tokai University,
Shimizu 424, Japan.
 Facies belts of the Middle America Trench and forearc region,
southern Mexico: results from Leg 66 DSDP

J. Casey Moore, Joel S. Watkins, Kenneth J. McMillen, Stephen B. Bachman,

Jeremy K. Leggett, Neii Lundberg, Thomas H. Shipley, Jean-Francois Stephan,
Floyd W. Beghtel, Arif Butt, Borys M. Didyk, Nobuaki Niitsuma, Les E.
Shephard & Herbert Stradner
SUMMARY: The Middle America Trench SE of Acapulco is flanked by a steep
canyon-incised slope and narrow shelf, showing one of a variety of sedimentary facies
patterns possible at convergent margins. Piston and drill cores from this region define eight
facies belts including: (1) a pelagic facies of brown clay, (2) an outer slope mud facies, (3) a
trench sand facies, (4) a foraminiferan-free facies on the lower slope, (5) a foraminiferan-
bearing facies on the mid-slope, (6) a laminated mud facies on the upper slope, (7) a shelf
facies of sand and mud, and (8) a canyon facies of sand and gravel. The superposition of
trench and lower slope sediment during accretion results in a fining upward sequence
reflecting a gradual uplift of the seafloor through the trench sediment-plume. The lower
limit of the foraminiferan-bearing facies is defined by the absence of in situ calcareous
foraminiferans and is controlled by the calcite compensation depth. The upper slope
laminated mud facies probably reflects the depth range of the oxygen minimum zone.
In the Leg 66 area sedimentation rates are high in the trench and on the outer and lower
slope, decrease on the mid-slope, and increase again on the shelf. On the inner shelf,
waves and currents concentrate sand which funnels through a prominent submarine
canyon, bypassing the mud-dominated slope and accumulating in the trench. A ter-
rigenous sediment-plume generated by trench turbidity flow causes accelerated sediment
accumulation to about 500 m above and 40 km seaward of the trench. The volume of
material transported by the trench sediment-plume is five or six times greater than that
moved by the shelf sediment-plume which supplies detritus to the shelf, upper slope and
mid-slope environments.

Facies models for trench and trench-slope de- bathymetric data (Shipley 1981). The piston
posits are essential for the palaeogeographic cores show the distribution of modern facies,
interpretation of ancient convergent margin whereas the drill cores record the distribution
sequences, especially those involved in collision of facies belts back to early Miocene.
zones. Development of field criteria for the For convenience we have subdivided and
recognition of various sub-environments is informally named the marine sedimentary en-
especially desirable. vironments off southern Mexico as follows: (1)
Available data from modern convergent shelf environment (0-200 m); (2) upper slope
plate boundaries and probable ancient equiva- environment (200-1000 m); (3) mid-slope en-
lents indicate substantial variability in vironment (1000-3400 m); (4) lower slope en-
sedimentary deposits (Schweller & Kulm 1978; vironment (3400 m - t r e n c h floor or about
Underwood & Bachman 1981; Underwood et 5000 m); (5) trench environment (any sediment
al. 1980 and references therein). The resultant ponded in trench axis); (6) outer slope environ-
facies mosaics reflect complex depositional sys- ment (area from the outer margin of trench
tems developed in response to the diversity of sediment pond, or the trench axis if unfilled, to
morphotectonic patterns. We present a detailed seaward limit of hemipelagic sediment accumu-
facies model for a type of convergent margin lation); (7) pelagic environment (seaward of
that consists of a juvenile subduction complex limit of hemipelagic sediment accumulation).
emplaced against truncated older continental
basement terrane.
Our facies model for the Middle America Tectonic setting
Trench and trench-slope off southern Mexico
stems from the study of 21 piston cores (McMil- The Pacific margin of southern Mexico is char-
len & Haines 1981; McMillen et al. 1981), eight acterized by a rugged coastal mountain range, a
Deep Sea Drilling Project Sites (Moore, Wat- narrow continental shelf, and a steep trench
kins et al. 1979), and associated geophysical and slope which lacks a forearc basin (Fig. 1). The

77
78 J . C . Moore et al.
_99o -98°
~o.,-,.~._ ' " ". {" :'e~ ~ "`~ ~ J - [ -...R,o o~,,,~,: I ' I I

i
,,. - --'~ " ~'~- - ~ ', "%\ ~.3 ! ,o ,

/ / /
/ "-- -
,, ,,,L - ---44. . . . . ,
,/ " ~ 6 t _ _ ~ - -
.
]
.
~----~
~
:-2e

l "" ...... O O "'\ ---- -" ....... ~. . . . . .

1 * / DSDP SITES ~ ""' P~';i~i~"ie~'--~"i " / - ~ ---- i

• PISTON CORES ~ 7 "~" '\',, ~.....--__ ___
SEISMIC .... -"'" \ .~
./ REFLECTION LINES \~ - / / \ ~,,i . ~ t . J _ : ~ : : : - : ~1
/ 0 50km 2,/' \ ~ ~.~s_~) - - -~-?U;_..~ /

-99° -98"
F16. 1. Bathymetry and core locations. Bathymetry from Shipley (1981).

Middle America Trench in the Leg 66 area is Plate tectonic reconstructions and the
about 5 km deep and marked by a series of magmatic history of the continental margin
discontinuous sediment ponds with a maximum both suggest that subduction has been occurring
of 625 m of turbidite fill overlying about 200 m at least intermittently beneath southern Mexico
of pelagic and hemipelagic sediment which rests about 100 Ma (Karig et al. 1978). An extensive
on Miocene oceanic crust (Shipley 1981 and zone of off-scraped deep-sea sediments and a
Fig. 2). In the Leg 66 drilling area the Cocos broad forearc region might be expected along
plate converges with the North American plate this margin in view of its long history of inferred
at about 7 cm yr -1 along an azimuth of 038 °, or subduction. However, Mesozoic and Palae-
about 20 ° from perpendicular to the strike of ogene intrusions and associated basement
the trench (Minster & Jordan 1978). rocks (Mejorada 1976) occur at the shoreline

0
493
489 _ _-

- ~ " < " " ~ "~-~-~" ~ "~. ~'~'~\ \. LANDWARD . . . . . . . . . . . .

OCEANIC C R US]"

0 10 20 km
10
VE--2 1

FI6.2. Generalized cross-section along Deep Sea Drilling transect. Portion from Site 490 landward
and 486 seaward taken along seismic line OM-7N, remainder along seismic line MX-16 (locations in
Fig. 1).
 Facies belts o f the M i d d l e A m e r i c a Trench 79

and are shown, by magnetic, seismic and drill- with tectonic setting and sedimentary facies
ing data, to extend within 35 km of the trench similar to southern Mexico may have occurred
(Karig et al. 1978; Shipley et al. 1980; Moore, along western North America during the
Watkins et al. 1979). Neogene rocks both over- Mesozoic; the Cretaceous subduction zone
lie the basement or are faulted against its bordering the Salinian block in California may
seaward boundary, suggesting removal of any be a prime example (Page 1981).
pre-existing Mesozoic and/or Palaeogene fore-
arc and accretionary terrane. The present phase
of accretion apparently began in the Neogene I n t e r p r e t a t i o n o f D S D P c o r e s in t h e L e g
(Moore, Watkins et al. 1979) at the same time 66 a r e a
as magmatism in the Trans-Mexican volcanic
belt (Cantagrel & Robin 1979). The interpreted drilling results are summarized
Adjacent to the Leg 66 area coastal moun- below from the seaward-most to the landward-
tains of crystalline basement and intrusive rocks most site (Fig. 3). We interpret only those
supply coarse quartzofeldspathic sediment to portions of the DSDP holes that directly pertain
the marine environment while blocking volca- to the facies model; complete descriptions are
niclastic detritus from the Trans-Mexican volca- provided elsewhere (Watkins, Moore et al.
nic belt situated 240kin NE of the trench 1981).
(Enkeboll 1981; Bachman & Leggett 1981). Drilling on the outer trench slope at Site 487
The narrow shelf, steep trench slope cut by penetrated Quarternary hemipelagic mud
submarine canyons, and lack of a forearc basin which covers Pliocene and upper Miocene
all permit transport of large volumes of coarse brown pelagic clay that in turn overlies basaltic
clastic sediment into the trench. oceanic basement. Movement of the Cocos
Thus the tectonic elements controlling clastic plate toward a terrigenous source apparently
sedimentary facies along southern Mexico in- accounts for the superposition of hemipelagic
clude substantial sediment supply, a relatively over pelagic sediments. Lack of calcareous ooze
rapid convergence rate, and direct input of directly above the oceanic crust suggests it
sediment into the deep sea. Subduction zones originally formed below the local CCD.

~ _
487

- LITH.
_
-.-_ 0
_-___- EN, ~
486 488
LITH.

...... -5
- ----: :
-NV ,~

2
o:
491 492

,,z. kN4
490

=_

_
489
LITH
_
_

=_- = - -
_-=

_-
_ _
_

W
493
o
meters

g,, --::-.:-
i:/i:: !--:-----:
-:_=- . . . .
"~
.
-i-:-i-:z:

-:- -::-- _z_ _ _ : I00

___
o,; _=-__ _ -- ::_-:1_ :::- -7:
- _- . _=~___

i-::: - - le -_.__ : : z ......

:-.-. o i :--_:
0 o c

. . . .
200
_ ._
~:_~-_==-

: - L _z:: I

. . . . .

zr:__
-:-Z-:-
ENV. = ENVIRONMENT OF DEPOSITION ::.:._.:-:.._r. - -. . ...... .
:_- z _ _ : _ ~ "~'---:-:
z Z--:-: :
--- ~ .....
c-=-=-=-_ ~_z_:::z--_-
a,vvv= HIATUSOR UNCONFORMITY

::~.:::::::..,
. . . . .

LITHOLOGIG SYMBOLS
,;.-i._.:,!?-::::
i PELAGIC CLAY ~ THICK SANDBEO~ _ : _ _ : =

~ FORAM-FREE
MUD ANDMUDSTONE ~ BASALT

g----].H,N SA.OBEDS ~ O,O.,T.

~ LAMINATED
M U D AND MUDSTONE

FIG. 3. Lithologic summary of DSDP sites. Interpretation of depositional environment locally

omitted where controversial.
80 J.C. M o o r e et al.

The upper Quaternary fine- to coarse- their depositional environments and defined by
grained sands and muds at Site 486 confirm the their lithology, sedimentary structures, fossil
sand-rich character of the modern trench fill as content, and sedimentation rates. The essential
indicated by piston coring (McMillen & Haines characteristics of each facies are summarized in
1981; McMillen et al. 1981). The sequence at Table 1 and augmented in the text by detailed
the lower and mid-slope sites 488,491 and 492 references to specific DSDP sites and piston
consists of hemipelagic mud underlain by mud- cores. To normalize for effects of compaction
stone and sand. Surface muds at Site 488 lack in we quote sedimentation rates in units of mass
situ calcareous foraminiferans indicating de- per area per time (van Andel et al. 1975).
position below the CCD. On the mid-slope at
Sites 491 and 492, hemipelagic mud, respective-
ly below 44 and 80 m sub-bottom, was probably Pelagic facies
uplifted from below the CCD, as indicated by
the absence of in situ calcareous foraminiferans The Pliocene to upper Miocene brown clay
(McMillen & Bachman 1981). The thick- in the lower portion of hole 487 constitutes an
bedded sand and mudstone units encountered open-ocean facies having typical abyssal pelagic
at depth at Sites 488,491 and 492 lithologically characteristics: a siliceous microfossil assemb-
resemble the modern trench sediments, but lage, authigenic phillipsite, a lack of terrigenous
could also have accumulated in a basin or silt, and basal metal-enrichment (Leggett
submarine canyon on the lower slope. At each 1981). Accumulation rates range from 0.4 to
site the depth of the sand-rich units coincides 1.7 gm cm-2/1000 yr and are the lowest in the
with the depth of the zone of dipping reflectors Leg 66 area. In contrast to the pelagic facies at
(Fig. 2) which shows acoustic characteristics Site 487, modern pelagic deposits seaward of
similar to the modern trench deposits (Shipley the Leg 66 area consist of siliceous and cal-
1981). Neither buried submarine canyons nor careous ooze (Horn et al. 1972).
slope basins are evident on a network of seismic
lines across the trench lower slope. As such, we
prefer to interpret the sand-rich unit as uplifted Outer slope mud facies
trench deposits.
In the mid-slope environment, drilling at Site The outer slope mud facies is distinguished
490 penetrated 589 m of mud and mudstone by a green-grey colour, a terrigenous silt com-
with local ash and thin silt and fine sand beds. ponent, and a sedimentation rate higher than
At Site 490, rocks below 450 m lack in situ that of the pelagic facies. Pyrite, plant frag-
calcareous foraminiferans suggesting accumula- ments and a modest organic carbon content
tion below the CCD (Fig. 3 and McMillen & (0.4-1.7%) reflect the reduced nature of the
Bachman 1981). We interpret the whole section outer slope muds relative to the pelagic brown
as a slope deposit though its lower part clays which lack pyrite and plant fragments and
(accumulated below the CCD) apparently was have organic carbon contents of 0.1-0.2%.
uplifted from the lower slope environment (Fig. Faint colour laminations and ash beds occur
3). locally in cores from Site 487. Several 1 to 2 cm
Sites 489 and 493 are now respectively in the thick terrigenous sand beds were recovered
mid- and upper slope environments. Forami- from a piston core located 12 km seaward of
niferan- and glauconite-bearing mud charac- and 700 m shallower than the trench axis
terize surface deposition at Site 489 although at (McMillen et al. 1981), but such sand beds were
the base of the hole we cored a sand-rich unit not drilled at Site 487.
that included articulated bivalves (Fig. 7C), While the concentration of terrigenous silt
scaphopods, echinoids, and shallow-water fora- and sand-sized material in the outer slope mud
miniferans. Coring at Site 493 initially pene- is far greater than that in the pelagic clay, it is
trated laminated muds but terminated in a less than that of the inner slope mud (see
sequence of muddy siltstone and sandstone below). The outer slope mud contains radio-
overlying quartz dioritic basement. The muddy larians but no foraminiferans except a probable
silt and sandstone contains foraminiferans in- redeposited fauna similar to that recovered in
dicative of a deep shelf environment. the trench. The sedimentation rate of the outer
slope mud is 10 gm cm-2/1000 yr, markedly
higher than that of the pelagic mud but less than
Facies belts in the L e g 66 area recorded on the lower slope. The outer slope
mud is about 100 m thick at Site 487, 7 km
The facies belts described below are named for seaward of the trench floor.
 Facies belts o f the M i d d l e A m e r i c a T r e n c h 81

Trench sand facies Site 486 lying on seismic line OM-7N. The
gradient of the trench from the mouth of
We sampled the trench sand facies in three Ometepec Canyon towards Site 486 is 1/125
piston cores and two DSDP holes (Site 486) in (0.5 °) which is at the upper limit of slopes
the modern trench and presumably at depth at observed on submarine fans (Nelson et al. 1978)
DSDP Sites 488, 491, and 492 on the trench and an order of magnitude steeper than those
inner slope. In the Leg 66 area the trench sand from the continuous trench wedges of the Peru-
facies occurs in a series of discontinuous sedi- Chile and eastern Aleutian trenches (Schweller
ment ponds whose localization is clearly related & Kulm 1978; von Huene 1974).
to major submarine canyons as well as irregu- A second sediment pond occurs in the trench
larities on the oceanic plate (Fig. 4). off Rio Verde where a series of small canyons
cut the trench slope (Fig. 4). One piston core
from the north-western edge of this accumula-
NW SE tion recovered mud with local graded beds
showing Bouma Tc-e sequences (McMillen et
al. 1981).
The inferred trench deposits comprising the
basal units at Sites 488, 491, and 492 include
sand beds locally coarser than those in the
modern trench but overall the sequence is
50km
' ' 'VE,'20,' ' [] Trench Sand Facies richer in mud. Sand ranges from fine- to very
D Pelagic and Outer Slope Facies coarse-grained and granular; clasts locally reach
• Oceanic Crust 7 mm in diameter. At Site 488 sand layers
inferred from accelerated drilling rates are as
much as 11 m thick and appear to thicken
FI6.4. Distribution of trench sand, pelagic upwards over an interval of 100 m. A semi-
and outer slope mud facies longitudinally
consolidated unit more than 6 m thick from Site
along trench. Thickness of trench sediment
taken at base of inner slope with correc- 492 represents the thickest single sand bed we
tions made for any thickening due to mild recovered; overall this bed appears massive
deformation (after Shipley 1981). (Fig. 5A), although it does grade up-section
from very coarse to medium sand. The mud-
rich sequences interlayered with the massive
The major trench sediment pond occurs at sand beds at Sites 488, 491, 492 locally show
the mouth of Ometepec submarine canyon in thin graded and laminated sand and silt beds
the Leg 66 area (Fig. 4). Two piston cores (26 (Fig. 5C).
and 27, Fig. 1) on the small fan at the mouth of Foraminiferal faunas in both modern trench
this canyon consist predominantly of fine and deposits (Site 486) and their inferred ancient
coarse sand with minor mud. Similarly, the two equivalents (Sites 488, 491, 492) include dis-
holes drilled at Site 486 recovered predomi- placed inner shelf, outer shelf, upper slope, and
nantly fine- to very coarse sand (Fig. 5B) with mid-slope forms. Measurement of the sediment
minor mud. accumulation rate in the trench is precluded by
The morphology of the lower slope where poor biostratigraphic resolution. Using the
Ometepec Canyon enters the trench suggests dynamic steady-state model (e.g. Moore 1979)
levee development and a recent submersible the calculated rate is about 322 gm cm-2/
dive in this area indicates that boulder-filled 1000 yr (about 2.6 km Ma -1) for the area near
channels cut the fan (Lonsdale pers. comm. seismic line OM-7N.
1980). One seismic profile (MX-15, Shipley
1981; McMillen et al. 1981) shows a clearly Foraminiferan-free mud facies: lower slope
defined axial channel about 400 m wide and environment
40 m deep bordering a fault-line scarp. Two
small channels appear in the trench on seismic The modern inner slope sediments include
profile MX-16 9 km NW (down gradient) from foraminiferan-free mud on the lower slope,
MX-15; no channels cut the trench fill on foraminiferan-bearing mud on the mid-slope,
seismic profile OM-7N 18km NW of profile and laminated mud on the upper slope (Figs 6
MX-15 (Shipley 1981; McMillen et al. 1981). & 7). Each of these sediment types is defined as
Thus, a transition from channelized to non- a separate facies although their distinction may
channelized flow occurs within 27 km from the be subtle.
mouth of Ometepec Canyon to the vicinity of The foraminiferan-free mud facies of the
 0~
L~

t~

~ii~ill ..............~ ii~i~ii~i~i,~i

...............
ii,,,ii~ii~i'i~ii~iii~i~i~i~i~ii~iii~ii~,~,~
.... ~ i~

qJ ~
 Facies belts of the Middle America Trench 83

Fu6.6. Foraminiferan-free mud facies. (A) Thin sand bed from foraminiferan-free mud facies, Site
491, 167 m. This sand bed represents one of the thicker though photogenic sand beds of the
foraminiferan-free mud facies. Most sand beds were less than 3 cm thick. (B) Disrupted silt beds,
Site 488, 238 m.

modern trench lower slope has been sampled in 170 m of the foraminiferan-free mud here con-
five piston cores and at Site 488 and also sists of mud with local green and blackish-green
encountered at depth at Sites 489, 490, 491, laminations with isolated concentrations of
492, and 493. Piston core 1 (Fig. 1) just upslope plant matter. From 170 m to just above the
from the trench includes mud-rich turbidities trench sand facies at 310 m, locally graded fine
with Tc-e B o u m a sequences and a maximum sand and silt beds up to 3 cm thick punctuate
sand bed thickness of 3 cm. Other piston cores the otherwise muddy section. Thin beds of silt
show colour lamination and fine-grained sand and fine-grained sand, nowhere greater than
beds up to 1 cm thick. Notably, not even the 8 cm thick, make up a small percentage of the
two cores from slope basins encountered any foraminiferan-free mud facies at Sites 491 and
concentration of sand. 492 (Fig. 3).
D S D P Site 488 penetrated 310 m of the In summary, the foraminiferan-free mud
foraminiferan-free mud facies before coring facies is mud dominated with a few per cent of
probable trench deposits (Fig. 3). The upper fine-grained, thin, sand beds (Fig. 6A). It is

FIG. 5. Trench sand facies. (A). Semiconsolidated 'massive' sand bed from 492, 280.5-290 m.
Although the bed appears massive, it grades overall from medium to very coarse sand down-section
(from left to right). (B) Very coarse sand, Site 486A, 22 m. (C) Laminated, locally graded fine
sand and silt beds, Site 488,383 m.
,.,,
 Facies belts of the Middle America Trench 85
330

300
200
J --
@Measured Late Quaternary Sediment
Accumulation Rate
% 15o U Measured Sediment Accumulation in
N~ Lower Miocene Shelf Deposits at
E Sites 489 8~ 493
IOO
[]Calculated Sediment Accumulation Rate
50 ~ m i c Steady-State Trench

o
OUTER [TRENCHI LOWER I MID-SLOPE IUPPER I SHELF
SLOPE SLOPE SLOPE
SHELF PLUME
O] TRENCH PLUME 491/~-.-....,~we".~.--~
.o ?iL
::. "'"""

I ........... - o ,o
- .

~.o~
. . . . ". . . . '"'""II
I 0 ~ VE-2:I

Fro. 8. Sediment accumulation rates plotted across continental margin.

distinguished from the foraminiferan-bearing sample the foraminiferan-bearing mud facies.

and laminated mud facies by the absence of in Piston cores 30, 32, and 34 (Fig. 1) consist
situ calcareous foraminiferans. Both grading in predominantly of mud with local colour lamina-
the sand beds and rhythmic colour laminations tions and minor concentrations of forami-
with discernible grain-size variations suggest niferans, glauconite, radiolarians, and diatoms.
deposition in part by dilute turbidity currents. Core 30 penetrated a slope terrace, whereas
The thin sand beds occur in no recognizable cores 32 and 34 were taken from the slope.
megasequence although the foraminiferan-free At Site 491 the foraminiferan-bearing mud
mud facies does fine-upward overall. The sedi- facies is approximately 44 m thick and is
ment accumulation rate in the foraminiferan- characterized by a dominance of argillaceous
free mud facies varies between 4 and material, the common occurrence of forami-
44 gm cm-2/1000 yr and generally decreases niferans, and a complete lack of sand and silt
away from the trench (Fig. 8). beds (Fig. 3). Drilling at Site 492 penetrated
The hemipelagic mud component of the fora- about 80 m of foraminiferan-bearing mud
miniferan-free mud facies is essentially indis- which includes concentrations of foraminiferans
tinguishable from the hemipelagic mud of the ranging up to 10% of smear slides. Here,
outer slope mud facies. Except for a higher glauconite occurs as a dispersed component in
percentage of thin-bedded turbidities in the the predominantly muddy section and in dis-
former, these two facies would be impossible to crete beds up to 10 cm thick (Fig. 3).
separate on the outcrop, though their associ- The foraminiferan-bearing mud facies com-
ated rock types might allow their identification. prises the initial 450 m of sediment cored at Site
490. Concentrations of foraminiferans locally
Foraminiferan-bearing mud facies: mid-slope reach 5-15% in rare, thin sandy beds, whereas
environment glauconite is a common dispersed component
and also occurs in discrete beds up to 10 cm
Three piston cores and four DSDP sites thick (Fig. 3). The foraminiferan-bearing mud

Fro. 7. Laminated mud and shelf facies. (A) Laminated mud, Site 493A, 0-10 m. (B) Laminated
mudstone, Site 493, 278 m. (C) Articulated bivalve in silt of shelf facies, Site 489,279 m.
86 J . C . Moore et al.
facies at Site 490 includes recrystallized lime- tic intrusive rocks and fines upward through a
stone beds and nodules. At Site 489 the forami- predominantly muddy section. At Site 489 the
niferan-bearing mud facies is marked by the basal sand-rich sequence also fines upward but
presence of both foraminiferans and glauconite, is finer grained and thinner than the similar
but is only 6 m thick, overlying a substantial sequence at Site 493. Small articulated bivalves
hiatus. occur in the basal sequence at Site 489 (Fig. 7).
Quaternary accumulation rates in the forami- The foraminiferans in the sand-rich sections of
niferan-bearing mud facies are low at Sites 491 both Sites 489 and 493 indicate deposition in a
and 492 (1.8 and 3.6 gm cm-2/1000 yr, respec- deep shelf environment. Sediment accumula-
tively) seaward of the mid-slope high, but mod- tion rates in the sand-rich sequences are 4 and
est at Site 490 (10 gm cm-2/1000 yr) landward of 6 gm cm-2/1000 yr at Sites 489 and 493 respec-
the mid-slope high. tively. The shelf facies deposits at both sites
indicate an overall history of subsidence
Laminated mud facies: upper slope environment (McMillen & Bachman 1981).

Five piston cores from about 400 to 1000 m Canyon facies

water depth and DSDP Site 493 at 645 m water
depth all sampled laminated muds. Laminations Ometepec Canyon and other smaller sub-
in cores 5, 7, 8 and 9 occur as contrasting colour marine canyons are the principal conduits sup-
bands, but in core 35 as thin-bedded, fine- plying sediment to the trench in the Leg 66
grained, clean sand that locally shows channels area. Although we have not sampled these
and ripples (McMillen & Haines 1981). These features in detail a summary of available data
thin sand layers may have been deposited by for Ometepec Canyon is warranted.
overbank flow from Ometepec Canyon. The Piston cores 2, 3 and 35 penetrated deposits
near absence of bioturbation in all cores pre- on the slope flanking Ometepec Canyon. Cores
sumably accounts for the preservation of 2 and 3 contain layers of angular gravel clasts.
lamination. Core 35, unique among upper slope surface
Drilling at Site 493 penetrated Quaternary samples, contains numerous thin sand laminae,
and upper Pliocene laminated sediments from possibly deposited in response to density cur-
0 to 148 m (Fig. 3). Colour laminations are rent flow down Ometepec Canyon. Cores 26
most common (Fig. 7A, B) though foram- and and 27 from a small fan-shaped structure where
shell-rich sand laminations also occur. Between the Ometepec Canyon intersects the trench
234 and 372 m upper Miocene rocks display contain predominantly fine- to very coarse-
conspicuous colour laminations. The benthonic grained sand. Very coarse sediment up to boul-
foraminiferal fauna in the upper Miocene rocks der size occurs in channels at the canyon mouth
indicates accumulation under conditions of low (Lonsdale pers. comm. 1980).
dissolved oxygen (in the oxygen minimum Four seismic lines crossing Ometepec Can-
zone). Except for a trace occurrence in one yon show a narrow rough axis with thin fill
smear slide, glauconite is notably absent from (100 m or less). A fifth seismic line displays a
the laminated muds. Sediment accumulated at poorly stratified fill a maximum of 550 m thick
about 4 gm cm-2/1000 yr both in the Quater- and about 4 km wide that is cut by a channel
nary-upper Pliocene and upper Miocene lamin- approximately 300 m deep. The seismic data
ated mud sections at Site 493. indicate that Ometepec Canyon cuts into both
the crystalline basement complex and associ-
Shelf facies ated Neogene rocks.

Modern shelf deposits were recovered in

four piston cores and probable ancient shelf
sequences were penetrated at DSDP Sites 489 Discussion of facies descriptions in the
and 493. Piston cores 6, 36 and 37 (Fig. 1) Leg 66 area
contain massive clean sand with molluscan Non-diagnostic sedimentary components
shells, coral, and calcareous algae. A hemipela-
gic mud layer 1.8 m thick overlies an equal Definitions of sedimentary facies described
amount of shelly sand in piston core 6, and herein hinge on certain diagnostic sedimentary
piston core 4 consists of mud with one thin sand components. Non-diagnostic sedimentary com-
bed. ponents include siliceous microfossils, volcanic
A lower Miocene massive sand section at the ash, and limestone beds and nodules. Radiolar-
base of Site 493 occurs unconformably on grani- ia and diatoms occur in varying amounts in all
 Facies belts o f the M i d d l e A m e r i c a Trench 87

facies although they are most conscpicuous in downslope disappearance of in situ calcareous
sediments deposited below the CCD. All foraminiferans defines both the calcite com-
sedimentary facies also include dispersed ash pensation depth (CCD) (McMillen & Bachman
fragments and ash beds although sand derived 1981) and the boundary between the forami-
from the crystalline basement directly onshore niferan-free and foraminiferan-bearing mud
is far more abundant. Limestone beds occur facies. Clearly foraminiferans may be reworked
locally in facies accumulated above the CCD below the CCD. As such, study of sediment
and small calcareous nodules of probable dia- deposited between turbidite flows would best
genetic origin are present in sediments depo- define the boundary between the forami-
sited both above and below the CCD. niferan-free and foraminiferan-bearing mud
facies. The brown clays comprising the pelagic
facies in the Leg 66 area undoubtedly accumu-
Boundary between the foraminiferan-bearing
lated below the CCD but seaward of their
and foraminiferan-free mud facies
present location, removed from the terrigenous
The foraminiferan-free mud facies occurs influence of the continent. At other convergent
down-hole at Sites 490, 491 and 492 where the margins one might encounter a pelagic facies
sediments lack in situ calcareous forami- composed of other types of open-ocean sedi-
niferans. Diagenetic effects may have induced ments including carbonate oozes (e.g. Site 495,
this down-hole transition from foraminiferan- Leg 67, von Huene, Aubouin et al. 1981).
bearing to foraminiferan-barren mud. How- In addition to the CCD, the oxygen minimum
ever, at Sites 489 and 493 foraminiferans are zone (OMZ) exerts a critical chemical control
abundant in surface cores, initially diminish on the sediments accumulating in the Leg 66
down-hole, and then become common again at area. The substantially diminished bioturbation
depth. This distribution of foraminiferans is in the O M Z results in the preservation of finely
probably not simply due to burial diagenetic laminated sediments (Ingle 1975). In the Leg
effects but can be explained more simply by 66 area laminated sediment in piston cores and
vertical motions of the drilling sites relative from shallow cores from Site 493 apparently
to the CCD (McMillen & Bachman 1981). records the approximate depth range of the
A similar interpretation is preferred for the modern OMZ.
absence of foraminiferans at depth at Sites 490, The concentration of glauconite in the foram-
491 and 492. mud facies occurs at greater water depths than
The presence or absence of foraminiferans those at which glauconite normally forms
may be questioned as a criterion to separate (McRae 1972). Since neither the laminated
middle and lower slope sedimentary environ- mud nor the shelf sand facies include significant
ments. Sediments associated with the lower amounts of glauconite, simple down-slope re-
trench slope in the Leg 66 area might be defined working of this mineral is apparently precluded.
alternatively by the presence of the included We therefore infer that glauconite is forming
thin sand and silt beds. This definition is un- actively in the mid-slope region. The low sedi-
doubtedly a more useful field criterion although mentation rates and moderately oxidizing bot-
whether or not the thin sand beds would be tom waters in the mid-slope environment would
preferentially associated with the lower, mid- favour the in situ formation of glauconite
dle, or upper slope sediments would depend on (McRae 1972). Ross (1971) also noted a con-
the dynamics of clastic sediment distribution centration of glauconite between 1 and 2 km
along a particular trench slope. We use the water depth in the northern Middle America
absence of calcareous foraminiferans as an in- Trench slope off Manzanillo.
dex to lower slope sediments since it is probably
controlled by the CCD (see below) and less Control of facies by physical processes and
affected by trench slope physiography. physiography
The concentration of coarse material in the
C o n t r o l s on facies d i s t r i b u t i o n shelf sand facies undoubtedly results from the
winnowing effects of waves and currents in this
Oceanographic factors affecting facies
shallow environment. However, the occurrence
distribution
of sand or mud in piston cores is not simply a
Lithological differences between the facies function of water depth or distance from the
belts in the Leg 66 area are controlled by shoreline and therefore must record the com-
variations in oceanic chemical parameters as plex dynamics of this narrow shelf.
well as physiography and current patterns. The The presence of massive, coarse-grained sand
88 J.C. M o o r e et al.

in the trench but not on the slope indicates troduced by the trench plume. Material in the
sediment bypassing of the slope via Ometepec shelf sediment-plume may be transported in
and other submarine canyons. Calcareous sediment gravity-flows, by wave- and current-
material is nearly absent in the modern trench induced traction processes, or by grain-by-grain
sediments but common in the modern outer settling.
shelf sands. Dissolution below the CCD prob- The presence of hemipelagic sediment con-
ably does not account for the lack of calcareous siderably seaward of the trench axis (piston
detritus in the trench sand since deposition core 25) and the high sedimentation rate at Site
occurs in pulses which shield the bulk of the 488 attest to the widespread influence of the
sediment from the corrosive sub-CCD bottom trench sediment-plume beyond the trench
waters. Rather, much of the trench sand may be floor. At Site 488 the trench sand facies is
derived from the inner shelf or littoral zone succeeded upsection by a sequence of thin-
with any contribution from outer shelf cal- bedded turbidites which is in turn covered by
careous sand being severely diluted. Ometepec mud at 172 m sub-bottom (c. 425,000 yr BP). We
Canyon does head in the inner shelf area and believe that the sand represents deposits of the
could trap sand being transported in this region. trench floor and that the thin-bedded turbidites
Hence, the submarine canyons in the Leg 66 accumulated as Site 488 was progressively up-
area constitute a conduit for coarse sediment lifted through the trench sediment-plume to its
that bypasses the outer shelf as well as the present water depth of 4259 m. By using an
slope. uplift rate of 400 m Ma -1 (McMillen & Bach-
The thin fine-grained sand beds that occur in man 1981) we deduce that the cessation of
piston core 25 seaward of the trench axis prob- thin-bedded turbidite deposition (425,000 yr
ably represent 'contourite' deposits formed ago) occurred at a depth of 4429 m. If at this
through winnowing by bottom currents. time the trench floor was at its present depth of
Although the sediment bypassed into the trench 5000 m, then the thin-bedded turbidite-bearing
probably creates a substantial terrigenous plume rose 500-600 m above the trench floor.
plume (see below) the absence of sand beds at Within our limits of error, the apparent 500-
Site 487, 7 km seaward of the trench, would 600 m rise of the thin-bedded turbidite-bearing
seem to preclude trench-plume origin for sand plume above its distributary channel is compa-
beds in core 25, 12 km from the trench. rable to the 400 m upper limit of this phe-
nomenon as determined elsewhere (Damuth &
Embley 1979; Shipley 1978).
Effects of the trench sediment plume and the
magnitude of bypassing Alternatively the cessation of thin-bedded
turbidite deposition at Site 488 could be due to
Submarine canyons facilitate sediment bypas- the uplift isolation of this locality on a high
sing of the Middle America Trench slope as a shielded from a down-slope flux of fine sand.
whole (Underwood & Karig 1980). Data from However, this explanation does not account for
the Leg 66 area provide a basis for quantifying the lack of thin-bedded turbidites in the upper-
the bypassing process in a specific locality. most portions of Sites 491 and 492 (Fig. 3). The
Sediment accumulation rates plotted along the absence of sand beds at Site 487, 340 m above
generalized cross-section (Fig. 8) show a max- and 7 km seaward of the trench, apparently
imum in the trench, a minimum in the mid- precludes the extension of the sand-bearing
slope region (Sites 491 and 492) and a secon- plume to that point.
dary high on the shelf. This sediment distribu- By back-tracking at 6.5 cm yr -~ perpendicu-
tion pattern may be simply modelled by a lar to the trench (Minster & Jordan 1979) we
modest sediment-plume that emanates down- deduce that the initial appearance of terri-
slope from the shelf area and a more significant genous detritus 2.5 Ma ago occurred when Site
plume that originates from the inner shelf, 487 was about 160 km seaward of its present
reaches the trench via Ometepec Canyon, and location. For comparison, Horn et al. (1972)
then flows laterally down the trench axis. map the offshore limit of modern hemipelagic
Herein, sediment-plume refers to an influx of sedimentation about 500 km seaward of the
sediment of a particular environment irrespec- trench. The sediment accumulation rate in-
tive of transport process. The trench sediment- creases sharply in the hemipelagic sediment
plume probably is initiated as a sediment about 0.5 Ma ago when Site 487 was approxi-
gravity flow which may be characterized by mately 40 km seaward of the trench axis. Thus,
low-density turbidity currents near its outer we believe that the trench-plume strongly influ-
margins. Bottom currents may assist in re- ences sedimentation within 40 km seaward of
distributing terrigenous sediment once it is in- the trench axis, with secondary effects extend-
 Facies belts of the Middle America Trench 89

Shelf Sediment-Plume
l l l

02 Minimum
Zone

Trench Sediment-
Plume
1 A ,m =

tier

CONTINENTAL
CRUST
CEANIC CRUST

FIG. 9. Schematic representation of facies belts in Leg 66 area with possible controls. Note that facies
boundaries are approximately parallel to bathymetric contours.

ing several hundred kilometres or more larger than that affected by the shelf sediment-
offshore. plume.
The relative influence of the trench versus The volume of sediment accumulated under
the shelf sediment-plume can be established by: the shelf sediment-plume also can be estimated
(1) integrating the sediment accumulation-rate by measuring the cross-sectional area (on Fig.
curve across the continental margin, and (2) 2) of the sediment younger than 10Ma that
measuring the volumes of material accumulated accumulated in the shelf, upper slope, and
beneath the two plumes during the past 10 Ma. middle slope environments. Deposits of trench
The first approach involves determining the and outer slope have been structurally accreted
area under the sediment accumulation-rate beneath the lower slope. Similarly, the long-
curve from the seaward edge of the cross- term contribution of the trench sediment-plume
section to the lower-mid-slope boundary can be estimated from the cross-sectional areas
(trench-plume influence) and comparing this (on Fig. 2) of sediment younger than 10 Ma
value to the area under this curve in the shelf, that accumulated in the lower slope, trench and
upper slope and mid-slope environments (shelf- outer slope environments. The cross-sectional
plume influence). As such, the measured area area of accreted deposits can be approximated
of the sediment accumulation rate curve influ- by a triangle with corners at the base of the
enced by the trench plume is about six times trench slope, Site 492, and at the intersection of
90 J.C. Moore et al.

convergent margins. Hence, facies belts similar

DEPOSITIONAL DEPOSITIONAL
LITHOLOGY ENVI RONMENT PROCESSES to those of the Leg 66 area may occur at a range
O-
Foram-beoring Mid-slope of depths or even be absent elsewhere. For
Mud
CCD --
example, the CCD in the non-equatorial Pacific
Foram-tree Grain by has varied more than 1 km in depth since
Mud
Eocene times (van Andel et al. 1975) and thus
200-
Forom-free probably shifted any boundary between the
Mud with Thin, Lower Slope
Hne Groined foraminiferan-free and foraminiferan-bearing
Sand Beds mud facies along the Middle America Trench
during the past 40 Ma. The oxygen minimum
m 400--
I--
lad
zone may vary in depth as well as intensity, thus
allowing for changes in depth and width of, or
Maesive Fine
to Very Coarse even elimination of the laminated mud facies.
Groined Sand
and Mud Trench Sediment Finally, the bypassing of the trench slope in the
600- Gravity
Flows Leg 66 area constitutes but one of the many
possible clastic sedimentation patterns along
convergent margins (Underwood et al. 1980;
Underwood & Bachman 1981), patterns that
800--
are primarily controlled by sediment supply and
physiography.
FIG. 10. Fining-upward sequence caused
by uplift of seafloor from trench through
lower slope to mid-slope environment. Interpretation of ancient rock
sequences
Predicted outcrop-scale characteristics of facies
the oceanic crust and a plane projected down belts
along the dipping reflectors from Site 492 to the
oceanic crust (Watkins et al. 1981). If sediments One of our principal goals is the articulation
are accreted in the proportion that they enter of a facies model for the interpretation of
the subduction zone, then 90% of this triangle ancient subduction complexes. Our descrip-
is trench or outer slope sediment. As such, we tions may partially accomplish this end; in order
find that the cross-sectional area of post-10 Ma to define further each facies we have outlined
accreted and overlying lower slope sediments is their expected outcrop characteristics (Table
about five times larger than the cross-sectional 1). The proposed outcrop-scale characteristics
area of existing shelf, upper slope and mid- list lithology much as observed although we
slope sediments less than 10 Ma old. Note that have inferred some types of sedimentary struc-
the calculated volume of the accreted trench tures which may have been destroyed by drill-
and outer slope sediments is a minimum since ing or not recorded in a single drill core. We use
no adjustments were made for dewatering dur- Walker's (1979) turbidite facies terminology
ing accretion, nor for subduction of any of this (classical turbidites, massive sandstones, pebbly
material (Watkins et al. 1981). sandstones, etc.) which implies certain
Our two estimates of the magnitude of sedi- sedimentary structures and bedding characteris-
ment bypassing are essentially equal and are tics not listed in Table 1.
probably valid for the cross-section along which While Table 1 lists the essential outcrop
they were taken. The magnitude of bypassing is features of the various facies belts, some discus-
large due to the influence of Ometepec Can- sion of these characteristics is warranted. The
yon. However, the volume of material in the blind application of turbidite fan models to
shelf sediment-plume may also be large relative trench and trench slope settings is clearly
to other areas along the northern Middle Amer- inappropriate; nevertheless, the sedimentary
ica Trench due to the sediment input of Rio lithological associations and vertical sequences
Ometepec near the landward end of the cross- generated by the depositional elements of the
section. fans may also occur within trenches and on
trench slopes. For example, where Ometepec
Secular and regional variation of the controls of Canyon enters the trench the submarine chan-
nels would probably deposit coarse sedimentary
facies distribution
sequences (possibly thinning- and fining-
Controls on the distribution of sedimentary upward) that would be interbedded with finer
facies may vary in both time and space along grained levee and overbank deposits. However,
 TABLE 1.
SEDIMENT
ACCUMULATION
RATE
POSSIBLE OUTCROP CHaRaCTERISTICS
FACIES BELT LITHOLOGY DIAGNOSTIC FOSSIL/TRACE FOSSIL ASSEMBLAGES gm/cm2/lO 3 yr. (Not covered under lithology)

Pelagic Brown pelagic mud with authigenic Radiolarians, sponge spicules, fish teeth, some
phillipsite and lacking terrigenous planktonic foraminiferans near base of section
silt. suggesting water depth n e a r CCD.

Outer Slope Hemipelagic mud with rare thin un- Reworked benthonic foraminiferans from bathyal Erosion surfaces and lag deposits
graded sand or silt beds ("contourites"). environment. ("contourites") associated with re-
working by bottom currents.

Trench Sand Hemipelagic mud and massive sand Reworked benthonic foraminiferans predominantly from 322 Thinning and fining upward sequences
beds, local gravel concentrations, neritic environment, some reworked foraminiferans where axial channel development occurs.
thln-bedded turbidites. from bathyal environment. Thin-bedded turbldites as levee and
overbank deposits. Thickening coarsen-
ing upward sequences associated with
axial lobes. Classical turbidites as
sheet deposits down-current from lobes.
Axial paleocurrent directions except-
ing where submarine canyons build
small fans into trench.

Foramini- Hemipelagic mud, with thin-bedded In-situ qalcareous foraminiferans absent, radiolarians 4-44 Slump folds, olistostromes, slope
feran-free turbidites. and diatoms present, diagnostic trace fossil assem- decreasing sequence of fines upward from contact
mud blage: Chondrites-Planolites-Zo__oophycus-Teichichnus. upslope with trench sand facies. Possible
away from concentrations of turbidites in slope
trench basins, showing predominantly axial
paleocurrents.

Foramini- Dominantly hemipelagic mud with minor Benthonic foraminiferans of middle and lower bathyal 2-10
feran-bear- glauconite, foraminifera conspicuous zones, foraminiferans ooorly preserved n e a r lower"
ing mud relative to radiolarian mud facies, limit of water depth of facies belt due to
rare sand beds, locally graded and/or corrosion below lysocline.
glauconitic.

Laminated Hemipelagic mud commonly with color, Benthonic foraminiferans of upper and upper middle
Mud shell rich, and foram rich lami- bathyal zone, foramini ferans in@Icarive of low
nations, thin sand and silt beds, levels of dissolved oxygen.
locally graded.

Shelf Shelly sand, mud. Articulated and fragmented bivalves, coraline frag- 14-19 Sand both massive and cross-bedded,
ments, calcareous algae, benthonic foraminiferans of graded sand beds with hummocky
outer neritic zone. cross-stratification.

Canyon Sand, gravel, mud. Mud with thin Cbannelization conspicuous, very
sand beds on canyon wall. coarse-grained deposits in contrast
to fine-grained sediments of
adjacent slope and canyon wall.
92 J.C. M o o r e et al.

instead of extending radially out across the muddy portion of the trench sediment plume,
seafloor the channels veer to the NW and die and finally up through the CCD. This vertical
out in the trench axis. Thus, the vertical se- sequence should develop at other convergent
quence produced by the channelized flow might margins with substantial clastic influx and
be similar to that observed in a submarine fan bypassing of sediment to a trench with axial
but the palaeocurrent and lithofacies pattern turbidite flow.
would be different. Palaeocurrents would be
predominantly axial rather than radial as in the
case of a fan. Any transition from channels to Conclusions
depositional lobes would occur not on the fan at
the canyon mouth but down the trench axis to Along the Middle America Trench in the Leg
the NW. Facies associations at variance to those 66 area we discriminate: (1) a pelagic facies, (2)
observed on submarine fans also could develop an outer slope mud facies, (3) a trench sand
on slopes and in slope basins (see Underwood et facies, (4) a foraminiferan-free mud facies on
al. 1980 for discussion). the lower slope, (5) a foraminiferan-bearing
facies on the mid-slope, (6) a laminated mud
The trench slope transition: a diagnostic vertical facies on the upper slope, (7) a shelf sand
sequence facies, and (8) a canyon facies. During accre-
tion superposition of the mid-slope, lower
Vertical stratigraphic sequences provide a slope, and trench deposits results in a fining-
critical tool for interpretation of ancient upward sequence. The boundary between the
sedimentary deposits. Indeed, a classic fining- foraminiferan-free and foraminiferan-bearing
upward transgressive sequence occurs at Sites muds is defined by the CCD. Similarly, the
489 and 493 in the upper and mid-slope region oxygen mimimum zone controls the extent of
of the Leg 66 area. A n o t h e r kind of fining- the laminated muds on the upper slope. In the
upward sequence generated by the superposi- Leg 66 area, a prominent submarine canyon
tion of lower slope and trench deposits consti- causes coarse sediment to bypass the outer shelf
tutes a vertical succession of particular interest and slope. The volume of sediment bypassed is
to geologists investigating convergent margins. about five times that transported directly across
The vertical succession recording the shift the slope, which accounts for the high sediment
from a trench to slope environment consists of accumulation rates in and adjacent to the
coarse sand and mud topped by thin-bedded trench. Bypassed sediment also appears to in-
turbidites, covered by mud barren of forami- fluence hemipelagic sedimentation several hun-
niferans, and finally succeeded by foram- dred kilometres seaward of the trench.
bearing mud. Seismic profiles across the mod-
ern trench suggest that the basal sandy unit ACKNOWLEDGMENTS: We acknowledge the U.S.
should not exceed 500 to 700 m in thickness. National Science Foundation and similar agencies in
The thin-bedded turbidite unit ranges up to France, Great Britain, Japan, West Germany, and
several hundred metres thick in the Leg 66 the U.S.S.R. for support of the International Prog-
area, whereas the superjacent mud unit is a ram of Ocean Drilling. The marine crew, drilling
crew and technical staff of the Glomar Challenger
maximum of 172 m thick. were exceptionally cooperative in the collection and
This fining-upward sequence apparently shipboard examination of the drill cores on Leg 66.
accumulates as the accretionary process moves The University of California at Santa Cruz supported
a vertical section up off the trench floor, the preparation of the manuscript. We thank Roland
through the lower part of the sediment-plume von Huene and Rick Stanley for reviews of various
(that deposits thin sand beds), then through the drafts of this manuscript.

References
BACHMAN,S. B. & LEGGEIT,J. K. 1981. Petrology of eastern Mexican volcanic rocks--relations be-
Middle America Trench and trench-slope sands, tween the andesitic and alkaline provinces. J.
Guerrero margin, Mexico: results from Leg 66 Volcan. Geotherm. Res. 5, 99-114.
DSDP. In: WATK1NS, J. S., MOORE,J. C. et al. DAMUTn, J. E. & EMBLEY,R. W. 1979. Upslope flow
(eds). Init. Rep. Deep Sea drill. Proj. 66. U.S. of turbidity currents on the northwest flank of
Government Printing Office. Washington, in Ceara Rise: western Equatorial Atlantic.
press. Sedimentology, 26, 825-34.
CANTAGREL,J. M. • ROBIN, C. 1979. K-Ar dating on ENKEBOLL, R. H. 1981. Sedimentary petrology and
 Facies belts of the Middle America Trench 93
provenance of sands and gravels from the Middle 117-89. Dowden, Hutchinson & Ross, Strouds-
America Trench and trench slope. J. sediment. burg, Pennsylvania.
Petrol. in press, PAGE, B. M. 1981. Southern coast ranges. In: ERNST,
HORN, D. R., HORN, B. M. & DELACH, M. N. 1972. W. G. (ed.). Geotectonic Development in Cali-
Ferromanganese deposits of the North Pacific. fornia. Prentice-Hall, New York.
Techn. Rep. No. 1 NSF GX-33616. Office of the Ross, D. A. 1971. Sediments of the Northern Middle
International Decade of Ocean Exploration, America Trench. Bull. geol. Soc. Am. 82, 303-
National Science Foundation, Washington, 78 22.
PP. SCHWELLER, W. J. & KULM, L. D. 1978. Depositional
INGLE, J. C., JR 1975. Paleobathymetric analyses of patterns and channelized sedimentation in active
sedimentary basins. In: DICKINSON,W. R. (ed.). eastern Pacific trenches, In: STANLEY, D. J. &
Current Concepts of Depositional Systems with KELLING, G. (eds). Sedimentation in Submarine
Applications to Petroleum Geology. 11-1-12. Canyons, Fans, and Trenches. 311-24. Dowden,
San Joaquin Geol. Soc., Bakersfield, Hutchinson & Ross, Stroudsburg, Penn-
California. sylvania.
KAR1G, D. E., CARDWELL,R. K., MOORE, G. F. & SHIPLEY,T. H. 1978. Sedimentation and echo charac-
MOORE, D. G. 1978. Late Cenozoic subduction teristics in the abyssal hills of the west-central
and continental margin truncation along the North Atlantic. Bull. geol. Soc. Am. 89, 397-
northern Middle American Trench. Bull. geol. 408.
Soc. Am. 89, 265-76. - 1981. Regional geology of the southern Mexico
-

LEC6ETr, J. K. 1981. Geochemistry of Cocos Plate continental margin. In: WATKINS,J. S., MOORE,
pelagic and hemipelagic sediments, DSDP Site J. C. et al. (eds). Initial Rep. Deep Sea drill. Proj.
487, Leg 66. In: WATKINS,J. S., MOORE, J. C., et 66. U.S. Government Printing Office, Washing-
al. (eds). Initial Rep. Deep Sea drill. Proj. 66. ton, in press.
U.S. Government Printing Office, Washington, - MCMILLEN, K. J., WATKINS, J. S., MOORE, J.
-

in press. C., SANDOVAL-OcHoA, J. H. & WORZEL, J. L.

MCMILLEN, K. J. & BACHMAN, S. B. 1981. 1980. Continental margin of Guerrero and
Paleobathymetry and tectonics of the trench Oaxaca, Mexico. Mar. Geol. 35, 65-82.
slope, Leg 66, DSDP. In: WATKINS, J. S., UNDERWOOD, M. B. & BACHMAN, S. B. 1981.
MOORE, J. C. et al. (eds). Initial Rep. Deep Sea Sedimentary facies associations within subduc-
drill, Proj. 66. U.S. Government Printing Office, tion complexes (this volume).
Washington, in press. , & SCHWELLER, W. J. 1980. Sedimentary pro-
- - , ENKEBOLL, R. H., MOORE, J. C., SHIPLEY, T. cesses and facies associations within trench and
H. & LADD, J. W. 1981. Sedimentation in diffe- trench slope settings. In: FIELD, M. E., BOUMA,
rent tectonic environments of the Middle Amer- A. H. & COULBURN, I. (eds). Quaternary De-
ica Trench, Southern Mexico and Guatemala positional Environments on the Pacific Continen-
(this volume). tal Margin. Pacif. Sec. Soc. Econ. Palaeontol.
- - & HAINES, T. R. 1981. Late Quaternary sedi- Mineral. Tulsa.
ments of the southern Mexico margin. In: - & KARIG, D. E. 1980. Role of submarine can-
-

WATKINS, J. S., MOORE, J. C., et al. (eds). Initial yons in trench-slope sedimentation. Geology, 8,
Rep. Deep Sea drill, Proj. 66. U.S. Government 432-6.
Printing Office, Washington, in press. VAN ANDEL, T. H., HEATH, G. R. & MOORE, T. C., JR
MCRAE, S. G. 1972. Glauconite. Earth Sci. Rev. 8, 1975. Cenozoic history and paleoceanography of
397-440. the central Equatorial Pacific Ocean. Mem. geol.
MEJORADA, S. H. S. 1976. Carta Geologica de la Soc. Am. 143, 134 pp.
Republica Mexicana, 4th ed. Instituto de Geolo- VON HUENE, R. 1974. Modern trench sediments. In:
gia U.N.A.M. Ciudad Universitaria. BURK, C. A. & DRAKE, C. L. (eds). The Geology
MINSTER, J. B. & JORDAN, T. H. 1978. Present-day of Continental Margins. 207-11. Springer-
plate motions. J. geophys. Res. 83, 5331-54. Verlag, New York.
MOORE, J. C. 1979. Variation in strain and strain rate VON HUENE, R., AUBOUIN, J., AZEMA, G.,
during underthrusting of trench deposits. Geo- BLACKINGTON, G., CARTER, J., COULBOURN, W.,
logy, 7, 185-8. COWAN, D., CURIALE, J., DENGO, C., FAAS, R.,
- - , WATKINS,J. C., SHIPLEY, T. H., BACHMAN, S. HARRISON, W., HESSE, R., HUSSONG, D., LADD,
B., BEGHTEL, F. W., BUTT, A., DIDYK, B. M., J., MUZYLOV, N., SHIKI, T., THOMPSON, P &
LEGGE'rT, J. K., LUNDBERG, N., MCM1LLEN, K. WESTBERG, J. 1981. Leg 67 of the DSDP Mid-
J., NIITSUMA,N., SHEPHARD,L. E., STEPHAN,J. F. American Transect off Guatemala. Bull. geol.
& STRADNER, H. 1979. Progressive accretion in Soc. Am. in press.
the Middle America Trench, Southern Mexico. WALKER, R. G. 1979. Turbidites and associated
Nature, 281,638-42. coarse clastic deposits. In: WALKER,R. G. (ed).
NELSON, C. H., NORMARK,W. R., BOUMA, A. H. & Facies Models, 1, 91-103. Geoscience Canada
CARLSON, P. R. 1978. Thin-bedded turbidites Reprint Set.
in modern submarine canyons and fans. In: WATKINS, J. S., MOORE, J. C. et al. 1981. Initial Rep.
STANLEY, D. J. & KELLING, G. (eds). Sedimenta- Deep Sea drill. Proj. 66. U.S. Government Print-
tion in Submarine Canyons, Fans, and Trenches, ing Office, Washington, in press.
94 J.C. M o o r e et al.

WATKINS, J. S., MCM~LLEN,K. J. & BACHMAN,S.B. (eds). Initial Rep. Deep Sea drill. Prof. 66, U.S.
1981. Tectonic evolution of the Middle America Government Printing Office, Washington, in
Trench. In: WATKINS,J. S., MOORE, J. C. et al. press.

J. CASEY MOORE, Earth Sciences Board, University of California, Santa Cruz,

California 95064, U.S.A.
JOEL S. WATK1NS,Gulf Research and Development Company, Pittsburg, Pennsyl-
vania, U.S.A.
KENNETH J. MCMILLEN, Gulf Research and Development Co., P.O. Box 2038,
Pittsburgh, Pennsylvania 15230, U.S.A.
S'I'EPHEN B. BACttMAN, Department of Geological Sciences, Cornell University,
Ithaca, New York, U.S.A.
JEREMV K. LI~GGEEr, Department of Geology, Impcrial College of Science and
Technology, London, England.
NEH, LUNDBI-RC;, Earth Sciences Board, University of California, Santa Cruz,
California, U.S.A.
TrIOMAS H. SmPt,EV, Scripps Institution of Oceanography, La Jolla, California,
U.S.A.
JEAN-FRANCOm STEPHAN, D6partement de G6otcctonique, Univcrsit6 Pierre et
Marie Curie, Paris, France.
Fl~oYl) W. BF(iltTt'I,, Phillips Research Center, Bartlcsville, Oklahoma, U.S.A.
ARIV Burr, Univcrsit~it Tubingcn, Tubingen, Federal Republic of Germany.
BORVS M. DIDYK, Empresa Nacional del Petroleo Concon, Chile.
NO~UAKI NnrSUMA, Department of Gcoscicnccs, Shizuoka University, Shizuoka,
Japan.
Ll~s E. SHI-PHARD,Texas A & M University, College Station, Texas, U.S.A.
HERBE-RTSrRADNt':R, Gcologischc Bundesanstalt, Vienna, Austria.
Tectonic processes along the Middle America Trench inner slope

T. H. Shipley, J. W. Ladd, R. T. Buffier & J. S. Watkins

SUMMARY: The relationship between sedimentation and structural variations observed
in geophysical data provides insight into tectonic processes on the convergent Middle
America Trench margin. We observe that some of the hemipelagic and pelagic oceanic
sediments seem to continue relatively undeformed beneath the lowermost slope. Trench
sediment bodies restricted to ponds in the vicinity of submarine canyons are incorporated
into the lower slope by folding and thrust faulting. This offscraping (shallow accretion) at
the base of the Middle America Trench slope appears at least partly dependent on the
presence of trench fill. The internal structure of the so-called accretionary zone, the top of
which is usually identified on seismic reflection profiles by a prominent zone of
diffractions, is rarely resolved in the seismic reflection data. The shallow part of the
accretionary zone along much of the margin may consist of deformed slope sediments
since down-section increase in both folding and other deformation are observed in the
seismic reflection and drill data. Beneath the undeformed and deformed slope sediments,
offscraped trench fill may occur if trench fill was present in the ancient trench. The deeper
part of the accretionary zone representing the bulk of the volume is not resolved in the
reflection data but may consist of other sediments or crustal rocks added by an
underplating process at deeper structural levels.

Evolution of ancient convergent continental margin, necessary for understanding the de-
margins is difficult to confirm in modern ana- velopment of the accretionary zone, has been
logues. Exposures of ancient rocks provide difficult to decipher. Spreading centre re-
most of our information on evolutionary pro- orientation in the Miocene has complicated
cesses. Recent seismic reflection and refraction magnetic anomaly mapping and makes recon-
surveys of modern margins, hampered by the struction of earlier history difficult with the
local variability and abrupt transitions of major present data base (Lynn & Lewis 1976; Schilt &
structures, are interpreted as evidence for sedi- Truchan 1976). Magnetic anomaly interpreta-
ment subduction, sediment and/or oceanic crust tions predict that Miocene ocean crust is now
accretion or transitional (or steady-state) entering the trench with present plate motions
regimes. Every new investigation of a modern producing slightly oblique convergence. The
convergent margin reveals new variations. convergence rate increases southward from ab-
Perhaps one reason we have not been successful out 7 cm yr -1 off Mexico to more than
in completely understanding the variations be- 10 cm yr -~ off Costa Rica (Minster & Jordan
tween and within a single modern arc system is 1978). The modern trench is bounded by trans-
the difficulty in separating the accretionary verse oceanic plate structural features; the
processes from local complexities. Cocos Ridge on the south near Costa Rica and
A n investigation was designed for the mod- the Tamayo Fracture Zone off central Mexico.
ern continental margin adjacent to the Middle The University of Texas Marine Science In-
America Trench to address some of these prob- stitute undertook an investigation of the Middle
lems. This margin is suited for a comparative American margin which focused on three speci-
study because of differences in the gross con- fic areas for detailed study and included mul-
tinental margin structure, physiography and tichannel seismic reflection, refraction, coring,
sedimentation pattern along the arc (Fig. 1). A dredging, magnetics, bathymetry and sparker
narrow trench-to-shoreline distance north of the surveys. Each of the areas off Mexico, Guate-
North American-Caribbean plate boundary mala and Costa Rica are topics of other inves-
(near the intersection of the Tehuantepec tigations detailing the geology (Buffler & Wat-
Ridge trend with the shoreline in Fig. 1) con- kins 1978, 1981; Ibrahim et al. 1979; Ladd et al.
trasts with a wide trench-to-shoreline distance to 1978; McMillen 1980; McMillen et al. 1981;
the south (Molnar & Sykes 1969) while a wide Shipley et al. 1980; Shipley 1981). Deep sea
and deep forearc basin extends along the drilling was conducted along transects in the
shoreline from southern Mexico south to about Mexico and Guatemala areas based on these
the Nicoya Peninsula of Costa Rica. studies (Moore et al. 1979a, b; yon H u e n e et al.
However, the plate motion history of this 1980). This paper is a summary of these studies

95
96 T.H. S h i p l e y et al.

CR-7 "::':"".::

~ Oceanicor transitionalcrust(;') /
~ Continentalcrust / / /

500 km /
/ / / / c p v ¢.,d'/ /~b~/ / /

FI6. 1. Index map showing location of profiles in subsequent figures. Continental crust extends
offshore north of the Gulf of Tehuantepec, oceanic or transitional crust inferred to the SE.

and attempts to document the significant margin structure over the 150 km segment sur-
structural variations and their causes. A single veyed (Figs 2 & 5). Within this area the oceanic
multichannel seismic reflection section from crust is overlain by pelagic and hemipelagic
each area is presented in Figs 2-4. oceanic sediments, generally not more than
170 m thick. Recent normal faults, which most-
ly step down to the trench, offset the oceanic
crust and most of the overlying sediments. The
Structural styles trench contains turbidite ponds only near two
Mexico submarine canyons that are incised into and
traverse the shelf. The slope section consists of
The narrow central Mexico shelf and trunca- numerous diffractions and short discontinuous
tion of Mesozoic and older terranes is evidence reflections, while a zone of diffractions 0.5-1.0
for an unusually abrupt continental/oceanic sec sub-bottom in Fig. 2 defines the top of a
transition formed post-Cretaceous and pre-mid- slightly higher velocity section (the so-called
Miocene. Exposed Cretaceous batholiths near accretionary zone). Farther to the NW
the coast attest to an earlier arc which from near the Rio Ometepec Canyon, landward-
several lines of evidence was probably trun- dipping reflections generally 2-5 km in length
cated in the early Miocene although the down dip, occur in the section below the dif-
mechanisms responsible for this feature are fracting surface. Drilling data show they prob-
difficult to evaluate (e.g. erosion, oblique sub- ably represent dipping interbedded sands and
duction) (Karig 1974a; Karig et al. 1978; Kesler mudstones, presumably uplifted from the
1973; Moore et al. 1979a; Schilt & Truchan trench during the accretionary process (Fig. 6)
1976). Studies of the modern Mexico margin (Moore et al. 1979b). Dips of the presumed
reveal recent sediment accretion at the base of bedding planes reach high values (>8 °) within
slope (Karig et al. 1978; Moore et al. 1979a, b; 4 km of the trench and remain high and vari-
Shipley et al. 1980). able farther up slope (Fig. 6). Landward of
The Mexican area reveals little variation in 1640Z (Fig. 2) is an upper slope sediment
 9 MAY 1977
SW 1000 0900 0800 0700 0100 0000 FOREARC BASIN NE
0600 0500 0400 0300 0200

:,1
GUA14!

1,,
,
Ht tt 1f' '' ,,,
TOP OF OCEANIC!CRUSTAL SLICE
,,
.

LOPE SEDIMENTS
3 - ~

oo O3
K] £3
Z Z
O O
~!ii1, i!ii h/ii ii;il ~ til!l~ (J
ILl UJ
O0 ~i~I~11tII II II~I~IINIII~lh!l! lllllll O3
'~5
.,.

i~!!i;;i,,!ii!J~,; i!l
Ijj~i!~l!l,lli~'~''...... '
TOP
I
OF ' ACCRETIONARY
•1 ' I' '1 ' i • liill I ' III ~,LII '
ZONE
' ~'

IC SECTION
":,!~l!l]t!.
iiiiii}iiiiillli';iiliiiii!i!rrt{fi,ll.lll. Ihhll~=...~tlttl,rI ~hl..I,,l}rlll,ltlltrr~tlll}l..~jll~.~tlt';:i~i~
.......i q

I ,#11 II II1 II I.JJ II Illt,lrl,lf111 ,, * , , ~ , ,.'~ ..~ "~', "-i";~

,;" . ! '; . ,,, !, ,,,,,,. .,

~ """ ,~ ~ " • "" ~.-,'.'~',';-~"~?-".~'~,'j'~°:%~ ,,~'.'~''....~,~'1.'~'~"2~.~ "~'~:- ~'~'.~i"...'~.~"],":. ~.:.'Z'-?';'r'~:~-~.~,~-''"~.~ ..;~'~:..:"~;',~..~"~-'.~,'~'i~'.".-';~'? "-~" • ~.~.'-:"';~"~. ~'.~,'7" -~"~.":":~"~-~ -",".~"~'~ ~-" 'F..~T4';.~'.~,"-,?~:-~4~" '.- ,1":~,~ -''~ .~.'-."Z-'t.~,","~'..~.~'-.~:- : '-",'/'.-'~"i-'~,~ ~'k : ;'.".~','.~.~,.'~,.'~'~"~':~'~ :;"'~..'~ ,-,,,i'.~ ..~. ~.~'. "',~;" ~"~'~" :"~;-'." " "".'.','~'~;~:~.,: ~-~ "~-':~ ~;~"~'"~ ~ ~ ~. .... ~ ., ~4... ,, .,. . ,.,.,;~ ..... ~.. . . . . . .,~ , .~.~.. .,~ ~ . . ,~.- ~.,~.~,. ~,- ...,~,,., .,r~, .,,,.,..~ ~ ,~., ~" . . . . . . ' ' ,~ ' ' ~ ' ' ,, ' ~ ' '! . !'I "" ' I ' 'I~!I . . . . . i . . . . . . . I'"I' i ;,'~"~' "ir~i[i
 11 JUNE 1977
SW 2300 2200 o91 A N O~N~ 1 I~AA 1500 1400 NE

O9 U3
121
Z Z
o 4 4O
w IJJ
of)
 •ms!u~qoom ~u!l~ldaopun olqissod
'(8L6I) ~U!ll!S :~ u ~ o o ]o ,sou!i ~OlJ, Su!lo!dop s ~ o a a v "dip u ~ o p qlpI~, u! tu~I ff lnoqe ouoz ~ ougop
p l n o ~ suo.lpoljoa osoq) 'q3uoa) oql ol pol~odsueal oao~ luom!p~s jo somnlOA o$a~ 1 oaoqm (9 "~[.A)
o3.[x~I~ ]o sla~d JJO "ll!J q3uoal pod~a3sjjo 3!l~tuoq3s lu~soadoa suo.il3oljoa ~u.tdd!p-pae~pu~ I laoqs
"([96I) aoqs!3 ~) aoqs pu~ (gL6I) "1v l a £ouooIAI '(6L6[) "Iv Ia m!q~aqi ~o ~l~p uo!13~ajoa pu~ '(086[)
•I v l a ouon H UOA '(6L6[) £I°°S '(~6L6I) "1v Ia oaooIN jo ~l~p ~U!ll!ap 'SOl!load luo3e[p~ moaj po3npop
oanl3nals "L-~ID pu~ ~I-VrlO 'NE-IAIO suo!pos 3~tus~os ~UOl~ suop3os qldop OA!lOadaolUI 'ff "O~:l
AAS
/
I v ^ ~ v>~^ < ^< 4 "~ r~ 4 "r ]A L:9E I
l.,~-<.., <,. ,.v 4 < """<^"r <F ^,~ <-~ ~oL /
/> ,a " " L ^~ ^ ^ ~ ^ ¢ v . <~ < 4, L-U3 /
,c ^ v 4~.? <'~ <'IV ~" ~< ~ v^~ v c ~. ~> 4c < ;,-
VAODIN 3
&" t. '7 /*- #.. ) v ~ A A ~ ~" r" v ~ v A ? ~ K" V
.< ,- I < L -~ v
t~t
~ ~ .v ~ v r " L '~ ~ v > . , • '- v , . ~ ' v " "?-, " v ~ t,L-vno
~ .~^v>~>^ ~14~v ~ >v4 ^~ ~L
3L
I r~ u J' ~. 4 ,c > t.. ¢ O01 ;'.I ~ < ,4
a ~ ~. r ~. ^. a ^ 4 v ( " - v v v • r 1
/
~ a l e , l p A q se8 ~.o aseq 8
3N AAS
~L~ v ~
t~t
~. ,- < .,"., ..., • ,.." < "~'OT
~. ~~r .'~
~ ~-> ~~'~- ^r a~ u~ v ~
> 7 N~IAIO
~t
.~ < ~v ~ ~ ~ ~ A
". 4 ~" > < "~ ~ > ~ ~ ~ ~ ~ ~ V<
Ot
"~ • • V ~ V ~" 4 1A < ,i
9
~ ( ; "- "J C.';Z , ' , " - ' ) 0 C\ , , ~, " ~ [ - ~ ~
-..,-,v . . . . . .',-..-,. \ "--" : ;~-L. -~ ~,xv"....±
,-~.', ,,.-,-':,,.,,.m,,'. ~ ~ ~ z -.-~--~..~f g
L6 tlOUa~tI v:)!~tatu,¢"alpp~.lN atll @UOlV ~a~a:)o~td s~.uol)~,Z
98 T.H. Shipley et al.
2-
• 492
: ::MXle: ::: :
........... ....................................................................... :
.. . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. .
J, . ~

3 . . . . . . . . . . . . . . . .] .............. . [.~.[ ...... ~]... ........................................................................................................... ]. ......... , ....................... ~ ;

•~ ~: _~ . ~ , ~.T.~.-2~ .~ ~

.................................................................................................................................
~ ~ ~ / ~ .
;¢, ~.--~ ;m - ~ 4 ~ . . ;~ *:~.....~, ". -.. ~

~ ~

488 ~ : ~ ~ ~ ~ ~ e . ~ ~=, < - ~ :

V :~&~;~.~ ~;27~ 6 " . ~ " ~ ....... ;;=~: ........... ~ . . . . o~ ,

. ,;~" z:. , ;.~,,'~f'~,z'.~ x ~ ~,,..:','.~ --~.~,, ",.,,- ""::r'~"2r :,r.~_"-,~-~:,'..".i.:~:.e'-- f'..- ,='-,:- £':.,:,'-., ...-. :-:. ~.?".,.:~:-:.: " %.. ...:,r..~ ', 4....~'~ ..'..'~
.....:.:<- ~ r ~ ~:. r: ~:L~ :~< ~; ~ ;y;;~z:,~; ;/: ::~7~" ".= ,

~ 2 ~;, ~ ~ ~ ~ &.~Y2~<~;L;~;~ < ~ ;~T L~::z::~:~ L• =~; T :

Fl~. 6. Portion of migrated seismic reflection profile MX-16 near the Rio Ometepec Canyon. Three
DSDP drill sites are shown. Note landward (NE) dipping reflections which are correlated with
uplifted turbidites. Vertical exaggeration 5 x.

package of variable thickness overlying (at 1 sec and lack lateral continuity of more than 5 or
at 1440Z) the top of the continental crust which 10 km. Many small channels originating in the
was confirmed from refraction data and by upper or middle stope region are blocked by
correlation to dredging and drilling in the these ridge and basin structures. The zone of
northern part of the survey area. The shoreline most bathymetric relief or roughness, which
is less than 3 km from the end of the section. may be related to more intense deformation, is
Large shelf sedimentary basins are uncommon restricted largely to a width of less than 10 km
north of the Gulf of Tehuantepec and south of with structures becoming more subdued be-
the Tamayo Fracture Zone. In the survey area neath the slope sediments. In most profiles
shelf basins have widths up to 20 km and con- (Fig. 2) there are no discernible internal struc-
tain a maximum of 1 km of early Miocene or tures within the base of slope region and the
younger sediments (Moore et al. 1979a; Shipley slope sedimentary section is not seismically
1981). Similar basins are reported and observed defined. This is repeated along most of the
in other reflection data to the north (Karig et al. profiles in the area except near the Rio
1978; Ross & Shor 1965) and in unpublished Ometepec Canyon. This is the same area where
University of Texas data to the south of the significant trench turbidite deposits are found.
survey area. The presence of a large submarine canyon on
Refraction profiles discovered no high veloci- the slope within the survey area reveals both
ties which might suggest accreted slices of the influence of sedimentation parameters on
oceanic crust within the slope (Shipley et al. structural development and the ability to iden-
1980). The area between the seaward edge of tify such structures in seismic data. Profiles
continental crust 20-40 km offshore, and the normal to and along the base of slope (Fig. 7)
base of the trench inner wall defines an accre- reveal changes in degree of deformation and
tionary zone less than 50 km wide (Fig. 5a). structural style over short distances. In the
There is no evidence for large downfaulted vicinity of the Rio Ometepec (for a zone of
continental blocks, and on some seismic reflec- about 50 of the 150 km of margin surveyed) the
tion profiles the top of the oceanic crust extends lower slope departs from a fairly consistent
to within 10 km of the edge of continental crust slope angle to abrupt scarps and trench sub-
at a depth of 8 km (Shipley et aI. 1980; Shipley parallel ridges. This is also the only area where
1981). the internal structure is delineated by turbidite-
Small accretionary slope basins and ridges related reflections which reveal anticlines, sync-
are generally present within 20 km of the trench lines and thrust faults (Fig. 7).
 Tectonic processes along the Middle America Trench 99

-- OMTN .... ~!~ , |

:5: 2 k i n _ .,,

• - . . . . . . . . . . . . . . . .. : : ~ / , - ~ ; . - - ~ -.-

- - ~ ~

. i ~-~:~,.'~G',.,: ........... ,..~.

- . ~ - ~ : - . : ,.<~..~ ::% ~<,~ :::.:.~ ~:.:~,,- ::
...:.,~.~-.~,,~... .... ..-...., .:..---.:-.,.,: : ~ % . : , ,.;,., . . ~ . , ~:-~.~:;~....:..~ ..'.,....~-~.~
• ,-f~:,.,;.~.,~,~_~;- .,; . : . : : , ~ . .... - :. ~ , : ...:,. -; -.~ :;:;

........... . ~.;,...~-"
-"~'~.%.~

, ~:,: ~ ~ .. :. ;:;:,.'~:i~.i"-'.~ ~-~,,~:".::".:':~:

'"~.~,!~',"~" . I ...... :,:"":"~.~'~.;:'.~".';'~:,"i~*"&-'~%';~¢,~"~'.~:~-~ ~"

•.;.~~
-'..~ ~ : ~:~-~.~:~ ~:.~.-:~~~ : ~ 2 ~ .:-"~5..~:
~:..,..'.',:~.~-...
. ]~.
'.7'..'~:~,-.;
~.-.,'~..;::.:
~. .'::.:-~....:-:.~.~:~
.-..... ;
:..~ ;,,~,.,: '; '..':,.'-~~;..~:::~,i.:-':':,:~,~L.'.-,,
. ..~,:...,.:. ~-,'-.,..:.~;.,~,.., ,..~.-

. . .. ~.....,-,::,~.:...,,:.;.,.,:-'--~...:..~ ...... ~ . - , ~ ~".... ..... . .,.

• . :.. :;.... ..-..~,: • ., .~,'..,,,.:,,:-..~....~.~.:..: ~,.y-:.- .~:~,e-~.~... -... '_ ~:~

..~,
.... ~ ~, a'~ ~ . .~ , . e .,-'-~
. . • :• ~ ~ ,.N. :','~"M:. ~*.

oM4N

~ :~'~:"~-~'~'-"

FiG. 7. Migrated 24-fold seismic reflection profiles normal to the trench axis in the vicinity of the Rio
Ometepec Submarine Canyon of Mexico. The distance between adjacent profiles OM-7N, MX-16,
MX-15, OM-6N and OM-4N is, respectively, 8, 8, 18 and 34 km. The Rio Ometepec Canyon reaches
the trench floor between profiles MX-15 and OM-6N. Most turbidity currents flow to the NW
(from MX-15 to MX-16 to OM-7N) and a higher portion of sand is expected in that direction.
OM-6N contains less turbidite fill while OM-4N contains what is interpreted as a slump deposit. The
oceanic section of pelagic and hemipelagic sediments are seismically distinguishable beneath the
trench fill and seaward (left) on most of the profiles. Note the changes in slope and internal
reflections, a result of deformation and incorporation of trench fill into the lower slope. Arrow notes
trench axial channel or deepest point on crossing where channel is absent. Compare with Fig. 2, a
profile approximately 60 km to the SE of the Canyon.
100 T. H. S h i p l e y et al.

While the sedimentological aspects of the fined to the Rio Ometepec region and that the
trench fill are not well known, seismically signi- supply to the trench has varied only slightly
ficant differences in the reflection response since the late Miocene.
indicate significant changes in grain size or While we do not deny that the lack of long
bedding thickness over short distances. For (>2 km) dipping reflections in the slope may
example, the difference between the acoustic result from a lack of impedance contrasts in
response of the trench fill on OM-7N and muddy turbidites and slumps, unless the stress/
OM-6N are striking (Fig. 7). OM-7N is 25 km strain regime changes markedly over tens of
down-axis (NW) from the Rio Ometepec entry kilometres along the trench, there is no reason
point into the trench while OM-6N is up-axis to expect that we would not have detected
(SE) about 8 km (Shipley 1981). OM-6N trench ancient trench sand bodies elsewhere in the
fill may consist of either coarse debris flows slope. This observation suggests long-term loca-
related to the canyon or finer-grained deposi- lization of the Ometepec submarine canyon,
tion from slumping. In either case the physical though the actual locus of deposition off the
properties are significantly different from the Rio Ometepec may have shifted back and forth
coarse turbidites being deposited south of the up to 10 km depending on incoming plate relief
canyon entry point. The production of gentle to and sediment and convergence rates. In the
tight folds and high angle to low angle thrusts modern trench only a small part of the exten-
might be explained by bedding competency sively surveyed area contains turbidite ponds
related to the sediment variations in the trench. thicker than 200 m or wider than 3 km, while
They do not seem related to an evolution of most of the trench extending from Acapulco all
deformational styles with time since all are the way to Costa Rica apparently contains very
presumably of nearly the same age. limited turbidite ponds based on a nearly con-
The variability of the lower slope structures tinuous University of Texas seismic line down
and shallow accretion at the base of the slope the trench axis. We observe evidence of folds
seems related to the presence of trench fill. The near the base of slope only in regions with
sedimentological and reflection data are not significant sediment ponds. We do not observe
dense enough to adequately determine whether these structures to the SE in Mexico. We do
the type of trench fill (e.g. sandy turbidites, not deny that offscraping may occur in the
muddy turbidites or slump material) or vari- absence of significant trench fill but it may be
ations in the oceanic sediment physical prop- volumetrically of little consequence here.
erties influence the process or details of defor- Within this survey area the shallow
mational style. However, the presence of these offscraped section apparently represents only a
structures are striking evidence for incorpora- small portion of the accretionary zone. This is
tion of the trench sediments into the slope also supported by drilling on the Mexico slope
(accretion). which recovered slope deposits of early
More than half of the existing profiles in the Pliocene age within a few hundred metres sub-
Mexico area show no base of slope bathymetric bottom directly overlying the accretionary
irregularities, no significant trench fill and thus zone. A significant thickness of accretionary
little evidence for modern offscraping and shal- material (several kilometres) is inferred below
low accretion of trench fill. Where significant the slope deposits and beneath the zone of
trench sediments are not present in the modern landward-dipping reflections here (Fig. 5).
trench much of the oceanic sediments seem to
continue relatively undeformed at least initially Guatemala
beneath the slope (Fig. 2; Shipley 1981 figs 8 &
12). This implies that even in the absence of Investigations including the Guatemalan
trench fill, offscraping (shallow accretion) of margin led Seely et al. (1974) to development of
the pelagic and hemipelagic oceanic section is an imbricate thrust model for lower slope accre-
slow or inactive, or at least not volumetrically a tion of sediments and oceanic crust. Dickinson
very important process here. & Seely (1979) and Seely (1979) have gone on
Preserved landward-dipping reflections with- to develop a model for the entire slope and
in the accretionary zone correlating with up- forearc structure off Guatemala. Studies by
lifted sands and muds suggest that ancient Ladd et al. (1978) and Ibrahim et al. (1979) with
offscraping has been occurring near the Rio additional seismic reflection and refraction data
Ometepec since at least late Miocene (Fig. 6) confirmed the existence of oceanic crust at
(Moore et al. 1979b). The absence of landward- shallow depths within the slope. Recent drilling
dipping reflections elsewhere on the margin data at the base of slope do not support a
implies that ancient offscraping has been con- presently active imbricate thrust regime but
 Tectonic processes along the Middle America Trench 101

reveal significant subduction of sediments (von reached but is presumed to be transitional

Huene et al. 1980). ocean crust or Nicoya-type basement based on
Figure 3 is the seaward portion of a margin velocity characteristics (Ibrahim et al. 1979;
profile off Guatemala excluding the substantial Seely 1979).
forearc basin to the NE. Normal faults on the The continental slope off Guatemala is
oceanic crust, usually separated by 3-5 km with smoother than that off Mexico but small-scale
a throw of about 75 m down to the trench, benches and ridges are common within 20 km
appear to offset recent sediments. Horst-graben of the trench (McMillen 1981; McMillen et al.
structures are rare near the trench. The pelagic 1981). Locally, near the San Jose Canyon,
and hemipelagic oceanic section of Miocene, benches extend farther upslope (Ibrahim et al.
and younger sediments are about 450 m thick 1979, fig. 7A). Reflection data reveal no
(von Huene et al. 1980). The muddy turbidite evidence of internal structure at the base of
trench fill is notably restricted to the mouth of a slope, where the slope sediment cover is thinner
canyon system and even then with less than and more discontinuous than off Mexico. While
150 m of fill in localized axial basins (McMillen a single seismic line in this vicinity was inter-
1981; McMillen et al. 1981). The oceanic crust is preted by Seely et al. (1974) as containing
not traceable landward of the trench in Fig. 3 landward-dipping reflections near the base of
but other profiles in the area reveal crust 18 km slope, none of our extensive survey data readily
from base of slope. support this observation.
The slope region, from the trench to about The lack of reflections (and thus definition of
0500Z (Fig. 3) is characterized by a thin slope lower slope structure) could result from an
blanket of low reflectivity sediments of variable absence of impedance mismatches. However,
thickness underlain by a strong zone of diffrac- McMillen (1981) has argued, based on the
tions below which little structure is observed bathymetry, present sedimentation regime, and
(accretionary zone). The upper slope region bathymetric reconstructions related to the
between 0500Z and 0100Z is more complex evolution of the outer high of Seely (1979) that
than off Mexico, consisting of a thickened slope little coarse detritus was able to broach the
blanket containing more short and discon- outer high and reach the trench for much of the
tinuous reflections. history of this portion of the margin. Most
A landward-dipping reflection about 8 km in present-day canyons do not traverse the wide
length beneath the upper slope apron is similar forearc, and most fluvial sediments are trapped
to reflections found on other profiles in the near shore. Sparsity of trench fill as off Mexico
Guatemala area (Fig. 3). The dipping reflec- may limit the significance of sediments accreted
tions correlate with magnetic anomalies, high by offscraping versus some other type of deeper
refraction velocities and in some places with mass addition process. Further, the Guatemala
near surface exposures of basalts, serpentinite drilling programme discovered bedded deposits
and chert. These reflections are interpreted as as old as Cretaceous near the base of slope,
oceanic crustal slices (Ibrahim et al. 1979; Ladd while Miocene oceanic crust is presently being
et al. 1978; McMillen 1981). Other sections subducted (yon Huene et al. 1980).
show evidence for two or more such reflections
suggesting small oceanic fragments. If these Costa Rica
fragments are emplaced during initial or later
subduction phases then the sediment overlying Seismic reflection studies of the Costa Rican
them may be older slope or rise sections. margin along and just north and south of the
Within the upper slope apron a strong reflec- Nicoya Peninsula reveal a complex slope mor-
tion at 0.6 sec sub-bottom between 0230Z and phology (Buffler & Watkins 1978, 1981). The
0330Z with inverted polarity correlates with the peninsula, which is surrounded by a narrow
base of a gas hydrate layer (Shipley et al. 1979). shelf, contains Late Cretaceous oceanic crust
Gas hydrated sediments were encountered in and accretionary complex (Dengo 1962; Galli-
drilling on this transect but in an area where this Olivier 1979; Lundberg 1981). A seismic sec-
reflection (the so-called BSR) is not present tion beginning within 3 km of the shoreline of
(von Huene et al. 1980). the Nicoya Peninsula and terminating in the
The shelf section at the outer high (about trench axis is shown in Fig. 4. Seaward of the
0010Z) consists of about 3 km of landward- trench (not shown) normal faults with a throw
dipping reflections which thickens toward the down to the NE offset a highly reflective
NE. Drilling data reveal that this section is as oceanic sediment section of pelagic sediments.
old as Late Cretaceous near the base (Seely The reflective oceanic sedimentary section and
1979). The underlying basement was not associated normal faults are identified up to
102 T. H . S h i p l e y et al.

15 km beneath the lower slope (to 1630Z) and Discussion

oceanic crust extends to the end of the profile
(at 5.7 sec). The differences in major structural features
Most profiles show a thick slope blanket within the upper slope and shelf region along
extending down the entire inner slope with the Middle America Trench margin represent a
internal reflections becoming more discon- spectrum of convergent continental margin
tinuous down-slope. While the slope is fairly structures. While there have been changes in
smooth and free of basins in Fig. 4 other rates of convergence and obliquity of conver-
profiles to the south reveal significant variations gence along the margin with time, the entire
in the slope bathymetric profile. Between 0.8 margin has probably been subject to a conver-
and 2.5 sec sub-bottom beneath the slope sedi- gence component since at least Miocene if not
ments the top of the accretionary zone is de- the Cretaceous.
fined by diffraction hyperbolas similar to the The narrow shelf north of the Gulf of
sections off Guatemala and Mexico. Again no Tehuantepec is an unusual but not unknown
structure is observed in the zone. The only feature of convergent continental margins. For
complexities exist near the base of slope example, parts of the Peru-Chile Trench have a
(1600Z-1700Z). Because the oceanic crustal similar narrow shelf (Karig 1974a; Kulm et al.
reflection continues without significant inter- 1977), where Hussong et al. (1976) favour a
ruption and with little relief in the time section, model of erosion of the continental crust by
shallow occurrence of high velocity oceanic down-faulting in the middle slope region lead-
slices or blocks are probably not involved in this ing to a retreat of the continent. Our data do
structure (Fig. 5c). The reflective sequence may not resolve the mechanism of removal of the
be slope sediments and not uplifted oceanic Cretaceous arc along the northern portion of
sediments. Buffler & Watkins (1981) have not the Middle America Trench but the process
found evidence for shallow sediment accretion ended by early Miocene and no continental
at the base of slope. crust remains in the middle slope region (Moore
The origin of the thick slope sediment et al. 1979a, b; Shipley et al. 1980; Shipley
sequence is not obvious because the geography 1981).
precludes a significant present near-shore sedi- Dickinson & Seely (1979) and Seely (1979)
ment source. The present absence of turbidites suggest that forearc basins are formed on tran-
in the trench precludes deposition on the slope sitional or oceanic crust. The differences in
of sediments during passage of recent down- basin development along the Middle America
slope turbidity currents. The erosion and highly margin may be related at least partly to the
reflective nature of the shallow section (up to presence or absence of continental crust be-
0.3 sec sub-bottom) between 0000Z and 0200Z tween the original trench suture and the conti-
could be interpreted as evidence of erosion and nent. Variation in the regional sedimentation
redeposition related to shallow-water con- pattern does not seem large enough in itself to
tour(?) currents. If the slope deposits are older account for the basin evolution. A speculative
then the modest deformation of the 2.0 sec possibility is that subduction along Central
thick sediment interval is anomalous consider- America and Mexico may have begun or re-
ing the high convergence rate and expected started after significant left-lateral movement
high strain rate. on the Motagua-Polochic Fault Zone. Such
Well-defined oceanic sediments extending at motion would produce an offset in the margin
least 15 km beneath the slope of Costa Rica with oceanic crust trapped behind the suture
reveal little evidence for compressional tecto- south of the Gulf of Tehuantepec.
nics. Lower slope structures would be well- The oceanic crust within the accretionary
defined by the highly reflective oceanic sedi- zone of Guatemala at first appears unusual in
ment and mildly reflective slope apron if they terms of modern margins but will probably be
existed. Reflection data in the slope section more commonly recognized as more extensive
appear to confirm the absence of offscraping investigations such as in the Middle America
(shallow accretion) within the base of slope Trench are undertaken. The presence of
region. The gradual downslope decrease in oceanic crust within m61anges is well known;
reflection continuity in the slope sediments thus, the observations in Guatemala are not
suggest that some sort of shallow deformation is surprising except that there they seem to occur
more intense near the base of slope. at shallow structural levels in an older(?) part of
the accretionary zone. The incorporation of
oceanic crust at this level might be related to
initial ruptures of the oceanic crust. Sediments
 T e c t o n i c p r o c e s s e s a l o n g the M i d d l e A m e r i c a Trench 103

trapped above this old(?) crust may contain an Huene 1979). Thus transfer of pelagic and
earlier accretionary section or old rise sedi- hemipelagic oceanic sections at the base of the
ments pre-dating subduction. The presence of slope to the accretionary zone may not always
old oceanic crustal slices would suggest that be an important process in outbuilding of the
large-scale accretion of oceanic crust is episo- slope. Where trench fill is significant and has
dic. A non-steady state regime might also ex- been for some time then offscraping becomes
plain a number of observations, particularly the more important.
old age of sediments at the base of slope in If our observations are valid in the Middle
Guatemala and thick slope sediment in the America Trench, then shallow trench fill even
trench off Costa Rica. where present is volumetrically a minor accre-
The Nicoya Peninsula Late Cretaceous tionary component in the Middle America
m61ange sequence is essentially an exposed Trench and the accretionary zone consists
outer arc high with a much smaller forearc mainly of material derived by a different pro-
sedimentary basin. The elevation of the outer cess. One process might be deformation of the
high above sea-level may in part be related to slope deposits. The accretionary zone has been
increased uplift due to the shallow and presum- defined seismically as the zone capped by dif-
ably thickened oceanic crust of the Cocos Ridge fractions and which contains few distinctive
currently being subducted (a hot spot trace, internal reflections (except near the Rio
Hey 1977). Ometepec Canyon). It is often associated with a
The presence or absence of base of slope distinct increase in velocity. The diffractions
structures, amounts of trench fill, the thickness result partly from easily recognized large-scale
of the slope sediments and internal deformation irregularities in the slope. Diffractions may also
together reveal some of the active processes be produced by a number of other geological
within the accretionary zone. Observations in affects dealing mainly with surface roughness or
the Middle America Trench off Mexico suggest abrupt termination of reflecting horizons,
that offscraping and shallow accretion of the neither of which appear applicable here.
type proposed by Seely et al. (1974) is control- Perhaps the diffraction 'horizon' is not a
led by the presence of trench fill. Four adjacent significant structural feature but an upward-
profiles near the Rio Ometepec Canyon (an moving deformational front within the slope
average of 10 km apart) reveal a shallow sub- sediments. Carson (1977) has shown that signi-
bottom zone of dipping reflections (less than ficant dewatering and strain hardening occur
5 km in width) with inconsistent to constant within the subduction zone. The gradual in-
dips over most of the upper slope (Fig. 6). crease in fissility and other evidence of compac-
These reflections are correlated with uplifted tion and strain are observed progressively
sands and muds and define a shallow zone of downhole in data within the slope deposits of
trench sediment fill offscraping that has occur- the Nankai Trough (Karig et al. 1975); Japan
red since the late Miocene (Moore et al. 1979b). (Arthur et al. 1980), and Mexico (Lundberg &
The canyon has been a sediment source since at Moore 1981; Moore et al. 1979b). Perhaps the
least the late Miocene. As the sections in Figs zone of diffractions is related to some signifi-
2-4 show, significant trench fill has not been cant changes in physical properties resulting
introduced in some other areas along the mod- from the incipient strain. Drilling to date,
ern trench and are not observed as reflections however, has shown no obvious correlations,
within the slope. though it is hampered by poor recovery and
If accretion at the base of slope is dependent shallow penetration.
on trench fill, then some accretionary complex While we observe a mechanism for incor-
m61anges, both modern and ancient, may con- poration of slope sediments into the seismically
tain large portions of sediments added in defined accretionary zone, much of the volume
another fashion. Scholl & Marlow (1974) and must still be related to sediment addition at a
Scholl et al. (1977) have emphasized the small much deeper structural level. Another possible
amount of oceanic sediments in ancient subduc- tectonic process which may be responsible for
tion zones. Studies of modern and ancient the large acoustically unresolvable portion of
subduction zones (Karig 1974b; Moore 1975) the accretionary zone is some type of mass
suggest selective shallow accretion of turbidites addition by underplating. Watkins et al. (1981)
or indicate that oceanic sediments continue have convincingly shown that underplating is a
beneath the deformational front where the necessary process for Mexico mass balance
trench sediments are deformed and added to calculations and fits a wide variety of other
the base of slope (Karig & Sharman 1975; Kulm geological observations. Sediment budget cal-
& Fowler 1974; Moore & Curray 1980; von culations by Watkins et al. (1981) reveal that
104 T. H. S h i p l e y et al.

significant addition of material is required at The composition of the seismically defined

depth beneath the slope section and landward- accretionary zone, particularly in the absence of
dipping reflections. Cowan & Silling (1978) landward-dipping reflections and shallow drill-
have conducted model experiments which show ing penetration, make it difficult to resolve.
that with a buttress, material will begin to flow Diffraction hyperbolas at the top of the zone
upward in such a situation. Such a process does correlate with an increase in velocities. Careful-
not preclude a zone of deformed slope sedi- ly documented changes in structure and inci-
ments and a zone of offscraped landward- pient deformation observed in the drilling data
dipping trench fill. The underplating process show that deformation slowly increases down-
may act beneath this section and in fact helps core within slope deposits (Arthur et al. 1980;
explain the observed uplift without further rota- Lundberg & Moore 1981). Perhaps on some
tion of bedding planes. margins the shallow part of the accretionary
zone represents increasingly deformed slope
sediments which reach a stage where significant
physical property changes produce a ragged
Conclusions deformational front responsible for the charac-
teristic diffraction patterns.
As Dickinson & Seely (1979) and Seely (1979) While some sort of incorporation of slope
have proposed, the presence of a large shelf sediments into the accretionary zone is likely,
basin along the margin of the Middle America the processes of uplift and total volume of the
Trench might perhaps be explained by the line accretionary zone suggest some sort of mass
of initial suture occurring seaward of the ocean- addition at deeper levels. The underplating
continent boundary trapping oceanic crust (as proposed by Watkins et al. (1981) is probably a
from the Gulf of Tehuantepec to Nicoya). In fundamental process. The absence of acoustic
other areas the suture may have occurred right return from these zones could result from the
at the continental margin. Though highly spe- mode of emplacement and associated bedding
culative, perhaps the pre-Miocene history in- disruptions. Direct sampling will be required to
volved lateral movement in the Gulf of Tehuan- verify and describe the details of the mass
tepec which altered the margin configuration addition process. Underplating, offscraping and
moving Mexico westward relative to Central inplace slope deformation represent three diffe-
America. The complexities in the southern part rent processes of tectonic accretion which will
of the trench are probably partly related to vary from place-to-place in volumetric signifi-
subduction of thickened shallow oceanic crust. cance even within one arc depending on the
Along the Middle America Trench margin at sedimentological and tectonic setting.
least three different tectonic processes may be
responsible for accretion. At the base of slope ACKNOWLEDGMENTS:This programme was conducted
by the University of Texas Marine Science Institute in
offscraping seems dependent on the presence of 1977 and 1978 and encompasses the effort of many
trench fill. Oceanic sediments are not signifi- individuals in the marine geophysical group including
cantly involved in this process. A shallow imbri- K. Griffiths, J. Kunselman and A. Roberts and their
cate thrust model generally fits such a sitution. staffs. A number of colleagues, M. Houston, A.
The regional significance of such a model de- Ibrahim, G. Latham, K. McMillen, J. Moore, J.
pends on the amounts of trench fill and appears Shaub, R. yon Huene, and J. Worzel, were involved
fairly minor along most of the Middle America in shipboard operations and later analysis which
Trench south of Acapulco where trench fill is contributed greatly to the success of this programme.
The paper was improved with thoughtful reviews by
thin and discontinuous. Where offscraping is
J. Leggett and anonymous Geological Society re-
believed to have continued for any length of ferees. This investigation was supported by the
time (off the Rio Ometepec of Mexico) dips of Oceanography section of the National Science
presumed bedding planes reach high values Foundation Grant OCE-76-2330 and NSF-IPOD sub-
within 4 km of the axis and remain high and contracts CU-TEX 25907-2, -3. Contribution to
variable at shallower water depths. Scripps Institution of Oceanography New Series.

References
ARTHUR, M. A., CARSON,]3. ~¢,VONHUENE, R. 1980. 56-57, 569-613. U.S. Govt Printing Office,
Initial tectonic deformation of hemipelagic sedi- Washington.
ment at the leading edge of the Japan convergent BUFFLER, R. T. & WATKINS, J. S. 1978. Geologic
margin. In: Initial Rep. Deep Sea drill. Proj., Leg structure of the continental margin off the
 Tectonic processes along the Middle America Trench 105
Nicoya Peninsula, Costa Rica. Abstr. Prog. geol. Arcs, Deep Sea Trenches and Back-Arc Basins.
Soc. Am. 10, 374. Am. Geophys. Union, M. Ewing Ser. 1, 285-
& - 1981. Geologic structure of the con-
- 301.
tinental margin off the Nicoya Peninsula, Costa LADD, J. W., IBRAHIM, A. K., MCMILLEN K. J.,
Rica, based on multifold seismic reflection data LATHAM, G. V., VON HUENE, R. E., WATKINS, J.
(in prep.). S., MOORE, J. C. & WORZEL, J. L. 1978. Tecto-
CARSON, B. 1977. Tectonically induced deformation nics of the Middle America Trench. Int. Syrup.
of deep sea sediments off Washington and north- Guatemalan Earthquake and Reconstruction Pro-
ern Oregon: mechanical consolidation. Mar. cess, Feb. 4th, 1976, vol. 2.
Geol. 24, 289-307. LUNDBERG, N. 1981. Evolution of the slope landward
COWAN, D. S. & SILLING, R. M. 1978. A dynamic, of the Middle America Trench, Nicoya Peninsu-
scaled model of accretion at trenches and its la, Costa Rica (this volume).
implications for the tectonic evolution of subduc- - & MOORE J. C. 1981. Structural features of the
-

tion complexes. J. geophys. Res. 83, 5389-96. Middle America Trench slope off southern Mex-
DENGO, G. 1962. Tectonic-igneous sequences in Cos- ico, Deep Sea Drilling Project, Leg 66. In:
ta Rica. In: ENGEL, H. L. et al. (eds). Petrologic MOORE, J. C. & WATKINS, J. S. et al. (eds). Initial
Studies: a volume in honor of A. F. Buddington, Rep. Deep Sea drill. Pro]., Leg 66. U.S. Govt
131-61. Geol. Soc. Am. Printing Office, Washington, in press.
DICKINSON, W. R. & SEELY, D. R. 1979. Structure LYNN, W. S. & LEWIS, B. T. R. 1976. Tectonic
and stratigraphy of forearc regions. Bull. Am. evolution of the northern Cocos Plate. Geology,
Assoc. Petrol. Geol. 63, 2-31. 4, 718-22.
GALL1-OLIVIER, C. 1979. Ophiolite and island-arc MCMILLEN, K. J. 1981. San Jose Submarine Canyon,
volcanism in Costa Rica. Bull. geol. Soc. Am. 90, Middle America Trench, Guatemala: Its influ-
444-52. ence on slope and trench sedimentation, and
HEY, R. 1977. Tectonic evolution of the Cocos-Nazca tectonic accretion. Mar. Geol. (in press).
Spreading Center. Bull. geol. Soc. Am. 88, 1404- --, ENKEBOLL, R. H., MOORE, J. C., SHIPLEY, T. H
20. & LADD, J. W. 1981. Sedimentation in different
HUSSONG, D. M., EDWARDS, P. B., JOHNSON, S. H., tectonic environments of the Middle America
CAMBELL, J. F. & SUTlrON, G. H. 1976. Crustal Trench, southern Mexico and Guatemala (this
structure of the Peru-Chile Trench: 8° S-12 ° S volume).
latitude. In: SUTTON, G. H. et al. (eds). The MINSTER, J. B. & JORDON, T. H., 1978. Present-day
Geophysics of the Pacific Ocean Basin and its plate motions. J. geophys. Res. 83, 5331-54.
Margin. Geophys. Monogr. Am. Geophys. Un- MOLNAR, P. & SYKES, L. R. 1969. Tectonics of the
ion, 19, 71-86. Caribbean and Middle America regions from
[BRAHIM, A. K., LATHAM, G. V. & LADD, J. W. 1979. focal mechanisms and seismicity. Bull. geol. Soc.
Seismic refraction and reflection measurements Am. 80, 1639-84.
in the Middle America Trench offshore Guate- MOORE, G. F. & CURRAY,J. R. 1980. Structure of the
mala. J. geophys. Res. 84, 5643-9. Sunda Trench lower slope off Sumatra from
KARIG, D. E. 1974a. Tectonic erosion of trenches. multichannel seismic reflection data. Mar. Geol.
Earth planet. Sci. Lett. 21,209-12. Res. 4, 319-40.
- - 1974b. Evolution of arc systems in the western MOORE, J. C. 1975. Selective subduction. Geology, 3,
Pacific. Earth planet. Sci. Ann. Rev. 2, 51-76. 530-2.
, CARDWELL,R. K., MOORE, G. F. & MOORE, D. -- et al. 1979a. Middle America Trench. Geotimes,
G. 1978. Late Cenozoic, subduction and con- 24, 20-2.
tinental margin truncation along the northern - et al. 1979b. Progressive accretion in the Middle
-

Middle America Trench. Bull. geol. Soc. Am. America Trench, southern Mexico. Nature, 281,
89, 265-76. 638-42.
& SHARMAN, G. F. 1975. Subduction and accre- MOONEY, W. M., MEYER, R. P., HELSLEY, C. E.,
tion in trenches. Bull. geol. Soc. Am. 86, 377-89. LOMNITZ, C. & LEWIS, B. T. R. 1975. Refracted
et al. 1975. Initial Rep. Deep Sea drill. Proj. Leg waves across a leading edge: observations of
31. U.S. Govt Printing Office, Washington, 927 Pacific shots in southern Mexico. Trans. Am.
PP. geophys. Union, 56, 452.
KESLER, S. E. 1973. Basement rock structural trends Ross, D. A. & SHOR, G. G. 1965. Reflection profiles
in southern Mexico. Bull. geol. Soc. Am. 84, across the Middle America Trench. J. geophys.
1059-64. Res. 70, 5551-72.
KULM, L. D. & FOWLER, G. A. 1974. Oregon con- SCHILT, F. S. & TRUCHAN, M. 1976. Plate motions in
tinental margin structure and stratigraphy: a test the northern part of the Cocos Plate. Trans. Am.
of the imbricate thrust model. In: BURK, C. A & geophys. Union, 57, 333.
DRAKE, C. L. (eds). The Geology of Continental SCHOLL, D. W. & MARLOW, M. S. 1974. Sedimentary
Margins, 261-83. Springer-Verlag, New York. sequences in modern Pacific trenches and the
~ , SCHWELLER, W. J. & MASIAS, A. 1977. A deformed circum-Pacific eugeosyncline. In:
preliminary analysis of the subduction processes DOTI', R. H. & SHAVER, R. H. (eds). Modern and
along the Andean continental margin. In: Ancient Geosynclinal Sedimentation. Spec. Publ.
TALWANI, M. & PITMAN, W. C. (eds). Island Soc. econ. Paleont. Mineral, 19, 193-211.
106 T.H. S h i p l e y et al.

, (~ COOPER, A. K. 1977. Sediment sub- spread occurrence of possible gas hydrate hori-
duction and offscraping at Pacific margins. In." zons on continental slopes and rises. Bull. Am.
TALWANI, M. & PITMAN, W. C. (eds). Island Assoc. Petrol. Geol. 63, 2204-13.
Arcs, Deep Sea Trenches and Back-arc Basins. - - , MCMILLEN, K. J., WATKINS, J. S., MOORE, J.
Am. Geophys. Union, M. Ewing Ser. 1, 199- C., SANDOVAL-OCHOA,H. & WORZEL, J. L. 1980.
220. Continental margin and lower slope structures of
SEELY, D. R. 1979. The evolution of structural highs the Middle America Trench near Acapulco,
bordering major forearc basins. Bull. Am. Mexico. Mar. Geol. 35, 65-82.
Assoc. Petrol. Geol. 29, 245-60. SHOR, G. G., JR. & FISHER, R. L. 1961. Middle
, VAIL, P. R. & WALTON, G. F. 1974. Trench America Trench: seismic-refraction studies.
slope model. In: BURK, C. A. & DRAKE, C. L. Bull. geol. Soc. Am. 63, 2204-13.
(eds). The Geology of Continental Margins, 249- YON HUENE, R. 1979. Structure of the outer conver-
60. Springer-Verlag, New York. gent margin off Kodiak Island, Alaska, from
SnIPLEY, T. H. 1981. Seismic facies and structural multichannei seismic records. Mere. Am. Assoc.
framework of the southern Mexico continental Petrol. Geol. 29, 261-90.
margin. In: MOORE, J. C. & WATKINS,J. S. et al. et al. 1980. Leg 67: The Deep Sea Drilling
(eds). Initial Rep. Deep Sea drill. Proj., Leg 66. Project Mid-America Trench transect off Guate-
U.S. Govt Printing Office, Washington (in mala. Bull. geol. Soc. Am. 91,421-32.
press). WATKINS, J. S. et al. 1981. Accretion, underplating,
, HOUSTON, M. H., BUFFLER,R. T., SHAUB,F. J., subduction and tectonic evolution of the Middle
MCMILLEN K. J., LADD, J. W. (~ WORZEL, J. L. America Trench, southern Mexico: Results of
1979. Seismic-reflection evidence for the wide- Leg 66 DSDP, Oceanologica Acta (in press).

THOMAS H. SHIPLEY, Scripps Institution of Oceanography, University of California,

La Jolla, California 92093, U.S.A.
J O H N W . LADD, Lamont-Doherty Geological Observatory of Columbia University,
Palisades, New York 10964, U.S.A.
RICHARD T. BUFFLER, Marine Science Institute, University of Texas, Galveston,
Texas 77550, U.S.A.
JOEL S. WATKINS, Gulf Science & Technology Company, P.O. Box 2038,
Pittsburgh, Pennsylvania 15230, U.S.A.
 Sedimentation in different tectonic environments of the Middle
America Trench, southern Mexico and Guatemala

Kenneth J. McMillen, Robert H. Enkeboll, J. Casey Moore, Thomas H. Shipley

& John W. Ladd
SUMMARY: Late Pleistocene to Holocene sediment facies and composition within and
bordering the Middle America Trench offshore southern Mexico and Guatemala reflect
two distinctly different tectonic provinces. The truncated Mexican margin with crystalline
rocks onshore and a narrow shelf exhibits a locally thick sandy trench fill. Sand mineralogy
of quartz, feldspar, and biotite matches the onshore source terrane. The Guatemalan
margin with a volcanic terrane onshore and a wide forearc basin show smaller amounts of
predominantly muddy trench fill. Sand composition of volcanic rock fragments, plagio-
clase, and heavy minerals accurately reflects the volcanic source. Trench fill in both areas
corresponds to submarine canyon location, and extensive bypassing of the slope occurs.
The outer shelf contributes little sediment to the Mexican slope and trench, the main
source being in the littoral zone and inner shelf. Conversely, most Guatemalan slope and
trench sand has come from the shelf. Trench fill correlates best with onshore geology, with
similar slope sediments in both areas. The Mexican margin with locally thick trench fill
displays clear evidence of accretion, the Guatemalan margin with meagre trench fill
apparently has not accreted lower plate sediments recently. Possibly, larger volumes of
trench fill encourage accretion:

Studies of modern sediment distribution and Laboratory analyses consisted of general de-
depositional processes on active margins pro- scriptions of split cores noting colour, general
vide an understanding of controls on the type texture, structures, and composition of the wet-
and amount of sediment present in deep sea sieved sand fraction. Micropalaeontological
trenches, the sediment pathways to the trench, samples, taken roughly every 20 cm, were
and trapping effects of the shelf and of slope boiled in peroxide to remove organic matter
basins. Description of sedimentary petrology and to clean fossil tests, wet-sieved through a
can be related to source terrane and to varia- 63 ~ m mesh screen and mounted as strewn
tions in tectonic setting of the margin. slides. Petrological samples were taken from
In this paper, we report on piston cores and sand beds, soaked in peroxide to break down
seismic reflection profiles collected from two organic matter, wet-sieved through a 6 3 p m
portions of the Middle America Trench on mesh screen, mounted in epoxy, and cut and
surveys by the University of Texas Marine ground for thin-section point-count analysis
Science Institute for I P O D drilling offshore of 300-800 grains using the methods of Dickin-
southern Mexico on Leg 66, and offshore son (1970). Slides were stained with con-
Guatemala on Leg 67 (Fig. 1). We describe centrated sodium cobaltinitrate solution for
trench, slope, and shelf sediments, infer deposi- potassium feldspars and amaranth solution for
tional processes, and relate sedimentological plagioclase feldspars. Radiography followed
and petrological variations to differences in standard techniques of Bouma (1969).
tectonic setting.
Cores were collected with a Ewing-type pis-
ton corer of 20 or 40 ft (6.7-13.3 m) length and Southern Mexico regional setting
2.5 in (6.4 cm) inside diameter fitted with poly-
carbonate liners. Cores were sectioned and The southern Mexico margin north of the Gulf
stored aboard ship. Satellite navigation was of Tehuantepec differs from most active mar-
used for all underway data collection and cor- gins in several ways. The distance from trench
ing. Bathymetry, which was recorded on to shoreline is small, generally less than 70 km
3.5 kHz records, was corrected with Matthews (Fig. 2). The shelf is narrow and, as the land
Tables (Matthews 1939). Seismic reflection surface rises steeply from shore, there is no
data were collected and processed using techni- forearc basin. The trans-Mexican volcanic arc
ques described in Shipley et al. (1978). In the trends at an angle to the trench, and is located
text, core numbers are preceded by an 'M' for about 200 km landward of the Leg 66 drilling
Mexican cores and a 'G' for Guatemalan cores. area. The area between the arc and shoreline

107
 ( g 6 l v 9 # ) suo!leaOl al!s 99
~aq d(1S(I pup. suo!leao I oaoa uols!d ~u!~oqs ul~.letU o3!xalAI u.lgqlnos aql jo ~.llam3~qle~ (e)';~ "91j
o86-
.gl
/,, , . ~ #
.- / . ~ ..... ~... . . . \\ \
F --
.91
. /
i
f
/
/
t
t
1 ,
<,66-
"(696I ~u!~) lza!JOtUV qlJoN jo di~tu o!uolool uioJj
•OU!lO.mqs u e a ! x o ~ oql Iv, luotuose.q ue!.iqtui,'Oaad u! u!p,.,~ le.lmanals pue qao.i ai.uolnld jo uo!leaunal
olo N "(poq31eq-ssoa3) sp.oJe. ~pnts jo uo!lp,3o I aql ql!~ lza!.ioulv a l p p ! ~ jo ~OlOa~ leUO!~O8 "I ".,'Jlj
.~/I sl!sodeo u!eldImSeo9 , ]
~"Y s o,uealOA p u e
.~...~%,,,.~,o,,
-.o N#" '
~ . , , ~ . . ~ . . . . o ~ . ~ ~'-~'~1
' "
.. -......;/
)2" "
.. :?:';-!i:'(<.i i': : ;.':/i
"F lo Uall!lNOl~ 7 "X 80I
 S e d i m e n t a t i o n in the M i d d l e A m e r i c a Trench 109
oj~O0'
~4'~3' I --- 14~(~ '
SANe, JOSE

1_ _

SHELF

~--_ ,3 \

13oo{

v,%.

CONTOUR INTERVAL 2OOm

LANO - - k

L 20 KM I

BATHYMETRY OFFSHORE GUATEMALA

91'~N3'

FI6.2.(b) Bathymetry of the Guatemala margin showing piston core locations and DSDP Leg 67 site
locations (494-499). Contour interval is 200 m.

contains a variety of continental rocks including Guatemala regional setting

Precambrian crystalline rocks, Palaeozoic and
Mesozoic shallow-water and platform The Guatemala Pacific margin typifies many
sedimentary rocks, and Mesozoic plutonic active margins. Its wide ( - 1 2 0 kin) forearc
rocks (de Czerna 1965). Structural grain of basin traps a large volume of sediments from
many of these rocks trends offshore, crystalline the volcanic arc (Seely et al. 1974). The arc
outcrops occur at the shoreline (Fig. 1), and parallels the trench and is closer than in the Leg
these continental rocks underlie the shelf and 66 area (Fig. 1). The forearc basin is probably
upper slope (Fig. 3a, Shipley et al. 1978, 1981) floored by trapped oceanic crust (Seely et al.
indicating margin truncation. Two theories to 1974; Seely 1978; Lundburg 1981) and filled by
explain this truncation are 'tectonic erosion' volcanic sediments to comprise a broad shelf
(Karig 1974) and removal by strike-slip faulting and coastal plain. Oceanic crust probably
(Karig et al. 1978). Results from Leg 66 drilling occurs within the inner trench slope (Fig. 3b,
(Moore et al. 1979a) did not resolve the Ibrahim et al. 1979). The actual zone of recent
mechanism of truncation but show that it occur- accretion is probably very small or non-existent
red prior to early Miocene. (yon Heune et al. 1980, 1981).
110 K.J. McMillen et al.

a.

(km SOUTHERN MEXICO

"Transition .. Shore
Slope Z o n e" .,,,,~._ . . , , , , ~ ' J ' +
Apron .,,,,.,~'l~-'~" + + +
• '~ ~ ¢/"
I + + + + +4 •0 - 4.6
Trench]. AXlS ~ ~ . ~ " ~? + Con line2ta, +
" 2,2-3.0? + + ~Crust + +
" - A ~ c r e t f o n a r y . . . . .
4 ~ ~ ~ Wedge - Landward
~
6.8 4
~
^v,~ ,4~'-..
Dipping R e f l e c t o r s
NNE
Oceanic ¥ V~
h /,, Crust I¢ A ~O 10 km
I MX16 I
A ^
V

bo Forearc
o GUATEMALA Basin "
(kin)
2 Slope Apron ~

4
Q ~ - ~ o \ j3\\ \ < .
KM6 ^ \-~"- "-3.2
2:.-.-.-.-.-.-.-.-.-~ 4 £ ' - 4 7
- - "k
~A
.v \
..~, \.v
s.6~
v
" "

8 ., ~ "J v\ "-..v " v"',.~

10
°.'~ , ~ ; ~ O~ean,cOr.st? "J
"~'Eo" - q "" ,, ., .aUL~ "J
Oceanic B.1 v Y ~ 10 km t -I
14 /

Crust ",' v ,,,, GUA 13 /

v "4 v I
i
,/ ¢ ~ •

FiG. 3. Interpretive geological sections across (a) the Southern Mexico and (b) the Guatemala Pacific
margins showing major reflectors, refraction and interval velocities, and underlying basement.

Southern Mexico bathymetry Guatemala bathymetry

Margin morphology and core locations are The margin morphology offshore Guatemala is
shown in Fig. 2(a) for the southern Mexico simpler than offshore Mexico (Fig. 2b). The
margin. Major indentations of the upper slope shelf edge is fairly straight except for a major
may be due to extensions of continental struc- embayment associated with San Jose submarine
tures offshore (Shipley et al. 1978). The lower canyon. The upper slope is steep facing the
slope is steep with a few small terraces. Several trench but is gentler within the embayment.
submarine canyons cross the slope but only The gentler gradient of the middle slope may be
one, Ometepec Canyon, reaches the trench; structurally controlled as it is underlain, at least
others are blocked by uplifted or slumped mas- in part, by emplaced oceanic crust (Ibrahim et
ses. Trench sediment occurs as isolated basins al. 1979). San Jose Canyon deflects to the east
with no apparent interconnection. Two large as it crosses the middle slope and here shows
trench sediment basins are related to the loca- levee-like depositional structure adjacent to the
tions of Ometepec Canyon and a series of canyon. Several small terraces interrupt an
canyons offshore from Rio de la Arena and Rio otherwise steep lower slope. San Jose Canyon
Verde (Fig. 2a). Multichannel profiles across breaks up into smaller channels which are de-
the trench fill show evidence of fan channels, flected around topographic ridges, The trench
levees, and deformation of trench sediment on offshore Guatemala essentially lacks sediment
the inner trench wall (Fig. 4a-c). Landward- except in one small pond fed by San Jose
dipping seismic reflectors beneath the inner Canyon (Fig. 4d).
trench slope probably represent older, tilted
and uplifted trench sediments (Moore et al.
1979b).
 Sedimentation in t h e M i d d l e A m e r i c a Trench 111
-a- material with virtually no clay or silt matrix
,2 OM-6N
TRENCH
INNER
Older uplifted
WALL
¢--~'~ -.~
(Fig. 5a). Two other trench cores recovered
sandy turbidites with c-e Bouma sequences
trench- __ ~ , ~ (Fig. 6a).
t ur bidit e s . ~ , / . _ i ~ _~~ A predominance of mud characterizes the
Channels , ~
slope sediments. Outer slope hemipelagic sedi-
ments, cored in M25, consist of hemipelagic
mud with zones of 1-2 cm thick sand beds
having sharp upper and lower contacts (Figs 5a
& 6). These may be contourites deposited by
2 km
bottom currents. The 700 m elevation of core
8 M25 above the trench probably precludes sand
-b-
deposition by turbidity currents, unless rapid
5 vertical uplift has occurred.
MX-15 INNER W A L L
(sec) Older uplifted Lower inner slope core M28 includes thin
TRENCH t r e n c ~ u r b i d i t e s .~
sand laminations (1 cm). Except for one thin
.-. Deformed , , ~ ~__L_~.-=~
6-
"~-~" Channel trench ~x" "/ sand bed in core M29, all other lower slope
"~"',~.~ Levee I~, fill~ , . . ' ~ " ~_ cores are devoid of sands. Middle slope cores
~'.,~-~._-~---._E(- __ ---~" ..'-_'-= are mud only except for gravels in Ometepec
Canyon which may be derived from upslope or
onshore. Upper slope sediments are also mud-
dy except near Ometepec Canyon where rhyth-
2 km
mically bedded sand and mud laminations occur
with rippling and local channelling in core M35
(Fig. 6b). All other upper slope cores consist of
-C-
5 laminated mud where laminations are recog-
MX-16 INNER W A L L
(sec) TRENCH Older uplifted,~ C h a ° t i c Z nized by colour contrasts (Fig. 5a). Outer shelf
t,ench sediments recovered in cores M36 and M37
6- x ,,r .... ",~, consist entirely of shelly sand, shelly sand cap-
Deformed ¢' / ~ /%.'~. "~"
Channels trench fill '×/ ."~'~'~ ped by hemipelagic mud in core M6, and mud
--.__i" ~ ~ ........ <% ~ - - only in core M4. Laminations were not
;-- . ~ . - ~_-.'-_-~-2~.~z/ " ~ , . , ~ - ~-~ observed in any outer shelf cores.

-d- Guatemala
4
(krr H.I.G. INNER W A L L Sand bed deposition at core G16 is similar to
TRENCH
sand deposition at Mexican core M25: the sands
~"-.z".,
~"-~",, "'",,. Channels ?
have been deposited by bottom currents, or
~i', ~ // f=- turbidity currents above the level of the trench.
" ,, ~'~,-..'--~ t ~ J _--
Subsequent rapid uplift analogous to uplifted
turbidites in the Peru trench (Prince et al.
1974), or overtopping by large turbidity flows
2km
can explain sand deposition above trench level.
Trench sediments offshore Guatemala are
Fro. 4. Profiles across the Middle America mostly muddy to the 11 m cored depth (Fig.
Trench axis showing sediment fill: (a-c) 5b). Coarse clastic deposits consist of thin sand
offshore southern Mexico, and (d) Guate- and ash beds and thicker muddy ash horizons
mala. Profiles (a-c) are from the Univer- with convolute structures representing ashy
sity of Texas survey, profile (d) is from turbidites in cores G14, G17 and G13. Most
the Hawaii Institute of Geophysics. trench cores contain a great deal of interstitial
gas which cracked open clayey sediment (Fig.
6d).
Sediment distribution Inner and outer slope sediments are also
nearly all muddy. Outer slope core G16 nearly
Mexico
700 m above the trench axis, contains thin-
Trench sediments recovered in cores M26 bedded sands in the lower half of the core and
and M27 consist of sand- to granule-size core G15 on the outer rise contains ash beds.
112 K . J . McMillen e t al.

MIDDLE SLOPE OMETEPEC

i,,o:o
I

OUTER SLOPE TRENCH LOWER SLOPE (> 300Ore) CANYON

25 26 27 1 31 28 29 30 33
"~-'-'~
i

~'~ ii~.:!:t ! '_'-i.,

(m) ---~ "=-:.-_.
, ;
-°---i
. . : , . , . '.! ;oi;, ~'_=:= . --°---.t
~ .
- - - - i
~
~ii ?.;i:!
. . . . i

6 . . . . 7 "___V"---i
---_.--
~ . ~
y'_~
. . . . i . . . . - - °.---i
~ ° ~ . . . . =
;?., :q~,,,

~..:~.,:~!,
7 . . . . .
--.-- '_--. "__~., 'LT.'__-? "_-:'--~ ._.-----' ....-..-.... - ~ I ~__KL_
.... I °¢~1--"- -
. . . .
~" --T._~ . . . . . ° . . . . .-'---~-
.

° ~ _ ~.--.~ . . . . . _.° . . . . . . . . .
,~--~ . . . . . ° . . . .. . .. .

~'-_--7"Z - - . ~
- "---~ T" .---1 -'_T--_'
8 . . . .

~~ _ .....~ ~ - ~ - °• _- _- ,° -_

. . . .
~ .... -_.~'--
_ _ . ~
. . . . . .

. . . .. . . .. . . . . -_-_°°-_-
_ _ ° - -

4 •

:---~_ -.--__~ _.__~

_ . _ _

_ _ . _

_ ° _- -_

!----,-.._...2"
~0-

UPPER SLOPE (200 - 1000 m) SHELF (0 - 200 m)

35 .~-2~ 7 8 9 36 37
o . . . . ~_° . . . . I . . . . . . . _.__'.

:....--:1: ---::_, _--..-1. .... •.r ~ . ; .~

_. ---.~
........ ,--.~ ~ .... ;"-
...~.~.
........ __ .... --. _ __.__ ,,...~ ,, ,~,.: .. '~--".'..~ !.... ~ Algal, Shelly Sand
(m) ........ :--.--= . . . . . . . . . . . . . ~-,&.~ .~ ~ ~, "z-:'-" --:'__.~
........... -4 ~ . . . . . . •

........
........
.--.~ -'--:-'_: .__°.
_._ :_-:-- ....
......... - ..~.. ~ ~.~
"~ '~ k. ~ '
_:___
. . . .
..... sand

'.2:':::: i ] •- - " _ - - z-:z- ~ ~,.d,o,~ ,o , , ° . s . ° °

. . . . .
.... ~ 1
~ ,~-~ ~
T'_-:---." ,-" .----" ~ Mud
:S:;:r:: I-"-i "~7--."
"--- "-----~ LUJ ~C
......
....... , ,, 2
- , - - °
- - - "__--"j
t--:--q Clay

'--.-----
. . . .
":.U~s._~ '<
~ ^ah
. . . . . . . . 7--°
. . . . o a0.,.

.... .-e~ L.
.... ~ "~l: ~ L a m i n a t e d Sand and
I ,
"~-~-*~ ~.~; .-~..~ t , , . , . , ~ Clay or M u d B e d s

~ Laminated Clay(color) or
. ,
--,- L I Clay and Mud
....

. - - . ----;

FIG. 5. Logs of all piston cores recovered from (a) the Mexico margin and (b) the Guatemala margin.
Cores are grouped by environments, labelled in Fig. 2. In the text, cores from Mexico are prefixed by
the letter 'M', and cores from Guatemala by the letter 'G'. The N/S (nasselarian-spumellarian)
minimum is shown as a bar to the left of some core logs.
 Sedimentation in the M i d d l e A m e r i c a Trench 113

OUTER TRENCH LOWER SLOPE (>3000m)

SLOPE I SAN JOSE
CANYON
15 16 14 17 13' 21 13 12 11 10 9 8 7 12" 18
o -,--'~
----Z_~." . . . . .
(m)
.... 7---~
....
~.__ n . . . . . 7"-_-~-
....
___
2- . . . .
....

....
N
....
3-.--.- Z2~----
-.___-- ~
----- .__ == -_.-~-___- -__K--- - - "
....
.... _°-- -..-~-..
4..... _.__~ .... ....
_.o_
....
.... t: . J --o.--

N
5 -
. . . .

....
a- .__
....... -i.
-.-~
.... _o_
.---.-2
o-__0~

8-

LOWER UPPER SLOPE

SLOPE MIDDLE SLOPE (1000-3000m) (200-1000m) SHELF (O-2OOm)
LEVEE SAN JOSE CANYON
o 19 11' ~.----"
6 "~--~-"-J
5 4 22 2o 1 2 3 23 24

7 -" - - :
(m) _ _ . _ _
. . . . . . ,.__.,
--__~ ~ . . . . . .
.----._~ • --. .. . . . .
_-- --.-- "~-T-- . . . . . .
7-/__T2 T- "____~ - - .--" ~2_.---~ "_.-':-~ -- .~ ~ Muddy Gravel
2--2-..-- . 7-'._._~ --.-- "~T-. "---.i . . . . . .
- - ..___~ .-- "__.~
~ Sand

3_ ,--.

~ Mud with Sand Laminations

• - - .--_,
a-.__,
I~:~1 Laminated Mud
" - - L_.~
7 -"-- - .__..~_
~'~7 M ud
__'__~ . . . . .

--"TT
Muddy Ash
~ (redeposited)

Ash

F~. 5 (continued)

Inner slope cores contain four types of coarse and from San Jose Canyon itself (cores G4
clastic deposits: and G12');
1. thin-bedded, sandy turbidites (Fig. 6e) in 3. ash beds; and
lower slope terraces (in cores G8, G10, 4. coarse gravels (in cores GT, Gg, G l l , and
and G12); G22 from lower slope scarps) and gravel
2. thin-bedded sandy turbidites from the San clasts (from San Jose Canyon cores G4
Jose Canyon mid-slope levee (core G l l ) and G12') (Fig. 5b).
 MEXICO GUATEMALA
TRENCH UPPER SLOPE OUTER SLOPE TRENCH LOWER SLOPE
Cm cm cm
OUMA 5 3 r- 3 0 0 cm
JNITS 19
10

410-
CHANNEL

130-

~ ~ 54

20~

780-
42 0 -
t~

140--

55 SH 3

430 - 7 c"
,'~ b c d e
bl 3 1 M35 M-25 G 17 G - 1<2

FIG. 6. Radiographs of cores showing distinctive structures: (a) turbidites from Mexican trench, (b) Mexican upper slope laminated sand, silt, and
clay with ripples and channelling, (c) Mexican outer slope sand and clay deposits (contourites), (d) Guatemalan trench mud with gas cracks, and
(e) Mexican inner-slope terrace thin-bedded turbidites. Photographs are contact prints of actual radiographs so density response is reversed.
 Sedimentation in the Middle America Trench 115

Shelf sediments are sandy with much cal- & Emery 1968), as the N/S ratio is low in
careous debris such as shells. modern nearshore environments (Casey &
Bauer 1976). Since this N/S minimum is prob-
ably isochronous, it can be used to compare
Micropalaeontology and correlation relative rates of sedimentation in different en-
vironments and is plotted in Fig. 5.
Calcareous microfossils are unsatisfactory for Sedimentation rates estimated using the
Quaternary biostratigraphic zonation on the 15 000 yr N/S minimum are seen to be highest
Middle America Pacific margins due to carbon- in ponded areas of trench sediment and lowest
ate solution in response to the relatively shallow on the slope. Near submarine canyons such as
East Pacific calcite compensation depth (CCD) Ometepec and San Jose Canyons, the rate
of 3.4kin (McMillen & Bachman 1981). increases. Upper slope sediments off Mexico
Siliceous microfossils such as radiolarians were deposited at a much higher rate than
should provide biostratigraphic control, but the comparable sediments off Guatemala. Sedi-
commonly used Quarternary equatorial zona- mentation rates in portions of the trench with
tion of Nigrini (1971) is unsuitable because no sediment ponds are low off Guatemala
upwelling in this area has excluded the critical (cores G18 and G21) and similar to slope rates.
surface-dwelling index species (McMillen Holocene to late Pleistocene slope sediments
1981). The extinction level of Axoprunum ange- are generally muddy and thin sand beds show
linum (Campbell and Clark) at 0.40 Ma (Hays up only in the Pleistocene section. A notable
1970) provides a good datum in the NE exception is upper slope core M35 with lamin-
Pacific (McMillen 1981; Dinkelman 1973). ated sand and mud in the Holocene or late
None of the piston cores we recovered con- Pleistocene.
tained A. angelinum so all cores are of late
Quaternary age less than 0.40 Ma old. The Petrology
base of core M-32 contains mud with late
Miocene-age radiolarians. This sediment could We have investigated coarse-fraction consti-
be older slope deposits or reworked. tuents of sands and muds, and mineralogy of
Recognition of the Pleistocene-Holocene the terrigenous portion of sand beds. Coarse-
boundary, based on the presence of radiolarian- fraction constituent study gives clues as to the
rich sediments above the boundary and forami- origin of lower slope and trench sands. Sedi-
niferal-rich sediments below, has been attemp- ments delivered by river to the Middle America
ted by Duncan et al. (1970). As Bernard & margins consists of mud, terrigenous sand, and
McManus (1973) point out, however, this plant material. Shelf sediments receive a signifi-
change from foraminiferal- to radiolarian-rich cant contribution of calcareous debris in the
sediments up-core is controlled by changes in form of shell fragments, foraminiferal tests, and
the CCD, and is time-transgressive up-slope. algae. Trench and lower slope sands offshore
Nevertheless, the foraminiferal-radiolarian Mexico contain mostly terrigenous grains, espe-
boundary, which can be identified offshore cially in the thick-bedded, coarse-grained sands
Guatemala (McMillen 1979), does provide a at the base of Ometepec Canyon. These coarse-
rough indication of the Pleistocene-Holocene grained sands were derived from a littoral drift
boundary. or fluvial source, with little contribution of shelf
The ratio of nassellarian (cone-shaped) to sands. Sand beds on the Guatemala slope and
spumellarian (spherical) radiolarians, or the trench often contain calcareous shelly debris
N/S ratio, has been shown to correlate roughly from molluscs, ostracodes, and foraminifers;
with the down-core foraminiferal increase this indicates a source on the shelf.
offshore Guatemala (McMillen 1979). Since Terrigenous petrology of sands differ con-
variations in the N/S ratio represent faunal siderably between the two areas (Fig. 7, En-
changes due to changing oceanographic condi- keboll 1981). Mexican sands contain abundant
tions rather than solution changes, maxima and quartz, feldspar, and metamorphic rock frag-
minima in the N/S ratio should be isochronous ments. The modal composition for fine-grained
over a small area. A N/S minimum, which sands is Q31F50L19, whereas it is Q41F22L37 for
occurs just below the level of foraminiferal coarse-grained sands. The chert to total quartz
increase, is correlative from core to core and is ratio is 0.05 in fine-grained sands and 0.04 in
usually located near a time-parallel ash horizon. coarse-grained sands. The plagioclase to total
McMillen (1979) reasoned that this N/S mini- feldspar ratio is 0.53 in fine-grained sands and
mum was due to the late Pleistocene lowering 0.43 in coarse-grained sands. Biotite is the
of sea-level, occurring at 15,000 yr bp (Milliman commonest mafic constituent.
116 K.J. M c M i l l e n et al.

QUARTZ canyons. Sandy trench fill reflects the crystal-

line source terrane. Guatemala has a wide fore-
arc basin underlying a wide shelf and fluvial
coastal plain, volcanic highlands onshore, and
few submarine canyons. Trench fill is muddy
due to sand trapping on the shelf and break-
~ SOUTHERN down of unstable volcanic products. Middle and
lower slope sediments are similar in both areas,
consisting mainly of slowly deposited hemipela-
gic mud with a few thin sand beds. Upper slope
sediments differ betweeen Mexico and Guate-
mala. Laminated sediments deposited at a fairly
high rate by slope currents occur offshore Mex-
FELDSPAR ROCK
FRAGMENTS ico in or near the oxygen-minimum zone.
Hemipelagic upper slope sediments offshore
Fro. 7. Compositional comparison of sands Guatemala have been deposited at a slow rate
from the trench, trench slope, and shelf and no effect of the oxygen-minimum zone is
from the Mexico and Guatemala margins.
seen. Erosion of the Mexican slope at slightly
greater depths at DSDP Site 489 (Moore et al.
Guatemalan sands, all fine-grained, are rich 1979a) suggests that considerable variability can
in pyroclastic material and poor in quartz; the exist in upper slope deposition. Shelf sediments
average composition is Q1F43L56. The plagio- are similar in both areas, consisting of ter-
clase to total feldspar ratio is 0.99. Volcanic rigenous and bioclastic calcareous sand.
rock fragments constitute 99% of all rock frag- The volume of trench fill differs between
ments. Pyroxene is the dominant mafic mineral. both areas as well. While neither area has
In addition to these sands, thick gravel beds especially thick fill, the portion of the Guate-
also occur on the lower Guatemalan slopes. malan trench studied has considerably less fill
These gravels have unusual lithic constituents than the Mexican area studied. The trench fill
consisting of chert, mudstone, zoelite-facies offshore Guatemala is both thinner and less
argillite, and prehnite-pumpellyite-facies extensive laterally. The wide forearc basin of
metabasite. the Guatemalan margin may capture more sedi-
ment and result in a smaller quantity actually
being delivered to the trench than offshore
Discussion Mexico.
On the other hand, the higher convergence
Sediments of the southern Mexico and Guate- rate across the Guatemalan margin (10 cm yr -~,
malan portions of the Middle America Trench Molnar & Sykes 1969) than across the Mexican
margin can be compared lithologically, petrolo- margin (about 7 cm yr -1, Minster & Jordan
gically, and on the basis of their accumulation 1978) would also tend to reduce the volume of
rates. Trench sediments differ most between trench sediment at any given moment (see
the two areas. The Mexican Trench contains Schweller & Kulm 1978).
Holocene coarse- to medium-grained, clean Sediment bypassing of the slope is a signifi-
sand or thick-bedded sandy turbidites, whereas cant feature of both margins; however, the role
the Guatemalan trench contains muddy sedi- of the shelf in supplying sediment varies.
ments with late Pleistocene or early Holocene Offshore Mexico, trench sediment is derived
thin-bedded sandy or ashy turbidites. Sediment mainly from fluvial or possibly littoral-drift
fill within the trench offshore Mexico is thicker sources with a little contribution from the outer
and more extensive, showing evidence of large shelf, as is shown by differing coarse fraction
channels and levees, and local tectonic de- composition between terrigenous trench sands
formation. Guatemalan trench fill is much less and terrigenous and bioclastic sands of the
extensive with only a few small channels and no outer shelf. Trench sediments here bypass both
compelling evidence for deformation within the slope and shelf with little residence time on the
trench. shelf. Offshore Guatemala high sedimentation
Trench sediment lithology and texture corre- rates in the trench indicate that the muddy
late well with nearshore and onland geology. trench-fill largely bypasses the slope, although
The Mexican margin lacks a good forearc basin, local slope deposition does occur in a levee near
has a narrow shelf, highlands of crystalline rock San Jose Canyon and in lower slope basins.
immediately onshore, and numerous submarine Similarities in the coarse-fraction composition
 Sedimentation in the M i d d l e A m e r i c a Trench 117

of trench, slope, and shelf sands (all sands Long-term extrapolation of the amount of
contain bioclastic as well as terrigenous debris) trench fill offshore Guatemala is possible using
suggests that these sands have been derived Seely's (1978) model for development of the
from the shelf. Most sands accumulate in the forearc basin. The basin initially formed from
forearc basin. Sediments are suspended during oceanic crust trapped landward of a tectonic
flood or by wave action and carried over the ridge of oceanic crust. Sediments sampled in
shelf edge. The San Jose Canyon embayment the Esso Petrel #1 well within the basin docu-
probably acts as a catchment basin for this ment continued shallowing since the
muddy sediment and funnels it down slope and Palaeocene. It is probable that nearly all
into the trench. coarse-grained sediments delivered to the mar-
Before considering the possible tectonic gin were trapped within this forearc basin so
effects of varying trench wedge size, we need to that the trench has remained virtually sediment-
evaluate possible variations in the trench wedge free throughout the Tertiary. In fact, now that
through time. Factors which affect the volume the forearc basin is essentially filled at the
of a trench wedge include: the convergence rate present time, sediments might even be spilling
(which determines the rate that the trench over into the trench at a greater rate than in the
wedge is deformed or consumed), and the rate past.
at which sediments are supplied to the trench to The relative amounts of trench fill do seem to
build up the trench wedge (Karig & Sharman be related to the amount of accretion taking
1975; Schweller & Kulm 1978). Convergence place on the margins. The Southern Mexico
varies along the Middle America Trench with a margin, with the sizable accretionary wedge
rate of roughly 7 cm yr -1 off Southern Mexico abuting on to continental basement, has been
(Minster & Jordan 1978) and roughly accreting trench material since the Miocene
10 cm yr -1 off Guatemala (Molnar & Sykes (Moore et al. 1979a). The Guatemala margin,
1969). Sediment input to the trench is control- on the other hand, seems to have very little or
led by sediment load of rivers (reflecting cli- no active accretion taking place (von Heune et
mate and topographic relief), the presence of al. 1981), perhaps a direct consequence of the
sediment-trapping areas such as forearc and meagre trench fill. In both areas, the thickness
slope basins, and relative sea-level, which con- of pelagic cover on the subducted oceanic plate
trols whether rivers discharge into an estuary, is small (Fig. 4) so differences in the amount of
on to the shelf, or directly on to the shelf edge accretion are probably due to the different
and the heads of submarine canyons. amounts of trench fill present in these two
With these variables in mind, estimates of areas.
past amounts of trench fill based on present-day
amounts can be hazardous. The Pleistocene
Middle America Trench could have had more Conclusions
fill because: (1) the sea-level was lower and
rivers discharged nearer the shelf edge, and (2) We can make the following generalizations
the Pleistocene climate could have been wetter about sedimentation along the margins associ-
and resulted in greater runoff. ated with the Middle America Trench:
Even with these possible causes for increased 1. Trench sediment type and amount corre-
sedimentation, the Pleistocene trenches lates well with the tectonic environment of
offshore Guatemala and Mexico are inferred to the margin. Relatively thick, sandy trench
have not contained a great deal more sediment fill occurs where truncated crystalline
than at present. Current convergence rates highlands occur close to shore and where
result in roughly 1 km of underthrusting at the the shelf is very narrow offshore Mexico.
trench over the last 10 000 yr, or since the Thinner, muddy trench fill occurs where
beginning of the Holocene, a time when trench volcanic highlands occur far from shore
fill would have been greater than at present. and where a large forearc basin intervenes
Yet the width of trench fill offshore both Mex- between the volcanic arc and shelf edge
ico and Guatemala is 5-6 km, so most trench offshore Guatemala.
fill should consist of relict Pleistocene sedi- 2. Trench fill correlates well with the posi-
ments. tion of major submarine canyons that
Therefore, we infer that trench fill has been supply sediment to the trench, a point
small offshore Guatemala during the late Qua- already emphasized by Underwood &
ternary, and that there have been appreciably Karig (1980).
greater amounts of fill offshore Mexico than 3. Slope sediments are similar in both areas
offshore Guatemala. and consist mainly of muddy, hemipelagic
118 K . J . McMillen et al.
sediments. Bypassing of the slope occurs result in t h e m u c h g r e a t e r a m o u n t of
in b o t h areas, a l t h o u g h t h e r e is local de- o b s e r v e d a c c r e t i o n t h e r e t h a n off G u a t e -
p o s i t i o n n e a r large s u b m a r i n e c a n y o n s . m a l a , w h e r e t r e n c h fill is small a n d prob-
4. S e d i m e n t a t i o n o n a n a r r o w shelf, such as ably has b e e n small in t h e past.
occurs offshore Mexico, does not provide
significant d e t r i t u s for r e m o v a l to t h e
t r e n c h , as m u c h g r e a t e r a m o u n t s of sedi- ACKNOWLEDGMENTS: Cores and seismic data were
m e n t are d e l i v e r e d t h e r e directly f r o m collected on NSF contract CU-TEX-2S907-3 to the
rivers. C o n v e r s e l y , t h e w i d e shelf o f f s h o r e University of Texas Marine Science Institute. We
G u a t e m a l a c o n t r i b u t e s m o s t of t h e sedi- wish to thank the captain and crew of the RV Idea
m e n t to s u b m a r i n e c a n y o n s a n d the Green and the assistance of Helen Kim, John
Kumselman and staff, and Fran Sebak. Acknowledg-
trench.
ment is made to the donors of the Petroleum
5. R e l a t i v e a m o u n t s of m a r g i n a c c r e t i o n m a y
Research Fund, administered by the American Chem-
be c o n t r o l l e d by t h e a m o u n t of t r e n c h fill ical Society, for partial support of Enkeboll and
in t h e M i d d l e A m e r i c a T r e n c h , as o c e a n Moore's participation in this study. Don Hussong
plate s e d i m e n t t h i c k n e s s c o m i n g into t h e allowed us to use the line drawing from the Hawaii
s u b d u c t i o n z o n e is so low (Fig. 4). T h e Institute of Geophysics profile across the Guatemala
t h i c k e r t r e n c h fill o f f s h o r e M e x i c o m a y trench.

References
BI-RNARD, W. D. & MCMANUS, D. A. 1973. Plankto- tinental margin truncation along the northern
nic foraminiferan-radiolarian stratigraphy and Middle America Trench. Bull. geol. Soc. Am.
the Pleistocene-Holocene boundary in the north- 89, 265-76.
east Pacific. Bull. geol. Soc. Am. 84, 2097-100. & SIIARMAN,G. F. 1975. Subduction and accre-
BOUMA, A. H. 1969. Method for the Study of tion in trenches. Bull. geol. Soc. Am. 86, 377-89.
Sedimentary Structures. Wiley, London, 450 pp. KING, P. B. 1969. Tectonic Map o]" North America.
CASl:,Y, R. E, & BAUER, M. A, 1976. A seasonal study U.S. Geol. Survey.
of Radiolaria and Foraminifera in the water LARSON, R. L. & CHASE, C. G. 1970. Relative
overlying the south Texas outer continental velocities of the Pacific, North America and
shelf. Abstr. Prog. geol. Soc. Am. 8-1, 11. Cocos Plates in the Middle America Trench
DE CZE-:RNA,A. 1965. Reconocimiento geologico en la region. Earth planet. Sci. Lett. 7, 425-8.
Sierra Madre dei Sur de Mexico, entre Chilpa- LUNDBERG, N. 1981. Evolution of the slope landward
ningo Y Acapulco, Estado de Guerrero. Bol. of the Middle America Trench, Nicoya Peninsu-
Inst. Geo, Univ. Nac. Aut. Mexico, 62, 76 pp. la, Costa Rica (this volume).
DICKINSON, W. R. 1970. Interpreting detrital modes McMU,LEN, K. J. 1979. Radiolarian ratios and the
of graywacke and arkose. J. sediment. Petrol. 40, Pleistocene-Holocene boundary. Trans. Gulf
695-707. Coast Assoc. geol. Soc. 29, 298-301.
DINKELMAN, M. G. 1973. Radiolarian stratigraphy: 1981. Radiolarians from the Southern Mexico
Leg 16 Deep Sea Drilling Project. Initial Rep. active margin, DSDP Leg 66. Initial Rep. Deep
Deep Sea drill. Proj. 16, 747-813. Sea drill. Prof. 66, in press.
DUNCAN, J. R., FI,OWER, G. A. & KULM, L. D. 1970. - - & BACHMAN, S. B. 1981. Bathymetry and ver-
Planktonic foraminiferan-radiolarian ratios and tical tectonics of Leg 66 DSDP sediments. Initial
Holocene-late Pleistocene deep-sea stratigraphy Rep. Deep Sea drill. Proj. 66, in press.
off Oregon. Bull. geol. Soc Am. 81,561-6. MA/THEWS, D. J. 1939. Tables of the Velocity of
ENKEBOLL, R. H. 1981. Sedimentary petrology and Sound in Pure Water and Sea Water for Use in
provenance of sands and gravels from the Middle Echo-sounding and Sound-ranging. The Hydrog-
America trench and trench slope: Guatemala raphic Department, Admiralty, London, 52 pp.
and southern Mexico. Bull. Am. Assoc. Petrol. MILLIMAN, J. D. & EMERY, K. O. 1968. Sea levels
Geol., in press. during the past 35 000 years. Science, 162, 1121-
Hays, S. D. 1970. Stratigraphy and evolutionary 3.
trends of Radiolaria in North Pacific deep-sea MINISTER, J. B. & JORDON, T. H. 1978. Present-day
sediments. Mem geol. Soc. Am. 126, 185-218. plate motions. J. geophys, res. 83-Bll, 5331-54.
IBRAHIM, a . K., LATHAM, G. V. & LADD, J. W. 1979. MOLNAR, P. & SYKES, L. R. 1969. Tectonics of the
Seismic refraction and, reflection measurements Caribbean and Middle America regions from
in the Middle America Trench offshore Guatem- focal mechanisms and seismicity. Bull. geol. Soc.
ala. J. geophys Res. 84-10, 5643-9. Am. 80, 1639-84.
KhRm, D. E. 1974. Tectonic erosion of trenches. MOORE, J. C., WATKINS, J. S., BACHMAN, S. B.,
Earth planet. Sci. Lett. 21,209-12. BEGHTEL, F. W., Bua~r, A., DIDYK, B., LEGGETr,
, CARDWELL, R. K., MOORE, G. F., & MOORE, J. K., LUND~EaG, N., McMILLEN, K. J.,
D. G. 1978. Late Cenozoic, subduction and con- NIITSUMA, N., SHEPHARD,L. E., SHIPLEY,T. H.,
 S e d i m e n t a t i o n in the M i d d l e A m e r i c a T r e n c h 119

STEPHAN, J. F. & STRADNER, H. 1979a. Middle slope model. In: BURK, C. A. & DRAKE, C. L.
America Trench. Geotimes, 24, 20-2. (eds). The Geology of Continental Margins,
- - , SHIPLEY,Z. H., BACHMAN,S. B., BEGHTEE, 249-60. Springer-Verlag, New York.
1~. W., BUTT, A., DIDYK, B., LEGGETY, J. K., SHIPLEY, T. H., MCMIEEEN, K. J., WATKINS, J. S.,
LUNDBERG, N., MCMIEEEN, K. J., NIITSUMA,N., MOORE, J. C., SANDOVAL-OCHOA,H. & WORZEL,
SHEPHARD, L. E., STEPHAN, J.-F. & STRADNER, J. L. 1978. Continental margin and lower slope
H. 1979b. Progressive accretion in the Middle structures of the Middle America Trench near
America Trench southern Mexico. Nature. Lon- Acapulco, Mexico. Mar. Geol. 65-82.
don, 281, 638-42. , LADD,J. W., BUFFEER,R. T. & WATKINS,J. S.
NIGRIN~, C. A. 1971. Radiolarian zones in the Quater- 1981. Tectonic processes along the Middle
nary of the equational Pacific Ocean. In: America Trench inner slope (this volume).
FUNNEEL, B. M. & RIEDEE, W. R. (eds). The STEWART, R. 1978. Neogene volcaniclastic sediments
Micropalaeontology of Oceans, 443-61. Cam- from Atka Basin, Aleutian Ridge. Bull. Am.
bridge University Press. Assoc. Petrol. Geol. 62, 87-97.
PRINCE, R. A., RESIG, J. M., KUEM, L. D. & UNDERWOOD, M. B. & KARIO, D. E. 1980. Role of
MOORE, T. C. 1974. Uplifted turbidite basins on submarine canyons in trench and trench-slope
the seaward wall of the Peru Trench. Geology, 2, sedimentation. Geology, 8, 432-6.
607-11. VONHEUNE, R., AUBOIN,J., AZEMA,J., BLACKINTON,
SCHWELLER, W. J. & KULM, L. D. 1978. Depositional G., CARTER, J., COULBOURNW., COWAN, D.,
patterns and channelized sedimentation in active CURIALE, J., DENGO C., FAAS, R., HARRISON,
eastern Pacific trenches. In: STANLEY, D. J. & W., HESSE, R., HUSSONG, D., LADD, J.,
KEELING G. (eds). Sedimentation in Submarine MUZYLOR, N., SHIKI, T., THOMPSON, P. &
Canyons, Fans, and Trenches, 311-24. Dowden, WESTBERG, J. 1980. Leg 67: The Deep Sea Drill-
Hutchinson & Ross, Stroudsburg, Pennsylvania. ing Project Mid-America Trench Transect off
SEEEY, D. R. 1978. The evolution of structural highs Guatemala. Bull. geol. Soc. Am. 91,421-32.
bordering major forearc basins. In: WATKINS,J.
S., MONTADERT, L. & DICKERSON, P. W. (eds).
Geological and Geophysical Investigations of ' ' '& ~ , 1981. A summary
Continental Margins. Mere. Am. Assoc. Petrol. of Deep Drilling Project Leg 67 shipboard results
Geol. 29, 245-60. from the Mid-America Trench transect off
, VALE, P. R. & WALTON, G. G. 1974. Trench Guatemala (this volume).

K. J. MCMILLEN, Gulf Science and Technology Co. P.O. Box 2038, Pittsburgh,
Pennsylvania 15230, U.S.A.
R. H. ENKEBOEE,H. Esmaili and Associates, Inc., 2718 Telegraph Avenue, Suite 200,
Berkeley, California 94705, U.S.A.
J. C. MOORE, Earth Sciences Board, University of California, Santa Cruz, California
95064, U.S.A.
T. H. SHIPLEY, Scripps Institution of Oceanography, La Jolla, California 92093,
U.S.A.
J. W. LADD, Lamont-Doherty Geological Observatory, Palisades, New York 10964,
U.S.A.
 A summary of Deep Sea Drilling Project Leg 67 shipboard
results from the Mid-America Trench transect off Guatemala

Roland von Huene, Jean Aubouin, Jacques Azema, Grant Blackinton, Jerry A.
Carter, William T. Coulbourn, Darrei S. Cowan, Joseph A. Curiale, Carlos A.
Dengo, Richard W. Faas, William Harrison, Reinhard Hesse, Donald M.
Hussong, John W. Ladd, Nikita Muzylov, Tsunemasa Shiki, Peter R. Thompson
& Jean Westberg
SUMMARY: The Middle America Trench off Guatemala was transected by 24-channel
seismic reflection surveys, seismic-refraction surveys, and drilling with the Glomar
Challenger. The drilling was done at three sites on the oceanic Cocos plate and four sites
on the Caribbean plate. These plates converge at about 10 cm yr -1. At all drill sites
sediment of upper Miocene to Quaternary age is almost entirely hemipelagic mud with
interbedded thin volcanic ash, except in the trench where mud and fine sand turbidites less
than 400 000 yr old are ponded. However, the underlying rocks are very different. On the
oceanic Cocos plate a basal chalk sequence of lower and middle Miocene age is overlain
by a thin section of abyssal clay. At a site only 3 km landward of the trench axis where
drilling penetrated the slope deposits we recovered a Cretaceous to lower Miocene
claystone sequence resting on a section containing igneous rock of continental affinity. A
large net subduction of sediment along with ocean crust has occurred during the present
(Miocene-Quaternary) episode of subduction and perhaps parts of the continental
framework have been subducted as well. However, no current model satisfactorily
explains the surprising occurrence of Cretaceous-Miocene claystone at the foot of the
trench slope.

The Mid-America Trench transect off Guate- Geological setting

mala is the second of two geophysical and
drilling transects recommended for the Interna- Investigations of global plate motions and
tional Program of Ocean Drilling ( I P O D ) geophysical studies provide a convincing case
across the convergent margin of southern for normal convergence of the Cocos and
Mexico and Central America. The first, a trans- Carribbean plates along the Mid-America
ect off Oaxaca, Mexico, was drilled on Leg 66 Trench off G u a t e m a l a (Fig. 1) (Minster &
of the G l o m a r C h a l l e n g e r to investigate a con- Jordan 1978; Jordan 1975; Molnar & Sykes
vergent margin where the leading edge is a 1969). The rate of convergence is estimated at
subduction complex tectonically accreted dur- 10 cm yr -1 since the late Miocene and the
ing the Neogene, and where a truncated Pre- compressional tectonism implied by plate con-
cambrian and Palaeozoic continental vergence is compatible with the geology of the
framework extends oceanward to the mid-slope trench slope deduced largely from geophysics
(Moore, Watkins et al. 1979, 1981; McMillen et (Seely et al. 1974; Ladd et al. 1978; Ibrahim et
al. 1981). Off Guatemala our principal objec- al. 1979). The oceanic crust continues beneath
tive was to study a margin where geophysical the trench landward slope and is overlain by a
data indicated continuous accretion and im- thick sequence of landward-dipping reflections
brication (Seely et al. 1974). Seely and his (Fig. 2). Near the top of the trench slope strata
colleagues made a case for the imbricate thrust deduced from landward-dipping reflections be-
model that is widely accepted, and this model is come the seaward flank of a forearc basin that is
supported by the multichannel seismic reflec- filled with as much as 8 km of sediment. This
tion site surveys (Ladd et al. 1978; Ibrahim et al. thick sediment probably underlies the coast
1979). above which tower the volcanoes of the magma-
This paper is patterned after a more complete tic arc. From a drill hole at the edge of the shelf,
report prepared on the preliminary results Seely (1979) documented Palaeocene and
of Leg 67, where the data from the cores Eocene uplift of the shelf edge and showed that
are discussed in greater detail (von Huene, the Cenozoic section is broken by a series of
Aubouin et al. 1981). hiatuses, one of the most visible ones in seismic

121
122 R. yon Huene e t al.

Mid-

,.. SEABEAM " ~ - . ~ ~

el ~'LArE ": / l

91 o I 0 ' 91o00 o 90050 , 90°40 '

1 ! I ! 1 I I

oo _..~,\~.C x
12050'

50OO
12°40'

iO0

• IPOD S I T E S
LEG 57

IZo30'
5 I 0 K I L O M E T ERS
t I

I i I I I I " I

Fl6. 1. Plate tectonic diagram of the Central American area showing the relation of the Caribbean
and Cocos plates along the Mid-America Trench. The Caribbean and the Cocos plate are converging
at 10cmyr -1 across the trench. The lower part of the figure is a bathymetric map made with a
Seabeam instrument (after Renard et al. 1980).

records c o r r e s p o n d i n g to a M i o c e n e angular s t a c k e d and tilted i m b r i c a t e slices, based on

u n c o n f o r m i t y . Seely et al. (1974) i n t e r p r e t the pervasive l a n d w a r d dips in seismic records, as
structure of the G u a t e m a l a n m a r g i n as o n e of well as a series of b e n c h e s on the lower slope
 A summary o f D S D P Leg 67 results 123

2t- ~ ~__4----------V-~0 ~ 33"--..

LJ-~L,, e ~ ~ 3.0 -.., - ~ Z53, :---.>._. , ~

6 7 " '~-'-.~:~ :'~ - ~ t / _ - 5 /~ / ~ . . -
4.1 47.
8 o., --. ', \// 6.5 -
Foreorc
~, Landward basin
"~ , ~ dipping reflective
6-.5 _ "" "-- -.. ,,, reflections sequence
6.8 "~- -,..

Reflection ~/f'~ ""

8-.0 8.1 from top
of i g n e o u s
o c e a n i c crust
I0 km
! NNE
- V.E.=x2.5
8.1

FI6. 2. Summary of seismic-reflection and refraction studies showing velocity structure and major
reflectors. The location of sites drilled by Glomar Challenger on Leg 67 are indicated.

which may be likely places for the surfacing of The trench landward slope is stepped into a
large thrust faults. sequence of benches and the lowest was well
surveyed. It continues for at least 30 km with
minor relief. The next higher bench is shown
Seabeam bathymetry for 20 km of its extent. Site 494 on the first
bench appears to be on a secondary low nose
The first results from a Seabeam survey by the trending diagonally across the bench.
RV J. Charcot (Renard et al. 1980) made after
the Leg 67 drilling, revealed an unexpected
topographic trend of horst and graben on the Drilling results
trench seaward slope. The Seabeam instrument
Cocos Plate
installed on the Charcot maps bathymetry in a
swath with width that is about 75% of the water The oceanic Cocos plate forms the seaward
depth (Renard & Allenou 1979). Individual slope of the Mid-America Trench off Guatema-
swaths are not distorted by inter-ships-track la. Ocean basement is overlain by a thin sedi-
navigational discrepancies and the effects of ment blanket that is uniformly 200-300 m thick.
diffractions from steep slopes are minimized. As the ocean crust is flexed down into the
The preliminary map shown in Fig. 1 will later trench the crust breaks into a sequence of horst
be adjusted to the limit of resolution which is and graben. The downward flexure forms a
less than 10 m (Renard & Allenou 1979). trench slope that drops about 2000 m into the
The Seabeam bathymetry indicates that trench in the area of the drilled transect. Site
structure on oceanic crust strikes about 30 ° 495 is on a horst about 22 km seaward of the
from the trend of the Mid-America Trench axis. trench axis and 1925 m above it (Fig. 2). The
This is also the trend of regional oceanic sediment follows topography (except where
magnetic anomalies (Fig. 1). A typical horst faulted) with uniform thickness, a feature often
and graben topography, commonly seen as interpreted in seismic reflection records as char-
ocean crust is flexed down into the trench, is acteristic of pelagic sediment.
very apparent in the Seabeam bathymetry. The sediment section penetrated at Site 495
Ridges formed by horsts are not completely (Fig. 3) records the early and middle Miocene
covered by sediment ponded in the trench axis path of this site as it passed northward through
and these ridges divide the trench fill into small the equatorial carbonate belt, subsiding down
isolated basins (Fig. 1). The trench fill basins the east flank of the East Pacific Rise, into a
terminate landward against a linear escarpment zone of normal slow pelagic deposition. Then
that forms the beginning of the trench landward during the late Miocene it drifted into a
slope. Site 500 on the trench floor was drilled hemipelagic environment influenced by a
along one of the ridges. terrigenous source (rather than the pelagic
124 R. v o n H u e n e et al.

Sub-bottom Depth (meters)

, o , ~o , o° , o,

LOWER MIOCENE JM M,OCENE U MIOCENEI PUOCENE,d QUATERNARY ! ~

-
~'.'H H ' ~ F ' H ; ' I - I- I" "1- "t- f I-i_~1-1..~1 I- I- ~,li'}l!II ' I t[: }!':' [if
,,;~:'i ~- $~I'
. I -t:~:tPHHHLF.
.I',+,I'.t. . . ,. ,. -. - : , I-, I'[,.~:, Li :,iJ] i! i i~ ! IJl:, ti~ i,i- ~ "or_
' ~ i--} " ~ OLIVE GRAY DIATOMACEOUS
NANNOFOSSIL-FORAM. OOZE AND CHALK HEMIPELAGIC MUD
l,i
_~ I ~ ~,~ m
LM,OCENE ~=.'U
m _ L.
~:~u~'P"-- I U.
QUATERNARY
' . . . . !
:)['i i ''t 7 ' i~=l;i'i:i! I I ', ,}:'~lli,::!i,
r : o)

t[FIIIF
IFII mz~o ~
m--m
I'M
im I O Z
Z
LOWER MIOCENE QUATERNARY
El l l l l l l l l l • L atom l i l l e mm
"" ~ " *t +~:i! i{:' :~:I ~
m-.-t
m~
r-¢')
- ~c

.~1 MIOCENE
• I I I I II
I~
•
PLIOCENE - QUATERNARY j
! ~!i'i~i ~'ii'iIIi~ii ~
i~l,i~i:i!iii:i:,, I, t , f
DIATOMACEOUS MUD AND MUDSTONE

U. ]UCl I'. MIOC ~ I [

.m CRETACEOUS ME] M EOCENE ~ OLI~L~ ] LOWERQUATERNARY U. QUATERNARY
j , = ,,l=ll=l L I' 'I,~1 . . . . . . . . . . .
~ ~ ril,i!i,i,il~!~:,,,:,:~,l,~:,
~'
i~i:i:l:>i-w,~ , i ~,,ill g,l!:,!l;iiil
;i : i:,~,~l ,,L,~:
'::~" ~,~-
ii ii ~'i
ill !!ti 'I ~' I!,I," I,~'
,,l/i~ifi,<!,'t:~:,i',l
.z ~ =0 :o o >""4
O=Z~I ~ P O
~ CE DIATOMACEOUS
m
Z
~L L. u. 1 ,0PL,OCENE,
- QUAT. [
m
- PLIOCENE
" ' . . . . . . . . . . .
QUATERNARY
Z
~cnp-,
'l:!jJ °: :,1 Ii :r~i %~ 'i Fii:;:i: :, lii;i;lii.i illl:li
SAND, PEBBLY MUD, LIMESTONE& MUDSTONE 1 DIATOMACEOUSNANNOEOSSIL MUD "Io
m

~o L. MIOCENE 'Mp
I MIOC PILIOC. QUATERNARY
l• . m=-- I J"-- el' I ! ? =l l .===.=

:ili~il,:~ i!i~ ' !t ':i' .!:=l!=l, .,-, ,.,~,,,~.,~.:t:..,

'" !'J';',' i I .... !!!
SANDY MUD. PEBBLY MUD DIATOMACEOUS NANNOFOSSIL MUD
&MUDSTONE
....
Fro. 3. Simplified biD- and litho-stratigraphy of sites drilled on Leg 67 ( f r o m y o n H u e n e , A u b o u i n et
al, 1981),

environment previously inferred from seismic the hemipelagic section the lowest beds were
records). The upper hemipelagic section indi- deposited when the site was about 900 km
cates that a surprisingly large amount of sedi- seaward of the trench. Such great distances of
ment was transported a great distance seaward sediment transport are not seen in the present
across the trench. If present rates of converg- pattern of hemipelagic sedimentation as re-
ence have been constant during deposition of corded at other D S D P sites in the area (van
 A s u m m a r y o f D S D P L e g 6 7 results 125

Andel, Heath et al. 1973; Hays, Cook et al. formed by the emergence of a large thrust fault
1972). at the seafloor. Below the seafloor is a weak
sequence of reflections that dip gently land-
ward. These in turn overlie a strong reflection
Trench floor
from the top of the igneous oceanic crust. When
The trench floor has ridges separated by the vertical axis of the seismic reflection record
intermittent flat areas which are underlain by is converted from time to depth the beds of the
horizontal seismic reflections (Fig. 2). These reflective sequence are nearly horizontal and
reflections are presumed to represent ponded have a thickness of about 1200 m. At Site 494
sediment which is in turn underlain by weak on the bench just above the trench floor, we
reflections that dip gently landward and are recovered a 366 m sequence of rock in normal
presumed to be the deep ocean sediment sec- stratigraphic succession that ranges in age from
tion riding into the trench axis on the Cocos Late Cretaceous to Quaternary (Fig. 3). The
Plate. Upper Cretaceous rock is about 900 m above
Sites 499 and 500 were drilled on the trench the top of igneous oceanic crust (Fig. 4).
floor and they include seven holes most of
which bottomed in basalt or basalt rubble at
various depths (Fig. 3). The trench fill, as seen
in seismic records, is represented by an upper
Quaternary (slightly less than 400 000 yr old)
sequence of alternating muddy and sandy turbi-
dites with a microfossil assemblage transported
from the shelf and slope. This turbidite
sequence is underlain by a lower Miocene to
Quaternary oceanic sequence like the one
recovered at Site 495 on the Cocos plate, but
about half as thick. At Site 500 this section is
cut by a normal fault that probably formed
during the development of the horst and graben
structure of the trench seaward slope. The
sediment beneath the trench floor is compacted
normally and shows few signs of compressional
deformation even against the trench landward
slope.
The drilling results indicate that the trench is
more complex than would be expected from the
seismic records. Basalt and basalt rubble occur
at a variety of levels, probably from a basalt
basement structure not revealed by the seismic
records. The trench fill has no pronounced
lateral facies differences transverse to the
trench axis as might be expected from axial
turbidity current channels (Piper et al. 1973;
yon Huene 1974).

Trench lower slope

Fl6. 4. Diagram of interpreted subsurface
The trench lower slope has a series of geology across the Mid-America Trench
benches. The lowest of these is 3 km from the off Guatemala based on geophysics and
trench floor, about 500 m above it, and extends Leg 67 drilling. The topography is after the
more than 30 km parallel to the trench. In our Seabeam survey in Fig. 1 and was drawn by
Tau Alpha.
previous article (von Huene et al. 1980) we
refer to bathymetry from a detailed survey by
the Kana Keoki that suggests re-entrants in the At Site 494 upper Miocene to Quaternary
terrace. A Seabeam survey (Renard et al. 1981) slope deposits cover a sequence similar to that
has shown that the bench is a linear feature reported by Seely (1979) from the edge of the
without re-entrants. This type of topographic continental shelf. The environments repre-
feature is often interpreted as having been sented by this sequence, the nature of hiatuses,
126 R. von Huene et al.

and their relation to the sequence of the shelf are part of an imbricated stack resulting from
are interpreted in down-hole order as follows: subduction as has been inferred by Seely et al.
1. A Pliocene and Quaternary deposit con- (1974), Ladd et al. (1978) and Ibrahim et al.
sisting principally of sediment transported (1979). Drilling was terminated short of the
from the shelf and upper slope to this site. primary objectives for safety considerations be-
2. An upper Miocene unconformity coinci- cause we encountered gas hydrate. Slope de-
dent with the widespread upper Miocene posits, as interpreted from seismic reflection
unconformity on the shelf. records, unconformably drape the rock that
3. A sequence of distal, terrigenous, makes up the bulk of the continental margin
hemipelagic clay that accumulated during (Fig.2). The slope deposits thin toward the
the early Miocene on seafloor near the trench lower slope.
CCD and at rates an order of magnitude Site 497 was drilled in olive-grey mud with
less than age-equivalent deposits on the varying biogenic and vitric tuffaceous compo-
adjacent shelf (3 versus 100 m my-l). nents (Fig. 3). It is essentially a uniform section
4. An upper Eocene-Oligocene hiatus which of mixed hemipelagic and terrigenous muds
is the age-equivalent of a widespread un- interspersed with a thin pebbly mudstone that
conformity in the adjacent shelf section represents mass movement. A similar lithology
but of unknown origin in the section was penetrated at Site 496 where the upper
drilled. sequence is biogenic mud of Quaternary age
5. An Eocene hemipelagic clay deposited underlain by biogenic sandy mudstone of
below the foraminiferal CCD but above Pliocene and Miocene age.
the nannofossil CCD at rates at least five Sediments recovered from the mid-slope area
times less than age-equivalent shelf de- are rich in terrigenous detritus and contain
posits. volcanic ash beds, and yield a microfossil
6. A hiatus of unknown origin but of an age assemblage displaced from upslope. The fora-
that is represented by thick, widespread miniferal assemblage at Site 496 indicates subsi-
Palaeocene sediments below the adjacent dence of the site in Miocene time from shelf to
shelf. lower bathyal depths. This subsidence is coinci-
7. An Upper Cretaceous claystone that dent with a subsidence reported by Seely (1979)
accumulated in an open ocean environ- from the adjacent shelf edge.
ment above the CCD at depths about
equivalent to, but at rates apparently less
than age-equivalent rocks below the
Conclusions
present shelf.
8. A contact of unknown origin between Geophysical data from the Mid-America
claystone and igneous rocks. Trench transect off Guatemala contain nothing
9. Igneous rock originally of basaltic and unusual when compared with geophysical data
andesitic composition and atypical of from other convergent margins although the
igneous oceanic crust; the nature of the records do not resolve a great deal of structure
rock body sampled is unknown. at depth along the trench landward slope, as is
This sequence shows microfracturing from seen for instance along the companion transect
tectonic stress beginning with rock at the base off Mexico (Moore, Watkins et al. 1979; Shipley
of the Pleistocene, but if there are thrust faults 1981). The results from drilling off Guatemala
cutting the drilled section they are not large also reveal nothing unexpected in the post-
enough to give major discontinuity in rock ages. middle Miocene sediment. The unexpected re-
At Site 498, only the slope section (olive-grey sults occur in the one drill hole that sampled
mudstone) appears to have been penetrated. below the cover of slope deposits, returning
We interpret the difference in thickness of the rock that is much older than that anticipated
Miocene sections of olive-grey mudstone as from previous geophysical work. Perhaps
reflecting relief on the upper Miocene uncon- another unexpected aspect is the scant evidence
formity. for compressional deformation. However, it is
difficult to detect structure geophysically that
Trench middle slope has less than a few hundred metres amplitude in
very deep water and structure defined from the
Sites on the trench middle slope were posi- study of isolated drill cores is no longer than
tioned to penetrate the slope deposits and to about 10 cm. Thus the lack of more compres-
sample the underlying landward-dipping reflec- sional structure can be attributed to a lack of
tions. We planned to test whether the reflectors information at the appropriate scale.
 A summary o f D S D P Leg 67 results 127

The ocean crust entering the Mid-America provenance near a terrigeneous source. The
Trench off Guatemala has a Miocene section Miocene section at Site 494 originated seaward
that clearly records northward passage of the of those deposits recovered from the mid-slope
Cocos Plate first beneath the equatorial carbon- at sites 496 and 497. The 494 section was
ate belt and then into proximity with an upper deposited close enough to land or entirely at
Miocene to Present terrigenous source. The latitudes north of the carbonate zone of high
distance of Site 495 from land when the ter- productivity not to be inundated by carbonate
rigenous sediment first reached it is problema- ooze.
tical, but is certainly hundreds of kilometres if
plate reconstructions for this area are correct.
As the ocean plate was flexed down into the Tectonic interpretation
trench, the crust was tensionally faulted into
horst and graben that persist beneath the trench An interpretation of our shipboard data in the
until hidden from view under the trench land- context of previous geological studies is difficult
ward slope. in particular because the data from Site 494
The trench floor is distinguished from the do not fit the commonly applied steady-state
trench seaward slope by the ponded turbidites. model of tectonic accretion (Fig. 4); nor do any
The ponded sediment contains a microfauna of the less often invoked models seem to apply
less than 400 000 yr old, attesting to the youth without much modification. Three end-member
of the present trench floor. The youthful sedi- models or explanations, all of which may apply
ment together with active seismicity and con- in part, include the slumping of a large block,
tinental volcanism, are strong arguments for subduction of sediment without accretion, or
rapid convergence of oceanic and continental tectonic erosion.
crust. Nonetheless, the failure to find any in- A slump that displaced the large bench on
crease in compressional structure or even initial which Site 494 was drilled would have origin-
tectonic consolidation of trench fill and under- ated in an environment seaward of the shelf
lying oceanic sediment at the foot of the trench edge (Seely 1979) and of Sites 496 and 497
landward slope is puzzling. based on the slope sediment facies. Since there
On the trench landward slope, a cover of are no slump scars as large as the proposed
lower Miocene to Present sediment drapes the slump block, and since the shallowed uncon-
underlying rocks which are generally obscured formity is of early Pliocene age, early Pliocene
in seismic reflection records. Few DSDP holes or previous slumping seems to be required.
in any convergent margin have penetrated Although a slump block would alleviate the
deeply into the zone of obscured reflections so problematical absence of compressional struc-
commonly observed in convergent margins. In ture at the foot of the trench slope and provide
seismic reflection records, the upper layer con- a means of transporting an older accreted sec-
sists of reflections paralleling the seafloor and tion from upslope to the present trench floor,
the obscured zone occasionally reveals land- slumping does not explain the absence of accre-
ward-dipping reflections (see for instance, tion in the Pliocene and Quaternary. During
Moore, Watkins et al. 1979). The upper layer is the minimum period of accretion allowable by
presumed to be slope deposits. These reveal slumping 4 Myr ago, 280 km of ocean crust has
more small-scale deformation at depths of been subducted and the totally dewatered
about 200 m than was seen in the sediment of volume of oceanic sediment on 280 km of ocean
the trench at Site 500 almost at the trench land- crust greatly exceeds the maximum space avail-
ward slope. able for undrilled accreted material. Thus
At Site 494, only 3 km landward from the slump explanation also requires considerable
trench floor, we penetrated the slope deposits subduction of sediment.
and found rock that is surprisingly old and The concept of sediment subduction implies
overconsolidated. The sedimentary section disposal of a great amount of sediment some-
rock here was first deposited in water shallower where down the Benioff zone since the present
than the CCD and then at depths at or just convergence began (early Miocene). Perhaps
below the CCD. The section has distal con- the rather surprising section at Site 494 is part
tinental affinities, that is to say, it received of an elevated block brought in on the Cocos
terrigeneous sediment but probably not trench plate from the west. The block would have
slope deposits. From the shipboard studies we arrived in the pre-middle Miocene based on the
are unable to constrain the original environ- oldest slope deposits at Site 498 and the
ments of deposition much more than to infer youngest anomalous deposits at Site 494.
that they are of a base of slope or open ocean There is no evidence in the post-middle
128 R. v o n H u e n e et al.

Miocene slope deposits at Sites 494 and 498 for ble and tectonic erosion would require subduc-
collision between a crustal fragment and the tion of the continental framework (previously
Caribbean plate. Thus the total sediment sub- accreted rock) a difficult process to conceptual-
ducted must involve all of the post-middle ize. However, tectonic erosion is an efficient
Miocene sediment on the oceanic crust that was way to truncate the base of the slope despite the
subducted. Although possible, such a history conceptual problems of stuffing rock of less
raises the question of mechanism and the ulti- density beneath those of greater density or the
mate disposition of so much sediment. abrasion of the continental framework on a
Tectonic erosion is a third concept that might massive scale.
be applied to explain the drill data. Since the These explanations are briefly discussed here
vector of plate convergence is normal to the to provide some focus for subsequent study,
trench axis little strike-slip rifting seems possi- despite their problematical aspects.

References
HAYS, J. D., COOK, H. et al. 1972. Initial Rep. Deep STEPHAN, J.-F., BEGHTEL, F. W., BU~, A.,
Sea drill. Proj. 9, U.S. Govt Printing Office, DIDYK, B. M., NIITSUMA, N., SHEPARD, L. E. &
Washington. 1205 pp. STRADNER, H. 1981. Facies belts of the Middle
IBRAHIM, A. K., LATHAM, G. V. & LADD, J. 1979. America Trench and forearc region, southern
Seismic refraction and reflection measurements Mexico: results from Leg 66 DSDP (this
in the Middle America Trench offshore Guate- volume).
mala. J. geophys. Res. 84, 5643-9. PIPER, D. J. W., VONHUENE, R. E. & DUNCAN,J. R.
JORDAN, Z. H. 1975. The present day motions of the 1973. Late Quaternary sedimentation in the ac-
Caribbean Plate. J. geophys. Res. 80, 4433-9. tive Eastern Aleutian Trench. Geology, 1, 19-
LADD, J. W., IBRAHIM, A. K., MCMILLEN, K. J., 22.
LATHAM, G. V., VON HUENE, R. E., WATKINS, J. RENARD, V., AUBOUIN, J., LONSDALE, P. & STEPHAN,
S., MOORE, J. C. & WORZEL, J. L. 1978. Tecto- J. F. 1980. Premiers resultats d'une etude de la
nics of the Middle "America Trench offshore fosse d'Amerique centrale au sondeur multifais-
Guatemala. In: Int. Syrup. Guatemala February ceaqux (Seabeam). C. r. Seances Acad. Sci.
4 Earthquake and Reconstruction Process, Paris, 291, 137-42.
Guatemala City, May 1978. SEELY, D. R. 1979. Geophysical investigations of
MCMILLEN, K. J., EUKEBOLL, R. H., MOORE, J. C., continental slopes and rises. In: WATKINS, J. S.,
SHIPLEY, T. H. & LADD, J. W. 1981. Sedimenta- MONIADI-:RI, L. & DI('KI~RSON, P. w . (eds).
tion in different tectonic environments of the Geological attd Geophy.sical Investigations of
Middle America Trench, southern Mexico and Continental Margins. Mum. Am. Assoc. Petrol.
Guatemala (this volume). Geol. 29, 245-6(I.
MINSTER, J. B. & JORDAN, T. H. 1978. Present-day SEELY, D. R., VAIL, P. R. & WALTON, G. G. 1974.
plate motions. J. geophys. Res. 83, 5331-4. Trench slope model. In: BURK, C. A. & DRAKE,
MOLNAR, P. & SYKES, L. R. 1969. Tectonics of the C. L. (eds). Geology of Continental Margins,
Caribbean and Middle America regions from 261-83. Springer-Verlag, New York.
focal mechanisms and seismicity. Bull. geol. Soc. VAN ANDEL, T. H., HEATH, G. R. et al. 1973. Init.
Am. 80, 1639-84 Rep. Deep Sea drill. Pro]. 16, 949 pp. U.S. Govt
MOORE, J. C., WATKINS, J. S., BACHMAN, S. B., Printing Office, Washington.
BEGHTEL, F. W., BuIT, A., DIDYK, B. M., YON HOENE, R. E. 1974. Modern trench sediments.
LEGGE'Iff, J. K., LUNDBERG,N., MCMILLEN, K. In: BURK, C. A. & DRAKE, C. L. (eds). The
J., NnTSUMA, N., SHEPARD, L. E., SHIPLEY, T. Geology of Continental Margins, 20%11.
H., STEPHAN, J. F. & STRADNER, H. 1979. The Springer-Verlag, New York.
Middle America Trench off Mexico. Geotimes, VON HUENE, R. E. & AUBOUIN, J. 1981. Leg 67:
24, 20-2. The Deep Sea Drilling Project Mid-America
-- , MCMILLEN, K. J., BACHMAN, S. B., Trench transect off Guatemala. Bull. geol. Soc.
LEGGE'FI', J. K., LUNDBERG, N., SH1PLEY, T. H., Am. in press.

ROLAND VON HUENE, U.S. Geological Survey, Menlo Park, California 94025,
U.S.A.
JEAN AUBOUIN,D6partement de Geologic Structurale, Universit6 Pierre et Marie
Curie, 4, Place Jussieu, Paris 75230, France.
JACQUESAZEMA,D6partement de Geologic Structurale, Universit6 Pierre et Marie
Curie, 4 Place Jussieu, Paris 75230, France.
GRANT BLACKINTON, Department of Geology and Geophysics, University of
Hawaii, Honolulu, Hawaii 96822, U.S.A.
 A summary of DSDP Leg 67 results 129

JERRY A. CARIER,Department of Geology and Geophysics, University of Hawaii,

Honolulu, Hawaii 96822, U.S.A.
WILLIAMT. COULBOURN,Scripps Institution of Oceanography, La Jolla, California
92093, U.S.A.
DARRELS. COWAN,Department of Geological Sciences, University of Washington,
Seattle, Washington 98195, U.S.A.
JOSEPH A. CURIALE,School of Geology and Geophysics, University of Oklahoma,
Norman, Oklahoma 73019, U.S.A.
CARLOSA. DENGO, Center for Tectonophysics, Texas A & M University, College
Station, Texas 77843, U.S.A. and ICAITI, Guatemala.
RICHARDW. FAAS, Department of Geology, Lafayette College, Easton, Pennsylva-
nia 18042, U.S.A.
WILLIAM HARRISON,School of Geology and Geophysics, University of Oklahoma
Norman, Oklahoma 73019, U.S.A.
RE1NHARDHESSE, McGill University, Department of Geological Sciences, Montreal
M3A 2A7, Canada, and Technische Universit~t, Mtinchen, Federal Republic
of Germany.
DONALD M. HUSSONG, Department of Geology and Geophysics, University of
Hawaii, Honolulu, Hawaii 96822, U.S.A.
JOHN W. LADD, Lamont-Doherty Geological Observatory, Palisades, New York
10964, U.S.A., and Marine Science Institute, University of Texas, Galveston,
Texas 77550, U.S.A.
NIKITA MUZYLOV, Geological Institute of USSR Academy of Sciences, Moscow,
U.S.S.R.
TSUNEMASA SmKI, Department of Geology and Mineralogy, Kyoto University,
Kyoto 606, Japan.
PETER R. THOMPSON,Lamont-Doherty Geology Observatory, Palisades, New York
10964, U.S.A.
JEAN WESTBERG,Scripps Institution of Oceanography, La Jolla, California 92093,
U.S.A.
 Evolution of the slope landward of the Middle America Trench,
Nicoya Peninsula, Costa Rica

Neff Lundberg
SUMMARY: The Nicoya Peninsula of Costa Rica represents an uplifted portion of the
trench slope break landward of the Middle America Trench, and is composed of the
Nicoya Complex and its sedimentary cover. The Nicoya Complex is upper Mesozoic
oceanic crust, showing the effects of continued igneous activity and deformation probably
due to the Late Cretaceous initiation of subduction in an oceanic region. The sedimentary
cover reflects uplift and progressive deformation throughout the evolution of this
intra-oceanic arc-trench system.
Sediments of the Nicoya Complex are open-ocean deposits, comprising locally derived
sedimentary breccias made up of recycled Nicoya Complex material as well as more
typically pelagic radiolarian cherts, black shales and deep-water limestones. The basal unit
of the sedimentary cover is the mainly hemipelagic Campanian Sabana Grande Forma-
tion, which was probably deposited on the juvenile trench slope. In the Sabana Grande
Formation, basal siliceous mudstones give way up-section to the foraminifer-rich cal-
careous mudstones, representing the passage of the sediment surface up relative to the
CCD. Interbedded conglomerates and sedimentary breccias reflect recycling of Nicoya
Complex material, apparently by uplift and erosion of the forearc basement. The Sabana
Grande Formation is overlain by the ?Campanian to Palaeocene Rivas and Las Palmas
Formations which consist of thick, mainly volcanogenic, thin- to thick-bedded turbidites
and massive and/or pebbly sandstones characteristic of mid-fan facies associations. These
rocks were probably deposited in a forearc basin situated landward of a structural high.
In contrast, Palaeocene turbidites which crop out to the SW and seaward of the pro-
posed high are thin-bedded, associated with redeposited hemipelagic mudstones, and
interpreted as trench slope deposits. This thin-bedded turbidite/hemipelagic mudstone
unit is in turn overlain by Palaeocene or Eocene calcareous mudstones and Eocene
siliceous mudstones, both of which are interpreted to represent continued deposition on
the trench slope, respectively above and below the CCD. These ?Palaeocene and Eocene
mudstones are overlain unconformably by Eocene to ?Quaternary shallow-water clastic
and carbonate deposits, reflecting rapid Eocene tectonic uplift. The general trend of
shallowing in depositional environments through time is paralleled by an overall decrease
in intensity of structural deformation up-section, indicating progressive deformation in
this forearc terrane.

The Middle America Trench is becoming one of or all trench and pelagic deposits are subducted
the world's best known trenches, following two to at least shallow levels along with the oceanic
recent Deep Sea Drilling transects, yet the lithosphere, which itself may even tectonically
processes operating during the development of erode and subduct upper plate material (Karig
the inner trench slope remain controversial. 1974b; Scholl et al. 1980; Murauchi & Ludwig
The results of I P O D Leg 66 off southern Mex- 1980). These contrasting drilling results suggest
ico (Moore et al. 1979) indicate progressive considerable variation in processes associated
incorporation of trench deposits into the lower with subduction, even along the strike of a
trench slope, supporting the idea of sediment single trench. The critical questions concern the
accretion through some version of the popular evolution of the trench slope, and the fate both
imbricate-thrust model (Seely et al. 1974; Karig of sediments deposited in this environment and
1974a; Karig & Sharman 1975). However, Leg of the subducting oceanic plate with its mantle
67 results off Guatemala (von Huene, Aubouin of pelagic, hemipelagic, and trench deposits.
et al. 1980) are not readily explained by a simple While deep-sea drilling provides otherwise
imbricate-thrust model. Drilling off Guatema- unavailable data on active subduction zones,
la, as well as the transects off Japan (Scientific the information is limited to about the upper
Party 1980) and across the Mariana Trench kilometre of sediment, and drilling sites com-
(Hussong, U y e d a et al. 1978), suggest little or prise one-dimensional data sets which must be
no sediment accretion accompanying subduc- extended laterally by geophysical data. Onland
tion. These drilling results support instead a geological studies of presumed uplifted portions
process of sediment subduction, in which most of active trench slopes are therefore useful for

131
 L~
98 ° 84 °

/ t t~
..~ r~ ,.-,
,-,=~
...j ." N~TH AMERICAN PL l 1 cm/yr .. " ~v ~./
•,'i' 9,OOG~i:
tl / '; i t t
I °
!/c~. ~.o~.~,, c ~,~,_k . 1 "~ ~ " i'~ _ 1 8
G I~ "~ , ~ ~ __...~• .... :

MOTAGUA :1
o"¢~ f .. '
FAULT ' "" .
~... N CARAGUA ...

>~
SYSTEM
/. i.~'- ~ " RISE

~'-- ~'~7 ~'~' " "

~.=. ~ --... ,- CARIB q~PLATE
.~'~= -.:...°o:::,o~~...: i
7 cm/yr
.. , . . , , -
LEG 66 //~
f
IPOD /~,0
/
TRANSECT / ........ -':"?.~: ~~: , : .... " '.,. o~
g~
=, .
LEG 67 /
~1 0 °
IPOD
/ SANTA ELENA
T R A N S E C T PENINSULA
R _~.

10 °- NICOYA PENINSULA

COCOS PLATE 0 5OO

r.~ i i
km

/ I 9/8o t
 E v o l u t i o n o f the slope l a n d w a r d o f the M i d d l e A m e r i c a Trench 133

comparison, although by their emergent nature region (Donnelly 1975), and north along the
these terranes are somewhat removed from the Pacific offshore of Central America at least as
locus of tectonic activity within the lowermost far as Guatemala (Ladd et al. 1978).
trench slope.
The Nicoya Peninsula of Costa Rica, 600 km Nicoya Complex
SE of the IPOD Leg 67 transect off Guatemala
and 300 km NW of the Cocos Ridge (Fig. 1), The Nicoya Complex was originally de-
constitutes an emergent portion of the struc- scribed as 'an intensely folded rock series con-
tural high landward of the Middle America sisting of basalt, sedimentary units and intru-
Trench. This paper reports preliminary results sives' by Dengo (1962a), who interpreted it as
of an onland sedimentological and structural uplifted oceanic crust. Because of the complex
study of the Nicoya Peninsula, and the conse- history of continued volcanism, sedimentation
quent implications for the evolution of the slope and deformation in this region, however, more
landward of the Middle America Trench. recent studies have adopted a variety of defini-
tions of the Nicoya Complex (see Kuijpers 1980
for a review). In addition, the Nicoya Complex
Tectonic setting and geological has been shown to be more complicated than a
background simple piece of oceanic crust.
The Nicoya Complex has been divided into
Judging from the record of intermediate- two major units by several workers (De Boer
composition volcanic activity, rocks of the 1979; Kuijpers 1980; Schmidt-Effing et al.
Nicoya Peninsula apparently have been near 1981). The lower unit has generally been de-
the site of more or less continuous plate con- scribed as massive or pillow basalt, with little or
vergence since Late Cretaceous times. Thick no intercalated sediment, overlain by a se-
sequences of Upper Cretaceous and Palaeocene quence of siliceous sediments. The age of these
volcaniclastic strata were deposited on the sediments is predominantly Early Cretaceous
Nicoya Complex beginning in the early Campa- (latest Jurassic (Galli-Olivier 1977) or earliest
nian (Fisher & Pessagno 1965; Galli-Olivier & Cretaceous to Hauterivian or possibly Barre-
Schmidt-Effing 1977; Galli-Olivier 1979). mian, according to Schmidt-Effing et al. (1981);
Eocene to Quaternary volcanic rocks are cur- and Berriasian to Aptian according to Baum-
rently exposed in the Costa Rican magmatic arc gartner, in Kuijpers (1980)). Most workers
(Dengo 1962a; Pichler & Weyl 1973, 1975). have agreed that the lower unit of the Nicoya
The Cocos Plate is currently being consumed Complex represents the upper level of uplifted
beneath Costa Rica at 9 cm yr -1, along an oceanic crust, and some (Galli-Olivier 1979; De
azimuth of N29E (Minster & Jordan 1978; Fig. Boer 1979) have interpreted the serpentinized
1). The broad shelf landward of the Middle peridotite of the Santa Elena Peninsula, 30 km
America Trench south of the Tehuantepec north of the Nicoya Peninsula (Fig. 1) as the
Ridge (Fig. 1) comprises a large forearc basin, associated 'layer 3' ultramafic rocks.
about 100 km wide and at least 9 km deep, The upper unit of the Nicoya Complex is
bounded by a prominent structural high which more heterogeneous, however, and its inter-
underlies the shelf edge (Seely et al. 1974; Ladd pretation is disputed. De Boer (1979) has de-
et al. 1978; Seely 1979). This structural high is scribed an upper unit of pillow lava and volca-
similar to the antiformal structure of the Nicoya nic agglomerate, intercalated with chert,
Peninsula (Dengo 1962a), in which a cover of siliceous limestone, and tuffaceous sediment.
Upper Cretaceous to ?Quaternary marine He has interpreted this unit, along with gabbros
sediments are exposed on the landward and and diorites which intrude the lower unit of the
seaward flanks of a structural high (Fig. 2). This Nicoya Complex, as an early volcanic arc, built
antiformal high is cored by the basement unit of on oceanic crust represented by the lower
the Nicoya Peninsula, the ?Upper Jurassic to Nicoya Complex. Schmidt-Effing (1979) has
Upper Cretaceous Nicoya Complex. The subdivided an upper unit, composed of mainly
Nicoya Complex is thought to represent the submarine basalt flow material with irregular
basement of all southern Central America inclusions of sedimentary rocks, into six 'sub-
(south of Nicaragua) as well, upon which an complexes'. These are defined by differing ages
Upper Cretaceous through Quaternary magma- of sedimentary rocks associated with the more
tic arc has been built (Case 1974). Broadly prevalent volcanic rocks, both those incorpo-
similar terranes extend down the west coast of rated as xenoliths and those deposited on volca-
South America into Ecuador (Pichler et al. nic units, providing respectively maximum and
1974; Goossens et al. 1977), into the Caribbean minimum ages of volcanism. The minimum age
134 N. Lundberg
of latest volcanism in these six subcomplexes sists mainly of mostly basalt and ophitic dia-
ranges from Maastrichtian (latest Cretaceous) base, commonly including gabbroic rocks and
to Eocene, and the thickness of the upper less commonly 'plagiogranite'. He has reported
Nicoya Complex in general is estimated at a Cenomanian or early Santonian (late Cre-
2-3 km. Schmidt-Effing (1979) and Schmidt- taceous) age for a radiolarite intercalation in
Effing et al. (1981) have interpreted the upper the upper part of this unit, which he has inter-
unit as an oceanic plateau, similar to the preted as a nappe of younger oceanic crust
Ontong-Java Plateau, built on oceanic crust which has been thrust over the older oceanic
represented by the lower Nicoya Complex. crust of the lower Nicoya Complex. Galli-
Kuijpers (1980), on the other hand, has de- Olivier (1977, 1979) has also interpreted the
scribed a thin (300+ m) upper unit which con- Nicoya Complex as structurally complicated

~t

Oo Nicaragua /
%
Caribbean Sea
11 o
%
o .... .......
O00 0
~-~
.......~3n'~ Tip ),

\ y ~ lp
"°°°°
America J ' " ~ . / .A anama

Pacific Ocean ~ ~'~

85 ° 830 84°50'
9° 50'

N "
dKnc
+

\', , . ~ : ~ ~ o 1o
- ~ L I I
\ ~ km
85°00'
9° 40' + 9+40'
85 ° I0'.,~
"Ksg

A A'
i

85° I0' A
+
i ÷ l/JKnc % JKnc

No vertical exaggeration

Fro. 2. G e n e r a l i z e d geological m a p and cross-section o f the s o u t h - e a s t e r n p o r t i o n of the Nicoya

 Evolution of the slope landward of the Middle America Trench 135

oceanic crust, although he did not recognize of a multistage evolution of the Nicoya Com-
major subdivisions within it. He has interpreted plex (H. Wildberg, pers. comm. 1980) and may
the Nicoya Peninsula as an accretionary prism constrain its origin. Interestingly, there is no
formed of slabs of oceanic crust progressively clear distinction between a lithologically
sliced off a subducting lithospheric plate, im- homogeneous lower unit and a more heter-
plying the presence of imbricated thrust- ogeneous upper unit of the Nicoya Complex in
bounded packages of rock (Galli-Olivier & the south-eastern Nicoya Peninsula. In this area
Schmidt-Effing 1977; Galli-Olivier 1979). at least minor pieces of deeper, older levels of
These various interpretations of the upper the Nicoya Complex, including radiolarian
part of the Nicoya Complex are difficult to cherts of late Jurassic or early Cretaceous age
evaluate on the basis of field relations alone. (D. Jones, pers. comm. 1980) have been juxta-
Chemical analyses of the volcanic rocks are posed with the more widespread exposures of
potentially useful, but unfortunately few analy- shallower, younger portions of the Nicoya
ses are available. Analyses of three samples of Complex and of the sedimentary cover. Both
Nicoya Complex basalt by Pichler & Weyl younger and older terranes have relatively
(1975) suggest mid-ocean ridge origins, using abundant associated sediments, and both in-
Pearce & Cann's (1973) empirically-derived Ti clude radiolarian cherts, sedimentary breccias
versus Zr discrimination technique. More re- and black shales. Gabbroic rocks have intruded
cent compilations of chemical analyses have volcanic breccia as well as basalt; both layered
questioned the validity of this method, how- and non-layered gabbro are present.
ever, for ancient (and often altered) volcanic Notwithstanding the controversy over the
rocks (Garcia 1978; Hill 1979); and moreover, origin of the upper Nicoya Complex outlined
Pichler & Weyl (1975) did not differentiate above, the Nicoya Complex represents at least
major units within the Nicoya Complex. Addi- in part oceanic crust of late Jurassic and/or early
tional chemical analyses are currently in pro- Cretaceous age. Sediments of the Nicoya Com-
gress, in a study integrating more recent ideas plex document a history of highly variable

LEGEND
QuaternaryI--~ Alluvium
Contact
Pliocene ~ Montezuma Formation Dashed where approximated,
Miocene
dotted where inferred
Oligocene ~ Limestone and sandstone of Mal Pals
Eocene ~ Sandstone and mudstone of Cabo Blanco Fault
Dashed where approximated,
Paleocene~ has Palmas Formation dotted where inferred;
dot on downthrown side
~ Rivas Formation

Cretaceous ~ Sabana Grande Formation Thrust Fault

Dashed where approximated,
~ Nicoya Complex dotted where referred;
Jurassc(?) sawteeth on upper p/ate

I JKnc ~....Ksg KTr I v.IK,sg.~/KTr# KTr ~', KTrI$ Tip ,

I ~ ~ KTr
t Ksg
I [ JKnc -- "~ \ JKnc
' /

~'~'~"Bend in section

Peninsula. Alluvium and water not shown on cross-section.

136 N. Lundberg
Landward section
Qa Alluvium

~~7~:~i:/::i,/•
i))ili~i,!::iIii :
Seaward section i~,i~i:~j • ::iii
Las Palmas Formation

Alluvium
Montezuma Formation

Limestone and sandstone

of M a l P a i s
. 3.-"-2 2:
Sandstone and mudstone i ¢:.13111 =.,..:.:..:::,::
of C a b o B l a n c o
Rivas Formation
Sabana Grande Formation Rivas Formation

Nicoya Complex

metersI 500 Sabana Grande Formation

O
Nicoya Complex
[:(/'(.:1
h ::1
FIG. 3. Stratigraphic columns for seaward (SW) and landward (NE) flanks of antiformal high of the
SE Nicoya Peninsula. Symbols refer to legend to Fig. 2.

deep-sea sedimentation, and an analysis of the Because of abundant faulting, the recon-
sedimentary cover offers insights on the subse- structed thicknesses I report I consider to be
quent development of the terrane which now minimum values. Petrographic characterization
forms the structural high landward of the of units is based on reconnaissance study; more
Middle America Trench. detailed summaries of sedimentary petrography
and of measured sections are currently being
compiled.

Sedimentology NicoyaComplex
In order to expand the available data base I The sediments of the Nicoya Complex are
have mapped a strip along the SE coast of the mainly pelagic deposits and coarse sedimentary
Nicoya Peninsula, across the structural strike of breccias, and contain no significant continent-
the margin (Fig. 2). The study area includes derived quartzo-feldspathic detritus. Pelagic
considerable exposures of the sedimentary sediments are predominantly layers and pods of
cover and mainly shallow levels of the Nicoya radiolarian chert, with minor black shale and
Complex, apparently due to down-to-the-SE deep-water limestone, while the sedimentary
offset along a series of NW-trending faults breccias represent reworked Nicoya Complex
which cut the Nicoya Peninsula (De Boer material.
1979). Specifically I use lateral and vertical Although the Nicoya Complex includes
facies changes and regional unconformities, in- abundant volcanic breccias and locally tectonic
tegrated with major contrasts in structural style, breccias as well, the sedimentary breccias con-
to constrain the evolution of this trench slope. stitute a key lithology. They are typically
 Evolution of the slope landward of the Middle America Trench 137

Sabana Grande Formation

In several localities sedimentary breccias of
the uppermost Nicoya Complex are concor-
dantly overlain by the Sabana Corande Forma-
tion, a lithologically heterogeneous unit of early
to late Campanian age (Galli-Olivier &
Schmidt-Effing 1977; S. Hart & S. Percival,
written comm. 1980). The most complete sec-
tion of the Sabana Grande Formation (Fig. 5)
has a minimum thickness of 220 m. The true
thickness is probably somewhat more, as abun-
dant faults have apparently removed minor
sections. Green radiolarian-rich siliceous mud-
stone at the base of the Sabana Grande Forma-
tion is interbedded upsection with beds of vol-
canic ash (Fig. 5). Ash beds are commonly
5-10 cm thick, and rarely exceeded 30 cm in
thickness. Siliceous mudstone gives way upsec-
tion to calcareous mudstone rich in calcareous
foraminiferal tests, representing the passage of
the sediment surface up through the calcite
compensation surface. Volcanic ash beds re-
main prevalent, and distinctive interbedded
conglomerates contain clasts of Nicoya Com-
plex basalt and chert as well as smaller clasts of
the underlying siliceous mudstone, in a matrix
of foraminiferan-rich calcareous mudstone. The
calcareous mudstone is overlain by a 10 m
Fro. 4. Sedimentary breccia of the Nicoya
Complex, including principally light- section of red shale and very thin-bedded green
weathering basalt clasts and dark- volcanogenic turbidites. This thin elastic section
weathering chert clasts. grades up-section to a pink to white limestone,
indicating a gradual cessation of elastic input. In
one locality, however, sedimentary breccias
formed of angular to sub-rounded clasts of
similar to those of the underlying Nicoya Com-
basalt and lesser amounts of chert, with minor
plex are interbedded at the base of the lime-
limestone and gabbro (Fig. 4). The small
stone, suggesting continued uplift and erosion
amount of matrix generally comprises sand and
of the forearc basement. The limestone is up to
gravel apparently made up of broken clasts,
80 m thick and is very widespread in the south-
although occasionally the clasts are set in chert
eastern Nicoya Peninsula, marking the top of
or (very rarely) limestone. Beds of sedimentary
the Sabana Grande Formation. The limestone
breccia are locally interbedded with argil-
is late Campanian in age (Schmidt-Effing 1979),
laceous chert, and typically occur between
medium to thin-bedded, and composed largely
basalt and the overlying sedimentary units as
of planktonic calcareous foraminiferan tests,
well as interbedded with basalt at deeper levels
coccoliths, and fine-grained volcanogenic detri-
in the Nicoya Complex. Clasts are locally de-
tus. This white, pelagic limestone grades upsec-
rived, as demonstrated by a unique sedimentary
tion through pink argillaceous limestone to
breccia which overlies gabbro and is itself com-
shale and interbedded thin turbidites of the
posed largely of gabbroic clasts. These breccias
overlying Rivas Formation, reflecting a gradual
reflect considerable relief in a deep-sea en-
re-introduction of elastic detritus to this area.
vironment, which could be related to a spread-
ing ridge, an aseismic rise or seamount, a
fracture zone, or a trench. Pelagic sediments Rivas Formation
and sedimentary breccias are intercalated with
basalt on Isla Negritos Fuera, the easternmost The Rivas Formation is composed of elastic
island in the study area (Fig. 2), in a section turbidites which range in age from late Campa-
both underlain and overlain by basalt and vol- nian or early Maastrichtian (Late Cretaceous) to
canic breccia, indicating a complex history of Palaeocene (Galli-Olivier & Schmidt-Effing
volcanism and sedimentation. 1977; S. Hart & S. Percival, written comm.
138 N. Lundberg
1980). Reconstruction of numerous fault blocks They probably represent progradational pro-
along the best exposed and most complete cesses similar to those which form depositional
section of the Rivas Formation suggests a thick- lobes in mid-fan environments of submarine
ness of 1200+ m (Fig. 6). In general this unit fans (Walker 1978, 1980). Palaeocurrent data
begins with a shale-rich section with thin- (Fig. 6) and a preliminary reconstruction of
bedded turbidites, overlain by more sand-rich basin geometry, however, suggest deposition by
sequences of thick-bedded turbidites. These are axial flow down an elongate basin sub-parallel
overlain in turn by thick massive sandstones to the present-day trench, suggesting that a fan
which are commonly amalgamated and infre- geometry is probably not applicable to this unit.
quently pebbly, interbedded with shale and However, the identification of fan facies asso-
very thin-bedded turbidites. The thin turbidite ciations remains a useful exercise in order to
beds at the base of the Rivas Formation are characterize turbidite deposits (Nilsen 1980),
generally less than 30 cm thick with sharp bases and comparison with fan models allows an
and gradational tops. They do not form thin- initial interpretation of large-scale changes in
ning or thickening megasequences, and corres- depositional processes.
pond to Walker's (1978) basin plain and lower The massive sandstones which overlie the
fan facies associations, probably representing thick-bedded turbidites are up to 12 m thick
distal turbidites (Nilsen 1980). The overlying and generally show evidence of amalgamation
thick-bedded turbidites are generally 10- of sands deposited by more than one turbidite
120 cm thick, representing discrete depositional event. These sandstones and the interbedded
events, and in many cases form cycles of thick- shale and thin-bedded turbidites probably re-
ening- and coarsening-upward sandstone beds. present channel-fill and interchannel deposits

SABANA GRANDE FORMATION

i 1 I
200 I I EXPLANATION
I I
I I
I I
Limestone

m Shale and sandstone

% "°°°'0 ° ~c

Conglomerate
L

i
-".-".: = : .

Volcanic ash

Calcareous mudstone

i
; ........
Siliceous mudstone

o~ ~ Sedimentarybreccia
Flo. 5. Generalized section of the Sabana Grande Formation. Uppermost beds of the underlying
Nicoya Complex are shown to emphasize the concordant depositional contact and the similarity
between sedimentary breccias interbedded in Sabana Grande limestone and the Nicoya Complex.
Bed thicknesses not to scale.
 Evolution of the slope landward of the Middle America Trench 139

respectively (Walker 1978). Further upsection sandstones with clear evidence of channelling,
in the landward portion of the study area this overlain by coarse conglomerates interbedded
unit commonly exhibits thinning- and fining- with abundant shale and thin-bedded turbi-
upward cycles, in which massive sandstones dites. These constitute an upper fan facies
give way gradually to thin-bedded turbidites association, with the coarse material represent-
and shale, probably representing episodes of ing fill of feeder channels and the fine-grained
channel filling and abandonment (Walker & beds representing channel-levee deposits (Wal-
Mutti 1973; Walker 1978). The massive sand- ker 1978). Large, well-rounded extraformation-
stones with interbedded shale and thin beds are al clasts in the conglomerates (including basalt
more similar to mid-fan than upper fan channel clasts up to 2 m across) suggest that the coarse
deposits, in that they generally lack truly coarse detritus is fed by a major canyon which cut
material but clearly include thinning- and through any bathymetric highs present land-
fining-upward megasequences (Nilsen 1980). ward of this area.
Although upper fan facies associations were Turbidites of the Rivas Formation thus form
not identified in the study area, they are present an overall progradational sequence, with chan-
in Rivas Formation exposures along the west nelled sandy deposits overlying prograding lobe
coast of the Nicoya Peninsula, NW of the study deposits (Walker 1980). These overlie thin-
area. These exposures are located seaward of bedded, probably distal turbidites similar to
the antiformal high of the Nicoya Peninsula, those of lower fan and basin plain environ-
and the rocks dip predominantly seaward. The ments, which in turn overlie hemipelagic to
Rivas Formation beds in this area are composed pelagic deposits of the Sabana Grande Forma-
of thick (up to 30+ m), massive, amalgamated tion. The Rivas Formation is volcanogenic,

RIVAS F O R M A T I O N

I000

EXPLANATION

Flute casts
©
-~- Groove casts

500
Q Predominant paleocurrent
direction

O Massive sandstone

© Thin-bedded turbidites

BB Shale with very thin-

bedded turbidites

FIG. 6. Generalized section of the Rivas Formation. Individual bed thicknesses not to scale, though
sand/shale ratios are representative.
140 N. Lundberg
with both intermediate and mafic volcanic de- ated by thin- to thick-bedded turbidites which
tritus, and reflects substantial erosion of the locally form thickening- and coarsening-upward
volcanic arc which was presumably located to cycles, commonly ending in thick, massive
the NE (present coordinates). sandstones. The upper Las Palmas Formation is
composed of mainly thick, massive, and
Las Palmas Formation amalgamated sandstones, interbedded with
shale and thin-bedded turbidites. Thus prograd-
The Rivas Formation is overlain by the Las ing depositional-lobe features are overlain by
Palmas Formation, a thick accumulation of channel-system facies, as in the Rivas Forma-
Palaeocene to possibly lower Eocene turbi- tion, although fan facies associations are less
dites (Fig. 7). These turbidites are rich in well-defined and the Las Palmas Formation
carbonate lithic fragments as well as volca- lacks deposits corresponding to basin plain and
nogenic detritus, probably reflecting reworking lower fan environments. Again, however, a fan
of shallow-water limestones as old as geometry may not be applicable to this unit.
Palaeocene which are exposed NW of the
study area (Dengo 1962b; Schmidt-Effing Sandstone and mudstone of Cabo Blanco
1979). The basal contact of the Las Palmas
Formation with the Rivas Formation is grada- In the seaward portion of the study area, the
tional, as minor carbonate detritus is present in thick turbidite sequences of the Rivas and Las
the upper Rivas Formation. The Las Palmas Palmas Formations are not present. Instead,
Formation is over 1000 m thick at the landward thin (200-300 m) sequences of thin-bedded
end of the study area, and composes a second turbidites overlie the Nicoya Complex and
turbidite wedge which prograded over the Sabana Grande Formation with probable angu-
generally similar Rivas Formation. The lower lar unconformity, although vestiges of the
portion of the Las Palmas Formation is domin- lowermost Rivas Formation are preserved

LAS PALMAS FORMATION

rn.

I000 ©
EXPLANATION
G
Q
~,- Flute casts

Groove casts

Q Predominant paleocurrent
direction

500

Pebbly sandstone

Massive sandstone

Thin-bedded turbidites

/ Shale with very thin-

bedded turbidites

Fl(~. 7. Generalized section of the Las Palmas Formation. Bed thicknesses not to scale, though
sand/shale ratios a~e representative.
 Evolution of the slope landward of the Middle America Trench 141

locally (Fig. 3; these remnants are too small to

S A N D S T O N E AND M U D S T O N E
be portrayed on the geological map in Fig. 2).
OF CABO BLANCO
The thin-bedded turbidites have yielded a
Palaeocene assemblage of nannoplankton (S. NORTHEAST OF
NORTHWEST OF
Hart & S. Percival, written comm. 1980) and so CABO BLANCO
are correlative with the Rivas and/or Las Pal- CABO BLANCO
mas Formations. These turbidites are generally
5-15 cm thick, rarely exceeding 30 cm in thick- m. 13'1.
ness, and the intercalated fine-grained sediment
300 400
is commonly rich in biogenic material. The
absence of thick Palaeocene turbidites corre-
sponding to down-current facies of the channel
systems found landward strongly suggests the
presence of a bathymetric high which shielded
this area from the bulk of the clastic detritus. At
the seaward edge of the study area these turbi-
dites are overlain by at least 150 m of radiola-
rian-rich siliceous mudstone, and along the
SE coast extending closer to the centre of
the study area they are underlain and over-
lain by at least 50 and 125 m, respectively, of EXPLANATION
foraminiferan-rich calcareous mudstone. The
turbidites and mudstones together make up the
unit informally called the sandstone and mud- / Siliceous mudstone
stone of Cabo Blanco (Fig. 8). The mudstones
are uniformly thin- to medium-bedded and
individual beds are subtly graded. They prob-
ably represent locally-derived hemipelagic
I-1 Calcareous mudstone

material redeposited by minor turbidity cur- Thin-bedded turbidites

rents. The siliceous and calcareous mudstones
are very similar except for the virtual absence of Shale with very
calcareous foraminiferan tests in the former.
The siliceous mudstone is of lower or middle
i thin-bedded turbidites

Eocene age (R. Schmidt-Effing, pets. comm. FIG. 8. Generalized sections of the sand-
1980) and was probably deposited below the stone and mudstone of Cabo Blanco
CCD. The calcareous mudstone is of (southernmost point of study area; see Fig.
Palaeocene or Eocene age (R. Schmidt-Effing, 2). Bed thicknesses not to scale, though
pets. comm. 1980) and was deposited above the sand/shale ratios are representative.
CCD. Although broadly similar in age as well
as appearance, the exact age relationship be- shallow-water deposits form two distinct units,
tween these two units must await more detailed the limestone and sandstone of Mal Pais and
dating. the Montezuma Formation (Fig. 3). The lime-
I interpret the hemipelagic mudstones of this stone and sandstone of Mal Pals is only exposed
unit as slope deposits, and the thin-bedded at the seaward (SW) end of the study area, and
turbidites as the basin fill of small slope basins. includes late Oligocene sandstones reported by
These slope basins were probably fed by minor Dengo (1962b). A distinctive white limestone
canyons which incised the landward high. of highly variable thickness (up to several tens
of metres) is present at the base of this unit in
several localities. The limestone is made up of
Limestone and sandstone of Mal Pals
larger foraminiferans and is Eocene in age (R.
In the seaward portion of the study area the Schmidt-Effing, pers. comm. 1980). The re-
generally deep-water deposits of the Nicoya mainder of this unit is mainly sandstone, the
Complex, the Sabana Grande and Rivas lower portion of which is fine- to medium-
Formations, and the sandstone and mudstone grained and unstratified. The lack of stratifica-
of Cabo Blanco are overlain by shallow-water tion is attributed to complete reworking of the
sediments. These range in age from Eocene to sediment by bioturbation. Upsection the sand-
probably Quaternary, and generally overlie the stone becomes gradually coarser grained and
older section with angular unconformity. The stratified, with planar-stratified sandstone giv-
142 N. Lundberg
ing way in turn to moderate-angle cross- ment, and overlying finer-grained bioturbated
stratified sandstone interbedded with con- sands a less energetic, probably offshore en-
glomerate. The upper portion apparently repre- vironment (Harms et al. 1975, chapter 5;
sents a higher-energy environment than does Seilacher 1978). Significant tectonic uplift must
the lower portion, but certainly the abundant be invoked to explain the superposition of these
larger foraminiferans and the prevalence of shallow-water deposits directly over siliceous
moderate-angle cross-stratification both indi- mudstones presumably deposited below the
cate a shallow-water origin (Bandy 1964; CCD, and deep-sea rocks of the Nicoya Com-
Walker 1979). plex.

Montezuma Formation Structural geology

The Montezuma Formation ranges in age The Nicoya Complex is structurally complex on
from Miocene (Dengo 1962a) to probably a large scale, although it is not a m61ange
Quaternary (S. Hart & S. Percival, written terrane. Folding and faulting, including some
comm. 1980) and unconformably overlies the thrust faulting, has resulted in a juxtaposition of
Nicoya Complex, the Sabana Grande and Rivas fault blocks which is generally difficult to recon-
Formations, and the sandstone and mudstone struct, and stratigraphic sections are difficult to
of Cabo Blanco. The Montezuma Formation piece together. However, I found no evidence
reaches a maximum thickness of 120 m along of large-scale nappe emplacement or of the
the sea-cliffs of the south-eastern coastline of imbricate stacking of thrust-bound slices of
the study area, and thins northward to taper out ocean floor. The Nicoya Complex may instead
against a series of hills composed of the under- represent in originally coherent piece of oceanic
lying Nicoya Complex (Fig. 2). Commonly a crust, of a somewhat atypical nature as outlined
1-3 m thick basal conglomerate of the Monte- earlier. The Nicoya Complex has suffered signi-
zuma Formation directly overlies basalt of the ficant shortening, however, especially prior to
Nicoya Complex. The conglomerate, composed deposition of the sedimentary cover.
of sub-rounded to sub-angular basalt and chert The overlying sedimentary units have largely
clasts in a coarse sandstone matrix, is unstrati- not been affected by this complex juxtaposition
fied, and generally matrix supported. The re- of fault blocks, although significant thrusting
mainder of the Montezuma Formation is pre- has involved the Sabana Grande Formation and
dominantly sandstone. Sandstone in the lower the lower part of the Rivas Formation. The
portion is typically medium- to coarse-grained, structural style of the lower part of the
pebbly in places, planar- or cross-stratified, and sedimentary cover, including all units strati-
is commonly interbedded with pebbly and/or graphically below the shallow-water deposits, is
shell-rich horizons. In contrast, sandstone in generally characterized by large-scale folding
the upper Montezuma Formation is generally with seaward vergence, with fold patterns be-
fine-grained, unstratified, devoid of pebbles coming tighter in more seaward exposures.
and shells, and well burrowed. Dips range up to steeply overturned in local
The lower portion of the Montezuma Forma- zones of intense deformation, and faulting is
tion represents a high energy depositional en- generally dip-slip along moderate to high-angle
vironment in which material as coarse as surfaces. Major Tertiary normal faults separate
medium- to coarse-pebbly sand was deposited large landward-dipping slabs floored by the
as migrating bedforms. The interbedded pebbly Nicoya Complex landward of the present anti-
and shelly horizons probably represent lag de- formal high (Fig. 2). The shallow-water de-
posits of intermittent storms (Johnson 1978). posits are considerably less deformed than the
The lack of stratification in the upper portion is older rocks. The limestone and sandstone of
probably due to biological reworking of the Mal Pals exhibits broad open folds with bedding
sediment, and this massive sandstone repre- dips up to 45-50 ° and normal faults, while the
sents a quieter, lower energy environment, in Montezuma Formation is virtually flat-lying
which the sedimentation rate was lower than and is affected by only very minor faults.
the rate of bioturbation. The Montezuma
Formation is clearly a shallow-water deposit,
probably a transgressive sequence which
Evolution of the trench slope
formed over a possibly subaerial erosional sur-
face. In this interpretation, the stratified sands Sediments of the Nicoya Complex reflect
with interbedded pebbly and shelly horizons predominantly pelagic deposition. The sedi-
represent deposition in a nearshore environ- mentary breccias intercalated with basalt prob-
 Evolution o f the slope landward o f the Middle America Trench 143

ably formed at a spreading ridge, an aseismic style, between the structurally complex base-
rise, or a fracture zone. Certainly the Nicoya ment and its generally intact sedimentary cover,
Complex does not represent a piece of simple which overlies these upper breccias concordant-
oceanic crust, with a straightforward stratigra- ly. Because there is no record of a pre-existing
phy of sediments over ridge-generated basalt, arc or continental massif in Central America
but the field relations seen in the south-eastern south of Nicaragua, nor of any other rocks as
portion of the Nicoya Peninsula could be ex- old as the Nicoya Complex, this subduction
plained in various ways: by a significant ridge zone apparently developed in the open ocean,
jump (e.g. Handschumacher 1976) or by prox- well removed from continental sediment
imity to a ridge-ridge transform, elevating sources. A major suture has been suggested by
portions of the sea-floor above the CCD and De Boer (1979) and Gose et al. (1980) between
extruding basalt over pelagic deposits, as well southern Central America (including Costa
as by the formation of an aseismic rise Rica), with its oceanic basement, and the con-
(Schmidt-Effing 1979) or by the initiation of the tinental 'Chortis' block of northern Central
island arc (De Boer 1979). Although drilling America. De Boer (1979) has proposed a Late
results from the Mariana Trench suggest that Cretaceous suture marked by the ultramafic
arc rocks may underlie most or all of an intra- rocks of the Santa Elena Peninsula of northern
oceanic forearc region (Hussong, Uyeda et al. Costa Rica (Fig. 1) on the basis of regional
1978; Meijer 1980), none of the Nicoya Com- geology and aeromagnetic data, while Gose
plex rocks have been shown to be arc related (1980) has suggested a Tertiary suture north
(Pichler & Weyl 1975). De Boer (1979) has also of southern Nicaragua using palaeomagnetic
reported younger volcanic rocks interbedded in data. The question of a pre-existing arc in
Palaeogene sediments of the Nicoya Peninsula, southern Central America remains open,
and suggested that they represent arc activity in however, because the basement of the magma-
this region. Alternative origins for near-trench tic arc in Costa Rica is largely covered by
magmatism, however, have been suggested for Cenozoic volcanic deposits (D6ndoli et al. 1968;
other forearc terranes. Echeverria (1980) has see also Schmidt-Effing 1979).
proposed the leaking of basaltic magma Schmidt-Effing (1979) has suggested that the
through extensional cracks in down-bending Sabana Grande Formation represents con-
oceanic lithosphere, and the intersection of a tinued deposition in an open ocean environ-
fracture zone and a trench as two possible ment, and the relative shallowing of the sedi-
origins of gabbroic sills in the Mesozoic Francis- ment surface is caused by submarine volcanic
can Complex of California. These gabbros in- build-ups which form an aseismic rise repre-
truded turbidite deposits after their initial in- sented by the upper part of the Nicoya Com-
corporation into an accretionary wedge but plex. This interpretation explains the complex
before their subduction to depths producing lateral variations in both the Sabana Grande
blueschist-grade metamorphic assemblages Formation and the Nicoya Complex, but it is
(Mattinson & Echeverria 1980). Reid & Gill difficult to reconcile with the concordant con-
(1980) have proposed a ridge-trench encounter tact between the Sabana Grande Formation
to explain the interbedding of basalt and ande- and the overlying Rivas Formation, and the
site with trench or near-trench turbidites of the deposition of a thin sequence of turbidites just
Palaeocene Ghost Rocks Formation of Kodiak before the upper limestone of the Sabana
Island. Thus near-trench volcanism may be Grande Formation. This interpretation would
related to oceanic heat sources rather than to also predict the presence of vesicular basalt in
arc processes, although the hypothesis of a the upper part of the Nicoya Complex, especial-
ridge-trench encounter is not supported in the ly under the sediments reflecting shallow-water
Nicoya Peninsula case by the remaining magne- deposition. However, the shallowest age-
tic anomalies of the Cocos Plate (Handschu- equivalent deposits on the Nicoya Peninsular,
macher 1976). upper Campanian Rudist-bearing reefal car-
I suggest that much of the deformation of the bonates, overlie non-vesicular basalt and
Nicoya Complex and the formation of the coarse-grained gabbro in at least one locality,
uppermost sedimentary breccias in this unit suggesting tectonic uplift into the photic zone.
were due to the Mesozoic initiation of the An alternative interpretation has been sug-
Middle America Trench in this region (Fig. 9, gested by De Boer (1979), in which the upper
stage 1). The deformation, and these breccias part of the Nicoya Complex represents the
post-date the bulk of pelagic deposition, and initial island arc built on oceanic crust, and the
pre-date arc-derived detritus. Additionally Sabana Grande Formation may thus represent
these breccias mark a contrast in structural back-arc basin sediments. Additional geochem-
144 N. Lundberg
SW NE
0 t L ote $onton/on- E o r l7 0 o m p o n / o n

",v.
c,
2. L o r e Compon/on Ksg

3. Eor/y Po/eocene KTr KTr

Tcb Tip
4 Lore P o l e o c e n e

1--<.
,
Tmp

i \ '.-N icoya Pen.-. 5

oI 6.PreseotMA T ~Tm

5O
I

8 V.E.= 4 X KM

FIG. 9. Proposed model for the evolution of the Nicoya Peninsula. Each diagram represents a time
slice recording the development of a major unit (see text for explanation). Symbols for units refer to
legend of Fig. 2. Present-day bathymetric profile off Nicoya Peninsula from seismic data courtesy of
R. T. Buffler, University of Texas Marine Science Institute.

ical data integrated with detailed mapping of recovered on the present-day slope landwards
field relations are needed to evaluate these of the Middle America Trench (Moore et al.
models for the origin of the Nicoya Complex. 1981). Thus the Sabana Grande Formation may
Nevertheless, I suggest that lateral variations simply represent deposition on the juvenile
can be expected on a juvenile trench slope as trench slope, with the newly formed island arc,
well, and that the Sabana Grande mudstones presumably now covered by the Cenozoic vol-
resemble hemipelagic muds that have been canic arc, providing the clastic component of
 Evolution of the slope landward of the Middle America Trench 145
the mudstones and turbidites, as well as the ash Conclusions
layers (Fig. 9, stage 2). The passage of the
sediment surface up through the CCD recorded The foregoing observations and interpretations
in this unit probably represents uplift of the suggest a straightforward scenario for the
early trench slope, because of the relative evolution of this portion of the Middle America
scarcity of calcareous sediment below this level Trench margin (Fig. 9). The Nicoya Peninsula
and the abundance of them above it. Alterna- is an uplifted trench slope break composed of
tively this may reflect sediment accumulation the Nicoya Complex and its sedimentary cover.
through the CCD on a stationary basement, or The Nicoya Complex is upper Mesozoic
a secular lowering of the CCD. The recycling of oceanic crust, upon which younger volcanic
Nicoya Complex and other material to form rocks of as yet uncertain tectonic affinity were
interbedded conglomerates and sedimentary extruded. I suggest that the deformation of the
breccias is probably due to uplift and erosion of Nicoya Complex was due to the initiation of
the oceanic basement in the forearc region. The subduction during the Late Cretaceous in an
widespread, thick, and relatively pure pelagic oceanic region. The sedimentary cover reflects
limestone at the top of this unit may represent uplift and progressive deformation of the trench
uplift of the Nicoya Peninsula terrane in slope through the evolution of this intra-oceanic
Campanian time, preventing clastic detritus arc-trench system. The schematic diagrams of
from reaching this area. selected time periods in Fig. 9 depict the forma-
The Rivas and Las Palmas Formations prob- tion of the major units in the study area, and
ably were deposited in a large forearc basin the area labelled 'Nicoya Peninsula' in the final
(Fig. 9, stage 3). The age-equivalent thin- diagram shows the lateral extent of onland data
bedded turbidites in the seaward portion of the on which this interpretation is based. The early
study area I believe represent deposition in stages illustrated in Fig. 9 depict a subduction
small basins on the trench slope, which must zone initiating in simple ocean crust, and are
have been shielded from the voluminous clastic based in part on Weissel & Anderson's (1978)
detritus of the forearc basin by a structural high model of an initiating subduction zone north-
in approximately the same position as the pre- east of the Caroline Plate. These early stages
sent-day anticlinal high of the Nicoya Peninsula may be especially subject to revision, if for
(Fig. 9, stage 4). Sparse palaeocurrent data example this present-day forearc region is
from the thick landward sections of the Rivas underlain in part by arc rocks as suggested by
and Las Palmas Formations suggest significant De Boer (1979).
longitudinal transport, supporting the existence Major differences in lithology and timing of
of a structurally-restricted elongate forearc events suggest the presence of a variety of
basin. In addition, the presumed angular un- terranes with differing geological histories even
conformity below the seaward Palaeocene beds within the Costa Rican forearc region (Galli-
suggests tectonism and perhaps uplift in the Olivier & Schmidt-Effing 1977; Stibane et al.
seaward area at this time. I interpret the 1977; Schmidt-Effing 1979). Thus caution
siliceous and calcareous mudstones which over- should be exercised in extrapolating this model
lie thin-bedded turbidites in the seaward region and the data on which it is based to the alleged-
to represent continued deposition on the trench ly correlative terranes in Columbia and
slope, respectively below and above the CCD, Ecuador, the Caribbean region, and the Guate-
in the absence of significant clastic input from malan trench slope, although it is noteworthy
the arc. This shut-off of clastic detritus probably that the sequence of events in the vicinity of the
reflects more efficient local shielding by bathy- shelf-edge high off Guatemala outlined by
metric highs, as Eocene volcanic and clastic Seely (1979) is broadly similar.
sequences have been reported elsewhere in this
portion of the margin (Dengo 1962a; Pichler &
Weyl 1973, 1975). ACKNOWLEDGMENTS;Acknowledgment is made to the
The Cenozoic shallow-water deposits indi- Donors of the Petroleum Research Fund (PRF 9219-
cate uplift of much of the area to shelf depths AC2), administered by the American Chemical Soci-
during the Eocene, and by this time the older ety, for the partial support of this research. I thank
strata had been folded and faulted (Fig. 9, stage the Mobil Foundation for supporting much of the
5). Palaeogene shallow-water deposits were field research, and S. Hart and S. Percival of Mobil
Exploration and Producing Services Inc. for nanno-
broadly folded and normal faulted by Miocene fossil and palynological age determinations. The In-
time, and shallow-water to emergent conditions stituto Geogr~ifico Nacional of Costa Rica provided
have prevailed through to the present (Fig. 9, significant logistical support, and the Patent Fund of
stage 6). the University of California provided partial travel
146 N. Lundberg
support. R. Schmidt-Effing kindly provided micro- the family of Sr. Ing. Mario Rodrfguez, Chap Porter
fossil age determinations and a preprint and D. Jones and Isabel Vfisquez de Porter, Fritz Schutt, and many
provided radiolarian age determinations. I thank C. other people of Costa Rica for their generous hospi-
Rodrfguez for enthusiastic and capable assistance in tality. Finally I thank J. C. Moore for his unflagging
the field, and H. Gursky, H. Wildberg, and M. encouragement and assistance throughout this pro-
Streben for stimulating discussions. I am indebted to ject.

References
BANDY, O. L. 1964. Foraminiferal biofacies in sedi- the Central North Atlantic Ocean. Proceedings of
ments of Gulf of Batabano, Cuba, and their a symposium at Louisiana State University,
geologic significance. Bull. Am. Assoc. Petrol. Baton Rouge, Louisiana, March 3-5 1980,
Geol. 48, 1666-79. pp. 51-4.
CASE, J. E. 1974. Oceanic crust forms basement of HANDSCHtJMACHER, D. W. 1976. Post-Eocene plate
eastern Panama. Bull. geol. Soc. Am. 85, 645- tectonics of the Eastern Pacific. In: StJrroN, G.
52. H., MANGHNANI, M. H. & MOBERLY, R. (eds).
DE BOER, J. 1979. The outer arc of the Costa Rican The Geophysics of the Pacific Ocean Basin and
Orogen (oceanic basement complexes of the its Margin: a volume in Honor of George P.
Nicoya and Santa Elena Peninsulars). Tecto- Woollard. Monogr. Am. Geophys. Union, 19,
nophysics, 56, 221-59. 177-21)2.
DENGO, G. 1962a. Tectonic-igneous sequence in Cos- HARMS, J. C., SOUTHARD, J. B., SPEARING, D. R. &
ta Rica. In: ENGEL, A. E. J., JAMES, H. L. & WALKER, R. G. 1975. Depositional environments
LEONARD, B. F. (eds). Petrologic Studies: a interpreted from primary sedimentary structures
volume to Honor A. P. Buddington, 133-61. and stratification sequences. Short Course, Soc.
Geol. Soc. Am., Boulder. Econ. Paleontol. Mincral., Tulsa, 2, 161 pp.
1962b. Estudio Geol6gico de la Regi6n de HILL, M. D., 1979. Volcanic and plutonic rocks of the
Guanacaste. Costa Rica. Instituto Geogr~ifico Kodiak-Shumagin Shelf, Alaska: subduction de-
Nacional, San Jos6, Costa Rica, 112 pp. posits and near-trench magmatism. Thesis, PhD,
D6NDOI.I, C., DENGO, G. & MALAVASSI, E. 1968. Univ. California, Santa Cruz. 259 pp. (unpubl.).
Mapa geol6gico de Costa Rica. Scale 1:700,000. HUSSONG, D., UYEDA, S., BLANCliET,R., BLEIL, U.,
lnstituto Geogrfifico Nacional, San Josd, Costa ELI.IS, C. H., FRANCIS, T. J. G., FRYER, P.,
Rica. HORAI, K., KLING, S., MEIJER, A., NAKAMURA,
DONNEt.i.V, T. W. 1975. The geological evolution of K., NATLAND,J., PACKHAM,G. & SHARASKIN,A.
the Caribbean and Gulf of Mexico---some critical 1978. Leg 6(I ends in Guam. Geotimes, 23, 19-22.
problems and areas. In: NAIRN, A. E. M. & JOHNSON, H. D. 1978. Shallow siliciclastic seas. In:
STEHI.I, F. G. (eds). The Ocean Basins and READtN~i, H. G. (ed.). Sedimentary Environ-
Margins, vol. IIl. The Gulf of Mexico and the ments and Facies, 2(17-58. Elsevier, New York.
Caribbean, 663-89. Plenum Press, New York. KARIG, D. E. 1974a. Evolution of arc systems in the
ECHEVERRIA, L. M. 1980. Oceanic basaltic magmas in western Pacific. Ann. Rev. Earth planet. Sci. 2,
accretionary prisms: the Franciscan intrusive 51-75.
gabbros. Am. J. Sci. 280, 697-724. - - 1 Tectonic erosion at trenches. Earth
9 7 4 b .

FISHER, S. P. & PESSAGNO, E. A. 1965. Upper planet. Sci. Lett. 21,209-12.

Cretaceous strata of northwestern Panama. Bull. - & SHARMAN, G. F. III. 1975. Subduction and
-

Am. Assoc. Petrol. Geol. 49, 433-44. accretion in trenches. Bull. geol. Soc. Am. 86,
GALLI-OLIVIER, C. 1977. Edad de emplazamiento y 377-89.
periodo de acumulaci6n de la ofiolita de Costa KING, P. B, 1969. Tectonic Map of North America.
Rica. Cienc. Tech. Univ. Costa Rica, l, 81-6. Scale 1:5,000,000. U.S. geol. Surv. Washington.
1979. Ophiolite and island arc volcanism in KUIJPERS, E. P. 1980. The geologic history of the
Costa Rica. Bull. geol. Soc. Am. 90, 444-52. Nicoya Ophiolite Complex, Costa Rica, and its
- - & Schmidt-Effing, R. 1977. Estratigrafia de la geotectonic significance. Tectonophysics, 68,
cubierta sedimentaria supra-ofiolftica cret~icica 233-55.
de Costa Rica. Cienc. Tech. Univ. Costa Rica, l, LADD, J. W., IBRAHIM, A. K., MCMILLEN, K. J.,
87-96. LATHAM, G. V., VON HUENE, R. E., WATKINS, J.
GARCIA, M. O. 1978. Criteria for the identification of S., MOORE, J. C. & WORZEL, J. L. 1978. Tecto-
ancient volcanic arcs. Earth Sci. Rev. 14, 147-65. nics of the Middle America Trench offshore
GOOSSENS, P. J., ROSE, W. I. JR. & FLORES, D. 1977. Guatemala. Int. Symp. Guatemala 4 February
Geochemistry of tholeiites of the Basic Igneous Earthquake and Reconstruction Process, 2,
Complex of northwestern South America. Bull. Guatemala City, May 1978.
geol. Soc. Am. 88, 1711-20. MA'ITINSON, J. M. & ECHEVERRIA,L. M. 1980. Ortiga-
GosE, W. A., Scow, G. R. & SWARTZ, D. K. 1980. lita Peak gabbro, Franciscan Complex: U-Pb
The aggregation of Mesoamerica: paleomagnetic dates of intrusion and high-pressure--low-
evidence. In: PILGER, R. H. JR. (ed). The Origin temperature metamorphism. Geology, 8, 589-
of the Gulf of Mexico and the Early Opening of 93.
 Evolution of the slope landward of the Middle America Trench 147
MEIJER, A. 1980. Primitive arc volcanism and a la (Costa Rica). Trans. 9th Caribbean Geol.
boninite series: examples from western Pacific Conf., Santo Domingo, Dominican Republic.
island arcs. In: HAYES, D. E. (ed). The Tectonic SCHOLL, D. W., VON HUENE, R., VALLIER, T. L. &
and Geologic Evolution of Southeast Asian Seas HOWELL, D. G. 1980. Sedimentary masses and
and Islands. Am. Geophys. Un. Monogr. 23, concepts about tectonic processes at underthrust
271-82. Washington. ocean margins. Geology, 8, 564-8.
MINSTER, J. B. & JORDAN, T. H. 1978. Present-day SCIENTIFIC PARTY 1980. Initial Rep. Deep Sea drill.
plate motions. J. geophys. Res. 83B, 5331-54. Proj. 56, 57. U.S. Govt Printing Office,
MOORE, J. C., WATKINS, J. S., BACHMAN, S. B., Washington.
BEGHTEL, F. W., BUTT, A., D1DYK, B. M., SEELY, D. R. 1979. The evolution of structural highs
LEGGETr, J. K. LUNDBERG, N., MCMILLEN, K. J., bordering major forearc basins. In: WATKINS, J.
NnTSUMA, N., SHEPHARD, L. E., SHIPLEY, T. H., S., MONTADERT, L. & DICKERSON, P. W. (eds).
STEPHAN, J. F. ~ STRADNER, H. 1979. Progres- Geological and Geophysical Investigations o f
sive accretion in the Middle America Trench, Continental Margins. Mem. Am. Assoc. Petrol.
southern Mexico. Nature. London, 281,638-42. Geol. 29, 245-60.
MOORE, J. C. et al. 1981. Facies belts of the Middle -- , VALE, P. R. & WALTON, G. G. 1974. Trench
America Trench and forearc region, southern slope model. In: BURK, C. A. & DRAKE, C. L.
Mexico: results from Leg 66 DSDP (this (eds). The Geology of Continental Margins, 249-
volume). 60. Springer-Verlag, New York.
MURAUCHI, S. & LUDWIG, W. J. 1980. Crustal struc- SEILACHER, A. 1978. Use of trace fossils for recogniz-
ture of the Japan Trench: the effect of subduc- ing depositional environments. In: BASAN, P. B.
tion of ocean crust. In: Scientific party. Initial (ed.). Trace Fossil Concepts. Short Course Soc.
Rep. Deep Sea drill. Proj. 56, 57, 463-9. U.S. econ. Palaeontol. Mineral. Tulsa, 5, 185-201.
Govt Printing Office, Washington. STIBANE, F . R . , SCHMIDT-EFFING, R. & MADRIGAL,R.
NILSEN, T. H. 1980. Modern and ancient submarine 1977. Zur stratigraphisch-tektonischen Entwick-
fans: discussion of papers by R. G. Walker and lung der Halbinsel Nicoya (Costa Rica) in der
W. R. Normark. Bull. Am. Assoc. Petrol. Geol. Zeit yon Ober-Kreide his Unter-Tertiar. Gies-
64, 1094-101. sener geol. Schriften 12, 315-58.
PEARCE, J. A. & CANN, J. R. 1973. Tectonic setting of YON HUENE, R., AUBOUIN, J., AZEMA, J., BLACK-
basic volcanic rocks determined using trace ele- INTON, G., CARTER, J. A., COULBOURN, W. Z.,
ment analyses. Earth planet. Sci. Lett. 19, 290- COWAN, D. S., CURIALE, J. A., DENGO, C. A.,
300. FAGS, R. W., HARRISON, W., HESSE, R.,
PICHLER, H., STIBANE, F. R. & WEYL, R. 1974. HUSSONG, D. M., LADD, J. W., MUZYLOV, N.,
Basischer Magmatisms und Krustenbau im sud- SHIKI, T., THOMPSON, P. R. • WESTBERG, J.
lichen Mittelamerika, Kolumbein und Ecuador. 1980. Leg 67: The Deep Sea Drilling Project
Neues Jahrb. Geol. Palaontol. Monatshefte, 2, Mid-America Trench transect off Guatemala.
102-26. Bull. geol. Soc. Am. 91,421-32.
& WEYL, R. 1973. Petrochemical aspects of WALKER, R. G. 1978. Deep-water sandstone facies
Central American magmatism. Geol. Rdsch. 62, and ancient submarine fans: models for explora-
357-96. tion for stratigraphic traps. Bull. Am. Assoc.
--& WEYL, R. 1975. Magmatism and crustal Petrol. Geol. 62, 932-66.
evolution in Costa Rica (Central America). -- 1979. Shallow marine sands. In: WALKER, R. G.
Geol. Rdsch. 64, 457-75. (ed.). Facies Models, 75-89. Geol. Assoc. Can.,
REID, M. & GILL, J. B. 1980. Near trench volcanism, Toronto.
Kodiak Island, Alaska: implications for ridge- - - 1980. Modern and ancient submarine fans: rep-
trench encounter. Abstr. Pros. geol. Soc. Am. ly. Bull. Am. Assoc. Petrol. Geol. 64, 1101-8.
12, 148-9. -- & M u ~ t , E. 1973. Turbidite facies and facies
SCHMIDT-EFFING, R. 1979. Alter und Genese des associations. In: MIDDLETON, G. V. & BOUMA,
Nicoya-Komplexes, einer ozeanischen Palao- A. H. (eds). Turbidites and Deep-water Sedi-
kruste (Oberjura bis Eozan) im sudlichen mentation, 119-57. Short Course, Pacif. Sect.
Zentralamerika. Geol. Rdsch. 68, 457-94. Soc. econ. Paleontol. Mineral, Tulsa.
, GURSKY, H.-J., STREBIN, M. & WILDBERG, H. WEISSEL, J. K. • ANDERSON, R. N. 1978. Is there a
1981. The ophiolites of southern Central Amer- Caroline Plate? Earth planet. Sci. Lett. 41, 143-
ica with special reference to the Nicoya Peninsu- 58.

NEIL LUNDBERG, Earth Sciences Board, University of California, Santa Cruz,

California 95064, U.S.A.
 Cenozoic structure, stratigraphy and tectonics of the central
Peru forearc

L. D. K u i m , J. M. Resig, T. M. T h o r n b u r g & H.-J. Schrader

SUMMARY: The central Peru forearc (6°-14°S) is characterized by a variety of structural

and tectonic styles. Continental shelf basins, Sechura, Salaverry, and East Pisco, are
floored by a block faulted Precambrian to Palaeozoic massif and accumulated as much as
3 km of Cenozoic and probably Mesozoic sediment. The massif rises to form an Outer
Shelf High (OSH) covered with disrupted Neogene strata. Structural and stratigraphic
relationships suggest several pulses of uplift and subsidence for the massif with the most
recent uplift detected on the positive massif features.
Upper slope basins are cradled between the OSH and a seaward Upper Slope Ridge
(USR) of deformed sediment. Late Eocene and younger clastic sediments, with minor
limestone beds, occupy the eastern Trujillo Basin, whereas late Miocene to Pleistocene
dolomicrite, micrite and glauconitic micrite are prominent in the western part of the basin.
Both the Trujillo and Yaquina Basins (7°-9°30'S) exhibit severe internal disruption caused
by compressional stresses. To the south, the Lima Basin (9°30'-13°S) contains similar
carbonate lithologies, but this basin and underlying massif have subsided 275 to 500 m
Ma-l since Pliocene-Pleistocene time, indicating tensional stresses.
Geological and geophysical data show the subducting plate is rupturing by thrust
faulting of layer 2 basalts which may form a sediment-basalt m61ange in the overlying
subduction complex. Because the seaward boundary of the arc massif is not well-defined,
several forearc models are presented which allow the subduction complex to range in
width from 15 to 60 kin.
The Peru forearc off Lima (12°S) exhibits many of the tectonic characteristics of the
Japan forearc off Honshu, especially the subsiding slope basin and contemporaneous
accreting subduction complex.

The A n d e a n convergence zone along the west aseismic ridges and fracture zones which collide
coast of South America is often considered with the margin (Nur & Ben A v r a h a m 1981).
the classical continental arc. It displays all the This study describes the geology of the Peru
structural-tectonic elements (subduction zone, forearc and shows the variability and complex-
arc massif, forearc basins and volcanic arc) that ity that exist within one of the most extensively
characterize an arc (Seely 1979) as well as the studied areas of the forearc extending from 6°
modern-day seismicity and volcanism. The rate to 14°S latitude (Fig. 1). We also contrast and
of convergence is 10 cm yr -1 (Minster et al. compare the crustal structure and stratigraphy
1974). of the major forearc features and relate t h e m to
Recent studies show that the Peru continen- the tectonic framework of the region.
tal margin and trench are characterized by a
variety of crustal structures, stratigraphies, and
tectonic regimes (Johnson & Ness 1981; Thorn- Crustal structure
burg & Kulm 1981; Kulm et al. 1977, 1981a,b; Basement structures and distribution of basins
Shepherd & Moberly 1981; Schweller et al.
1981; Coulbourn & Moberly 1977; Hussong et Continental margin
al. 1976; Mordojovich 1974; Scholl et al. 1970). Recent geological and geophysical studies
Precambrian and Palaeozoic strata crop out in indicate that the sedimentary basins on the
the coastal regions and may, in some areas, continental shelf are underlain by Precambrian
occur in close proximity to the Peru-Chile and Palaeozoic crystalline rock and metasedi-
Trench, giving the impression that the margin ment (Kulm et al. 1977, 1981a,b; Thornburg &
is truncated. A n u m b e r of forearc basins of Kulm 1981). This crystalline b a s e m e n t b e n e a t h
various sizes are distributed over the margin, the shelf exhibits compressional velocities
some resting upon this ancient arc massif and ranging from 5.7 to 6.1 km s -a and densities
some resting upon a d e f o r m e d subduction com- from 2.72 to 3.0 g cm -3. Mantle velocities
plex farther seaward. The geology of the con- (8.2 km s -l) occur at a depth of 30 km in the
vergence zone may be further complicated by Salaverry Basin (Fig. 2; Hussong et al. 1976).

151
152 L . D . Kulm et al.

oo

RIDt ECUADOR ANDEAN MARGIN

~E~%~.,~ t ,~..--
STUDY AREAS

ABYSSAL CONTOURS -
I KM C O N T O U R S

o O~3 SHELF -
200M CONTOURS

llmm TRENCH AXIS

(FROM MAMMERICKX AND

SMITH, 1978)

io ° PERU LIMA - BASIN

~.~I
f
f
f
f
15 ° m
f
f
BASIN

2o~

, cHILE i

25 ° ©

0
o

30 ° L
~p
90 ° 85 ° 80 o 75 ° 70 °
 Cenozoic structure of the central Peru forearc 153

82 ° 79"W
4°SF-
BASEMENT CONTROL
OF COASTAL AND

6° , '/~~1 MARGIN BASINS

t %'£ ONSHORE
GEOLOGY

ANDEAN
BATHOLITH

I _
I PRECAMBRIAN-
I '--
% PALEOZOIC

II° X I I1"
o
OFFSHORE
STRUCTURE -'%5

OUTER SHELF
HIGH (OSH)

UPPER SLOPE
%
RIDGE (USR) 1

"~%

%
t 1 1 l 116"
80" 75"W
Fro. 2. Geological features onshore and offshore Peru (after Thornburg & Kulm 1981). See text for
explanation.

FIG. 1. Location map of Peru continental margin and adjacent Nazca plate. Heavy-lined box is study
area.
154 L . D . Ku lm et al.
o

~, '. .... 1 I I I I I

STRUCTURE AND MORPHOLOGY

OF T H E PERUVIAN MARGIN

OUTER SHELF HIGH

SLOPE RIDGES

\ SLOPE BAStNS
SUBMARINE CANYONS
GEOPHYSICAL PROFILE
o

.... BATHYMETRIC PROFILE

~i- I N D U S T R Y DRILL HOLE

\ CONTOURS INTERVAL ,IO

"O ' OOm
20Ore "SHELF BREAK"
8°
2~T RUJILLO

9°

%
, ,,0~' i0 °

I I I I ii °
82" 81 ° 80 ° 79° 78°

FIG. 3. Crustal structure and morphology of Peru margin. Note petroleum industry drill holes (stars)
on line 9.0 (after Thornburg & Kulm 198l). Refer to Fig. 2 for names of features. Numbered profiles
are locations of geophysical profiles displayed in following figures.

Structural ridges form an Outer Shelf High Thornburg & Kulm 1981). Three upper-slope
(OSH) which is traced along the continental basins, Trujillo, Lima and West Pisco, lie be-
margin and rises above sea-level as mountain tween OSH and a prominent Upper Slope
ranges in north-western (6°S) and southern Ridge (USR) of deformed sediments (Figs 2 &
(14°S) Peru (Figs 2 & 3). Foliated metamorphic 3). A fourth and smaller basin, the Yaquina
rocks were recovered in drill holes on the OSH Basin, cradles between the USR and another
at 9°S and from exposures on small islands that structural ridge to the west. Numerous smaller
occupy the outer shelf. These basement ridges basins occur within an anastomosing network of
form the boundaries of the continental shelf structural ridges on the middle and lower con-
basins (Thornburg & Kulm 1981). From north tinental slope (Fig. 3).
to south, these basins are the Sechura,
Salaverry and eastern Pisco Basins (Fig. 2).
Structural ridges also form the foundation for Trench
the basins on the continental slope (Figs 2 & 3; The Peru-Chile Trench is characterized by
 C e n o z o i c structure o f the central P e r u f o r e a r c 155
II°
I 1
1. de Solinos /o Lachoy

12°

15 °
\Z.'5 \ \
%
)
14 °

15 °

I I I 1 /I 1 \
80 ° 79 ° 78 ° 77 ° 76" 75 °

FIG. 3. (continued).
two different systems of faulting on the de- indicating uplift of 4-22 cm yr-1 of the oceanic
scending oceanic slab (Schweller et al. 1981). crust along reverse faults inclined from 30 ° to
Bending of the oceanic plate produces exten- 60 ° (Prince & Schweller 1978).
sional stress and brittle fracture of the basaltic To the south, from 9°30'S to 14°00'S, the
crust, creating step faults and grabens on the compressional regime changes to an extensional
seaward trench slope. Compressional stress re- regime with step faults predominating in the
sulting from plate convergence can be transmit- oceanic crust of the trench (Fig. 6). The step
ted seaward from beneath the overriding con- faults are simple offsets with the downward
tinental plate through the oceanic plate, emerg- displacement on the landward side of the fault
ing as thrust faults within the oceanic crust near with apparent steep dips landward (Schweller et
the trench axis (Fig. 4). Large ridges consisting al. 1981).
of upper layer 2 basalts occur between 7°40'S Comparison of two multichannel depth sec-
and 9°30'S and signify the presence of the thrust tions across the two tectonic regimes in the
regime in this area of the trench off central Peru trench at a low vertical exaggeration (Fig. 7),
(Fig. 5; Schweller et al. 1981; Kulm et al. 1981a; shows the large scale of crustal rupture that
Prince & Kulm 1975). Axial turbidites are characterizes the 60 km long thrust ridge from
disrupted and occur on the 900 m high ridge, 8°55 ' to 9028 ' (Fig. 5) and the relatively small
156 L. D. Kulm et al.

8"55' S
0CE_.q.~lC ~,, ...... ~o~ ~"
~ L 4 ' ~
,/
/ b ( L N L , Id / "~-

WEA/~.~..
,N OCEAN,C~ER 2 " '\~~... SyEARZONE
AcT,VE

DREDGE
- ~ ~ . . ACTIVE

CORE

C o

~ ~ . THRUST ~x
• • ;"--~ RIDGE /

,o

"':":':'..~....,......... ...,.:..,::.
V.E. 15X

FtG. 4. Schematic diagram showing forma-

tion of a basaltic thrust ridge within trench
0 28'S
L__.
due to locking of portion of main under- t
thrust surface. Uplift and landward tilting
of axial turbidites caused by upward curv- P-~ BASALT ~ PELAGIC SEDIMENTS L-~ TURBIDITES
ing of leading edge of thrust fault in upper
oceanic layer (after Schweller et al. 1981).
FIG. 5. Schematic diagram showing struc-
ture and morphology of basaltic ridge in
trench axis between 8o55 ' and 9°28'S.
Cores are numbered on ridge. Diagram
from Kulm et al. (1981a).

11o35'S 7
PLATE ,~¢v/
...._- - .

0 I0
I I I
KM
V.E. 12:1
I0
;EC

FIG. 6. Line drawing of single channel reflection record on ocean plate showing step faults as
described in Schweller et al. (1981) for the Peru-Chile Trench.
 C e n o z o i c structure o f the central P e r u f o r e a r c 157
0 59 67

A -%%!" ,/' O.EO~E

5 "l PLAte I RmG~-~.~--.~I

~ ~ ". ~- "... TRENCH pC

~. 6 ~" • "~ ~ " "

I " • ~" ~ "~" ~ t. .~ < v ~ .:.::.::::'::.

PELAGIC SEDIMENTS ~ TURBIDITES ~ THOLEIITIC BASALT z.

KIM ? V,E, 2:1

PLATE sLO PE
x
TRENCH

7 < l,.- .:::::.

< /" ' ~

I I V.E, 2:1
KM

FIG. 7. Geological cross-sections constructed from multichannel seismic records CDP-2 (A) and
CDP-1 (B) across the Peru Trench at 9°S (see Fig. 8A) and 12°S (see Fig. 8B), respectively. See Fig.
3 for location of lines.

amount of sediment within the trench axis in the overlying velocity structure cannot remove
both trench sections. About 200-250 m of pela- these offsets of rupture structures underlying
gic and hemipelagic sediment is entering the the complex in either depth section (Kulm et al.
trench from the oceanic plate with 170-850 m 1981a; Hussong et al. 1976). The broken basal-
of continent derived turbidites being deposited tic slabs may be inherited as a result of thrust
over these sediments. faulting within the trench (Kulm et al. 1981a) or
apparently rupture later within the subduction
complex from the oceanic crust underthrusting
Massif and subduction complex
itself (Hussong et al. 1976).
While the arc massif on the continental shelf The generalized velocity structure of the
is identified in drill holes and in island expo- central Peru margin is shown in Fig. 11 and
sures, the nature of the subduction complex on can be divided into three different structures.
the continental slope is virtually unknown in the The lowest-velocity material ( 2 . 0 - 3 . 6 k m s -1)
Andean forearc. Only geophysical data and one forms a 15 km wide wedge-shaped mass across
set of geological observations indicate that a the lowermost continental slope. Velocities in
complex is present. Multichannel seismic reflec- this range are commonly found in the most
tion records trace the subducting oceanic crust recently formed dewatered part of the subduc-
landward beneath the continental slope at least tion complex (Kulm, von Huene et al. 1973).
50-60 km (Figs 8, 9 & 10) before the signal The only geological evidence of continental
becomes obscured by the bottom-multiple re- accretion is documented opposite the Nazca
flection. The slab dips initially from 8° to 10° as Ridge (Fig. 1) where pelagic sediment from the
it slides beneath the outer arc. A wedge-shaped ridge and turbidites from the trench are
diffracting mass overlies the oceanic crust; this accreted to the lowermost continental slope
is a typical geophysical signature of a disturbed (Kulm et al. 1974). High-velocity material (5.0-
subduction complex (Seely 1979). 5.6 km s -1) occurs landward of this wedge and
The basaltic ocean crust is rupturing within extends to the mid- to upper-slope region. Such
the trench (Fig. 7A) and beneath the subduc- material may comprise dewatered and
tion complex as shown in the multichannel highly cemented trench-slope turbidites and
depth sections (Figs 8, 9 & 10). Manipulation of hemipelagic deposits and a m61ange of these
 x
~.~ ro 5 o0 m .~ ro o I ~ I ~ 1 ~ I ' I ' 1 ,
O~
a 'I I I I I l I ~ I I I ~ | I -0
-- 0

, I ~ I , I i I , I I 1 i 1 i I
~-~ ~ ~ oo o~ .~ ro o
:~ I I I ~ I L I , I l I ~ I
~ro 5 oo o~ ~ ro 0
 Cenozoic structure of the central Peru forearc 159
0 I00 200 KM
i i 1 i i I i i i i 1 i

,RU~,LLO ~AL~V~RRY

.,, -,._,'-..:,:,:::? i>i: :'-:' .c ,7.':'::.;::-.7

V.E. 3.2:1

B ~ ~ i i ~ °

'~ s4 L ? ~ / :~ >~7".'# i;(',(-!J 7~ ",'_;~, ,~,:'-i;,:/"-i-":'?-?,7'.:! "',7::@;-'7'

~ $4 L~4 8 . ~ , " - " : "' . . . . . " - " ": % :":"'F

KM

Fro. 9. Alternative explanations of geological cross-section at 9°S. Data from multichannel depth
section (Fig. 8A), seismic refraction line at 9°S (Fig. 3; Jones 1981; Hussong et al. 1976), drill holes
and dredge data (Kulm et al. 1981b). Dotted pattern is inferred Mesozoic strata based on seismic
refraction data. Lined pattern is Cenozoic strata in forearc basins. See text for explanation.

sediments mixed with slivers of the underthrust- 12) (Kulm et al. 1981a). Accretion is also
ing basaltic slab (Fig. 12; Kulm et al. 1981a). suggested for the same material at 12°S (Hus-
Basalt slivers should substantially increase the song et al. 1976), especially since the basalt slab
velocities of this portion of the complex. Land- appears to be underthrusting itself (Figs 8B &
ward, the deeper structure is characterized by 10).
velocities of 5.7-5.9 km s -t or higher material There is also little doubt that the highest
which is known to consist of foliated metamor- velocity material is the Palaeozoic arc massif
phic strata typical of the arc massif (Kulm et al. which extends from the shoreline to at least the
1981a). These structures are mantled by slope edge of the continental shelf (Figs 2, 3 & 11).
sediment in the range of 1.5-3.0 km s -~. But the massif's position between the small
If largely geophysical criteria are used to accretionary-sediment wedge and shelf edge is
identify the position of the arc massif- uncertain. If we assume that the 5.0-5.6 km s -1
subduction complex boundary, several models material is the subduction complex (Fig. 11),
are feasible. There is little doubt that the the massif terminates in the vicinity of the
low-velocity mass occupying the toe of the upper continental slope. On the other hand, if
continental slope consists of sediment (Figs 9, this material is part of the foliated metamorphic
10 & 11) and accretion by some sort of thrust strata associated with the massif elsewhere,
faulting is strongly implied at 9°S during the then the massif extends to the lower slope and
most recent subduction history (Figs 4, 7A & the subduction complex is indeed quite small

FIG. 8. (A) Non-migrated multichannel depth section CDP-2 of trench and continental slope at 9°S
(after Kulm et al. 1981a). Vertical exaggeration 3:1. Note position of Yaquina Basin on upper
continental slope. See Fig. 3 for location.
(B) Non-migrated multichannel depth section CDP-1 of trench and continental slope at 12°S. Note
location of Lima Basin on the upper continental slope. Vertical exaggeration 3:1. See Fig. 3 for
location. This is a depth section of the time section shown by Hussong et al. (1976, fig. 8). Record
processing described in Kulm et al. (1981a).
160 L. D. Kulm et al.
0 80 160 KM
I 1 I 1 I 1 1 t I

TR
EN
CH~ ~ !!iI

^ . ^,j ~.
II

Fro. 10. Alternative explanations of geological cross-section at 12%. Data from multichannel depth
section (Fig. 8B), seismic refraction data at 12°S (Hussong et al. 1976) and dredge data (Kulm et al.

-l°
1981b). Lined pattern is Cenozoic strata in Lima forearc basin. See text for explanation.

TRENCH sLOPE SHELF Peru (Travis et al. 1976; Kulm et al. 1977;
Thornburg & Kulm 1981), it is clear that these
basins have undergone different depositional
~------~' I . . . . / 5.7- 6.0 -5 and tectonic histories which are related, in part,
~'-'-*~_ - - 7 " ~ ° ~ - ° ° / KM,sEc
-" sz40 ~ ~5tO
to the basement features described previously.
Extensive normal faulting characterizes the
subaerial outcrops of the coastal basins, and
0 40
t ~ = V,E. 3:1 20 geophysical data (Travis et al. 1976) indicate
KM KM
that irregular horst and graben basement fea-
tures occur on the adjacent continental shelf.
FIG. 11. Generalized velocity structure of Cenozoic subsidence of the Sechura, Salaverry
central Peru margin. Data compiled from
multichannel velocity spectra (Kulm et al. and East Pisco Basins (Fig. 2) on this block-
1981a) and seismic refraction information faulted arc massif is indicated by the deposition
(Hussong et al. 1976; Jones 1981). of approximately 1000-3000 m of marine clastic
sediments (shale, siltstone, and sandstone) with
(Figs 9 & 10). At 9°S, Kulm et al. (1981a) minor interbedded carbonate, phosphate, and
favour the intermediate position (i.e. upper glauconite deposits. Diatomite is especially
slope, Fig. 9B) and at 120S Hussong et al. (1976) prominent in the East Pisco Basin. Large-scale
favour the lower position (i.e. lower slope, Fig. gravity slumping and sliding and shale flowage
10A) for the arc massif. Additional discussion in the younger strata are detected on seismic
of these models is presented in the following reflection sections (e.g. Fig. 13, 9°S [east]) in
section. the Salaverry Basin.
The southern end of the Salaverry Basin
shallows on a feature called the Lima Platform
Geological history of forearc basins (Fig. 2), which has either undergone only minor
Continental shelf basins
subsidence or has experienced uplift and sub-
aerial erosion during late Cenozoic time be-
From the limited information available on cause only a few hundred metres of sediment
the geology of the forearc basins off central currently rest upon this massif block (Fig. 14).
 Cenozoic structure o f the central Peru forearc 161

DETACHED ATTACHED
BASALTIC RIDGE BASALTIC RIDGE

"°g47-E" TRENCH

THOLEIITIC BASALT
TRENCH TURBIDITES
SUBDUCTION COMPLEX
(SEDIMENT/SEDIMENT-BASALT
MELANGE)

FIG. 12. Transfer of ocean plate basaltic slivers to subduction complex by thrust faulting (detached
basaltic ridge) forming sediment-basalt m61ange. Evolution of two basaltic ridges as shown in
cross-section (Fig. 9). They remain essentially coupled to the subducting slab (attached basaltic
ridge). See Fig. 4 for origin of thrust features.

Pleistocene for the shallow deposits and about

Upper slope basins
1100 m during the Pliocene-Pleistocene for the
The upper slope Lima Basin is located im- deeper rocks. Minimum rates of subsidence are
mediately seaward of the Lima Platform and calculated at 500 and 275 m Ma -~, respective-
contains a maximum of 2 km of sediment rest- ly, for the metamorphic arc massif and over-
ing largely upon the strongly reflecting and high lying sedimentary deposits (Kulm et al. 1981b).
velocity arc massif (Figs 8B, 10 & 14). The In the Lima Basin strata are truncated on the
Lima Basin lithologies consist of late Miocene seafloor in water depths ranging between about
to Pleistocene siltstone, calcareous siltstone, 100 and 1800 m (Fig. 14). A poleward-flowing
micrite, dolomicrite and sheared dolomicrite undercurrent presently impinges on the outer
which were dredged at two locations near continental shelf and upper slope in water
11°30'S (Fig. 15, D-1 and D-46, Kulm et al. depths ranging between 100 and 400 m (Thorn-
1981b). These same strata are shown in the burg & Kulm 1981). If the undercurrent was
upper 1300 m of the multichannel section present in the same position during the Pliocene
(CDP-1) located in the vicinity of the dredge and Pleistocene, truncated strata found deeper
sites (Figs 3 & 15). All of these rocks were than 400 m today also suggest subsidence of the
originally deposited in water depths of 100- continental margin. Whereas subsidence char-
500 m and now lie at water depths of 837- acterizes the most recent tectonic movements in
1201 m (D-I) and 1639-2263 m (D-46) which the southern Lima basin, compressional stress
indicates subsidence of about 500 m during the is also inferred (Thornburg & Kulm 1981).
162 L.D. Kulm et al.

9.0°S (EAST)
w BALLENA E
__ ~ SALAVERRY BASIN

"---. . . . . ~ --~ ~ - ~ ~ .

I ~ _ --
SEC
o .
I K~ J V.E. 8.5:1

TRUJII_ Z.O DELFIN

o/~_~ BASIN ~ o

9.0°S(WEST)~~" ~'/'/-~"
~)1/1~:~~ //- /// 0 ~"~"%~'-8.x/E5:.1~'..~--S
-----Ic
~.:
START "" " ~
-
' : . : . ~ 1,~ ' ,
KM
5
,

FI6. 13. Line drawing of single channel seismic record across Salaverry and Trujillo Basins at 9°S (see
Fig. 3 for location). Line pattern is Outer Shelf High and dotted pattern deformed sediment of
Upper Slope Ridge (Fig. 3). Dredge (D-54) taken across fault blocks. Note slumping and movement
of shales in 9°S (east) younger sediments (Travis et al. 1976). Drill holes shown by towers.

Normal faults penetrate the sediment surface ered from the reverse (?) fault block at dredge
along the seaward edge of the metamorphic sites 54 (Fig. 13) and 59 (Fig. 17) west of the
Outer Shelf High (OSH), but high-angle re- drill holes. An orientated brecciation suggests a
verse faults in the basin deposits suggest com- compressive origin for these rocks. Unde-
pressional stress. This stress pre-dates or is in formed Neogene dolomicrite, glauconitic mic°
part synchronous with basin subsidence (Thorn- rite, and phosphorite were also recovered in
burg & Kulm 1981; Johnson & Ness 1981). these dredges. The intense deformation of the
To the north, the upper slope Trujillo Basin sheared rocks occurred during late Pliocene-
contains more than 3 km of post-Eocene shale, early Pleistocene time if radiometric dating of
siltstone, marlstone, minor thin limestone beds, the breccia and enclosing cement are correct
and in the basal part, late Eocene sandstone (Kulm et al. 1981b). The limited amount of
(Travis et al. 1976). This section was initially palaeodepth information obtained at dredge
called the Salaverry Basin section by Travis et sites 54 and 59 (Figs 13 & 17) indicates deposi-
al. (1976) but Thornburg & Kulm (1981), using tion in today's water depths.
additional structural data, define it as the Tru-
jillo Basin section. The latter author's inter- Discussion
pretation of the Trujillo Basin section is shown
in Fig. 16; the Delfin drill hole penetrated the The central Peru margin is characterized by its
section described above but cannot be shown diverse structural, stratigraphic, and tectonic
for proprietary reasons. Seismic data suggest elements. Andean subduction occurred
older lower Tertiary strata at greater depth in throughout most of Mesozoic and Cenozoic
this basin (Thornburg & Kulm 1981). time, and no doubt has created and destroyed
Single channel reflection data indicate that numerous forearc elements in its 150Ma
the Trujillo Basin deposits are more structurally history. Portions of the Cenozoic record are
disturbed than the Lima Basin deposits (cf. Figs examined in this study and compared along
13, 14, 15 & 17) which accounts for the large approximately 900 km (6°-14°S) of the Peru
volume of Neogene sheared dolomicrite recov- margin.
 Cenozoic structure of the central Peru forearc 163
w E
-O
o ~ 7""- J
I0.1 S

~ ~ : , - , r " ~ ,. ,.. o.,

..~.~...-':~....:.:.:.':'.'.
..'~T.:............:.:.:,:.. \ "-, , ~ ~ - . -~. ~ ~
~ -- ~ / . . ,' ; % - ' " t ~ I '-"
~.. . . . . " " " " ~ - KM

-0
IO'I°S ~ ~

0 I0 SEC
I ~ I V.E. 8 : 1
KM

-0
I I.O* S

~ "
START
~,'-, ~ ~ .- ~
~ / ~
- ~
~ ~ ,

___ . _ o ,o

-='" --~'-~.~/ s~c

-o

CDP- I ~ ; : - - I

~"~ ~ -L~ ~.-F_.~ ~ . ' - " : " ~ " - ; - 4,

'L--~. " 7-.~,~ " ~;~. • ..-~._ .-'.~'..o. ..: .. . -,..

0 I0 KM
I ~ I V.E. 3:1
KM

FIG. 14. Line drawings of single channel seismic records across Lima Basin (10. l°S, ll.0°S) and single
channel record (10.1°S) and multichannel depth section CDP-1 showing Lima Basin deposits near
ll.0°S single channel line. Line pattern is OSH and dotted pattern deformed sediment of U S R (Fig.
3). See Fig. 3 for location.
 164 L.D. Kulm et al.
LIMA BASIN SECTION (12"S)
A DEPTH AGE LITHOLOGY LITHOLOGIC DESCRIPTION
(m)
6
Oil ~ua~ HEMIPELAGIC MUO
z W-ZZ.,(--Z./J-4DOLOMICRITE WITH INTERBEDOEDCALCAREOUS

t
500 ~ EI ~ ~
~ -W-- I ~
SILTSTONEAND SL
l TSTONE
i000~ o - - I~3TYxtrtrtlMICRITE,DOLOMICRITE,BRECCIATEDDOLOMICRITE
CALCAREOUSSILTSTONE,ANDSILTSTONE
I
15001 ~__ ~ UNKNOWN (DOLOMICRITE, BRECCIATEDOOLOMICRITE,AND SILTSTONE,}

20ooq AceUsT'c °ASE"E", *'ETA'OR "'c S'"ATA')

2500 -~ .J
w E P0
projected D-, /
II.2°S I I I- I
/
projected D- 46 ~ 1 - - 2

- ~ ~ - - ------ ~ ~ ~ ~'-"~"-'~- I
_~-[ -~- ~ - ~ - ~ - - -~_ ~ - -~ -
~ ~/- . ./~ - f- - - oi ,
5i V.E. 2:1
// ~
KM L5
SEC
Fl(;. 15. (A) Stratigraphic section of Lima Basin near 12°S (modified from Kulm et al. 1981b).
Section based upon dredge data described in text. (B) Line drawing of time section of multichannel
seismic line CDP-I across Lima Basin. See Figs 2 & 3 for location.

SEA L E VEL

TRUJL
ILOBAsN
t J~--~.,~.nCF~ -NF- I i O

\ ~ / :PRECAMBRIAN/ ;,,:
\ LOWER - " ~ / OSH '(',ij
TERTIARY / ":ii:! 21.! 3

I 5KM J ,'. V.E. 7:1 '

4
KM

FIG. 16. Reconstruction of stratigraphic section at 9°S in Trujillo Basin (Figs 2 & 3). See Figs 3
(stars) & 13 (towers) for location of drill holes. Dredge data (Figs 13 & 17) also used in section.
Diagram after Thornburg & Kulm (1981).

Mesozoic strata probably occur in the deeper

Massif tectonics and basin deposition
portions of the shelf basins, particularly in the
The block-faulted arc massif underlies the Salaverry Basin where 4.6-5.2 km s -1 material
shelf and probably upper slope forearc basins. (Fig. 9, dotted pattern) is sandwiched between
The Outer Shelf High (OSH) and Lima Plat- the high velocity massif and much lower veloc-
form are the elevated expressions of the massif, ity (1.5-2.6 km s -l) Cenozoic deposits (line
and form the seaward boundaries and con- pattern) (Jones 1981).
figurations of the shelf forearc basins (Figs 2 & Correlation with onshore stratigraphy sug-
3). Limited sampling and extrapolation with gests that the region of the shelf basins was
onshore geology suggest these foliated meta- emergent and positive until late Eocene or
morphic strata are Palaeozoic-Precambrian post-Eocene time, when the Cenozoic marine
and contain a complex history of deformation sediment record begins in the shelf basin, un-
that is difficult to decipher. conformably overlying the deformed Mesozoic
 Cenozoic structure o f the central Peru forearc 165

w E
7.8°s I-°

YZIOUINA -~ ~ ~ ' ~ I " I

BASIN" ~ -- < "%'-- I "
~_~ TRUJIL L 0 [
~_,~i- 8,~snv LI

~ ~ 0
I KM
I0
I V.E. 12.4 : I 4

-5
SEC

Fro. 17. Line drawing of single channel seismic record at 7.8°S across the Trujillo and Yaquina Basins
(see Fig. 3 for location). Dredge D-59 shown over rock outcrops in Trujillo Basin.

sediments or pre-Mesozoic OSH (Travis et al.

Nature of subduction complex
1976). Sedimentation during the early Tertiary
may have been more continuous in the upper The structure and stratigraphy of the con-
slope Trujillo Basin, however, and the exposed tinental slope is complex. As shown previously,
massif (OSH) may have provided a sediment the arc massif extends seaward some distance to
source (Figs 2 & 16). One can speculate that the abut against a subduction complex of largely
OSH was a Mesozoic volcanic arc (Masias unknown composition. In both the 9°S and 12°S
1976): Mesozoic stratigraphy records the influ- areas, the minimum size of the subduction
ence of such a volcanic source (references cited complex is a 15 km wide wedge (Figs 9A &
in Thornburg & Kulm 1981). Mesozoic dykes, 10A). Considering this relatively small accre-
which cut the Precambrian/Palaeozoic complex tionary prism (25 km2), Hussong et al. (1976)
of correlative coastal Cordillera outcrops calculated that during 150Ma of subduction
(Travis et al. 1976), may represent the core of most of the sediment from the oceanic plate
such a feature. (750 km 2) and the trench (unknown volume)
In middle Cenozoic times, the OSH subsided was subducted beneath rather than accreted to
along with the shelf basins to allow the accu- the continental margin. Using a wider prism of
mulation of several hundred metres of sedi- 40 km (Fig. 10B) or 60 km (Fig. 9B), it is
ment. Subsequent uplift in middle to late still unlikely that a large volume of the potent-
Miocene time deformed and truncated the ially available plate and trench sediment has
Palaeogene sediments (Thornburg & Kulm been accreted to the margin during this
1981). The bulk of the stratigraphic section in period.
the Trujillo Basin appears to pre-date this This subduction of plate and trench sediment
event, implying that this deformational episode and vague evidence of downfaulting in the
caused a pronounced reduction in the Trujillo multichannel seismic line at 12°S (Fig. 8B) led
sedimentation which followed. In the neigh- Hussong et al. (1976) to conclude that the
bouring upper slope Lima Basin, however, continental slope and outer shelf is actually
1.3 km of the 2 km thick sedimentary section subsiding, which suggests tectonic removal of
may be of late Miocene or younger age (see the continental plate (i.e. the subduction com-
Geological History of Forearc Basins), implying plex and the arc massif). Kulm et al. (1981b)
that inception of the predominantly Neogene confirmed this large-scale subsidence on the
deposition may have resulted from these same upper to middle slope at 12°S during Neogene
middle to late Miocene tectonic movements. time as will be discussed later.
Thus, a major change in structural style near
9.5°S apparently separated the basins at this Rupture of subducting plate versus evolution of
time. This zone of structural transition has subduction complex
persisted throughout younger episodes of de-
formation also, causing the neighbouring basins While the subduction complex may not have
to respond in distinctly different ways of subse- historically swept up all the sediments enter-
quent tectonic movements. ing the subduction zone, it has nonetheless
166 L.D. Kulm et al.

played an important role in recent subduction- al to extensional regime in its relatively recent
related tectonics. Crustal rupture of the sub- geological history.
ducting oceanic slab is well documented off Neogene strata in the Trujillo and Yaquina
central Peru (Figs 5, 7 & 8A). Using the Basins (Figs 2 & 3) show abundant large-scale
10 cm yr -~ convergence rate (Minster et al. structural evidence of compression including
1974), all the rupture zones have evolved faults and folds (Figs 8A & 13), as well as the
within the trench and beneath the first 15 km of highly sheared dolomicrite. The entire Yaquina
the subduction complex during the past 0.5 Ma, Basin sequence is disturbed to a depth of 2 km,
with the trench rupture occurring during the including the acoustic basement. Age dating of
past few thousands or tens of thousands of the Trujillo Basin dredged lithologies (Kulm et
years. Inclusion of basaltic material into the al. 1981b) indicates that the Pleistocene section
subduction complex is strongly implied from is thinner here than in the Lima Basin;
geological and geophysical considerations (Fig. sediment trapping in the Salaverry Basin to
12; Kulm et al. 1981a). These data suggest that the east of the OSH (Fig. 2), and subsequent
the complex is actively accreting material, even deposition in the upper slope basins may be
at 12°S, where tectonic removal of the continen- responsible.
tal plate is postulated by Hussong et al. (1976). We cannot rule out a moderate amount of
However, this relationship is similar to the subsidence in the Trujillo and Yaquina Basins
Japan forearc where accretion along the lower because our palaeodepth information was
continental slope with large-scale subsidence obtained only at water depths between 202 and
immediately landward on the middle to upper 430 m, the approximate depth of the original
slope is well documented (von Huene, Nasu et environment of deposition. (Several attempts
al. 1978). The data presented in this paper were made to recover lithologies at deeper
suggest the Peru margin, between 9o30 ' and water depths, but each one failed. The dolomic-
14°S, is experiencing a tectonic regime of accre- rites are probably present in the Yaquina Basin
tion (uplift) and subsidence similar to that of because the dredges--one was lost--contacted
the Japan forearc. very resistant outcrops there, typical of the ones
Continental accretion seems even more likely dredged successfully.) The disturbed structures
at 9°S, because of the large basaltic thrust ridge of these basins, however, are more typical of
in the trench axis and slab rupture beneath the compressional than extensional tectonics, at
subduction complex. least in the upper slope region.
A major structural change occurs near lati-
tude 9°30'S between the highly disturbed Trujil-
1o and relatively undisturbed Lima Basins
Tectonics of upper slope basins
(Thornburg & Kulm 1981). The two basins are
The sedimentary deposits in basins on the offset here as is the Upper Slope Ridge of
continental slope offer important clues about deformed sediments (Figs 2 & 3). This bound-
the tectonic framework and processes affecting ary coincides with the strike of the Mendana
the forearc. The Lima Basin (Figs 2 & 3) has Fracture Zone on the Nazca Plate which in-
undergone from 500 to 1100 m of subsidence tersects the margin at this geographical location
during the past 4 Ma at 12°S (Kulm et al. 1981b), (Fig. 1). The basaltic ridge (Fig. 5) also termin-
probably over much of its 520km length ates near the fracture zone. Unfortunately, we
(Thornburg & Kulm 1981). Subsidence occur- have no geophysical data parallel to the con-
red without intensive deformation of the basin tinental slope to determine whether the fracture
deposits (Fig. 14) in contrast to the Trujillo and zone does influence the structure of the margin.
Yaquina Basins to the north (Figs 8A, 13 & 17).
Nevertheless, the Lima Basin exhibits subsur-
face high-angle reverse faults (Johnson & Ness Comparison of Peru and other forearcs
1981; Thornburg & Kulm 1981) possibly caused
by compression which uplifted the Upper Slope The central Peru forearc has many similar-
Ridge (Figs 2 & 3) forming the seaward bound- ities to other circum-Pacific forearcs, but it is
ary of the basin. The compression either pre- quite dissimilar in other aspects. The Peru
dated or was, in part, synchronous with basin continental arc massif consists of metamorphic
subsidence. Our geographically limited strata similar to that drilled off southern Mexico
palaeodepth information indicates that the (Moore, Watkins et al. 1979), which places very
greatest rates of subsidence ( 5 0 0 m M a -1) old crystalline rocks beneath the continental
occurred during the Pleistocene, suggesting that shelf and upper continental slope. Truncation
the Lima Basin has shifted from a compression- of the continental block has been proposed on
 C e n o z o i c structure o f the central P e r u f o r e a r c 167

the basis of Palaeozoic and older rocks found Conclusions

near the shoreline and at the edge of the
margin. The seaward limit of the massif can be The central Peru margin and trench are char-
defined by drilling where the sedimentary sec- acterized by a variety of structural, strati-
tion is relatively thin, but it is often difficult to graphic, and tectonic styles. The area studied
define solely from geophysical measurements. extends from 6 ° to 14°S latitude and covers the
Subsidence of the inner shelf massif is well indentation in the coastline lying between the
documented off Mexico and strongly inferred Palaeozoic and Precambrian coastal mountains
off central Peru in the Salaverry Basin on the of southern and northwestern Peru (Fig. 2).
inner shelf (Fig. 2). Interestingly, the outer Three forearc basins, Sechura, Salaverry,
edge of the massif off Mexico subsided perhaps and eastern Pisco Basins, rest upon a
as much as 3000 m and then began a pulse of Palaeozoic-Precambrian metamorphic massif
uplift about 19 Ma ago (Moore, Watkins et al. whose surface is undulatory and block faulted.
1979). The Outer Shelf High on the Peru Portions of the massif, especially the Outer
margin also appears to be experiencing a cur- Shelf High (OSH) and Lima Platform are ele-
rent cycle of uplift which followed the subsi- vated above the deeper parts of the massif and
dence of this former topographic high. Data apparently have been uplifted in Neogene time
from both of these margins indicate that the creating folds and faults in the overlying strata.
massif is an unstable block that periodically While the massif must have subsided to allow
undergoes vertical isostatic adjustments. The the accumulation of as much as 3.0 km of
block-faulted basement of the Peru massif (Tra- Mesozoic and Cenozoic sediment, it has prob-
vis et al. 1976) and numerous unconformities ably been influenced by several pulses of uplift
reflect these movements. and subsidence during its long history.
The Lima Basin on the upper continental Three additional forearc basins, Yaquina,
slope is subsiding at the rate of 275- Trujillo and Lima, occupy the middle to upper
5 0 0 m M a -~ since late Miocene time. Geo- continental slope and are bounded on the east
physical data suggest the underlying basement by the OSH and on the west by an upper slope
is the metamorphic massif (Fig. 10). A similar ridge of deformed sediment. Drill holes in the
rate of subsidence was first reported for the eastern Trujillo Basin contain late Eocene
Japan forearc basin deposits laid down on an sandstone with overlying post-Eocene shale,
Upper Cretaceous, steeply dipping, black silici- siltstone, marlstone, and minor thin limestone
fied claystone (Arthur et al. 1980). The Japan beds, whereas the western Trujillo Basin and
forearc subsided from 18 to 4 Ma ago with possibly the Yaquina Basin to the west contain
subsequent uplift. The subsidence was in part late Miocene to Pleistocene micrite, dolomic-
contemporaneous with the formation of a small rite, and glauconitic micrite dredged from rock
accretionary wedge at the toe of the continental outcrops. The Lima Basin has similar carbonate
slope off Japan (Langseth, Okada et al. 1978; lithologies.
von Huene, Nasu et al. 1978). Low-velocity The Trujillo Basin and particularly the Ya-
material at the toe of the slope at 12°S off Peru quina Basin are experiencing severe internal
(Fig. 11) and the underthrusting of the basaltic disruption which is expressed by fault and fold
slab upon itself (Figs 9 & 10) show that a similar structures in seismic records and by sheared
subduction complex probably exists here. Both dolomicrite obtained in the dredges. Compress-
forearcs are characterized by fast convergence ional stresses best explain the structural style of
rates (8-10 cm yr-X). The similarities of the these basins. In contrast, the Pliocene to Pleis-
Neogene tectonic histories of these two forearcs tocene sediments in the Lima Basin indicate
is striking and warrants further study. subsidence at a rate of 2 7 5 - 5 0 0 m M a -~ and
The forearc basins on the continental slope thus tensional stresses. An earlier phase of
off central Peru exhibit various degrees of uplift and compression is proposed for the
internal deformation with the Yaquina Basin Lima Basin because of the presence of subsur-
being the best example of extreme deformation face reverse faults.
(Fig. 8A). In fact, all slope basins are faulted to Multichannel seismic records and geological
some extent which indicates the entire con- data at 9°S and 12°S show that the subducting
tinental slope is undergoing structural adjust- oceanic plate within the trench (9°S) and be-
ments in response to compressional or tensional neath the overriding plate (9 °, 12°S) is rupturing
stresses. We do not have enough structural and by thrust faulting of the upper layer 2 basalts.
stratigraphic control to determine whether Slivers of basalt may be incorporated into the
these stresses co-exist or follow one another in overlying sediments of the subduction complex
some rational fashion. forming a sediment-basalt m61ange.
168 L . D . Kulm et al.
Several models are proposed which place the f o r m a t i o n of the P e r u u p p e r slope basins at 9°S
subduction complex-massif interface from 15 to s e e m s to be r a t h e r u n u s u a l with respect to o t h e r
60 km from the trench axis depending upon circum-Pacific f o r e a r c basins in a similar posi-
which set of geophysical criteria are used to tion.
define these features. T h e s u b d u c t i o n c o m p l e x
is at least a 15 k m wide w e d g e a n d the massif
e x t e n d s s e a w a r d to the u p p e r m o s t c o n t i n e n t a l ACKNOWLEDGMENTS:This research was sponsored by
slope in all m o d e l s . the Office of International Decade of Ocean Explora-
T h e central P e r u f o r e a r c at 12°S has m a n y of tion, National Science Foundation, under grants GX
the tectonic characteristics of the J a p a n forearc 28675, IDOE71-04208, and OCE76-05903. Hawaii
off H o n s h u , J a p a n . T h e severe internal de- Institute of Geophysics Contribution No. 10601.

References
ARTHUR, M., VON HUENE, R. & ADLESECK, C., JR. T. M., MASIAS, A. ~9. JOHNSON, L. 1981b. Late
1980. Sedimentary evolution of the Japan fore- Cenozoic carbonates on the Peru continental
arc region off Honshu, legs 56 and 57, Deep Sea margin: Lithostratigraphy, biostratigraphy, and
Drilling Project. In: SCIENTIFICPARTY (eds). Ini- tectonic history. In: KULM, L. D., DYMOND, J.,
tial Rep. Deep Sea drill. Pro]. 56 & 57, Part I, DASCH, E. J. & HUSSONG, D. M. (eds). Nazca
521-68. U.S. Govt Printing Office, Washington. Plate: crustal formation and Andean converg-
COULBOURN, W. T. & MOBERLY, R. 1977. Structural ence. Mem. geol. Soc. Am. 154, in press.
evolution of forearc basins off southern Peru and , YON HUENE, R. et al. 1973. Initial Rep. Deep
northern Chile. Can. J. Earth Sci. 14, 102-16. Sea drill. Proj. 18. U.S. Govt Printing Office,
HUSSONG, D. M., EDWARDS, P. B., JOHNSON, S. H., Washington, 1077 pp.
CAMPBELL, J. F. & SUTTON, G. H. 1976. Crustal LANGSETH,M., OKADA, H. et al. 1978. Near the Japan
structure of the Peru-Chile Trench: 8°S--12°S Trench: transects begun. Geotimes, 23, 22-6.
latitude. In: SorroN, G. H., MANGHNANI, M. H. MASIAS, A. 1976. Morphology, shallow structure, and
& MOBERLY, R. (eds). The Geophysics of the evolution o f the Peruvian continental margin, 6 °
Pacific Ocean Basin and its Margin. Geophys. to 18°S. Thesis, MS, Oregon State Univ., Corval-
Monogr. Am. geophys. Union, 19, 71-86. lis, 92 pp.
JOHNSON, S. H. & NESS, G. E. 1981. Shallow struc- MINSTER, J. B., JORDAN, T. H., MOLNAR, P. &
tures of the Peruvian margin 12°S-18°S. In: HAINES, E. 1974. Numerical modelling of instan-
KULM, L. D., DYMOND, J., DASCH, E. J. & taneous plate tectonics. Geophys. J. R. astron.
HUSSONG, D. M. (eds). Nazca Plate: crustal Soc. 36, 541-76.
formation and Andean convergence. Mem. geol. MOORE, J. C., WATKINS,J. S. et al. 1979. Off Mexico:
Soc. Am. 154, in press. Middle American Trench. Geotimes, 24, 20-2.
JONES, P. R. 1981. Crustal structures of the Peru MORDOJOVlCH, D. 1974. Geology of a part of the
continental margin and adjacent Nazca plate, 9°S Pacific Margin of Chile. In: BURK, C. A. &
latitude. In: KULM, L. D., DYMONO, J., DASCH, DRAKE, C. L. (eds). The Geology o f Continental
E. J. & HUSSONG, D. M. (eds). Nazca Plate: Margins, 591-8. Springer-Verlag, New York.
crustal formation and Andean convergence. NUR, A . & BEN-AVRAHAM,Z. 1981. Consumption of
Mem. geol. Soc. Am. 154, in press. aseismic ridges and volcanic gaps in South Amer-
KULM, L. D., PRINCE, R. A., FRENCH, W., JOHNSON, ica. In: KULM, L. D., DYMOND, J., DASCH, E. J.
S. & MASlAS, A. 1981a. Crustal structure and & HUSSONG, D. M. (eds). Nazca plate: crustal
tectonics of the central Peru continental margin formation and Andean convergence. Mem. geol.
and trench, ln: KULM, L. D., DYMOND, J., Soc. Am. 154, in press.
DASCH, E. J. & HUSSONG, D. M. (eds). Nazca PRINCE, R. A. & KULM, L. D. 1975. Crustal rupture
Plate: crustal formation and Andean converg- and the initiation of imbricate thrusting in the
ence. Mere. geol. Soc. Am. 154, in press. Peru-Chile Trench. Bull. geol. Soc. Am. 86,
, RESIG, J. M., MOORE, T. C., JR & ROSATO, V. 1639-53.
J. 1974. Transfer of Nazca Ridge pelagic sedi- - - & SCHWELLER, W. J. 1978. Dates, rates and
ments to the Peru continental margin. Bull. geol. angles of faulting in the Peru-Chile Trench.
Soc. Am. 85, 769-80. Nature, London, 271,743-5.
--, SCHWELLER, W. J. & MASIAS, M. 1977. A SCHOLL, D. W., CHRISTENSEN,M. N., VON HUENE, R.
preliminary analysis of the geotectonic processes & MARLOW, M. S. 1970. Peru-Chile Trench
of the Andean continental margin, 6 ° to 45°S. In: sediments and sea-floor spreading. Bull. geol.
TALWANI, M. & PITMAN, W. C. (eds). Problems Soc. Am. 81, 1339-60.
in the Evolution o f Island Arcs, Deep Sea Tren- SCHWELLER, W. J., KULM, L. D. & PRINCE, R. A.
ches, and Back-arc Basins. A m . geophys. Union, 1981. Tectonics, structure and sedimentary
M. Ewing. Ser. 1,285-301. framework of the Peru-Chile Trench. In: KULM,
- - , SCHRADER, H.-J., RESIG, J. M., THORNBURG, L. D., DYMOND,J., DASCH, E. J. & HUSSONG, D.
 C e n o z o i c structure o f the central P e r u f o r e a r c 169

M. (eds). Nazca Plate: crustal formation and THORNBURG,T. M. & KULM, L. D. 1981. Sedimentary
Andean convergence. Mem. geol. Soc. Am. 154, basins of the Peru continental margin: structure,
in press. stratigraphy, and Cenozoic tectonics from 6°S to
SEELY, D. R. 1979. The evolution of structural highs 16°S latitude. In: KULM, L. D., DYMOND, J.,
bordering major forearc basins. In: WATKINS,J. DASCH, E. J. & HUSSONG, O. M. (eds). Nazca
S., MONTADERT, L. & DICKERSON, P. W. (eds). Plate: crustal formation and Andean converg-
Geological and Geophysical Investigations of ence. Mem. geol. Soc. Am. 154, in press.
Continental Margins. Mem. Am. Assoc. Petrol. TRAVIS, R. B., GONZALES, G. & PARDO, A. 1976.
Geol. 29, 245-60. Hydrocarbon potential of coastal basins of Peru.
SHEPHERD, G. L. & MOBERLY, R. 1981. Coastal In: HALBOUTY,M. T., MAHER, J. C. & LIAN, H.
structure of the continental margin, northwest M. (eds). CSrcum-Pacific Energy and Mineral
Peru and southwest Ecuador. In: KULM, L. D., Resources. Mem. Am. Assoc. Petrol. Geol. 25,
DYMOND, J., DASCH, E. J. & HussoNc, D. M. 331-8.
(eds). Nazca Plate: crustal formation and Andean VON HUENE, R., NASU, N. et al. 1978. Japan Trench
convergence. Mem. geol. Soc. Am. 154, in transected. Geotimes, 23, 16-20.
press.

L. D. KULM, T. M. THORNBURG• H.-J. SCHRADER, School of Oceanography,

Oregon State University, Corvallis, Oregon 97331, U.S.A.
J. M. RESIG, Hawaii Institute of Geophysics, 2525 Correa Road, Honolulu, Hawaii
96822, U.S.A.
 Forearc and other basins, continental margin of northern and
southern Peru and adjacent Ecuador and Chile

R. Moberly, G. L. Shepherd & W. T. Coulbourn

SUMMARY: The continental margin between the Peru-Chile Trench and the west coast
of South America is the type example of the Andean margin, where one of the two
converging blocks of lithosphere is oceanic and the other is continental. Along Peru and
Chile, basement at the coast and under the landward side of forearc basins is dominantly
metamorphic and plutonic. Apparently, the edge of continental crust has been stoped
away during subduction. The seaward side of basins may be structurally high dams of
sediments accreted and deformed by subduction, or they may be horsts of continental
basement. Subsurface mapping in the on-shore and off-shore oilfields of the forearc basin
of north-western Peru has shown a mosaic of faulted blocks of tensional origin, overlain by
low-angle gravity glides. Details of this faulting, however, cannot be seen on single- or
multi-channel seismic records of forearc basins, even where they cross the fields. Off
northern and southern Peru, forearc basins are separated from one another along strike by
culminations of the basement; across strike there may be more than one forearc basin. The
pattern of culminations and depressions may be inherited from structures of the oceanic
crust and trench. Deeper seismic reflectors of the basins generally dip landward and
converge seaward, and show progressive migration shoreward of the centres of deposition.
Sediment movement into the basins and down the slopes is by turbidity currents and
slumps. Subduction is an efficient process, removing about four-fifths of the trench
sediments.
The Progreso Basin under the Gulf of Guayaquil owes its origin to a transform fault, the
Dolores-Guayaquil megashear, oblique to the trench. It is a pull-apart basin or rhombo-
chasm opened in the wake of the northward movement of a small plate. A forearc basin at
Talara and the adjacent Progreso Basin have substantial hydrocarbon accumulations.

The evolution of the Nazca oceanic lithospheric

plate was chosen as a topic for study in the U.S.
programme of the Geodynamics Project
(Drake 1973), and a large part of that research
was conducted in the Nazca Plate Project of the
International D e c a d e of Ocean Exploration, a
part of the U.S. National Science Foundation.
The broad multidisciplinary programme in-
cluded participants from the United States as
well as South American institutions, and in
particular the universities of Hawaii and Ore- RICA
gon State.
Hawaii Institute of Geophysics concentrated
its marine surveys of the convergent margin in
two areas (Fig. 1), between 2 ° and 7°S, where
the coastline of South America is closest to the
Peru-Chile Trench (Shepherd 1979), and be- FIG. 1. Areas of the South American mar-
tween 18° and 23°S, where the coastline and gin described in this report.
trench change their strike at the Arica Bight
(Coulbourn 1977). The purpose of this article is
to compare and contrast the tectonic and (1969) and Gansser (1973). The tectonic evolu-
sedimentary processes between the two areas, tion of the Nazca Plate in the Cenozoic has
and to show the degree to which our interpreta- been described by H e r r o n (1972), Handschu-
tions support or n e e d to modify some currently macher (1976), and Mammerickx et al. (1980).
held concepts of processes at convergent plate Geophysical investigators of the convergent
boundaries. zone between the A n d e s and the oceanic plate,
Comprehensive reviews of A n d e a n geology i.e. of the P e r u - C h i l e Trench and the continen-
are by Weeks (1947), Jenks (1976), Bellido tal margin, include Fisher & Raitt (1962),

171
172 R. Moberly, G. L. Shepherd & W. T. Coulbourn

82°W 80°
, I \ , x~iii~ I
Q D Quaternary alluvium and \
marine terrace deposits
TPv ~ Pliocene to recent volcanics I
_ -Tp ~ Pliocene sediments I Q
T m r ~ Miocene sediments / G~uKayaquil
TOITFTTF]Oligocene sediments I ~ K...ii?e:il
T e ~ Eocene sediments (Some I Sail PI
•Kp ~ P
Paleocene in Ecuador) I
K ~ Upper Cretaceous sediments I
Kp ~ Lower Cretaceous basic I
igneous complex: I
Pinon fm. or equiv. ~ I GUI F
PI ~ Undifferentiated Paleozoic I / OF
sediments and metamorphlcs ] /
KTg~ Cretataceous-Tertiary / \'IGUAYAQU/L .ISLA
intrusives-granodiorlte and / ( SA)NTA/
/CLARA
diorite / "~
Mg i';-++'~+
Mesoz°ic intrusives"granite / ~ Tumbe~,j~ ~ , <
~ Faults 4 / , f.. / Zorrit~.~i,~,\ '
Anticline ~ ~,,4 N C O "/"i.:;,,~E~.. . :"~ " . . _.l~.,~\ \ ~ \

~\\-
I

I
/
/
I

Taloro
(
3
l -'L
; LQ
-//

~//~"

I'~1
s~ ~ O ~'~ ~::~i!i!i!ii::i::ili::ili::iii::iii::i::i::~::.
: : ~!!ii!i rQ!i:i::':':'::::":-
. K ":~ ~. ~KTg "~)k~p~KTg~'//'
~

'~ 2 T ,
PI
' k 'Jura
%

5A S E C HU R A

5B - " - f
6
I ~' ) DESERT
Kp

~Pi K~
\ ~I .-,
\ t .

/-, .

I /oo ~ LOBOSDE
0 40 ~, TIERRA
NAUTICAL
KILO.METERS
MILES
0 8~0 .I
/ PI ISLA
~,LoBOSDE
I \ t r ") "~ AFuERA I
82°W 80*

Fl~. 2. Tectonic setting of NW Peru and adjacent Ecuador. Horsts of pre-Cenozoic metamorphic and
instrusive rocks in Cerros de Amotape, the headlands at Paita and Bayovar, and Lobos Islands.
Location of Figs 3 (oilfield subsurface), 4 and 5 (single-channel seismic lines), and 6 (multi-channel
seismic line) are shown. Map is simplifed from Shepherd (1979).
 Forearc basins o f Peru, E c u a d o r a n d Chile 173
Hayes (1966), Hussong et al. (1976), Kulm et al. igneous rocks are exposed within 80 km of the
(1977), and Lonsdale (1978). axis of the Peru-Chile Trench, and important
petroleum resources exist in a forearc basin.
Generally speaking, oilfields and crystalline
Basins of the convergent margin basement so close to a trench are unusual
features of arc-trench gaps.
Sedimentary processes in the basins of coastal
The economic importance of the coastal
western South America and the adjacent Pacific
basins of NW Peru has resulted in their intense
differ in degree and detail from similar proces-
study. Much information from the oil fields
ses in a number of more familiar areas. The
remains unpublished, but an introduction to the
climate is exceptionally arid, so there are no
geology of the Coastal Province and adjacent
major streams entering the Pacific along the
border of the High Cordilleran Province is in
Peru and northern Chile coasts. Recently up-
the local studies and regional reviews of such
lifted coastal terranes of unconsolidated sedi-
investigators as Travis (1953), Fischer (1956),
ment, and ash from Andean volcanoes provide
M. Paredes (1958), J. Paredes (1966), Cossio &
abundant detritus for the intermittent streams
Jaen (1967), Ruegg (1967), Martinez (1970),
that do exist. Upwelling and the Humboldt
Paz (1974), Morris & Aleman (1975) and
Current give waters too cold to allow reef
Shepherd (1979).
growth but with nutrients to support a high
The sediments of the forearc basins overlie or
level of organic productivity. A result is the
surround horsts, of a basement of Palaeozoic
layer of oxygen-depleted waters which inter-
and Mesozoic sedimentary, metasedimentary,
cepts the continental slope, allowing phos-
and igneous rocks. The oldest of these are
phorites to form (as they did in the Miocene of
exposed discontinuously in low massifs curving
Sechura) and organic matter is preserved in a
south from the Cerros de Amotape to Cerros de
band along the margin (Burnett 1977). The
Illesca and to islands offshore (Fig. 2).
steep slopes of the margin allow slumping at
Although not exposed here, similar Palaeozoic
various scales; abundant large and small earth-
metasedimentary rocks lie on Precambrian
quakes may trigger slumps and turbidity currents.
metamorphic and or intrusive rocks nearby in
Typical surface sediments between the coast southern Ecuador (Campbell 1975), the Cordil-
and the trench are diatomaceous muds. They lera Central (Hosmer 1959), and in southern
are green and grey as a result of the near- coastal Peru (Bellido 1969). Thick volcanic
reducing conditions, and have epiclastic and
rocks and volcaniclastic sediments of Triassic,
pyroclastic volcanic components. Glauconite Jurassic, and Late Cretaceous ages lie east of the
pellets are abundant, and phosphorite nodules coastal massifs. Within the massifs and in the
and early diagenetic pyrite are common locally. thick Mesozoic section are dioritic, granodiori-
Submarine canyons, basins, and the trench also tic, and granitic intrusions of the Andean coas-
contain turbidite sands, identified by modest tal batholith. The intrusions are mainly of
grading and benthonic calcareous foraminifers Palaeocene and Eocene age (Giletti & Day
displaced from shallower depth. On the oceanic 1968) and east of the massifs, but some Meso-
plate, because of the depth of water, dissolution zoic granites are exposed at the coast near
of microfossil tests has occurred, and the red- Bayovar and Paita.
brown to yellow muds contain planktonic nan- Massive normal faulting began in Late Cre-
nofossils and benthonic agglutinated foramini- taceous time and continued through the Ceno-
fers. The environment and the lithogenous, zoic into the Quaternary, providing the
biogenous, and hydrogenous sources of sedi- framework of the landward sides of the forearc
ment are given in more detail by Rosato et al. basins. They have filled with Cretaceous and
(1975), Coulbourn (1977), Shepherd (1979), Cenozoic detrital sediments of paralic and
and Burnett et al. (1980). marine facies. The section near Talara is about
The regions of this report, as well as the 9000 m thick (Travis 1953). Unconformably
Peruvian margin between them (Masias 1975; above 150 m of Aptian-Albian limestones are
Kulm et al. 1981), have forearc, trench-slope, about 1000 m of Senonian-Maestrichtian shales
and trench basins to trap sediment. We describe with subordinate sandstones and conglomer-
mainly the forearc basins, but also include a ates, as much as 2000 m of Danian shales and
description of a 'pull-apart' basin and comment lenticular sandstones, and a thick Eocene sec-
on some other aspects of sedimentation. tion. The Eocene is predominantly silty,
Forearc basins of NW Peru micaceous shale, but also contains oil-bearing,
feldspathic sandstones and a few conglomer-
In coastal NW Peru, old metamorphic and ates. Surface mapping and subsurface data from
NW ~ SE
.-
Platform
,,,, Platform -~
o 1131/Toblazo-Pleist,
Sea level H32~12 H32-I ~ H3-4 H3-5 HI-I HI-5 HIP-I ~ H-I-6 ,CB 1722 1659 1547 ~ ....... 1547 1538
Sea bottom 1 ] ........ TalazSI/~ / b Ech I , ~1-,_o, ~ ~ o c mus-.IM, to L ! E o c e n e
Ech j ~ h ' 4-- ' CB I ,.D b ~,, i -. ; d
C B ~ / , - ~ -- _. I'~'C ~ h ,/-~.J ~ ~ ' i""" Ecn. i ~ ..... ~'~

t , . n - •

s ...-:::~:~....:~:~:~Niiiiiiii

:.
" / i L°' o, -:i::':':':o"'o~On ~..~ ........ :....

"i"""':':':':'""" ....
"':':':':': B'a'lc'.::::::::" Sc
::':':~"~'~- cene
~-L. Eocene
~-
iiiii;iii!ii;i!iii!ii!iiiiiiiiiiiii
0 500 IO00m "..:.:.?:.:.::.... .:: ~.o.(c one...:.:.:.::::::.-.:.:."
CROSS-SECTION THROUGH I , J
Horizontal & Vertical scale
RESTIN-CABO BLANCO OFFSHORE OILFIELDS

FIG. 3. Off-shore and on-shore section through an oilfieid 15 km NE of Talara. Peru, near Cabo Blanco. For location see Fig. 2. Eocene and
Palaeocene section is broken by high-angle normal faults, and later cut by low-angle gravity slides. This is the inner edge of a forearc basin.
Coastal oilfields extend 50 km either side of Talara. Fields between Zorritos and Isla Santa Clara (reference Figs 2 & 13) are in Progreso Basin.
Section courtesy of Belco Petroleum Co.
 Forearc basins o f Peru, E c u a d o r and Chile 175

nearly 10 000 wells show exclusively normal economic at the present time. Most of these
and low-angle gravity faulting, rare minor on-shore and off-shore fields and seeps are near
warpings, and no folds. Gravity slides with Talara in the landward part of the forearc basin.
rubble, resembling olistostromes, extend sea- Elsewhere in the world oilfields are uncom-
ward at angles of a few degrees (Baldry 1938). mon in arc-trench gaps. Those of Barbados and
Many of the master faults cut basement. The Timor are non-commercial. The St Elena fields
subsurface information has been extended to of coastal SW Ecuador (1.5 x 107 t cumulative),
the offshore oilfields (Fig. 3). The entire region and off-shore fields east of Trinidad and west of
is being broadly uplifted, amounting to more Palawan are of commercial value. Poor quality
than 300 m since the early Quaternary. of reservoirs, low geothermal gradients for
The seaward edge of the forearc basins is maturation of organic-rich sediments to hydro-
beyond the narrow continental shelf, at one of carbons, and complex geology have tended to
the topographic breaks of the slope. Typically, bias explorationists away from convergent mar-
the edge is not at the most prominent of the gins.
breaks. Also, the edge is the locus of the
inflection in free-air gravity-anomaly profiles,
suggesting that denser rocks may underlie the Forearc basins of southern Peru-northern Chile
trench-slope break. Examination of single-
channel (Figs 4 & 5) and multi-channel (Fig. 6) The region of southernmost Peru and north-
seismic profile records shows major faults, but ernmost Chile is characterized by a major
does not show the mosaic of block faults and change of strike of the Peru-Chile Trench, by
olistostrome-like slides known to exist from great ruptures in the oceanic lithosphere sink-
offshore drilling (Fig. 3). The minor structures ing under South America (Rodriguez et al.
on the geophysical profiles look like folds, and 1976) and by active volcanism, in contrast to
slides may be indistinguishable from packets of gaps in the line of active volcanoes to the north
sediment. We believe that the apparent differ- and south. The change in strike at Arica also
ence between coastal-zone structures and those marks a change in rupture length of great
farther offshore is due to the different investiga- earthquakes (Kelleher & McCann 1976). East
tive techniques (seismic versus subsurface). of Arica and Inquique the Andes are of their
The bathymetry, sub-bottom profiling, and greatest width.
bottom samples show mass movement by As is the case for NW Peru, rocks character-
slumps and slides down the slopes, and turbid- istic of continental crust are exposed along the
ity currents and mass movement in submarine coast of southern Peru and northern Chile
canyons to be the important processes of (Farrar et al. 1970; Paredes & Megard 1972;
sediment transport. Canyons generally head at Frutos & Feraris 1973; Pitcher 1974; Audebaud
about 100 m depth, and extend to the trench at et al. 1973). The Peruvian Coastal Cordillera
about 5000 m, and thereby cross different tec- from near Pisco (14°S) to within the study area
tonic domains. They have cut through consoli- (Fig. 7) has at the coast mainly thick
dated sedimentary rocks of the upper slope as is metasedimentary and metavolcanic sections of
evident on the sub-bottom reflection records Precambrian age. Farther south and beyond the
(Fig. 5). They can be traced bathymetrically area of study the Peruvian and Chilean coasts
across the lower slope, but sub-bottom reflec- are bordered chiefly by granitoid intrusives of
tions cannot be resolved there. If the lower Permian through Palaeogene age, and locally
slope is underlain by an accumulating wedge of by middle Palaeozoic to late Mesozoic
tectonized sediments, the structural grain of sedimentary and metasedimentary rocks, or
that rising wedge is crossed by canyons. Thus it Mesozoic (submarine) and Cenozoic volcanic
is apparent that the canyon-cutting processes rocks.
were sufficiently efficient to maintain a canyon Large forearc basins, with about 1.5 sec
in its course rather than allow it to be deflected (perhaps 2.9 km) of moderately deformed sedi-
by wedge morphology. Canyons are essentially ment in them, lie under the bight between 17°
free of sediment. Thus canyon-cutting and and 20.5°S. Farther south and farther off-shore
transport in canyons are active geological pro- the forearc and trench-slope basins are smaller,
cesses (Shepherd & Moberly 1981). thin, and discontinuous; some have moderately
The complexly faulted oil pools in the coastal deformed sediment but most show little evi-
fields of Peru are of 'giant' status, by virtue of dence of deformation (Fig. 7). Of 29 crossings
having produced more than 1.3 × 108 tonnes of of the trench, 13 show no sediment. Sediments
oil. Tar seeps and tar sands constitute an enor- are thin on the oceanic plate. Sediments of the
mous potential resource of energy that is sub- region have been studied by Schmalz (1958),
176 R. Moberly, G. L. Shepherd & W. T. Coulbourn
100 80 60 40 20 OKM
I I I I I,, I l I

/
.//
VE GX

40 20 0 KM
I I I I
0
c-
O
U

,,= ,.(
E
I--"
f . .
x

Q.
VE '-' 9.75 X IJ,J
I.U
rv"
I--"

--~LE ,2 ,_

140 120
I I i I ,4

•= 6
C

FIG. 4. Tracings of single-channel reflection records. For location see Fig. 2. Line A (top) shows no
faulting of sediments and basement of oceanic plate (left). Trench floor is flat with turbidites. Record
of inner trench wall is poor. Minimum free-air gravity anomaly is at about 80 km and maximum is at
50 km. Faults between 30 and 40 km have the same trend as faults on other profiles on SE edge
of Banco Peru. Basin between 20 and 30 km is NE of Talara Forearc Basin. Basin between 0
and 15 km is SW end of Progreso Basin. Line B (middle) is typical of many profiles in the area:
Oceanic basement and its pelagic cover are faulted; trench has an apparent mixture of pelagic,
turbidite, and slump deposits; inner trench wall offers unresolvable acoustic reflections suggesting
deformed sediments; sub-bottom structure and gravity inflections indicate that the small trench-
slope break at 63 km (rather than the break at 45 km) is the 'structural high' marking the outer limits
of the Talara Basin; canyons cut the upper slope; some reflectors of the Talara forearc basin are
broken indicating high-angle faults, but most reflections suggest folds or unconformities. The inner
end of line B overlaps Fig. 3 within the Cabo Blanco oilfields. Line C shows slumps at (and in?) the
trench. The trench-slope break at 60 km marks the seaward edge of the Paita forearc basin.

FIG. 5. Tracings of single-channel reflection records. For locations, see Fig. 2. Lines A (top) and B
(bottom) show Sarmiento Ridge, one of several fracture zones on the N.azca Plate that strike NE into
the trench. O'n a structural basis, the trench-slope break is picked at 60 km in each profile.
Apparently, the seaward edges of the sediments in the Sechura forearc basin are slumping away from
the upper slope. Internal reflections suggest folds or unconformity-bounded packets of sediment•
Eastern end of each line is within 15 km of test well V I R U 4X-1.
 150 120 90 60 30 0 KM
I I I I I I I I I I I I I I
' ~0
E
/ t.-.--
_,J
t.U

I---
13_ 2
LU
2 r"
Q~
LI.I
I---
SARMI ENTO 4 I
RIDGE / .. ik~ 0
. ..~/................?~.~
•
~;~. i ~_-',,.~ -
j.~
VE 13X
\
I,.--
~ ~ ~;r---,: ~---~i~' }[[.
30 0 KM 2"
I I I I
0

60
1 . I

SARMIENTO
RIDGE

150
I I 1.
-4
no record
time log '~.4o
1.4 km offset ,,,/' -; 1~:

VE 12.5X VE 10X , ",-~-~m

_
.
I
,..J
 oo

e,-
80 60 40 20 0 KM
i i J t i i J i i i
8
tilt
0 Plio- Pleistocene ~..
Oligo-Miocene (%
Eocene "~
Cret- Paleocene ~ '
,,*,, ~ leozoic
,v
lid

" L:: , ~ S E I S C O M - DELTA L I N E

" 24 Channel Reflection Profile - Time Section
(%

Modified from L.K. Wipperman, unpublished data =L

FIG. 6. Multi-channel, common-depth-point seismic reflection profile, obtained by Seiscom Delta Co. for Nazca Plate Project. Line nearly lies
along middle of Fig. 5(B), for comparison. Lowest reflector at left (about 9.5 sec) probably is oceanic Moho. Highest sub-bottom reflector at left
(at about 6.7 sec) is probably top of oceanic basement, and can be traced under trench and lower slope to 46 km, possibly 37 km (offset).
Trench-slope break is about 35 kin. Incoherent reflectors between trench and the break are accreted sediment. Coherent but faulted reflectors
between the break and the end of the profile are in the Sechura Basin. Interpretation of the age and thickness of the section is projected from
VIRU 4X-l, which had a velocity survey. Further discussion is in Shepherd & Moberly (1981).
 Forearc basins of Peru, Ecuador and Chile 179

Zen (1959), Bandy & Rodolfo (1964), and but some samples seemed anomalous in loca-
Rosato etal. (1975). tion, texture, or microfauna content. A set of
The three largest forearc basins overlap or 54 samples was subjected to a variety of statis-
are separated from one another by constric- tical methods to solve for relationships between
tions, as controlled by oblique trends of the sedimentological properties and physical en-
structurally high ridges of the more highly vironment. Variables, methods, computations,
deformed acoustic basement (Fig. 7). The and results are in Coulbourn (1980). In general
there are four clusterings of sediment. One
group of samples is characterized by agglutin-
ated foraminifers, nannofossils, red to yellow
colours, and dominantly clay size. Most sam-
ples come from the Nazca-plate side of the
trench, and the microfossils and seismic profiles
show them clearly to be pelagic. Two samples
clustered with them were recovered several
kilometres landward of the trench, however, at
depths of 4700 and 3700 m, and presumably
indicate incorporation of pelagic sediment in an
uplifted tectonic wedge.
The other three clusters are characterized by
grey-green colours and relatively high contents

?,
of fecal pellets, glauconite, biotite, glass, and

D
t
AHuvium and salt ashore;
\\ detrital grains. They are hemipelagic. One clus-
ter is found at a range of sites over forearc
basins and the inner trench slope, and repre-
sents unconsolidated hemipelagic mud that has
unresolvoble acoustic
returns E O~ t,er, ch; Pelagic
sods W of ~rgnch,
not been resedimented. Another cluster is also
~_~ Undeformed sed~ ma,nly
turbld,tes
found within the forearc basins, and in addition
~.'~ M~deratet't deformed seds
o} [ore - clfc basins
was recovered from the axis of a submarine
t ~ Volcan,c rock'L mainly
Cenozo,c
canyon, sites in the trench axis, and two sites
te,,d C e n o z o , c
slightly above the trench on faulted blocks of
E~ Pc31eozt,,c a . d Me$ozo,c rock, the Nazca Plate descending into the trench.
~ Metamt~rph,c rocks
(P~eccsmbr~on ~, Displaced benthonic faunas and abundant sand-
Faults
sized glauconite and detrital grains indicate
Naut,ca~ Miles 50
these to be turbidites. Presumably the two
, h t ,, E )r t , "
anomalous samples are of hemipelagic turbi-
0
Krlome,er~
~oo )A .... dites that crossed the trench axis where fault-
..... ~' • ~ 2,:Z2):I block ridges and depressions of the oceanic
__ ~L_ . . . . - L _ _ _ _
72* 70" W plate intercept the trench floor off Peru (Fig.
10), and thence were channelled to the south in
Fro. 7. Tectonic setting of the Arica Bight, trench-slope depressions that have southward
coastal southern Peru and northern Chile. gradients less than that of the axis of the trench
Coast is lined with Mesozoic batholiths and off Chile. The final cluster is a small one, of
other crystalline rocks characteristic of
continental basement. Shown offshore are olive-grey semi-indurated mudstone sampled
forearc, slope, and trench basins. Loca- on a traverse over a structural high in about
tions of Figs 9, 11 and 12 are shown. Maps 1000 m of water (Fig. l l A ) . It is an outcrop of
is simplified from Coulbourn (1977). the tectonic wedge, which in most places is
veiled by hemipelagic mud.
basins lie over gravity minima (Fig. 8). Sedi- Reflection records, bathymetry, and sam-
ments within the Arequipa, Arica, and Iquique piing indicate that whereas off northern Peru
forearc basins can be separated into packets slumps and turbidity currents seem to be of
between successive reflectors. The thickest subequal importance in sediment transport, off
parts of the packets shift landward with time southern Peru and northern Chile both are
(Fig. 9), showing that the structural high has present but turbidity currents appear to be
had a long intermittent history of growth (Coul- dominant. Some slumps are seen on sub-
bourn & Moberly 1977). bottom reflection profiles, especially on the
The origin of most cored sediment samples upper slope adjacent to structural highs (Coul-
from the Arica Bight could be deduced easily, bourn & Moberly 1977). Low-angle glides can
180 R. Moberly, G. L. Shepherd & W. T. Coulbourn
,~. ,2. ,,. w ,o;
! ~ \ t _ ~.1 t ' 7"

PERU

IICA BASIN

,
o 6o(~ c "\,\.,

, ~'~,,

FREE AIR
GRAVITY ANOMALY MAP OF
THE PERU-CHILE TRENCH
17"-22"S LAT,

Values are contoured to 20 regal

2O
intervals and are refered to the
International Reference Ellipsoid
0 20 40 60 80 I00

' I
73"
"
72"
(f//l(//(l((flt
i
71"
JI s
22"
W 70*

FIG. 8. Free-air gravity-anomaly map of the Arica Bight area. Gravity minimum lies virtually along
trench axis. The forearc basins lie along a band of alternating gravity maxima and minima, suggesting
that the basins are localized along fault-bounded depressions of a segmented basement, and that
some of the faults trend at high angles to the coast. From Coulbourn (1977).

be interpreted from single-channel records of south, at an acute angle to coast and trench,
the basins (Fig. 12), using the analogy from and cut into sediments of the Arequipa and
northern Peru (Fig. 3). Perhaps some of the Arica forearc basins (Fig. l l B ) . Canyons are
boundaries of the packets of sediments (Fig. 9 less conspicuous off Chile, and the trench is
and Couibourn & Moberly 1977) are low-angle deeper, perhaps due to less sediment supply.
glide surfaces rather than structurally deformed
turbidite bedding or unconformities. Progreso Basin, northern Peru-southern
Extensive submarine canyons, recordings of Ecuador
off-shore cable breaks (Benest 1899), and dis-
placed benthonic foraminifers indicate that tur- The eastern part of the Gulf of Guayaquil is
bidity currents fill the forearc basins. After the underlain by the Progreso Basin. From south-
basins overflow, transport continues down- ern Ecuador the basin exis plunges SE to the
slope. Off southern Peru the canyons strike eastern end of the Isla Puna and thence con-
 Forearc basins of Peru, Ecuador and Chile 181

ARICA B ~
CORES
ARICA BASIN . r

m <

~..-, ..-,-,~ \Z%,-~,%,~!,~ ~.-..--:;.-.,..,,, . . . . . -

. : - ~ .-.,..._ ......: \ , , ~ - -,~ %.<~-,:C.-::.~....~_....,, .. • . . . . § i°,~
~:-:': ~'-. \>, k ". ~-'._z.,~ . . . . . ,~
3 . -
• .. ~-~_~.:~
,.. L,0
\ '" "~":"~'~:"- o ~o

KM

FIG. 9. Tracing of a seismic reflection profile through the Arica Basin. For location see Fig. 7.
Packets of sediment (shown at right) may indicate successive migration of the centre of deposition
landward as the structural high (trench-slope break) is forced upwards by accretion of a tectonized
wedge of sediments between it and the trench (after Coulbourn & Moberly 1977, fig. 10).
Alternatively, reflectors may be a mixture of bedding and low-angle glide planes. Symbols indicate
location of cores that obtained sediments and faunas indicative of hemipelagic sedimentation.

~' --~0~
FIG. 10. Perspective of the Arica Bight,
looking NW. Ridges from the Nazca Plate 1
'3I~-
descend into the Peru-Chile Trench. --~. ~ ~ --'-~---~ ~ .~---

Although nearly parallel to the Chilean

part, they intercept the Peru Trench at an
acute angle and may channel turbidites into
the oceanic side of the trench slope. The J
principal sediment supply across the fore-
arc basins, slope, and trench apparently
comes from the Peruvian margin.
FIG. 11. Profiles with evidence of geologic-
al processes in the Arica Bight. Simplified
tinues plunging SW past Isla Santa Clara (Fig. from Coulbourn (1977). For location see
2). Gravity and seismic surveys suggest that the Fig. 7. Line A (top) crosses a thin forearc
centre of the basin lies 20 k m west of Isla Santa basin (here, the arbitrary division between
Clara (Fig. 13). T h e Zorritos oilfields o n its the Arica and Iquique basins). The main
south flank p r o d u c e from M i o c e n e sandstones, structural high was cored (symbols), and
yielded indurated sediment that probably is
and gas discoveries have b e e n m a d e in the Gulf the inner edge of the accreted tectonized
of Guayaquil. T h e basin holds m o r e than 6000 wedge. Line B (bottom) shows canyons
m of M i o c e n e and y o u n g e r sediments. Its north near the coast of Peru that are cutting
and south margins are m a r k e d by steep gra- sediments of the Arequipa forearc basin.
dients in the free-air gravity anomalies (Fig. Canyons with terraces and levees in the
13~ and the south margin has growth faults Arequipa Basin have been illustrated else-
along it. T h e eastward e m b a y m e n t of the basin where (Coulbourn & Moberly 1977).
182 R. Moberly, G. L. Shepherd & W. T. Coulbourn

I 09

9 O STRUCTURAL ~ % 5 ~ . ' ~ ~ g " J ]

u.i _ ~ ~ k - ~ ; ~ -~--'/ I I

m ...~r'i~
~ _~ ~__~-~'~~ = - ~ ~ -,--
I o ~ _ ~ .m..-~. - _ ~ _ ~ ~-----~,.~-~ ~ - , - /
I ~ _..--~-. ~ = --,..'::"-,-.-=--': -~ -~ - -_-- -.=L - ,~-- " /

q
. _ .

FIG. 12. Tracing of a seismic reflection profile of a small forearc basin seaward of the lquique Basin,
showing reflectors generally converging toward the structural high. For location, see Fig. 7. Below,
the tracing has been photographed through a strip camera so that the vertical exaggeration of the
topography is reduced to 2:1. Whether or not any reflectors are low-angle glide faults is an
unanswered question; a possible interpretation is shown below.

contrasts with the strike of other South Amer- ly and genetically similar to basins forming
ican coastal basins, which follow the trend of today between thrust slices on the inner slope of
the Andes and trench. According to Faucher & trenches. In this case, however, the basin has
Savoyat (1973) the basin continues landward received a thick fill, and has been sufficiently
into the Jambeli Graben, where about 9000 m uplifted that its landward part is exposed.
of Cenozoic sediments lie. Several investigators Although Henderson & Evans (1980) also term
have proposed that the Gulf of Guayaquil and the Progreso a trench-slope basin, they differ
Progreso Basin have formed in response to from Lonsdale in interpreting the present tec-
right-lateral motion on the Dolores-Guayaquil tonic division between it and the Guayas Valley
megashear (Marchant 1961; Malfait & Dinkel- Basin, which they term a forearc basin, as
man 1972; Goosens 1973; Faucher & Savoyat dating only from the late Pliocene.
1973; Shepherd & Moberly 1975), or to interac- The Dolores-Guayaquil megashear (Fig.
tion between the Tumbes-Guayana (left lateral) 14A) is traced on land from north-western
and Dolores-Guayaquil megashears (Campbell Colombia to the latitude of Guayaquil,
1975). Ecuador. It separates a terrane of Mesozoic and
In a different interpretation, Lonsdale (1978) Cenozoic mafic volcanic and marine
terms the Progreso Basin a mature example of a sedimentary rocks in coastal Colombia and
trench-slope basin, believing it to be structural- Ecuador, which probably formed as oceanic
 Forearc basins of Peru, Ecuador and Chile 183

crust, from a terrane of Mesozoic and other

82 ° 80' metamorphic and plutonic rocks, the continen-
tal crust of northern South America. Geological
evidence for the fault zone separating two types
of crust, and its right-lateral movement, is
supported by seismicity, earthquake first-
motion, and gravity studies, as summarized by
Case et al. (1971) and Campbell (1974), who
projected the fault SW under the south
edge of the Gulf of Guayaquil. Geophysical
evidence for the extension under the Gulf into
the Peru Trench was presented by Shepherd &
Moberly (1975, 1981).
According to our interpretation, the Dolores-
Guayaquil megashear is a transform fault end-
ing at a trench-trench-transform triple junction
near 4°S. It is the eastern boundary of a small
plate uplifted from oceanic crust in coastal
Ecuador and Colombia, trapped against the
continent when the convergent margin jumped
from a former position at the megashear to its
present location at the Colombia-Ecuador
trench. Progreso Basin is a rhombochasm-like
basin (pull-apart basin of Crowell 1974) caused
by right-lateral movement on the megashear
(Fig. 14B).
The river Guayas and its many tributaries are
the principal source of sediment for the Gulf of
Guayaquil and the Progreso Basin. Presumably
the ancestral Guayas was the past source as
well. The climate-controlled gradient of sedi-
ment supply probably existed during the Ceno-
zoic, favouring a northern source. Moreover,
the former subduction zone or suture which
became the Dolores-Guayaquil megashear
would have provided an intermontane lowland
trapping runoff (Potter 1978) and directing it
laterally.
A number of present-day island arcs and
other young orogenic belts are cut at acute
angles by major fault zones that are active
80"
today or have been in the recent past. Several
of the faults control the position of rivers.
F16. 13. Free-air gravity-anomaly map of
the South American continental margin off Pull-apart basins with thick sediments may be a
NW Peru and SW Ecuador. Gravity mini- common feature of ancient convergent margins.
mum lies along and landward of the trench
axis. High-angle faults bounding the Gulf Trench and trench-slope basins
of Guayaquil and Banco Peru are seen on
reflection records or inferred from the Canyons on the slopes as well as samples of
gravity. They are splays of the Dolores- displaced faunas and turbidite sands, and profil-
Guayaquil megashear, and outline the er records of slumps and of pelagic-capped fault
Progreso Basin. Other high-angle faults blocks that descend into the trench all show that
bound massifs or horsts of Palaeozoic
metasedimentary and Mesozoic plutonic the Peru-Chile Trench is a sedimentary basin.
basement rocks. They are inferred to con- The basin, however, does not accommodate
tinue seaward and mark the lateral limits of great thicknesses of sediment. Cross-sections of
the forearc basins along the coast. profiles rarely show as much as 1 sec (two-way
travel time; perhaps 800 m) of sediment in the
trench, and sediment is nil in many crossings.
184 R. Moberly, G. L. Shepherd & W. T. Coulbourn
~- , ?" , "?°

- 10°

........- -~ -....
-
I
I
J '4
Y.: L I g "
---ti Y:: )//co~o.,,

GALAPAGOS IS ""~.~.--.

/ PROGRE SO ~ ] ! , p" .~, /

. ~ . ~ . 7 ~ j ~"~II~e---PROG~ESO eAS~N BASIN ',, ~ ~,'~"- /j

~\~"~"/~~ : ~ RU . . . . . .

NAZCA PLAT~ "~

I l r I I
9O" 00" 7O°

FIG. 14. Regional tectonic trends of NW South America, showing Dolores-Guayaquil Megashear as
a transform fault separating a small plate between the Colombia Trench and the South American
Plate. The transform fault may be nearly inactive today, but its earlier right-lateral movement may
have left the Progreso Basin as a rhombochasm or pull-apart basin.

Probably more than half of the sediment is Sedimentation

pelagic, entering from the oceanic plate, and The mixture of sedimentary sources and tec-
less than half is hemipelagic, redeposited as tonic processes off Peru today shows that cau-
turbidites and slumps or deposited directly as a tion is necessary in palaeoenvironmental recon-
rain of suspended particles. Pyroclastic and structions of ancient subduction zones where
authigenic contributions are easily identified exposures are incomplete and samples are few.
but modest. About four-fifths or more of the Submarine canyons are efficient avenues of
trench sediment is removed by subduction sediment supply for forearc basins, trench-slope
(Shepherd & Moberly 1981), leaving a small basins, and trenches. As the forearc basins are
fraction to enter an accretionary prism land- lifted, canyons may cut them, reworking sedi-
ward of the trench. The accretionary prism may ment and carrying it to greater depths. Mass
be more substantial where a thicker section of movement in the form of small slumps and
pelagic sediment can be provided to the trench, immense low-angle slides lead to olistrostrome-
as for example the carbonate oozes preserved like structures. They may gravitate to the same
on the Nazca (Kulm et al. 1974) and Carnegie levels as the tectonic m61ange of the accretion-
(Lonsdale 1978) ridges. ary wedge, leading to a possible source of
confusion in interpretation.
The hydrocarbon accumulations in the fore-
Geological processes along the Andean arc basins and a pull-apart basin in northern
Peru and southern Ecuador show that not all
margin
arc-trench gaps are unsatisfactory tectonic and
For the most part, our observations off South sedimentary settings for large oil and gas fields.
America support current concepts of Perhaps they are exploited off Peru because
sedimentary and tectonic processes during sub- they are partly exposed on land, unlike most
duction. In some instances, however, we be- young forearc basins, so that their tar seeps are
lieve modifications of those concepts are plainly visible. Perhaps they have received
needed. The question is, how typical of recent enough sediment to generate and mature hyd-
and ancient arc-trench gaps is the A n d e a n rocarbons, but are not so old or deformed that
margin? the oil is lost.
 Forearc basins o f Peru, Ecuador and Chile 185
VOLCANIC ARC

VOLCANIC ARC
.

FIG. 15. Alternative hypotheses for the relative roles of accretion and erosion or attrition by
subduction. Legend: (1) structural high, (2) forearc basin, (3) block-faulted continental crust under
all or part of forearc basins, (4) tectonic m~lange of the accretionary prism, (5) trench-axis turbidites,
(6) deformed and dewatered hemipelagic sediments of the continental margin, (7) Mesozoic
sediment, (8) coastal batholith. An earlier version of this illustration appeared in the 1977 JOIDES
report on the Future of Scientific Ocean Drilling.

where pelagic and hemipelagic sediment supply

Shallow deformation
is so scanty, is that some ridges of the belt may
Across the Andean margin, evidence of in part be ridges of oceanic basalt crust clipped
young compressive tectonics is restricted to: off the top of the descending lithosphere (Coul-
(1) lithospheric convergence shown by earth- bourn & Moberly 1977).
quakes, (2) the Subandean Province between Tensional tectonics prevail in all other seg-
the Cordillera and the shields, and (3) a small ments of a wide traverse across this converg-
prism immediately landward of the trench. ence zone, as may be the general case for
In the case of (1), the planes and first-motion convergent plates (Elsasser 1971; Moberly
studies of the shallow to deep earthquakes 1972). Kanamori (1971) has suggested that the
dipping under South America show the con- landward and subducting plates nearly de-
vergence. Burchfiel & Davis (1976) believe that couple from time to time, along the thrust-fault
thermal weakening of the overriding edge of interface between. The topography of an outer
the continental plate led to its middle Miocene rise may result from flexure of the oceanic plate
to recent deformation, thickening, and uplift in without buckling from horizontal forces (Cald-
response to the horizontal compressive stress well et al. 1977, for a contrary view see Watts et
at the shallow-dipping (10-15 °) Benioff zone. al. 1976), even though an interpretation off
(2) Foreland-type folding and thrust faulting central Peru suggests buckling (Hussong et al.
of the sedimentary prism characterizes the 1976). Normal faulting in the shallow arch of
Subandean Province. the outer ridge clearly is tensional (Coulbourn
For (3), the inner wall of the trench, mor- 1977), and so is the step-like fault-block pattern
photectonics and interpretation of seismic re- from the outer ridge down the outer wall of the
flection profiles suggest a belt of tight folds and trench (Shepherd 1979; Coulbourn & Moberly
imbricate thrust faults, but there is no proof of 1977).
it. An alternative to a compressive-structure Although the single- and multi-channel seis-
interpretation of the profiles in the Arica Bight mic reflection records might suggest folds (Figs
186 R. Moberly, G. L. S h e p h e r d & W. T. C o u l b o u r n

5 & 6), the subsurface information (Fig. 3) and and indeed all mountain belts will have to be
one interpretation of seismic records (Fig. 12) re-examined to determine whether or not they
show gravity faulting rather than compressive have had a similar history. To date, no such
folding. Some faults are high-angle ones that exotic blocks have been discovered in the
cut into basement (not listric), and others are Andes, but if they are, the granitic rocks at the
immense low-angle glides. Farther on-shore coast might be explained as a result of a rifted
and through the central Andes broad uplift and margin: origin of a spreading centre within
normal faulting has prevailed into recent times some Mesozoic granitic crust, movement of a
(Cobbing & Pitcher 1972; Gansser 1973). block or blocks with a Pacific passive or trailing
margin to a subduction zone along nuclear
South America, and then welding of those
Tectonic erosion and accretion during blocks to South America and the initiation of a
subduction new subduction zone along the present trench.
Until there is clear field evidence of exotic
The efficiency of the subduction process has terranes from mapped suture zones and
an obvious control on the sediments and shal- palaeomagnetic-pole calculations in the Andes,
low structures in trench and forearc basins. We we prefer the alternative that subduction has
have pointed out that most of the sediment removed the edge of the crust.
entering the trench is subducted. We believe Fig. 15 shows two extreme interpretations for
that continental crust is removed as well. Coast- the tectonic control of basins in the arc-trench
al Peru and northern Chile have plutonic and gap. One is that the accretion of a wedge of
metamorphic rocks at the coast, exposed 80- tectonized m61ange makes the structural high
140 km from the trench axis, and interpreta- separating the forearc basin from the remainder
tions of the style of deformation, the position of of the slope. Imbricate slices of the m61ange
gravity-anomaly inflections, and velocities trap hemipelagic sediments in slope basins and
along multi-channel and A S P E R seismic lines may deform and dewater them. Moore et al.
suggest that the edge of continental crust may (1980) present a current interpretation of this
be as close as 20-40 km to the trench axis off tectonic type for Sumatra.
NW Peru (Shepherd & Moberly 1981), 50 km The figure also shows an interpretation
at 12% (Hussong et al. 1976), and about 50 km where accretion is slight and subcrustal erosion
off the Arica Bight. Moreover, structures trans- is dominant. The structural high is a horst of
verse to the coast strike seaward (and help to continental crust over which sediments are
delimit the lateral limits of individual forearc draped. The trench slope under which structure
basins; Shepherd 1979; W. S. Pitcher pers. is so difficult to resolve by seismic reflection
comm. to W.T.C. 1975). Mesozoic plutons now methods is shown as a mosaic of fault-sliced
at the coast could not have formed by any blocks of continental crust entering the trench.
reasonable pattern of geotherms and dipping Their movement, and a modest amount of
plate so close to the trench. All of these accreted pelagic sediment or oceanic crust
observations strongly suggest that subduction slivers moving up over them, deforms and de-
has removed by stoping the underside and edge waters the hemipelagic blanket. This example
of continental crust. may exist off Peru, Chile, and Mexico. Or,
Karig (1974) thought that tectonic erosion more likely, truth may lie somewhere between
could be by lateral translation. The terrane of the two extremes.
coastal Middle America, Colombia, and
Ecuador resembles old oceanic crust, not con-
tinental (Case et al. 1971; Schmidt-Effing 1979).
ACKNOWLEDGMENTS:This research was supported by
Mesozoic batholiths and Precambrian and National Science Foundation grant ID071-04207-
Palaeozoic metasedimentary rocks line the Chi- A04. We appreciate technical assistance by HIG staff
lean shoreline for 3000 km south of the region on board R/V Kana Keoki and in the laboratories.
studied. We do not know where lateral slices Conversations with Donald Hussong, the late George
could have gone. Woollard, LaVerne Kulm, Johanna Resig, William
The past decade has been one of revolution- Woodard, Larry Wipperman, Roland von Huene,
ary interpretations of the origin of much of Leo Ocola, Hector Ayon, Fernando Feraris, O.
western Canada, the U.S. Pacific North-west, Zevallos, and others have helped to shape the evolu-
and Alaska, as a collage of large and small tion of our thoughts about the South American
margin. So also have the results of the 1977-1979
blocks of 'exotic terranes' that have been swept Glomar Challenger drilling of Pacific active margins.
against the North American continent by plate Hawaii Institute of Geophysics Contribution No.
convergence (e.g. Davis et al. 1978). The Andes 1091.
 Forearc basins of Peru, Ecuador and Chile 187

References
AUDEBAUD, E., CAPDEVILA, R., DALMYRAC, B., MOBERLY, R. 1977. Structural evidence for
DEBELMAS, J., LAUBACHER, G., LEFEURE, C., the evolution of fore-arc basins off South Amer-
MAROCCO, R., MARTINEZ, C. MATTAUER, M., ica. Can. J. Earth Sci. 14, 102-16.
MEGARD, F., PAREDES, J. d~ TOMASI, P. 1973. Les CROWELL,J. C. 1974. Origin of late Cenozoic basins in
traits geologiques essentiels des Andes Centrales southern California. In: DICKINSON, W. R. (ed.).
(Perou-Bolivie). Rev. Geog. phys. Geol. dyna- Tectonics and Sedimentation. Spec. Publ. Soc.
mique, 15, 73-114. econ. Paleontol. Mineral. Tulsa, 22, 190-204.
BALDRY, R. A. 1938. Slip-planes and breccia zones in DAVIS, G. A., MONGER, J. W. H. & BURCHF1EL,B. C.
the Tertiary rocks of Peru. Q.J. geol. Soc. Lon- 1978. Mesozoic construction of the Cordilleran
don, 94, 347-58. 'collage' central British Columbia to central Cali-
BANDY, O. L. & RODOLFO, K. 1964. Distribution of fornia. In: HOWELL, D. G. & McDOUGALL, K.
foraminifera and sediments: Peru-Chile Trench A. (eds). Mesozoic Paleogeography of the West-
area. Deep Sea Res. l l , 817-37. ern United States. Pacif. Sec., Soc. econ. Paleon-
BENEST, H. 1899. Submarine gullies, river outlets and tol. Mineral. Tulsa, 1-32.
fresh water escapes beneath the sea level. Geog. DRAKE, C. L. (ed.) 1973. U.S. Program for the
J. 14, 394-413. Geodynamics Project: Scope and Objectives.
BELLIDO, E. 1969. Sinopsis de la geologia del Peru. Natn. Acad. Sci., Washington.
BoL Serv. Geol. Min. Lima, 22. ELSASSER, W. M. 1971. Sea-floor spreading as ther-
BURCHFIEL, B. C. & DAVIS, G. A. 1976. Compression mal convection. J. geophys. Res. 76, 1101-12.
and crustal shortening in Andean-type orogene- FARRAR, E., CLARK, A. H., HAYNES, S. J., QUIRT, G.
sis. Nature. London, 260, 693-4. S., CONN, H. & ZENTILLI, M. 1970. K:Ar evi-
BURNETr, W. C. 1977. Geochemistry and origin of dence for the post-Paleozoic migration of grani-
phosphorite deposits from off Peru and Chile. tic intrusion foci in the Andes of northern Chile.
Bull. geol. Soc. Am. 88, 813-23. Earth planet. Sci. Lett. 10, 60-6.
-- , VEEH, H. H. & SOUTAR,A. 1980. U-series, FAUCHER, B. & SAVOYAT,E. 1973. Equisse geologique
oceanographic, and sedimentary evidence in sup- des Andes de I'equateur. Rev. Geog. phys. Geol.
port of recent formation of phosphate nodules dynamique 15, 115-42.
off Peru. In: BENTOR, Y. K. (ed.). Marine Phos- FISCHER, A. G. 1956. Desarrollo geologico del
phorites. Spec. Publ. Soc. econ. Paleontol. noroeste Peruano durante el Mesozico. Bol. Soc.
Mineral. Tulsa, 29, 61-71. geol. Peru, 30, 117-90.
CALDWELL,J. G., TURCOTTE,D. L., HAXBY, W. F. & FISHER, R. L. & RAITr, R. W. 1962. Topography and
KARIG, D. E. 1977. Thin elastic plate analysis of structure of the Peru-Chile Trench. Deep Sea
outer rises (abs.). In: TALWANI, M. & PITMAN, Res. 9, 423-43.
W. C. III (eds). Island Arcs, Deep Sea Trenches, FRUTOS, V. & FERARIS, F. 1973. Mapa Tectonico de
and Back-arc Basins. Am. geophys. Union, M. Chile. Instit. de Invest. geol., Santiago.
Ewing Ser. l, 467. GANSSER, A. 1973. Facts and theories on the Andes.
CAMPBELL, C. J. 1974. Ecuadorian Andes. In: J. geol. Soc. London, 129, 93-131.
SPENCER, A. M. (ed.). Mesozoic-Cenozoic GILE'ITI, B. J. & DAY, H. W. 1968. Potassium-argon
Orogenic Belts. Spec. Publ. geol. Soc. London, ages of igneous rocks in Peru. Nature. London,
4, 725-32. 220, 570-2.
1975. Ecuador. In: FAIRBRIDGE, R. W. (ed.). GOOSENS, P. J. 1973. Characteristicas estructurales de
The Encyclopedia of World Regional Geology, la parte noroeste del margin continental de
Part 1, Western Hemisphere, 261-270. Dowden, America del Sur relaconadas con la tectonica de
Hutchinson & Ross, Stroudsburg, U.S.A. placas (abs.). Int. Union Geod. Geophys. Conf.
CASE, J. E., DURAN, L. G., LOPEZ, A. R. & MOORE, on Geodynamics, Lima.
W. R. 1971. Tectonic investigations in western HANDSCHUMACHER, D. W. 1976. Post-Eocene plate
Colombia and eastern Panama. Bull. geol. Soc. tectonics of the Eastern Pacific. In: SUTTON, G.
Am. 83, 2685-712. H., MANGHNAN1, M. H. & MOBERLY, R. (eds).
COBBING, E. & PITCHER, W. 1972. Plate tectonics and The Geophysics of the Pacific Ocean Basin and
the Peruvian Andes. Nature (phys. Sci) 240, its Margin [Woollard Volume]. Monogr. Am.
51-3. geophys. Union, 19, 177-202.
Cosslo, A. & JAEN, H. 1967. Geologia de los Cuad- HAYES, D. E. 1966. A geophysical investigation of the
rangulos de Puemape, Chocope, Otuzco, Trufil- Peru-Chile Trench. Mar. Geol. 4, 309-51.
lo, Salaverry y Santa. Bol. Serv. Geol. Miner. HENDERSON, W. G. & EVANS, C. D. R. 1980.
Peru, no. 17. Ecuadorian subduction system: discussion. Bull.
COULaOURN, W. T. 1977. Tectonics and sediments of Am. Assoc. Petrol. Geol. 64, 280-2.
the Peru-Chile Trench and continental margin at HERRON, E. M. 1972. Sea-floor spreading and the
the Arica Bight. Thesis, PhD, Univ. Hawaii, Cenozoic history of the east-central Pacific. Bull.
Honolulu (unpubl.). geol. Soc. Am. 83, 1671-92.
- - 1980. Relationship between the distribution of HOSMER, H. L. 1959. Geology and structural develop-
foraminifera and geologic structures of the Arica ment of the Andean system of Peru. Thesis, PhD,
Bight, South America. J. Paleontol. 54, 696-718. Univ. Michigan, A n n Arbor (unpubl.).
188 R . M o b e r l y , G. L . S h e p h e r d & W. T. C o u l b o u r n

HUSSONG, D. M., EDWARDS, P. B., JOHNSON, S. H., MORRIS, R. C. & ALEMAN, A. 1975. Sedimentation
CAMPBELL,J. F. & SUTTON, G. H. 1976. Crustal and tectonics of middle Cretaceous Copa Som-
structure of the Peru-Chile Trench: 8°-12°S lati- hero Formation in northwest Peru. Bol. Soc.
tude. In: SuTrON, G. H., MANGHNANI, M. H. & geol. Peru, 48, 49-64.
MOBERLY, R. (eds). The Geophysics of the Paci- PAREDES, J. 1966. Estratigrafia del Paleozoico del
fic Ocean Basin and its Margin [Woollard noroeste Peruano. Comision Carta Geologica
Volume]. Monogr. Am. geophys. Union, 19, Nacional del Peru (unpubl. report).
71-86. & MEGARD, F. 1972. Carte Structurale schemati-
JENKS, W. F. (ed.) 1956. Handbook of South Amer- que des Andes de Perou (privately distributed).
ican geology. Mem. geol. Soc. Am. 65, 378 pp. PAREDES, M. P. 1958. Terciario de la Brea y Patinas y
KANAMORI, H. 1971. Great earthquakes at island arcs area de Lobitos. Thesis, PhD, Univ. Sciences et
and the lithosphere. Tectonophysics, 12, 187-98. Techniques de Languedoc, Montpellier (un-
KARIG, D. E. 1974. Tectonic erosion at trenches. publ.).
Earth planet. Sci. Lett. 21, 209-12. PAZ, M. H. 1974. Estudio estratigrafico del subsuelo
KELLEHER, J. & MCCANN, W. 1976. Buoyant zones, de los yacimentos Zapatal-La Tuna-Coyonitas-
great earthquakes, and unstable boundaries of Ronchudo Petroles del Peru, operaciones
subduction. J. geophys. Res. 81, 4885-96. noroeste, Talara (abs.). IlI Cong. Peru de Geol.,
KULM, L. D., RESIG, J. M., MOORE, T. C., JR. & Soc. geol. del Peru.
ROSATO, V. J. 1974. Transfer of Nazca Ridge PITCHER, W. S. 1974. The Mesozoic and Cenozoic
pelagic sediments to the Peru continental mar- batholiths of Peru. Pacif. Geol. 8, 51-62.
gin. Bull. geol. Soc. Am. 85, 769-80. POTrER, P. E. 1978. Significance and origin of big
, THORNBURG,T. M. & SCHRADER, H. J. rivers. J. Geol. 86, 13-33.
1981. Cenozoic tectonics and stratigraphy of a RODRIGUEZ, R. E., CABRI~, R. & MERCADO, A. 1976.
fast-convergence zone--Peru forearc region (this Geometry of the Nazca Plate and its dynamic
volume). implications. In: Surror~, G. H., MANGHNANI,
, SCHWELLER, W. & MASIAS, A. 1977. A pre- M. H. & MOBERLY, R. (eds). The Geophysics of
liminary analysis of the subduction process along the Pacific Ocean Basin and its Margin [Woollard
the Andean continental margin, 6° to 45°S. In: Volume]. Monogr. Am. geophys. Union, 19,
TALWANI, M. & PITMAN, W. C. III (eds). Island 87-103.
Arcs, Deep Sea Trenches, and Back-arc Basins. ROSATO, V. J., KULM, L. D. & DERKS, P. S. 1975.
Am. geophys. Union, M. Ewing Set. l, 285- Surface sediments of the Nazca Plate. Pacifi Sci.
301. 29, 117-30.
LONSDALE, P. 1978. Ecuadorian subduction system. RUEGG, W. 1967. El margen suroriental la cuenca
Bull. Am. Assoc. Petrol. Geol. 62, 2454-77. Para-Andina de Sechura en el noroeste del Peru.
MALFAIT, B. I. & DINKELMAN,M. G. 1972. Caribbean Bol. Soc. geol. Peru, 40, 99-121.
tectonic and igneous activity and the evolution of SCHMALZ, R. F. 1958. A technique for quantitative
the Caribbean plate. Bull. geol. Soc. Am. 83, model analysis by X-ray diffraction and its ap-
251-71. plication to modern sediments of the Peru-Chile
MAMMERICKX, J., HERRON, E. & DORMAN, L. 1980. Trench. Thesis, PhD, Univ. Harvard, Cam-
Evidence for two fossil spreading ridges in the bridge (unpubl.).
southeastern Pacific. Bull. geol. Soc. Am. 91, SCHMIDT-EFFING, R. 1979. Alter und Genese des
263-72. Nicoya-Komplexes, einer ozeanischen Palaok-
MARCHANT, S. 1961. A photogeological analysis of ruste (Oberjura bis Eozan) in sudlichen Zentra-
the western Guayas Province, Ecuador. Q. J. lamerika. Geol. Rdsch. 68, 457-94.
geol. Soc. London, 117, 215-32. SHEPHERD, G. L. 1979. Shallow crustal structure and
MARTINEZ, M. 1970. Geologica de basamento marine geology of a convergence zone, northwest
Paleozoic en las Montanas de Amotope y posible Peru and southwest Ecuador. Thesis, PhD, Univ.
origen del petroleo en rocas paleozoicas del Hawaii, Honolulu (unpubl.).
noroeste del Peru. Primir Congreso Latinoamer. - - & MOBERLY,R. 1975. Southern extension of the
Geol. 2, 105-38. Dolores-Guayaquil megashear across the con-
MASIAS, J. A. 1975. Morphology, shallow structure, tinental margin of northwest Peru and the Gulf
and evolution of the Peruvian continental margin, of Guayaquil (abs.). Trans. Am. geophys.
6 ° to 18°S. Thesis, MS, Oregon State Univ., Union, 56, 442.
Corvallis (unpubl.). & -- 1981. Coastal structure of the con-
MOBERLY, R. 1972. Origin of lithosphere behind tinental margin, northwest Peru and southwest
island arcs, with reference to the western Pacific. Ecuador. Unpublished manuscript accepted for:
In" SHAGAM,R. (ed.). Studies in Earth and Space KULM, L. D. et al. (eds). Nazca Plate, Monogr.
Sciences [Hess Volume]. Mem. geol. Soc. Am. geol. Soc. Am.
132, 35-55. TgAvlS, R. B. 1953. La Brea-Parinas oil field, north-
MOORE, G. F., BILLMAN,H. G., HEHANUSSA,P. E. & west Peru. Bull. Am. Assoc. Petrol. Geol. 37,
KARm, D. E. 1980. Sedimentology and 2093-118.
paleobathymetry of Neogene trench-slope de- WATrS, A. B., TALWANI, M. & COCHRAN,J. R. 1976.
posits, Nias Island, Indonesia. J. Geol. 88, 161- Gravity field of the north-west Pacific Ocean at
80. its margin. In: SuTroN, G. H., MANGHNANI,M.
 Forearc basins of Peru, Ecuador and Chile 189
H. & MOBERLY,R. (eds). The Geophysics of the WEEKS, L. G. 1947. Paleogeography of South Amer-
Pacific Ocean Basin and its Margin [Woollard ica. Bull. Am. Assoc. Petrol. Geol. 31, 1194-241.
Volume]. Monogr. Am. geophys. Union, 19, ZEN, E. 1959. Mineralogy and petrology of marine
17-34. bottom sediment samples off the coast of Peru
and Chile. J. sediment. Petrol. 29, 513-39.

R. MOBERLY, Hawaii Institute of Geophysics, University of Hawaii, Honolulu,

Hawaii 96822, U.S.A.
G. L. SHEPHERD,Cities Service East Asia, Inc., 606 Cathay Bldg., Mount Sophia,
Singapore 9.
W. T. COtJLBOURN, Geologic Research Division, Scripps Institution of Ocean-
ography, La Jolla, California 92093, U.S.A.
The geology of the western part of the Borb6n Basin, North-west
Ecuador

C. D. R. Evans & J. E. Whittaker

SUMMARY: The western part of the Borb6n Basin, North-west Ecuador, contains some
5 km of predominantly thin-bedded Tertiary mudstones with rare shallow-water sand-
stones, resting on a basement of Cretaceous basalts. The basin is floored by thin,
discontinuous, turbiditic carbonates of middle Eocene age, which pass up into a 3 km
thick sequence of mudstones, cherty at the base, extending into the mid-Miocene.
Shallow-water sandstones accumulated during the late Miocene and early Pliocene and
were succeeded by deeper water mudstones, which in the Pliocene, are over 1 km thick
and contain resedimented sandstone units. The strata are folded into NE-SW-trending
structures, parallel to the mid-slope basement high centred on the Rio Verde area.
Stratigraphic evidence suggests that much of this folding, and the emergence of the high,
post-dates the late Pliocene (c.2.9 Ma).
Since the middle Eocene the area has occupied the arc-trench gap between the
Ecuador-Colombia oceanic trench, and the volcanic arc of the Andes; it now lies on the
oceanward side of the mid-slope basement high.

Along the west coast of Ecuador and Colombia, Stratigraphy

between the Andes and the Pacific Ocean, lies a
belt of lowlands varying in width from 60- A 5 km-thick section, extending from Cre-
200 km (Fig. lc). This is the site of the Bolivar taceous to upper Pliocene, is exposed in the
'geosyncline' (Nygren 1950), a thick accumula- western part of the Borb6n Basin. The older
tion of Tertiary sediments resting on a Meso- rocks are exposed only in the core of the Rio
zoic, largely basaltic basement. Nygren (1950) Verde anticline; the younger Miocene and
recognized within the 'geosyncline' a number of Pliocene strata crop out over most of the area
basins or 'deeps', three of which, the South with the youngest strata forming a belt along
Guayaquil (or Progreso), the Manta (or Man- the coast. The stratigraphic nomenclature fol-
abi) and the Borb6n, occur in the coastal lows Faucher et al. (1970); however, changes
lowlands of Ecuador (the Costa). This paper are introduced in the upper Miocene and
deals with the geology of the western part of the Pliocene units, mainly after Bristow (1976b).
Borb6n Basin (Fig. 2). The units were dated (J.E.W.) using planktonic
The International Ecuadorian Petroleum foraminifera recovered from over 150 samples
Company (I.E.P.C.) drilled three wells in the collected across the area.
Borb6n Basin, including Camarones-1 in the
mapped area, during the 1940s. These wells, Pifi6n Formation
and unpublished oil company reports provided
the basis for the regional stratigraphy (Stain- Thirty metres of green, basaltic breccia of the
forth 1948, 1968; Canfield 1966). Between 1966 Pifi6n Formation are exposed in the core of the
and 1972 the Institut Fran~ais du P6trole Rio Verde anticline, near Businga (Cuadro 82.
(I.F.P.) examined in detail some sections in the Faucher et al. 1970). Camarones-1 well
basin and data from their Rio Verde traverse (I.E.P.C. 1944) terminated in 248 m of grey,
(Faucher et al. 1970) are included in Fig. l(a). fine-grained, basaltic agglomerate, pumiceous
The area consists of undulating ground rising tufts and lavas which were assigned to the
up to 400 m above sea-level: much of the formation on lithological grounds.
northern coastal strip and valleys are covered Outside the mapped area, the nearest major
by grassland but primary forest persists inland. outcrops of the formation are 150 km SW of
Apart from the coast, new road cuttings and Esmeraldas in the Montafias de Jama (Fig. 2)
along the deeper valleys, exposures are few and where they form a complex, at least 2000 m
deeply weathered. The area was mapped in thick, of massive, fine-grained gabbros with
1976-77 as part of a collaborative project be- brecciated, altered, pillow basalts near the top
tween the Institute of Geological Sciences, of the section (Evans & ArgueUo 1977).
London, and the Direcci6n General de Geo- Feininger (1977) states that ' . . . the chemi-
logia y Minas, Quito. cal composition, lithology and structure of the

191
 T l,,,,,,a
T
IKil~rnet res
lc . . . . 5 10
Alluvium RECEN T i J L J J i J

Estero Platano == Pacific Ocean

1200m Memb er .I°N
uoper Onzole oLIOCENE
I F. . . . . ion
110"250m~ SSa Member Rfovq

30~ ~Om
} lower Onzole
Formation
1 Angostura
1
4.( 2,50m J Formation
:w , o' 2.7
~..
"Od
o

Viche Formation
Bouguet anomaly map of nortfl west Ecuador Punta
After Femmger 119771Contour =nterva110 miIhgals Gorda
Areas of pos~|ive Souguer s o g a l s [ ~ " 1
Gravity low ~ Grawtv h~gn
basalt (ba) .

Camarones- ! Well - O I / • Borehole ~U

l Pambil Formation
Elevallon 45m 91 " - - + Horizontal strata
u. Onzole Fro. to 212m
OLIGOCENE S~a Member to 412m zo× DiD and strike

ZaoaOnzol
ol"l e-PambilFm.to 1312mFmto 1lOOm 5: - Inferred contact
-- Exposed contact
PitOn Fro. to 1560rn = __
-~ Inferred fault
~= Svnclinal axis
Zaoallo Formation =~ Anticlinal axis
50m [~i~Eo
J Punta Ostiones M. 5_ _ . . . . . ",_- _~y
24~..n + r~a~=Cp ] Pit, On Formation
~= Punta Galera

~'-: :+M;;::i:i-:-.-:-:~l:_i'i":_:._

-~ ===========================================----:--::c:~~ .~- ~ -: - --':'~-*.z~--- -

~:!:!:Zi:!:! !:!-5:!! ~:!:!!:5: ~::~

~. ':: :::::-~::: -5: :-:-5::::::::::::z ::-:::::~: : -:-:- -:-:- :-z . . . . .
:* ,: ::::::::::::::::::::::::::::::::::: ;-:::::.:-i ~-

d odw 79 ° s~w 79' 40' W 79"3o'w ~'2o'~

L

FIG. 1.(a) Geological map of the western part of the Borb6n Basin.
(b) Gravity map of the Borb6n Basin, after Feininger (1977).
(c) Map showing the location of the mapped area in South America.
 The geology of the Borbon Basin, Ecuador 193
' '
81°W
bed area

05

Relief
0- 0m 1 i
400m-2000m
20(X)rn-4000m
over 4000m
Quaternary •
volcanoes
Isobaths in kilometres; after
)~ ;~ii! K=lometres 10o
PROGRESO Q.
Lonsdale and ,Kli,tgordf 1978. "NBASIN i ........... ..,... .......... " " ' 'z . . . . .

FIG. 2. Physiography of western Ecuador and the adjacent Pacific Ocean.

formation all point to its oceanic orgin' and occurs as irregular banks, nodules and veins of
suggests that it is ' . . . a remnant of ocean floor chert, and as a matrix in pseudo-breccias (A. F.
of Cretaceous age'. However, Henderson Wilkinson, pers. comm.). Rich faunas of plan-
(1979) believes that the formation represents tonic foraminifera of mid-Eocene age are found
the tholeiitic, primitive, oceanward part of a with radiolaria in the porcellanites, and cal-
volcanic arc more fully represented to the east careous algae and bryozoans occur in the
in the Western Cordillera of the Andes. biomicrites. Near Businga, the member consists
In southern Ecuador, the Pifi6n Formation is of 40 m of bioclastic limestone and it is absent
overlain by the Senonian-Maastrichtian, Cayo from the Camarones-1 well. The unsorted na-
Formation (Bristow 1976a) which, over part of ture of the biomicrites and the commonly worn
the northern Costa, has been removed by pre- condition of the fauna and flora, suggest that
middle Eocene erosion (Evans & Arguello the limestones were originally deposited in shal-
1977). The oldest radiometric date for the low water at the edge of the basin, and redepo-
formation is 110+10 Ma, Aptian (Goosens & sited by slumping into deeper water where the
Rose 1973). porcellanites accumulated. The member passes
transitionally up into the siliceous mudstones
Zapallo Formation which comprise the bulk of the Zapallo Forma-
tion.
Across much of the Costa the Zapallo The formation, which is well exposed in the
Formation, or its equivalent, has at its base a river north of Businga and in the road cuttings
thin discontinuous limestone unit, named in this east of Mutiles, is at least 400 m thick. It
area, the Punta Ostiones Member. At Punta consists of green-grey, hard, bioturbated, in
Ostiones it consists of about 50 m of brown- part silicified mudstones with thin sandstone
grey, siliceous biomicrites and porcellanites, in and tuffitic beds. Dykes, lenses and rare beds of
beds 5-50 cm thick, interbedded with thinner, dark brown chert are found, often fractured
green, calcareous sandstone and marl. Silica and traversed by clear quartz veins. In the
194 C. D. R. E v a n s & J. E. Whittaker

Camerones-1 well the formation, which is the overlying mudstones of the lower Onzole
212 m thick, contains in the lower part thin Formation. The poorly sorted sandstones are
cherty limestones (I.E.P.C. 1944). lithic-feldspathic greywackes and contain abun-
A re-interpretation of planktonic forami- dant, angular quartz, fresh, zoned plagioclase,
nifera from the formation, listed in Stainforth red-brown biotite and green hornblende grains.
(1948) and Bristow & Hoffstetter (1977), sug- The basal samples of the overlying formation
gests a mid-late Eocene age, and deposition in contain a late Miocene fauna and the age of the
bathyal depths. Angostura Formation is probably mid- to late
Miocene with a possible depositional hiatus
Pambil Formation indicated for much of the mid-Miocene.

In the Rio Verde the formation is 1600 m Onzole Formation

thick (Faucher et al. 1970) and consists of West of the Rio Esmeraldas, the Onzole
green-grey, bioturbated, silty mudstones with Formation is divided into a lower and an upper
local flaggy bedded, fine-grained sandstones. (= Punta Gorda Formation auct) unit sepa-
Near Mutiles, two basalt domes break the rated by sandstones of the Sfia Member. These
sequence. They are at least 50 m thick and over sandstones thin rapidly eastwards and are ab-
400 m long, with an outer crust, up to 8 m sent east of the river where the formation can
thick, of brecciated basalt traversed by a net- only be divided on faunal grounds. A second
work of zeolite veins. The domes are concor- sandstone sequence (Estero Platano Member)
dant and the absence of vesiculation suggests occurs towards the top of the formation in the
emplacement beneath a considerable depth of west of the area.
water. The lower Onzole Formation is between 300-
The Pambil Formation is entirely Oligocene 500 m thick and composed of uniform, silty
in age. Its lower part is developed in a primarily mudstones with thin, poorly graded, fine ribs
radiolarian facies; the overlying beds contain a and rare beds of sandstone. The late Miocene
rich foraminiferal fauna indicative of a late fauna found in the mudstones indicates, in the
Oligocene age, and deposition in outer shelf to lower part, deposition in shallow water but
bathyal depths. thereafter a gradually deepening environment.
At S6a the member consists of about 250 m
Viche Formation of well bedded greywacke sandstones which
thin eastwards to about 10 m in the Rio Esmer-
The Viche Formation, which has a thickness aldas. The orange to yellow-grey, bioturbated
in excess of 1000 m (Faucher et al. 1970), sandstones are coarse- to fine-grained, with
consists of a uniform sequence of green-grey to pebbly lenses, powdery siltstone intercalations
brown, well-bedded, blocky, bioturbated mud- and locally large, rounded, calcareous concre-
stone with thin ribs of sandstone, locally rich in tions. Lenses of worn and broken molluscs and
molluscs and fish scales. Palaeontological evi- lignite-rich bedding planes are common. The
dence indicates that the formation extends from lower contact varies from an undulating uncon-
the early to lower mid-Miocene. The absence of formity to a planar disconformity and the con-
lowermost Miocene sediments may be attri- tact with the overlying mudstones is gradation-
buted to incomplete sampling or a non- al. The member contains shallow water, benthic
sequence at this stratigraphic level. The benthic foraminifera and the rich, planktonic forami-
foraminifera suggest deposition in bathyal nifera in the immediately overlying mudstones
depths and give no suggestion of shallowing place the member within the early Pliocene.
towards the top of the unit. The upper Onzole Formation consists of
about 1200 m of mudstones with thin-bedded
Daule Group turbidites and rare, thicker, massive sand-
stones and conglomerates. In the west, the
Angostura Formation formation includes about 300 m of re-
North of Carlos Concha and east of the mouth sedimented sandstones and mudstones (the
of the Rio Verde, the Angostura Formation Estero Platano Member) which thin rapidly
rests with a sharp, conformable contact on the eastward and are not defined east of Stia. The
underlying Viche mudstones. The formation is formation is well-exposed along the coast where
between 40-250 m thick and consists of fine- to it forms cliffs of dark, green-grey, well bedded,
coarse-grained, rusty-orange sandstones with faintly graded mudstones, often intensely
lenses of bioclastic debris and conglomerates bioturbated. Discontinuous partings and thin
near the base, and a gradational contact up into beds of rusty, fine-grained sandstone and sharp-
 The geology of the Borbon Basin, Ecuador 195

ly defined beds of white tuff occur on a deci- into the submarine canyon immediately off-
metre scale. shore at present (Fig. 2).
In contrast to the lower part of the formation,
the microfaunas in the upper Onzole Formation Quaternary deposits
are very rich, well-preserved and wholly of
Pliocene age; the benthic foraminifera suggest Parts of the coastal area, at an elevation of up
deposition in bathyal depths. The minimum age to 60 m above sea-level, are covered by up to
for the formation is fixed at early Zone 21, 4 m of grey, fine sand with gravel and bioclastic
lower late Pliocene (c.2.9 Ma). lenses. These are typical of the Tablazo Forma-
Interbedded with the thin-bedded turbidites, tion (Bosworth 1922) a raised beach deposit of
especially in the east, are dark brown, medium- Pleistocene or younger age common along the
grained sandstones up to 1 m thick with uncon- coast of Ecuador and northern Peru.
formable bases and convoluted, internal Isolated patches of river gravel, stranded up
laminations. They contain a rich fauna of shal- to 25 m above the present river level, are found
low water molluscs and have been transported along the banks of the Rio Esmeraldas along
by turbidity flows into deeper water. Lenses of with a number of low river terraces (Fig. la).
pebbly sandstone and conglomerates up to Stratigraphical comparison of the western and
50 m thick with step-like, erosive, lower con- eastern parts of the Borb6n Basin
tacts occur near Esmeraldas. Locally the basal
1-3 m contains rounded mudstone boulders up Stainforth (1968), using data largely from the
to 1 m in diameter set in a pebbly mud matrix eastern part of the basin, identified an axial and
which passes up into an irregular sequence of a peripheral stratigraphic sequence and the
sands and pebbles with trains of mudstone study area is typical of the former. The
blocks. These coarse redeposited sediments peripheral sequence, developed along the east-
have flowed into the deep water basin, via a ern margin of the basin, near the Andean
submarine canyon, and are preserved in a series foothills, is distinguished by having a series of
of channels cut into the thin-bedded turbidites shallow-water formations interdigitated with
of the basin floor. the deep-water mudstones. The oldest three
Matrix-supported mudstone breccias up to shallow-water units, the Santiago, Playa Rica,
5 m thick are found in the formation east of and the 'lower Angostura' are not recorded in
Punta Galera. They contain rounded to angu- the western part of the basin. Along the
lar, commonly sheared blocks of local and Andean foothills the Santiago is developed
exotic mudstone up to 1 m across set in a above the basal Punta Ostiones Member; the
muddy matrix, locally rich in sandy lenses. Playa Rica separates the Pambil and Zapallo
These breccias are the products of slumps which Formations and the 'lower Angostura' rests
have slid into the basin floor from the sub- unconformably on the Pambil Formation. The
marine canyon walls. 'upper Angostura' sandstones of Stainforth
At the type locality, on the road between (1968), equivalent to the Angostura Formation
Estero Platano and Quinque (Fig. la), the in this paper, are a sequence ' . . . moving slow-
Estero Platano Member consists of about 300 m ly basinwards over the Viche Shales' (Stainforth
of rusty-brown, very regularly bedded, fine- to 1968); the base of the formation in the east is
medium-grained, well-sorted sandstones sepa- dated mid-Miocene but in the west it is mid- to
rated by thin mudstones. In the lowest expo- late Miocene. In the eastern part of the basin
sures the individual sandstone beds are about a the overlying Onzole Formation is of shallow
metre thick and form over 75% of the succes- marine character while in the west the analo-
sion. The thickness of the sandstone beds and gous lower Onzole Formation was deposited in
the sandstone/mudstone ratio decreases gra- a deepening basin.
dually up the unit which, at the top, passes The early Pliocene event, which led to the
transitionally into the normal, thin-bedded uplift and the deposition of the Sda Member, is
turbidite sequence. recognized across most of the Costa. Much of
The absence of coarse detritus and the well- the eastern Borb6n, the Manabi and Progreso
sorted nature of the sandstones suggest a source Basins remained positive areas after the uplift.
distinct from that of the conglomeratic lenses However, subsidence recommenced in the
which are found in the formation. The sands of western part of the Borb6n Basin where the
the Estero Platano Member are similar to the upper Onzole Formation accumulated in outer
beach deposits along the north coast: the con- shelf to bathyal depths. The post-late Pliocene
glomerates are similar to the material on the uplift of the western part of the basin was
floor of the Rio Esmeraldas which debouches associated with the emergence of the Rio Verde
196 C. D. R. Evans & J. E. Whittaker
anticlinal area and the development, to the subducting under the South American Plate
east, of a new basin where 500 m of sands and (Lonsdale & Klitgord 1978)--and the andesite
gravel (Cachabi Formation) accumulated dur- volcanoes of the Andes (Fig. 2). The gap is
ing the Pleistocene (Bristow & Hoffstetter about 240 km wide in Ecuador, compared with
1977). This situation persists with thick recent a characteristic value of 250 + 50 km for con-
alluvium accumulating in the eastern basin and tinental margin settings (Dickinson 1971).
the western part of the basin continuing to rise. Lonsdale (1978) divided the Costa into a
mid-slope basement high (the coastal range of
hills) with a forearc basin to the east (the
Structure Manabi Basin) and the trench-slope basin to the
west (the Progreso Basin). In the northern
The study area lies on the north-western margin
Costa (Fig. 2) this sub-division is applicable
of the gravity low which delimits the Borbon
only to the post-late Pliocene; the mid-slope
Basin (Fig. lb), but the youngest strata are
basement high had no major topographic ex-
exposed near the coast, well to the north of the
pression prior to this time.
area of thickest accumulation of Tertiary sedi-
The tectonic history of the northern Costa
ments, as defined by the gravity data. An
is divisible into pre- and post-mid-Eocene
ENE-WSW-trending, fault-controlled gravity phases with a further three-fold sub-division of
high, which separates the Borb6n Basin from
the latter into mid-Eocene-mid-Miocene, late
the Manabi Basin to the south, forms the Miocene-late Pliocene and post-late Pliocene
southern margin to the basin.
to Quaternary phases.
The strata are folded into a series of open,
NE-SW-trending structures with wavelengths
upwards of 2 km and sinuous axial traces which Pre-mid-Eocene
are parallel to the main structural feature of the
Feininger (1977) suggested that the basement
area, namely: the anticline which brings the of the Costa, the Pifi6n Formation, was a
older formations to crop in the Rfo Verde area.
remnant of ocean floor which ' . . . now crops
This feature, which extends the length of the
out owing to its having been isolated by a
Costa from the Cerros de Colonche-Chongon
westward jump of the continental border sub-
northwards to the Rio Verde area (Fig. 2),
duction zone in Early Tertiary time'. On gravity
obliquely crosses the axis of the Borb6n Basin
evidence (Fininger 1977), the boundary be-
as defined by the gravity low. Stratigraphic
tween the pre-early Tertiary continental and
evidence indicates that the feature, a mid-slope
oceanic crust lies approximately at the base of
basement high, formed a positive topographic
the western Andes, about 180 km east of the
feature in the Manabf and Borb6n Basins only
present axis of the Ecuador trench. However,
after the late Pliocene.
Henderson (1979) considered the Pifi6n a
The coastal sections (Fig. la) show that the
tholeiitic oceanward part of the arc and sug-
region is divisible into structurally simple areas
gested that the co-linearity of the Cretaceous to
(Esmeraldas to Punta Gorda) and discrete
Eocene and post-Eocene volcanic arc indicated
more complex areas (around Sfia). -The faults
that no change had occurred in the position of
mostly trend between 330 ° and 360 ° with some,
the oceanic trench in the early Tertiary.
east of Esmeraldas, parallel to the coastline. All
The event which resulted in uplift and ero-
are normal faults with no evidence of thrusting.
sion of the Costa and the establishment of the
In the Camarones-1 well slickensided frac-
present forearc regime is dated between late
tures are reported at a depth of 500 m and a
Palaeocene--the youngest date for basaltic
poorly developed cleavage, parallel to bedding,
activity in the Cayo Formation (Goosens &
below 1000 m. The Pliocene mudstones on the
Rose 1973)--and the early mid-Eocene--the
west coast, near Quinque, display a fine, slight-
age of the Punta Ostiones Member. Limestones
ly irregular pattern of dark, linear veins (de-
of the latter unit rest unconformably on the
watering structures?) developed at a steep
weathered top of the Cretaceous basement
angle to the bedding.
(I.E.P.C. 1944) with no evidence for a totally
tectonic contact between the two.
Tectonic history and palaeogeography
Mid-Eocene-mid-Miocene
The Pacific coastal lowlands of Ecuador and
Colombia occupy the 'arc-trench gap' (Dickin- During this interval the sequence was domin-
son 1971) between the offshore bathymetric ated by a thin-bedded turbidite facies, depos-,
trench--the site of the oceanic Nazca Plate ited in a forearc basin, whose main axis of
 The geology of the Borbon Basin, Ecuador 197

deposition trended ENE-WSW. The occur- Esmeraldas-Stia area, a gently dipping succes-
rence of cherts only in the older strata indicates sion over 1200 m thick is now exposed. The
an initially deep basin floor which became exposure of the Stla Member in this area repre-
shallower with time. In the late Eocene to sents uplift and erosion of at least 1500 m
Oligocene, sediment accumulation in the map- (water depth plus thickness of strata removed
ped area, which is on the north-western margin by erosion) since the deposition of the unit.
of the main deposition axis, was about Geomorphological evidence indicates that the
100 m/Ma, compared to about 60 m/Ma in main uplift of the Andes was contemporaneous
the present Panama Basin (Lonsdale 1978). As with the uplift of the coastal range of hills
the basin infilled, the shallow-water sequences (Campbell 1974). With the rise of the mid-slope
associated with the uplift of the basin floor, basement high the axes of deposition in the
extended further oceanward (westwards). Thus Costa moved eastwards into the Guayas Valley
the earliest shallow-water formations were res- (Fig. 2) and eastern Borb6n Basin where thick
tricted to the eastern, peripheral parts of the alluvium and shallow-water sequences accumu-
basin and the upper Miocene Angostura lated during the Quaternary (Bristow & Hoff-
Formation was the first to extend into the stetter 1977). This sequence of events is recog-
western part of the Borb6n Basin. The Ango- nized in the coastal lowlands northwards into
stura Formation was also the first shallow-water Panama, although parts of the forearc in Col-
sequence developed west of the Montafias de ombia remain submerged with the offshore
Jama in the Manabi Basin (Evans & Arguello islands (Gorgona Island), equivalent to the
1977). mid-slope basement high in Ecuador (Fig. lc).
The basaltic activity in the upper part of the
Pambil Formation was contemporaneous with
the breakup of the Farallon Plate into the Conclusions
Nazca and Cocos Plates 27 Ma ago (late Oli-
gocene) (Lonsdale & Klitgord 1978). The effect The Tertiary stratigraphy of the western part of
of the breakup of the plate on the local stra- the Borb6n Basin is typical of the forearc in the
tigraphy was minimal; the facies remained un- northern Costa and southern Colombia (Nyg-
changed although the rate of sedimentation ren 1950). The sequence can be summarized as
during the early Miocene was about double that follows: thin discontinuous limestones which
for the late Eocene-Oligocene. floor the basin pass up into thin-bedded turbi-
dites which alternate, in the upper part, with
Late Miocene--late Pliocene shallow-water sandstones. The sequence is
typical of that in an idealised residual forearc
During this interval, between 8 and 2.9 Ma, basin ' . . . where strata lie depositionally on
two shallow-water sequences transgressed oceanic or transitional crust trapped between
across the basin to its oceanward margin, inter- the arc massif and the subduction zone (Dick-
rupting the normal thin-bedded turbidite sequ- inson & Seely 1979). The Borb6n Basin is
ence. The start of the interval coincided with a interesting because of the alternation of shallow-
reduction in the rate of subduction (Lonsdale & and deep-water sedimentation. The fluctua-
Klitgord 1978) under the forearc and therefore tions in the depth of the basin floor are caused
changes in the stresses on the basement. by regional tectonic events, such as rates and
Sedimentation was more rapid than during direction of subduction which influence the
the earlier part of the basin's history with a rate tectonic stresses on the basement of the arc-
of 500 m/Ma in some areas. No facies varia- trench gap. The regional character of the events
tions in the Pliocene units can be attributed is indicated by the contemporaneous changes in
directly to the present mid-slope basement the Tertiary stratigraphy of the forearc from
high, though the very rapid accumulation of northern Peru to Venezuela (Stainforth 1968).
sediment, probably in localised fault-controlled Karig & Sharman (1975) propose a model of
basins, may have been complementary to the forearc accretion which includes the oceanward
initial rise which eventually resulted in the movement and vertical growth of the trench-
mid-slope basement high centred on the Rio slope break as the basin matures. However, in
Verde area. the Borb6n Basin the similarity of tectonic
styles and facies of the Eocene-Oligocene and
Post-late Pliocene--Quaternary late Pliocene strata implies no significant migra-
tion of the basin margin since the mid-Eocene.
The youngest Pliocene strata were deposited A similar situation occurs off Guatamala,
in outer shelf to bathyal depths and, in the where the position of the shelf-edge has re-
198 C. D. R. Evans & J. E. Whittaker

mained static since the end of the Cretaceous lier dominant trend. However, these basement
(Seely 1979). Growth of the trench slope break faults continue to influence the geology of the
as the basin infills appears to have been vertical area as is shown by the fault which forms the
in the Borb6n Basin. During the early part of southern boundary of the Borb6n Basin. There
the basin's history the trench slope break was is an abrupt change in the strike of the arc
obviously d e e p and the basin may have sloped, where it intercepts the A n d e a n Massif (Fig. 2)
almost without a break, from the shore into the and an offset in the alignment of the Quater-
oceanic trench. As the basin infilled and shal- nary volcanoes. Thus mega-structures in the
low water sequences edged oceanwards the basement, which cross the present arc-trench
trench slope break rose to become a pro- geometry at an oblique angle, may divide the
nounced topographic feature separating the arc and forearc into sectors with differing pat-
forearc basin from the lower trench slope into terns of sedimentation, uplift and igneous activ-
the oceanic trench. ity.
Gravity data indicate that the bulk of the
Tertiary sediments accumulated in a basin ACKNOWLEDGMENTS: The authors are grateful to the
whose shape was controlled primarily by faults personnel of the Direcci6n General de Geologia y
Minas in Quito for the support during the fieldwork
in the basement and not the presumed arc- especially Ing. L. Cevallos and St. C. Cardenas.
trench geometry. The balance of influence Thanks are also due to the rest of the staff, past and
changed during the Pliocene and the shape of present, of the British Mission in the D.G.G.M.
the later basins was related to the present especially Mr W. G. Henderson and Dr A. F. Wilkin-
arc-trench system which cut obliquely the ear- soN.

References
BOSWORTll, T. O. 1922. Geology of the Tertiary and FEININGER, T. 1977. Bouguer anomaly map of
Quaternary Periods in the Northwest Part of Peru Ecuador. Inst. Georg. Militar, Quito.
(with an account of the Palaeontology by H. GOOSENS, P. J. & ROSE, W. I. 1973. Chemical
Woods, J. W. Vaughan, J. A. Cushman, and composition and age determinations of tholeiitic
others). MacMillan, London, 434 pp. rocks in the Basic Igneous Complex, Ecuador.
BmsTow, C. R. 1976A. The age of the Cayo Forma- Bull. geol. Soc. Am. 84, 1043-52.
tion, Ecuador. Newslr. Stratig. 4, 169-73. HENDERSON, W. G. 1979. Cretaceous to Eocene
1976b. The Daule Group, Ecuador. Newslr. volcanic arc activity in the Andes of northern
Stratig. 5, 190-200. Ecuador. J. geol. Soc. London., 136, 367-78.
-- & HOFFSTETrER, R. 1977. Lexique Stratigraphi- I.E.P.C. 1944. Composite log, Camarones-1 Well,
que International, 4, Fasc. 5a 2nd ed. Centre International Ecuadorian Petroleum Company.
National de la Recherche Scientifique, Paris. Dir. Hidrocarburos, Quito (unpubl.).
CAMPBELL, C. J. 1974. Ecuadorian Andes. In: KARIG, D. E. & SHARMAN,G. F. 1975. Subduction
SPENCER, A. M. (ed). Mesozoic-Cenozoic and accretion in trenches. Bull. geol. Soc. Am.
Orogenic Belts, Data for Orogenic Studies. Spec. 86, 377-89.
Publ. geol. Soc. London, 4, 725-32. LONSDALE, P. 1978. Ecuadorian subduction system.
CANFIELD, R. W. 1966. Reporte Geol6gico de la Costa Bull. Am. Assoc. Petrol. Geol. 62, 2454-77.
Ecuatoriana. Ministerio de Industrias y Com- - - & KLITGORD, K. D. 1978. Structure and tectonic
ercio, Quito. history of the eastern Panama Basin. Bull. geol.
DICKINSON, W. R. 1971. Clastic sedimentary Soc. Am. 89, 981-99.
sequences deposited in shelf, slope and trough NYGREN, W. E. 1950. Bolivar geosyncline of north-
settings between magmatic areas and associated western South America. Bull, Am. Assoc. Pet-
trenches. Pacif. Geol. 3, 15-30. rol. Geol. 34, 1998-2006.
-- & SEELY, D. R. 1979. Structure and stratigra- SEELY, D. R. 1979. Evolution of structural highs
phy of forearc regions. Bull. Am. Ass. Petrol. bordering major forearc basins. In: WATKINS,J.
Geol. 63, 2-31. S., MONTADERT, L. & DICKERSON, P. W. (eds).
EVANS, C. D. R. & ARGUELLO, C. 1977. Geological Geological and Geophysical Investigations of
Map of Sheet 10 (Jama) 1:100,000. Dir. Gen. de Continental Margins. Mem. Am. Assoc. Petrol.
Geol. y Minas, Quito. Geol. 29, 245-60.
FAUCHER, B., JOYES, R., MAGNE, F., SIGAL, J., STAINEORTH, R. M. 1948. Applied micropalaeontol-
VERNET, R., GRANJA, V. (J.C.), GRANJA, B. ogy in coastal Ecuador, J. Paleontol. 22, 133-51.
(J.C.), CASTRO, R. & GUEVARA,G. 1970. Estudio 1968. Mid Tertiary diastrophism in Northern
general de la Cuenca de Esmeraldas. Instituto South America. In: SAUNDERS, J. B. (ed.).
Frances de Petrole. Servic. Nat. de Geol. Trans. Fourth Caribbean Geol. Congr. 159-77,
Miner, Quito (unpubl.). 1965. Port of Spain, Caribbean Printers, Arima.
C. D. R. EVANS,Institute of Geological Sciences, Ring Road, Halton, Leeds LS15
8TQ, England.
J. E. WHITTAKER, British Museum (Natural History), Cromwell Road, London
SW7 5BD, England.
 Ancient plate boundaries in the Bering Sea region

M. S. Marlow, A. K. Cooper, D. W. Scholl & H. McLean

SUMMARY: Plate tectonic models of the Bering Sea suggest that the abyssal Bering Sea
Basin is underlain by oceanic crust, a supposition supported by refraction and magnetic
data. The oceanic crust is thought to be a remnant of the Kula(?) plate that was isolated
within what is now the Bering Sea when the proto-Aleutian arc began to form between the
Alaska Peninsula and Kamchatka in late Mesozoic or earliest Tertiary times. Prior to the
formation of the Aleutian arc, the Kula(?) plate moved NW, directly underthrusting
eastern Siberia; the plate's eastern edge either obliquely underthrust or slid past the Bering
Sea margin along a transform boundary.
The Koryak Range in eastern Siberia is composed in part of mrlange units that include
Palaeozoic and Mesozoic allochthonous blocks juxtaposed within a matrix of Cretaceous
sedimentary rocks. Structural trends suggest that these blocks were accreted into the
Koryak area from the south along an ancient subduction zone formed by underthrusting of
the Kula(?) plate.
The base of the Bering Sea continental margin that extends from eastern Siberia to the
Alaska Peninsula--the so-called Beringian margin--is underlain by a thick (7-10 km)
sedimentary section along the base of the slope. Rocks dredged from the basement
exposed farther up the slope (1500-2000 m deep) include shallow-water Upper Jurassic
sandstone that is unconformably overlain by shallow-water Eocene to Miocene diato-
maceous mudstone. Fauna in the dredge samples indicate that the shelf edge has subsided
several kilometres since late Palaeogene time, perhaps in response to the cessation of
motion relative to the adjacent oceanic plate and subsequent sediment loading of the
oceanic plate.
Uplift of the former plate boundary exposed in the Koryak Range occurred principally
in late Cenozoic time, and collapse of the adjacent plate boundary, the Beringian margin,
began in earliest Tertiary time and has continued to the present. Both tectonic events
occurred after the site of active plate collision shifted south to near the present Aleutian
Trench. We are uncertain as to why these two ancient, yet adjacent former plate
boundaries should behave so differently, i.e. why one area was folded and uplifted while
the other was extensionally deformed and subsided, both apparently in response to the
cessation of convergent or strike-slip plate motion.

We examine two postulated Mesozoic plate Geological data from eastern Siberia
boundaries in the Bering Sea; a SE facing
convergent boundary through the Koryak A generalized cross-section that traverses the
Range in eastern Siberia, and a SW facing north-eastern Koryak Mountains (Fig. 2) is
obliquely underthrust or strike-slip boundary, adapted from McLean (1979a), who in turn
the Beringian margin, connecting Alaska and abstracted the section from the original work of
Siberia (Fig. 1). Aleksandrov et al. (1976). The eastern Koryak
A terrane of complexly deformed rocks that Range is underlain by structurally juxtaposed
include mrlange sequences and ultramafic blocks of different age and lithology. McLean
masses exposed in the Koryak Range of eastern (1979a) compares the complex rock fabric of
Siberia is thought to record Mesozoic to earliest the Koryak Range to the Franciscan assemb-
Tertiary convergence between the Kula(?) and lage in California as described by various work-
North American plates (Scholl et al. 1975). The ers (e.g. Bailey et aL 1964, p. 5; Hamilton 1969;
range was deformed and in part uplifted in Bailey & Blake 1969; Ernst 1970; Page 1970,
Cenozoic time. The adjacent Beringian margin, 1978) and to the McHugh and Uyak Complexes
striking at right angles to the Koryak margin, of south-western Alaska (Clark 1973; Connelly
has undergone substantial subsidence in Ceno- 1978; Tysdal & Case 1976; Tysdal et al. 1977).
zoic time. The Beringian margin is underlain Rock types common to all three areas include
by a broadly folded sequence of shelf and upper chert (red, grey, and green), siliceous shale,
slope sedimentary and volcanic (evidence is siltstone, fine-grained greywacke, tuff, spilite,
volcanic-rich detritus) units. These units are pillow basalt, blueschist, and ultramafic rocks.
Jurassic and Cretaceous in age and presumably The age of the greywacke in the Koryak Range
are not typical of subduction complexes. is estimated as either Triassic to Early Cre-

201
202 M.S. Marlow et al.
,o.,7,'E 162"66.

Chukotsk
8 Peninsula

ALASKA
Naverin basin
\ ~.C'

Istlnd
Island

BERING SEA
sr -

Aleutian basin

Pribilof Islands
s r I.~
1~4"E •- o--'d se"
le2`

1so- let
1~4.w

FIG. 1. Location of seismic-reflection profile A-A' across the Beringian margin shown in Figs 3 and 4.
Line B-B' shows location of the generalized structure cross-section shown in Fig. 2. Line C-C' shows
location of hypothetical section shown in Fig. 6. Encircled dots indicate sites where Jurassic (J) and
Cretaceous (K) sedimentary rocks were dredged from the margin (Marlow et al. 1979b; Vallier et al.
1980; Marlow & Cooper 1981). Albers equal area projection.
taceous (Valanginian) or Late Jurassic (Titho- day Bering Sea. Thrusting of the allochthonous
nian) to Early Cretaceous (Hauterivian) units apparently continued into early
(McLean 1979a). Palaeogene time (Aleksandrov et al. 1976;
Large slab-like rock bodies occur within the McLean 1979a).
Koryak Mountains. These slabs are separated Early Cretaceous magmatism in the Koryak
by northward-dipping thrust faults and include Mountains coincided with the accumulation of
olistostrome and m61ange sequences (Fig. 2). olistostromes to the south toward the Bering
At the base of many of the tectonic slabs and Sea, in front of rising tectonic slabs (Aleksan-
sheets are ultramafic masses consisting of drov et al. 1976). Compression and foreshorten-
pyroxenite, peridotite and serpentinite associ- ing of the Koryak Range apparently continued
ated with gabbro and diorite. These marie and into Palaeocene time as evidenced by isoclinally
ultramafic assemblages may be part of an folded rocks of Palaeocene age exposed on the
ophiolite suite (Aleksandrov et al. 1976). The SE side of the range (Aleksandrov et al.
assemblages are unconformably overlain by 1976; McLean 1979a). These younger deformed
either siliceous sedimentary rocks of Palaeozoic rocks were derived as reworked debris from
age or volcanic and terrigenous deposits of older, previously deformed units of the subduc-
Mesozoic age. Exotic blocks of serpentinite tion complex present in the Koryak Mountains.
m61anges and olistostromes ( s e n s u Silver &
Beutner 1980) include shallow-water Devonian
Carboniferous, and Lower and Upper Permian Seismic-reflection data from the
limestones. Aleksandrov et al. (1976) suggest Beringian margin
that all Palaeozoic deposits in the Koryak
Mountains are allochthonous units that by the A 1620 km long seismic-reflection profile, A-
end of the Mesozoic were tectonicaUy derived A' (Fig. 1), transects the north-western Bering-
from the south, in the vicinity of the present- Jan margin west of St Matthew Island. An
 B B'
NW SE
Y a g e l ' n a y a River Elgevayam River Snezhnaya River Ekonay Mr.

'000 i OOgr.e,e, "

V.E.=2.5:1
~,~°
Blocks of Palaeozoic chert and
volcanic rock. Gabbro, diorite and plagio-granite.
Upper Jurassic to Lower Cretaceous
volcanic and terrigenous deposits. Serpentinite melange.
e~
Upper Jurassic to Palaeogene volcanic-
terrigenous deposits. Faults, predominantly thrust faults.
Tectonic melange and olistostromes. r~

F16.2. Generalized structure cross-section through the north-eastern Koryak Range (after Aleksandrov et al. 1976, p. 294 and McLean 1979a, p.
1473). Line of section is located on Fig. 1.
 t~

4~

I~1~ Figure 4
A A'
"=--- S W N E ---="

ALEUTIAN BASIN NAVARIN BASIN

O[ __ ~ Sandy mudstone - E. . . . . . - ---" - ~ --- . . . . . . ~ " ~ - -
:~--~--~- - iD
"= t ~ ~-0
1Upper section: uurbioites, to Miocen~ " ..... ~--~ ~ : - " - - - - .--t
2 - [ /I m u d s t o n e a n d d i a t o m o o z e - ~-~-~- " - - ~ - - ~ - ~ ÷ -- _~- - ~ ~. ~ _ ~_ ~_~7~ ~ ±;-
J ~ Miocene and younger i-~ . . . ~'~- f~'~"~-"~-~: . . . . .
41 ~ .... ~ J .... Acoustic basement: ~_ )~" - ~ ~ _ + _ _
~_~ -- ~ - :-:7"~ . . . . ~ Cretaceous limestone _L..... -- ~ !: - 4 "
~ ~ - - - : ~ 3 5 : ~ - - ~ - and sandstone - - -~
6~--__ --~--~--~--J~- : ? ~Acoustically opaque o ~OKM "-Upper basin fill- -. . . . . . . ~ ,/ 6 12
:~ --~ ~ zone of folding(?) . , Cenozoic Poorly reflective -~T "~t
8~Pelagics(?)- ' . . . . . . KM
10 ~ Mesozoic(?) . '",,\
-- ~
" ~ LowerLate sectiOn:cretaceous(?)terrigenOUSand-
V E =2 5 1 I o w e r B a s i n Till -
M e s o z o l c t - .r ^) ".
' noer aa lr -D7a- ss m
ear(
12 k( ms)e c
e c Till
.... II 8
~-110
L: ~"
:~
SEC Top or oceanic ~ ,
crust (Layer 2)- ~'aleogene SEC
Mesozoic(?)

FIG. 3. Interpretative drawing of 24-channel seismic reflection profile A-A' across the northern Beringian margin. For location of profile see Fig.
1. Travel time, in seconds, is two-way time. Detailed segment noted is shown in Fig. 4.
 Plate b o u n d a r i e s in the B e r i n g S e a 205

interpretive drawing of profile A - A ' (Fig. 3) 'wipe-out' zones beneath short discontinuous
shows that the south-western end of the profile reflectors in the upper few hundred metres of
crosses the eastern edge of the Aleutian Basin the section. The acoustic 'wipe-out' zones are
in water more than 3000 m (4.0 s) deep, where thought to be produced by shallow accumula-
4-5 km of undeformed sedimentary strata over- tions of gas that mask the lateral continuity of
lie a distinct acoustic basement. Magnetic and deeper reflectors (Marlow & Cooper 1981).
refraction-velocity data indicate that the base- Dredge data indicate that the upper 3-4 km of
ment is oceanic crust of Mesozoic age (Kula(?) the beds in the basin are younger than early
plate: Cooper et al. 1976a, b). Immediately Eocene in age. Deeper strata are poorly reflec-
overlying the basement is a highly reflective tive, diverge in dip from the overlying strata,
section a few hundred metres thick. Other and near the base of the section may be Cre-
profiles (not shown) show that this layer taceous or older in age (Marlow et al. 1976;
blankets and in-fills the irregular surface of Marlow 1979).
igneous oceanic crust, suggesting that the infill-
ing layer is a pelagic deposit, possibly similar to
the cherty units that blanket and smooth the Dredge data
surface of Mesozoic igneous crust in the Pacific
(Ewing et al. 1968). Overlying the pelagic sec- Dredging in 1978 sampled rocks of the subshelf
tion are moderately to poorly reflective strata basement that crops out on the continental
that thin and dip gently to the SW away slope (Figs 1 & 3; Marlow et al. 1979a, b).
from the margin. On the basis of this geometry Rocks recovered near profile A - A ' include
and of close proximity to the margin, these muddy limestone deposited in shallow water
deposits are interpreted to be mainly Upper (based on fossil flora), siltstone, and mudstone
Cretaceous(?) and Palaeogene terrigenous de- of Late Cretaceous (Campanian and Maestrich-
bris. The upper section beneath the Aleutian tian) age. Farther south and west of the Pribilof
Basin along profile A - A ' may be correlative Islands, shallow-water Upper Jurassic sand-
with the section at DSDP Hole #190, where stone was dredged from the continental slope
Creager, Scholl et al. (1973) drilled through (based on the presence of Buchia rugosa, Mar-
more than 600 m of turbidites, mudstone, and low et al. 1979a, b; Marlow & Cooper 1981).
diatomaceous ooze of late Miocene and youn- All of the Mesozoic basement rocks recovered
ger age. Strata equivalent in age and lithology to date are generally highly indurated but are
presumably underlie the basin along the south- not significantly sheared or fractured.
western end of profile A - A ' . Additional rock samples from the continental
Below the lower part of the continental slope slope in the vicinity of profile A - A ' include
the rock sequence is acoustically characterized mudstone and sandy mudstone that yield di-
by scattered and discontinuous reflectors (Figs atoms and foraminifers of Eocene, Oligocene,
3 & 4). Oceanic basement was not resolved and Miocene age (Marlow et al. 1979a, b).
here. In contrast, a strong basal reflector, a These Tertiary samples were recovered from
sub-shelf basement, can be traced from the the acoustically layered section overlying rocks
NE seaward to the middle of the continental of the subshelf basement (Fig. 3). The fossils
slope where the basement crops out. Dredged recovered from both the subshelf basement and
rocks and the flatness of the subshelf basement the overlying rocks and the presence of wave-
indicate that beneath the slope the shelf base- cut terraces now 1-2 km deep attest that the
ment is a wave-base unconformity cut across outer-shelf and upper-slope region have sub-
broadly folded rock sequences of Jurassic and sided 1-2 km since Eocene time. The excep-
Cretaceous age. These dredge and seis- tionally thick sedimentary fill in the adjacent
mic-reflection data, discussed more fully below, Navarin basin implies that locally areas of the
attest that the subshelf basement beneath the shelf may have subsided 10-12 km since late
slope has subsided at least 1500-1700 m since Mesozoic time (Marlow 1979).
about Eocene time.
Basement beneath the shelf can be followed
to the NE below the thick sedimentary sec-
Discussion
tion filling Navarin basin (Fig. 3). Near the
north-eastern end of the profile strata in the A postulated plate model for the evolution of
basin are nearly 12 km (7 s) thick. Reflections the Bering Sea and the Beringian margin (Fig.
from the upper basin fill are strong, continuous, 5) proposed by Scholl et al. (1975) suggests that
and fiat. Apparent breaks in the continuity of the Beringian margin was initially the site of
these reflectors are associated with columnar oblique convergence or strike-slip (transform)
 1"-,9

~- SW NE"~
O-- ': . . .... . . . . . . . . . . . . . ..i ¸ ii -: !/ii, i ?i ~ ~i i~i ~ :::~!ii/~r ~ ~
~ii ~ ! 4~ :i ~¸ i!, i~¸!, '! . . . . :' , ~ ~:?'i! ~
!:~i~ i'i! ! ~!/~ i:~I~ ~ /' :" :~

~ :~i i!! ¸ i~ill ii :~ ii. i!~ ~i~i i~~

iiii~iiiiil~ :~ i~ ~

!~i!! i,! ;ii J~.i ~i ¸ : ', ' ~ ....

P
~
4

8-

10-
Sec 0 10 Km
[ , I

FIG. 4. Twenty-four channel seismic-reflection record used in constructing part of the interpretative drawing shown in Fig. 3.
 Plate b o u n d a r i e s in the B e r i n g S e a 207

motion between the Kula (?) and North Amer- (1975) and Cooper et al. (1976b) predicts that
ican plates. Scholl et al. (1975) further argue the Beringian margin could be underlain by a
that the formation of the Aleutian arc by ear- subduction complex of oceanic and slope de-
liest Tertiary time would have shifted the site of posits accreted to the North American plate by
plate interactions south to an ancestral Aleutian oblique underthrusting of the Kula(?) plate
Trench. The shift in the site of plate con- (Fig. 5). Even if the margin was formerly the
vergence would have tectonically isolated both site of substantial transform or strike-slip mo-
eastern Siberia (Koryak Mountains) and the tion between the two plates, the ancient margin
Beringian margin from plate-bounding interac- should nevertheless have been the site of de-
tions with the Kula(?) plate, and, at the same position of deep-water slope deposits (for ex-
time, trapped a piece of Kula(?) plate in a ample, as is true of the present Queen Char-
proto-Bering Sea. lotte transform boundary of the NE Pacific
off Chigagof and Baranof Islands; von Huene et
al. 1978). During transform motion between
Koryak Range the two plates, slivers of allochthonous crustal
The structural and lithological similarities fragments transported from the south could
between the m61ange units in the Koryak fold- have been tectonically incorporated into what is
belt and the McHugh and Uyak Complexes of now the bedrock framework of the Beringian
south-western Alaska noted by McLean continental slope.
(1979a) support Burk's (1965) original compari- Dredging in 1978 and earlier sampling work
son of the two areas. The similarities between by Hopkins et al. (1969) have not yielded
the Alaskan complexes and the Koryak Fold- evidence of a former active or underthrust plate
belt suggests that the deformed rocks of the margin (McLean 1979b; Marlow & Cooper
Koryak Mountains are a subduction complex 1981). Rocks recovered thus far from the sub-
formed by the underthrusting of the Kula(?) shelf basement beneath the margin are indu-
plate beneath the North American or Eurasian rated and broadly deformed but they are not
plate in Mesozoic to earliest Tertiary time. part of m61ange sequences nor are they perva-
Allochthonous blocks of Palaeozoic and sively sheared or fractured as might be expected
Mesozoic age, associated serpentinite m61ange, along an ancient plate boundary of convergent
and northward-dipping thrust faults all suggest character. Fossils in the basement samples indi-
Mesozoic to earliest Tertiary foreshortening of cate deposition in shallow water or upper slope
the Koryak Range by horizontal compression. environments.
The ultramafic complexes in association with Oceanic crust can be traced on seismic reflec-
cherts, spilites, and pillow basalts in the range tion profiles to within 15-20 km of the shallow-
may be remnants of either oceanic crust, water basement rocks of continental affinity
perhaps fragments of former Kula(?) plate, or that underlie the continental slope. If deformed
island-arc crust. Convergence along the con- lower continental slope or oceanic deposits
tinental margin had apparently ceased in or by exist beneath the margin, they must be small in
earliest Tertiary time when the ancestral Aleu- volume. Alternatively, large volumes of
tian arc began to form and the convergence oceanic beds were subducted or thrust to the
zone jumped to the south along an ancient NE beneath the continental slope in Mesozoic
Aleutian Trench (Scholl et al. 1975). We note, time (Scholl et al. 1975).
however, that the Siberian margin has under- We should note here that our sample and
gone repeated uplift and broad compressional seismic-reflection data attest only that the
deformation in the Cenozoic, the strongest Cenozoic Beringian margin has been a subsid-
episode of Cenozoic tectonism occurred in ing one since about Eocene time. In Late Cre-
middle Miocene through early Pliocene time taceous time the outer shelf was underlain by a
(Gladenkov 1964; Tilman et al. 1969; Drabkin structural high that included broadly deformed
1970). The Koryak area may, therefore, still be Jurassic and Cretaceous sedimentary and
part of a broad plate boundary, separating but igneous (based on detritus in Mesozoic sand-
continuing to reflect very limited (<100 km) stone) rocks that accumulated in shallow-water
interaction between North America and Eura- regions. Tectonic models that suggest strike-slip
sia(?) (Fig. 5; see fig. 1 of Zoneshayn et al. or oblique convergence along the Beringian
1978). margin in Mesozoic time are based on regional
geological reconstructions and plate reconstruc-
Beringian margin tions that are based on magnetic anomalies
(Scholl et al. 1975; Cooper et al. 1976a).
The plate model proposed by Scholl et al.
208 M . S . Marlow et al.

NORTH AMERICAN PLATE "~ALASKA

~,--c~ .............:"'~e,)v"" ,=,~L,'" Before

KULA PLATE

/ l ~ ~ , ...,,~...... ~,@,,-;,y~,...
~,~irZ 'ii'''ri~v KULA PLA .m#;":'~~ 6B0e f oM.Y.
re

SIBERIA~L"~~ ALASKA

/~ ~
Earliest
Tertiary

?
SIBERIA ~/~ / L ~ ~ , ,
.,~ ~,/~'-" '~t ~'-'~) ALASKA
q~l~ e .

...... .... .:-'::~":'

Present

~Eci-#)c #T_~¥~ 7 \\ \

\,
 Plate boundaries in the Bering Sea 209

Conclusions and speculations BERINGIAN

MARGIN NE
KULA PLATE
Serpentinite m61ange, olistrostromes, ultrama-
fic complexes, and compressionally deformed J
terrigenous sequences exposed in the Koryak
Mountains of eastern Siberia are presumed to
MESOZOIC
be a subduction complex of Mesozoic age
(McLean 1979a). The converging lithospheric C'
plates are thought to have been the Kula(?) and NE

North American or Eurasian plates.

Following cessation of rapid convergence be- J
tween the two plates in late Mesozoic or earliest
Tertiary time, the Koryak Range underwent
repeated uplift, principally during the middle L A T E S T M E S O Z O I C OR
EARLIEST TERTIARY
Miocene and early Pliocene. Parts of the coastal
C
area subsided deeply, especially in the early
NAVARIN BASIN NE
Tertiary. Cenozoic deformation, differential ALEUTIAN
ISLANDBA~EO,,A~/~
uplift, and subsidence of the Siberian margin
may in part be attributable to the buoyant uplift
of sediment subducted beneath the margin dur-
ing earlier periods of plate convergence.
In contrast to the uplifted Koryak Range in
CENOZOIC DETACHED
eastern Siberia, the adjacent Beringian margin EDGE

to the SE is a foundered continental margin that

appears to be underlain by shallow-water sedi- FIG. 6. Postulated plate motion between
mentary and igneous rocks of Jurassic and the Kula(?) plate and Beringian margin
Cretaceous age. The initial collapse of the (North American plate) during the Meso-
margin in Late Cretaceous or early Tertiary zoic. For location of cross-section see Fig.
time could be ascribed to subsidence along 1. By latest Mesozoic or earliest Tertiary
time, formation of the Aleutian arc isolates
wrench faults that are associated with obliquely a remnant of Kula(?) plate in the proto-
underthrust plate boundaries (Fitch 1972). But Bering Sea and also isolates the Beringian
the continued collapse of the margin in middle margin. The detachment of a piece of
and late Cenozoic time must be attributed to Kula(?) plate beneath the Beringian mar-
other factors. One mechanism for collapse gin and its subsequent sinking causes the
might be the cooling of continental lithosphere subsidence and extensional deformation of
that had been heated by the underthrusting the margin throughout the Cenozoic.
action of oceanic crust.
Underthrusting in the Mesozoic would have Another mechanism that could result in the
resulted in the juxtaposition of thick low- collapse of convergent plate boundaries is the
density continental crust against thin ocean detaching of a segment of subducted oceanic
crust; cessation of underthrusting would allow plate (Blanton 1977; Uyeda & Kanamori 1979).
tensional stresses to develop in the continental Fig. 6 is a sketch adapted from Blanton (1977)
crust (Bott 1971; Bott & Dean 1972). The illustrating the oblique subduction of Kula(?)
stresses in the continental crust cause an ocean- plate beneath the Beringian margin. Following
ward creep of lower crustal rocks beneath the formation of the Aleutian arc and isolation of
outer shelf toward the outer margin and Aleu- the margin, a slab of the subducted plate is
tian Basin. Outflow of the lower crust would mechanically detached from the rest of the
contribute to subsidence of the margin and to plate underlying the adjacent Aleutian basin.
extensional deformation of the outer shelf. According to Blanton, after detachment of the

FIG. 5. Generalized sketch map of late Mesozoic and earliest Tertiary plate motions along the
Beringian margin. Prior to 80 Ma, the margin was obliquely underthrust by the Kula(?) plate.
Between 80 and 60 Ma, oroclinal bending of western Alaska and subsequent clockwise rotation of
the margin resulted in change in plate motion to transform or strike-slip movement between the
North American and Kula(?) plates (W. Patton written comm. 1979). Formation of the Aleutian arc
by earliest Tertiary time isolated a piece of Kula(?) plate, thus forming the Bering Sea as the site of
subduction shifted south to near the present Aleutian Trench. After Scholl et al. (1975) and Marlow
& Cooper (1981). Speculative boundary between the North American-Eurasian plates shown by
diagonal hachures.
210 M.S. M a r l o w et al.

subducted slab the continental margin bowed margin during the Cenozoic could be ascribed
upward, then ruptured extensionally and began to wrench faulting associated with strike-slip
to subside. motion.
A final speculation as to the cause of Ceno- In summary, we note that our model for the
zoic collapse of the Beringian margin is based late Mesozoic to early Tertiary evolution of the
on plate reconstructions for the recent move- Bering Sea does not satisfactorily explain the
ments between the North American and Eura- continued Cenozoic tectonism along the deacti-
sian plates by Z o n e n s h a y n et al. (1978, fig. 1). vated and supposedly quiescent former plate
Their work indicates that the North American- boundaries within the Bering Sea. The con-
Eurasian plate boundary passes through the tinued uplift of the Koryak Range and the
Koryak Mountains (eastern Siberia)-Beringian steady subsidence of the Beringian margin
margin area. In their model, the Koryak Moun- remain difficult to explain in a plate-tectonic
tains (eastern Siberia) could have been the site model for the North Pacific and Bering Sea
of limited Cenozoic convergence along a broad during middle and late Cenozoic time. Our
and diffuse boundary separating the North explanations, problematical as they may be, are
American-Eurasian plates (Fig. 5). The Bering- offered here to stimulate discussion and further
ian margin could also have been the site of study.
limited (few tens of kilometres) Cenozoic trans-
form or strike-slip motion extending from the
Koryak Mountains south-eastward along the ACKNOWI.EI)GMENIS: We thank Michael Fisher and
margin and intersecting the Aleutian Trench Roland von Huene for their thoughtful and thorough
near Unimak Pass (Fig. 5). Subsidence of the reviews.

References
ALEKSANDROV, A. A., BOGDANOV, N. S., BYALO- , SCIIOt.L, D. W. & MARLOW,M. S. 1976b. Plate
BZtlESKIY, S. G., MARKOV, M. S., TII.MAN, S. M., tectonic model for the evolution of the Bering
KttAIN, V. YE. & CHEKHOV, A. D. 1976. New Sea basin. Bull. geol. Soc. Am. 87, I119-26.
data on the tectonics of thc Koryak highlands. CREAGER, J. S., SCHOLL, D. W. et al. 1973. lnit. Rep.
Geotectonics, 9, 292-9. Deep Sea drill. Proj. 19, 913 pp. U.S. Govt
BAILEY, E. H. & BLAKE, M. C. JR. 1969. Tectonic Printing Office, Washington.
development of western California during the DRAaKIN, I. YE (ed). 1970. Severo-Vostok SSSR,
late Mesozoic. Geotectonics, 3, 148-54. Geologischeskoye Opisaniye. In: Geologiya
BAILEY, E. H., IRWIN, W. P. & JONES, D. L. 1964. SSSR, Tom 30, Kniga, 2. Izd. Nedra, Moscow
Franciscan and related rocks, and their signifi- (Geology of USSR, northeast USSR).
cance in the geology of western California. Bull. ERNST, W. G. 1970. Tectonic contact between the
- Calif. Div. Mines Geol. 183, 177 pp. Franciscan melange and the Great Valley sequ-
BLANTON,S. L. 1977. Geology of the Bering shelf. In: ence--crustal expression of late Mesozoic Be-
SIssoN, A. (ed). The Relationship of Plate Tec- nioff zone. J. geophys. Res. 75, 886-902.
tonics to Alaskan Geology and Resources. E l - EWING, J., EWING, M., AITKEN, T. & LUDWIG, W.
E6, 1977 Symp. Proc. Alaska Geol. Soc. 1968. North Pacific sediment layers measured by
Anchorage, Alaska. seismic profilings. In: The Crust and Upper Man-
B o o , M. H. P. 1971. Evolution of young continental tle of the Pacific Area. Geophys. Monogr. 12,
margins and formation of shelf basins. Tecto- 147-73. Am. geophys, Union, Washington.
nophysics, 11, 319-27. FITCU, T. J. 1972. Plate convergence, transcurrent
& DEAN, D. S. 1972. Stress systems at young faults, and internal deformation adjacent to
continental margins. Nature (phys. sci.) 235, southeast Asia and the western Pacific. J. geo-
23-5. phys. Res. 77, 4432-60.
BURK, C. A. 1965. Geology of the Alaska Peninsula GLADENKOV, IU. B. 1964. O tektonike vostochnoi
island arc and continental margin. Mem. geol. chasti Koryakskogo Nagor'ya. Akad. Nauk
Soc. Am. 99, 250 pp. SSSR Geol. Inst. Trudy. 113, 7-23 (Tectonics of
CLARKE, S. H. B. 1973. The McHugh Complex of the eastern part of the Koryak Uplands).
south-central Alaska. Bull. U.S. geol. Surv. HAMILTON, W. 1969. Mesozoic California and the
1372-d, D1-Dll. underflow of Pacific mantle. Bull. geol. Soc. Am.
CONNELLV, W. 1978. The Uyak Complex, Kodiak 80, 2409-30.
Islands. Alaska: a Cretaceous subduction com- HOPKINS, D. M. et al. 1969. Cretaceous, Tertiary, and
plex. Bull. geol. Soc. Am. 89, 755-69. early Pleistocene rocks from the continental mar-
COOPER, A. K., MARLOW, M. S. & SCHOLL, D. W. gin in the Bering Sea. Bull. geol. Soc. Am. 80,
1976a. Mesozoic magnetic lineations in the Be- 1471-80.
ring Sea marginal basin. J. geophys. Res. 81, MARLOW, M. S. 1979. Hydrocarbon prospects in
1916-34. Navarin basin province, northwest Bering Sea
 Plate boundaries in the Bering Sea 211
shelf. Oil Gas J. October 29, 190-6. TILMAN, S. M., BELYI, V. F., NIKOLAEVSKII,A. A. &
& COOPER, A. K. 1981. Mesozoic and Cenozoic SHRED, N. A. 1969. Tektonika Severo-Vostoka
structural trends beneath the southern Bering SSSR. Akad. Nauk SSSR Sibirsk. Otdeleniye,
Sea shelf. Bull. Am. Assoc. Petrol. Geol., in Severo-Vostoch. Kompleks, Nauchno Issled.
press. Inst. Trudy, 33, 78 pp. Magadan (Tectonics of
, , SCHOLL, D. W., VALLIER, T. L. & the northeast of USSR).
MCLEAN, H. 1979b. Description of dredge sam- TYSDAL, R. G. ,~ CASE, J. E. 1976. The McHugh
ples from the Bering Sea Continental Margin. Complex in the Seward quadrangle, south-
Open-File Rep. U.S. geol. Surv. 79-1139, 5 pp. central Alaska. In: BLEAN, K. M. (ed). The
--, SCHOLL, D. W., COOPER, A. K. & BUFFINGTON, United States Geological Survey in Alaska:
E. C. 1976. Structure and evolution of Bering accomplishments during 1976. Circ. U.S. geol.
Sea shelf south of St. Lawrence Island. Bull. 751-B, B48-9.
Am. Assoc. Petrol. Geol. 60, 161-83. - - , WINKLE, G. R. & CLARK, S. H. B. 1977.
, , -- & JONES, D. L. 1979a. Mesozoic Sheeted dikes, gabbro, and pillow basalt in flysch
rocks from the Bering Sea; the Alaska-Siberia of coastal southern Alaska. Geology, 5, 377-83.
connection. Abstr. Prog. geol. Soc. Am. 11, 90. UYEDA, M. S. & KANAMORI, H. 1979. Back-arc open-
MCLEAN, H. 1979a. Review of petroleum geology of ing and the mode of subduction. J. geophys. Res.
Anadyr and Khgatyrka basins, USSR. Bull. Am. 84, 1049-61.
Assoc. Petrol Geol. 63, 1467-77. VALLIER, T. L., UNDERWOOD, M. B., JONES, D. L. &
1979b. Pribilof segment of the Bering Sea con- GARDNER, J. V. 1980. Petrography and geologic-
tinental margin: a reinterpretation of Upper Cre- al significance of Upper Jurassic rocks dredged
taceous dredge samples. Geology, 7, 307-10. near Pribilof Islands, southern Bering Sea con-
PA6E, B. M. 1970. Sur-Nacimiento fault zone of tinental shelf. Bull. Am. Assoc. Petrol. Geol. 64,
California. Bull. geol. Soc. Am. 81,667-90. 945-50.
1978. Franciscan melanges compared with olis- VON HUENE, R., SHOR, G. S. JR ,~ WAGEMAN, J.
tostromes of Taiwan and Italy. Tectonophysics, 1978. Continental margins of the eastern Gulf of
47, 223-46. Alaska and boundaries of tectonic plates. In:
ROEDER, D. H. 1975. Tectonic effects of dip changes WATKINS, J. S., MONTADERT,L. & DICKERSON, P.
in subduction zones. Am. J. Sci. 275, 252-64. W. (eds). Geological and Geophysical Investiga-
SCHOLL, D. W., BUFFINGTON, E. C. & MARLOW, M. tions of Continental Margins. Mere. Am. Assoc.
S. 1975. Plate tectonics and the structural evolu- Petrol. Geol. 29, 273-90.
tion of the Aleutian-Bering Sea region. In: ZONENSHAYN, L. P., NATAPOV,L. M., SAUOSTIN,L. A.
FORBES, R. B. (ed). Contributions to the Geolo- & STAVSKII,A. P. 1978. Recent plate tectonics of
gy of the Bering Sea basin and A djacent Regions. northeastern Asia in connection with the open-
Spec. Pap. geol. Soc. Am. 151, 1-32. ing of the North Atlantic and the Arctic Ocean
SILVER, E. A. & BEUTNER, E. C. 1980. Melanges. Basins. Oceanology, 18, 550-5.
Geology, 8, 32-4.

M, S. MARLOW, A. K. COOPER, D. W. SCHOLL • H. MCLEAN. U.S. Geological

Survey, 345 Middlefield Road, Menlo Park, California 94025, U.S.A.
 The Chugach Terrane, a Cretaceous trench-fill deposit,
southern Alaska

Tor H. Nilsen & Gian G. Zuffa

SUMMARY: The Chugach terrane of southern Alaska extends for approximately

2000 km along the margin of the Gulf of Alaska. A seaward flysch facies of the terrane, the
Chugach flysch terrane, represents the fill of a Late Cretaceous trench and consists of
structurally deformed turbidites with some mafic volcanic rocks. It is intruded by anatectic
granitic plutons of early Tertiary age. The Chugach flysch terrane in most places is
bounded to the north by a landward-dipping thrust fault separating it from m61ange of the
Chugach terrane at least as young as Late Cretaceous. To the south the flysch terrane
contains oceanic volcanic rocks and is bounded by faults that separate it from Palaeogene
turbidites or upper Mesozoic metavolcanic rocks.
Interpretations of folds and faults in SW Alaska suggest that the Chugach terrane was
deformed during NW-directed subduction. Palaeocurrents indicate primarily westward
flow along the axis of the outcrop belt, and secondary southward transport. Turbidite
facies associations indicate an east-to-west progression from inner-fan to middle-fan,
outer-fan, fan-fringe and basin-plain deposits down the axis of the outcrop belt, and a
bounding slope facies association to the north. Rock-fragment petrography of sandstone
samples from the Chugach terrane indicates derivation from a magmatic arc that was
increasingly dissected eastward and from an older subduction complex. The magmatic arc
and adjacent shallow-marine forearc basin deposits are located to the north in southern
Alaska. Palaeomagnetic data from the Chugach terrane and adjacent deposits indicate
original deposition far to the south of the present latitude of Alaska. The entire magmatic
arc-forearc basin-trench complex migrated northward in early Tertiary time, when it was
probably oroclinally bent and accreted as a mesoplate to Alaska prior to the middle
Miocene.

Although modern deep-sea trenches adjacent derived from various sources and transported to
to both continental margins and island arcs are the trench by various routes. Continental-
abundant and widely distributed, very few good margin trenches are most likely to contain thick
examples of well-preserved and extensive sediment fills because the magmatic arc and
ancient trench-fill deposits have been described uplifted older terranes form larger source areas.
in the geological literature. Most examples de- Both the thickness and geometry of trench fills
scribed are of obducted fragments or slices of depend upon the rate of convergence across the
trench-fill deposits rather than the entire fill of trench and whether sediment transport is longi-
a major trench system. The reason for the lack tudinal or lateral (Schweller & Kulm 1978).
of good examples is clearly the structural de- Some trenches may be starved of sediment
formation imposed on trench-fill deposits, derived from the continental margin because
which generally are partially or wholly sub- most if not all of the sediment may be deposited
ducted, disrupted by lateral faulting, or inten- in forearc basins (the arc-trench gap) or in
sively deformed by plate-margin tectonism. major basins on the trench slope (Scholl &
The purpose of this paper is to describe the Marlow 1974). Because of extensive tectonic
general setting, geological history, and deformation and submarine landsliding, con-
palaeogeographic framework of part of the tinuous submarine canyons may not extend
Chugach terrane, inferred to be a Cretaceous across the landward trench slope to funnel
trench-fill complex accreted to southern Alas- sediments from the magmatic arc to the trench
ka. The Chugach terrane may represent the floor (Underwood & Karig 1980). Some trench-
longest, most continuous, and best preserved fills may consist mostly of hemipelagic sediment
ancient trench-fill sequence in the world. After deposited along the continental margin and
more than 10 years of detailed study of various pelagic sediments carried into the trench by the
parts of the terrane, the overall size, character subducting oceanic plate rather than turbidites
and significance of this remarkable sequence derived from the continental margin.
are beginning to emerge. The Chugach terrane was named by Berg et
Modern trenches may extend for thousands al. (1972) for outcrops of greywacke, argillite,
of kilometres and can be filled by sediment slate, conglomerate, volcanic rocks, chaotic

213
214 T. H. N i l s e n & G. G. Z u f f a

m61anges, and granitic plutons in the Chugach late Mesozoic trench-fill.

Mountains of southern Alaska. They suggested The age of the Chugach flysch terrane has
that the terrane extended south-eastward to been established as Late Cretaceous, primarily
Baranof and Chichagof Islands and divided it Maestrichtian (Jones & Clark 1973), on the
into two units thought to be of different age. basis of scattered megafossils. Brew & Morrell
Their older Chugach terrane cropped out land- (1979) reported Early and Late Cretaceous
ward of the younger Chugach terrane and fossils from parts of SE Alaska.
consists of a polydeformed and regionally The Chugach flysch terrane is bounded by
metamorphosed assemblage of phyllite, meta- major faults both landward and seaward
greywacke, quartzite, metachert, greenstone, throughout most of its extent. The Eagle River
amphibolite, and ultramafic rocks originally fault, called the Uganik fault on Kodiak Island
thought to be as old as Pe~rmian and Triassic. It by Moore (1978), bounds the Chugach terrane
originally included some rocks currently on the north. It generally dips landward, except
assigned to Wrangellia, a separate accreted in SE Alaska, and separates the Chugach flysch
terrane to the NE (Jones & Silberling 1979). terrane from the Chugach m61ange terrane,
The outboard younger Chugach terrane con- known as the Uyak Complex on Kodiak Island
sists of an exterior belt of greywacke and slate, (Connelly 1978), the McHugh Complex in the
generally very thick and unfossiliferous, also Chugach Mountains and adjacent areas (Clark
extending south-eastward to Baranof and 1973), the m~lange facies of the Yakutat Group
Chichagof Islands. Fossils from the younger east of Yakutat, the Tart Inlet suture zone
Chugach terrane are chiefly of Late Cretaceous rocks in the Glacier Bay National Monument
age. The Chugach terrane was extended by (J. Decker, pers. comm. 198(I), and the upper
mapping to SW Alaska (PlaNer et al. 1977; part of the Kelp Bay Group near Baranof
Moore & Connelly 1979; Nilsen & Moore Island (Loney et al. 1975). The Chugach
1979), where the younger Chugach terrane had mdlange terrane contains blocks of many diffe-
previously been mapped over a wide area as the rent lithologies and ages. Blocks of chert from
'slate and graywacke belt' (Burk 1965). the McHugh Complex contain radiolarians of
As mapped at present, the Chugach terrane Triassic, Jurassic, and Cretaceous as young as
crops out continuously, except where covered Cenomanian (Karl et al. 1979: G. R. Winkler,
by large glaciers and marine waters of the Gulf pers. comm. 1980); megafossils in the matrix
of Alaska, for about 2000 km from Baranof are as young as Maestrichtian (Plafker et al.
Island in the SE to Sanak Island in the SW (Fig. 1977).
1). The outcrop belt is as wide as 100 kin, but is The Border Ranges fault forms the northern
highly variable where affected by Cenozoic margin of the Chugach flysch terrane where the
strike-slip faulting, particularly in SE Alaska. Eagle River fault is not present and extends for
Moore (1972, 1973a) suggested that the out- most if not all the outcrop length of the Chu-
crop belt of the Chugach terrane may extend gach terrane. The Border Ranges fault dips
NW for another 1200 km from the Sanak Is- landward and locally juxtaposes Palaeozoic and
lands along the edge of the shelf of the Bering Mesozoic metamorphic rocks with the Cre-
Sea to the Koryak Mountains of SE Siberia. taceous units (MacKevett & Plafker 1974; Plaf-
However, because Upper Cretaceous rocks ker et al. 1976, 1977; PlaNer & Campbell 1979).
have not been recovered by recent dredging or To the south, the Chugach flysch terrane is
detected by geophysical studies of the Bering separated from Palaeogene turbidites and mafic
Sea shelf edge, the extension of the Chugach volcanic rocks by the Contact fault, in most
terrane to the NW is now considered unlikely places a high-angle landward-dipping reverse
(Cooper & Marlow 1979; McLean 1979). fault. In SE Alaska, the Queen Charlotte and
Because faunal evidence now indicates that Fairweather faults form the SW margins of the
both the older and younger Chugach terranes of Chugach flysch terrane and related late Meso-
Berg et al. (1972) contain fossils as young as zoic(?) volcanic rocks (PlaNer et al. 1977; Plaf-
Late Cretaceous, the distinction of the two ker & Campbell 1979). These faults are right-
parts of the terrane by age has become less lateral strike-slip faults that record Cenozoic
desirable. For the purposes of this paper, I offsets.
divide the Chugach terrane into the Chugach North of the Chugach terrane are extensive
m61ange terrane, formerly part of the 'older outcrops of non-marine and shallow- to deep-
Chugach terrane', and the Chugach flysch ter- marine Upper Cretaceous sedimentary rocks
rane, formerly the 'younger Chugach terrane'. (Fig. 1). These units, of Campanian and Maes-
The Chugach flysch terrane will be of principal trichtian age, are common in parts of the Alas-
interest here, because it appears to represent a ka Peninsula, on the NW side and beneath the
 165 ° 160 ° 155 ° 150 ° 145 ° 140 ° 135 °
/ "~ / I / Eii::iil]!i/ ~ ! .'.~~_~_.. I ~. ~l>-I 1 I

/ ~.\ ~ , ~ " . 7 _ ( "

. ~~ ~_ ,
~" ~ ~ ~ - , ~ .::......... ~ .
,_-_
.j,~,'~.o.-, I "

60 ° .:::ii!iiY "v' C

uc-d ~.~o,,o-~I ~.~oo oo

"
C- ,°°°,°,.
~ -'~ ..~. " ~ ' ?"

",, ,,,,o...r , . - " - - ~ ~ ~ am

¢3

¢%
56 °
s o oo .oo .oo.
Is,one, .~ 4s~s..... -.'" ' ' ' '
~.4;,,
l I [ l J l PORT ALE X ANDs-r~'~]

Neogene shallow-marine deposits BRF BORDER RANGES

FAULT
Paleogene turbidites end mafic volcanic rocks
CSF CHATHAM STRAIT r.,m
Late Cretaceous and early Tertiary granitic FAULT
plutons (magmatic arc)
CF CONTACT FAULT
Late Cretaceous (Chugach terrane)turbidites and ¢3
ERF EAGLE RIVER
volcanic rocks %
FAULT
Late Cretaceous shallow- to deep-marine and F F FAI R W E A T H E R
nonmarine sedimentary rocks (forearc basin)
FAULT
Late Mesozoic mafic volcanic racks, metamorphosed QCF QUEEN CHARLOTTE
FAU LT
Cretaceous (Chugach terrane) m61ange

Late Mesozoic and Paleozoic metamorphic rocks

-- ---- F A U L T - - dashed where concealed

Ix.)
FIG. 1. Distribution of Chugach terrane and major faults along margins of the Gulf of Alaska. Geology modified from Beikman (1974a,b, L/I
1975a,b), PlaNer et al. (1977), and PlaNer & Campbell (1979).
216 T. H. N i l s e n & G. G. Z u f f a

Cook Inlet, and in the Matanuska Valley NE of the last few years as a collage of various ter-
Anchorage (Grantz 1964; Jones & Detterman ranes that have been transported northward
1966; Mancini et al. 1978; Detterman 1978; and accreted to Alaska. Many different Meso-
Magoon et al. 1978, 1980). These clastic zoic terranes have been recognized in British
sedimentary rocks, generally transported south- Columbia (Monger et al. 1972), SE Alaska
ward and derived from volcanic and plutonic (Berg et al. 1978), and central and southern
sources, are forearc basin or arc-trench gap Alaska (Jones & Silberling 1979; Csejtey 1979).
deposits coeval with and situated landward of Palaeomagnetic data, largely from rocks associ-
the Chugach terrane. ated with the Chugach terrane, indicate that it
North of the forearc basin deposits is a belt of was originally deposited far to the south of its
Upper Cretaceous to lower Tertiary quartz present position (Hillhouse & Gromme 1977;
diorite to granite plutons that form part of a Stone & Packer 1979); during the Cenozoic the
large magmatic arc (Fig. 1). The plutons, called terrane moved northward, eventually reaching
the Iliamna-McKinley phase of the Alaska its present latitude and accreting to Alaska.
Range-Talkeetna Mountains batholithic com- However, the process, timing, and tectonic
plex by Hudson (1979a,b,c), have yielded effects of the accretion process on the Chugach
potassium-argon ages of 83 to 55 Ma (Reed & flysch terrane and related rocks are not well
Lanphere 1973). Plots of K20 trends by Moore understood.
& Connelly (1979) from data presented by The Chugach flysch terrane has previously
Reed & Lanphere (1974) for the Late Cre- been considered to be a trench deposit on the
taceous and early Tertiary Alaska Range- basis of tectonostratigraphic considerations
Talkeetna Mountains batholiths in SW Alaska (Burk 1965; Moore 1972, 1973a,b; Budnik
suggest that these plutons were generated in 1974a,b; Nilsen & Bouma 1977; Connelly 1978;
response to NW-directed subduction. In SE Nilsen & Moore 1979; Moore & Connelly 1979).
Alaska, Cretaceous and early Tertiary granitic However, regional petrographic, palaeocurrent,
plutons are abundant landward of the Border and facies studies have never been completed, so
Ranges fault (Brew et al. 1978). The magmatic that most palaeogeographic and palaeotectonic
arc may be bounded on the north in Central and generalizations have not been substantiated by
SW Alaska by the Denali fault system, which geological data.
may be a major Cenozoic right-lateral strike-
slip system.
Within the Chugach flysch terrane and on its Stratigraphy
south flank are extensive but discontinuous
belts of volcanic rocks that are locally highly The Chugach flysch terrane includes a number
deformed and metamorphosed adjacent to plu- of stratigraphic units of somewhat differing
tons. On Kodiak Island, basalts and andesites characteristics. Most published descriptions
have been included in the Palaeocene and note that it consists of deformed, thick, repeti-
Eocene Ghost Rocks Formation (Moore 1969; tively interbedded greywacke and shale or slate
Nilsen & Moore 1979; Reid & Gill 1980). In the with local conglomerate, limestone, and volca-
Prince William Sound area diabasic sheeted nic rocks.
dykes, gabbros, and basalts crop out within On the Sanak and Shumagin Islands, the
both the Chugach flysch terrane and the Chugach flysch terrane consists of the Shuma-
Palaeogene Orca Group to the south (Tysdal et gin Formation. Moore (1973a) estimated 3-
al. 1977). In SE Alaska, late Mesozoic oceanic 4 km as a minimum thickness after having
basalts form a mappable terrane along the SW measured sections as thick as 1470 m on the
margin of the Chugach flysch terrane (Plafker Sanak Islands and 4100 m on the Shumagin
& Campbell 1979). Volcanic rocks are interbed- Islands. Scattered fossils indicate a Maestrich-
ded with the Sitka Graywacke on Chichagof tian age (Jones & Clark 1973). On the Sanak
Island (J. Decker, pets. comm. 1980). These Islands, pillow lava and bedded chert are locally
various volcanic and plutonic rocks are clearly in depositional contact with the turbidites.
of two distinct ages, the first coeval with deposi- On Afognak and Kodiak Islands, the Chu-
tion of the Maestrichtian Chugach flysch ter- gach flysch terrane consists of the Kodiak
rane and the second coeval with deposition of Formation. Nilsen & Moore (1979) estimated a
Palaeogene turbidites. These rocks, partly stratigraphic thickness of 5000 m and noted no
oceanic in character, indicate the involvement volcanic rocks or chert within it. Scattered
of ocean floor with deposition and deformation fossils indicate a Maestrichtian age (Jones &
of the Chugach flysch terrane. Clark 1973).
Southern Alaska has been reinterpreted in In the Chugach Mountains and Kenai Penin-
 The C h u g a c h Terrane, a Cretaceous trench-fill d e p o s i t 217

sula, the Chugach flysch terrane has been map-

!iii.
NW A $E

ped as the Valdez Group by Clark (1972) and

Tysdal & Plafker (1978), but no stratigraphic
thicknesses have been estimated for it. Maes-
trichtian megafossils are present at scattered o b a 3 4 ~.

localities (Jones & Clark 1973). The Valdez

Group includes local interbedded tuff and pil-
low basalt, and on the Resurrection Peninsula NW B SE
near Seward is intruded by basalt sills, sheeted ~RF ¢F

'i "
/ '

dykes, gabbro and serpentinized dunite (Tysdal

et al. 1977).
The Chugach flysch terrane SE of the Yaku-
tat area has been mapped as part of the Yakutat
Group (see Plafker & Campbell 1979) and as an
unnamed unit east of the Fairweather fault
(Brew & Morrell 1979). Much of the Yakutat C
Group crops out west of the Fairweather and NW SE

Contact faults, outside the main outcrop belt of

the Chugach flysch terrane (Fig. 1), and could
be considered to be a separate terrane from the
Chugach terrane. No thicknesses or fossils have
been reported for this area, which is regionally
metamorphosed. Flo. 2. Cross-sections showing general
On Chichagof, Baranof, and adjacent is- structure of Chugach terrane and relations
lands, the Chugach flysch terrane has been with bounding terranes. Map symbols and
mapped as the Sitka Graywacke (Loney et al. abbreviations as in Fig. 1 except for Tg,
Tertiary granitic plutons.
1975). No thickness for it has been estimated. (A) NW-SE cross-section of part of Shu-
Fossils include only Early Cretaceous pelecy- magin Islands (from Moore 1973b,
pods transported and redeposited in debris Fig. 8).
flows at one or two localities. The Sitka (B) NW-SE cross-section of Kodiak Is-
Graywacke contains local bodies of greenstone, land (from von Huene et al. 1978).
greenschist and chert. (C) NW-SE cross-section from Anchor-
age area to Prince William Sound
(from Plafker et al. 1977).
Structure Uyak Complex (Moore & Wheeler 1978),
Throughout its extent, the Chugach flysch ter- 334 ° _ 7 ° along the Uganik or Eagle River fault
rane is highly folded and faulted (Fig. 2). Strata (Moore 1978), and 340 ° for the Chugach flysch
generally dip landward, except in SE Alaska, terrane (Moore & Bolm 1977). These under-
and strike parallel to the regional trend of the thrust directions are perpendicular to the pre-
belt. Faults and axial planes of folds also dip sent general strike of the Chugach flysch ter-
landward and strike parallel to the regional rane, forearc basin sequence, and Alaska
trend. Moore (1973b) demonstrated that on the Range-Talkeetna Mountains batholithic com-
Sanak and Shumagin Islands the Chugach plex.
flysch terrane was initially deformed in a par- Granitic batholiths, informally referred to as
tially lithified state and developed an axial the Sanak-Baranof belt, intrude the Chugach
plane slaty cleavage. A last phase of deforma- flysch terrane in many areas, and have been
tion involved uplift and landward tilting along dated by the potassium-argon method as
high-angle faults. 59.9 Ma on the Sanak Islands, 54-64 Ma on the
The Chugach flysch terrane on Kodiak Island Shumagin Islands, 58 Ma on Kodiak Island,
consists chiefly of landward-dipping homoclinal 47-52 Ma in the Chugach Mountains, and 43-
sequences of turbidites wi.th stratigraphic tops 48 Ma on Baranof Island (see summaries in
toward land, separated by high-angle landward- Hudson 1979a,b). These batholiths have been
dipping reverse faults. Because of this geomet- interpreted to be products of anatectic melting
ry, Moore (1969) originally postulated a thick- of the subduction complex rather than arc
ness of 30,000 m for the Chugach flysch ter- magmatism (Hill & Morris 1977; Hudson et al.
rane. Directions of underthrusting on Kodiak 1979). Regional metamorphism of Late Cre-
Island were determined at 332 ° _ 11° for the taceous age with penetrative deformation typi-
218 T. H. Nilsen & G. G. Z u f f a

ties the Chugach flysch terrane in the Yakutat turbidites suggests the presence of channelized
area (Hudson et al. 1977). fan deposits and possibly slope deposits.
The structural framework of the Chugach On the Kenai Peninsula between Anchorage
terrane suggests early deformation in a trench and Seward, Nilsen & Bouma (1977) reported
setting followed by regional uplift and tilting, at the presence of a slope facies association to the
least in SW Alaska, along landward-dipping NW and a middle-fan facies association to the
reverse faults. In parts of SE Alaska, the ter- SE in the Valdez Group. They suggested that
rane was regionally metamorphosed. The entire part of the slope facies association included
belt of turbidites was intruded by anatectically large olistoliths and olistostromes previously
derived granitic rocks produced by syndeforma- mapped as part of the McHugh Complex. Re-
tional melting of the sedimentary pile. These mapping of the Chugach flysch terrane-
plutons decrease in age eastward. McHugh Complex contact by Tysdal & Case
(1977) had resulted in placing some of these
olistostrome deposits within the Chugach ter-
rane. Budnik (1974a,b) and Mitchell (1979,
Facies associations 1980) also described facies associations in the
Valdez Group, referring to slope, middle-fan,
Turbidite facies can be most simply grouped on outer-fan, and basin-plain deposits.
the basis of vertical sequence into five facies On NE Kodiak Island, Nilsen & Moore
associations: slope, inner-fan, middle-fan, (1979) divided the Kodiak Formation into two
outer-fan, and basin-plain (Mutti & Ricci Luc- facies associations, a landward slope association
chi 1972, 1975; Nilsen 1977). These facies asso- and a seaward basin-plain association. The
ciations have been delineated for parts of the slope association contains abrupt changes in
Chugach flysch terrane, particularly on Chicha- thickness and grain size of deposits, conglomer-
gof Island, the Kenai Peninsula, and Kodiak ate deposited in channels and canyons, small
Island (Fig. 3). trench-slope basins, abundant slumps and syn-
On western Chichagof Island in SE Alaska, sedimentary folds, and thick pelitic sections.
Decker et al. (1979) recognized an inner-fan The small slope basins contain well-developed
facies association toward the SE, a middle- thin accumulations of little-deformed deep-sea
fan facies association toward the NW, and a fan facies associations fed by small canyons and
slope facies association between the two and fan channels that are enveloped by more highly
along the NE margin of the belt of outcrops of deformed fine-grained slope deposits. These
the Sitka Graywacke. The inner-fan facies asso- little-deformed slope-basin deposits are typical-
ciation extends SE to Baranof Island, where it ly more coarse-grained than the adjacent
forms the type area of the Sitka Graywacke. It trench-fill deposits to the south and have
consists chiefly of massive sandstone and con- palaeocurrent directions more indicative of
glomerate with a maximum clast size of about lateral transport. The basin-plain association
25 cm. The middle-fan facies association con- contains, in contrast, classical fine- to medium-
sists of channelized bodies of massive sandstone grained sandstone turbidites that extend for the
surrounded by thicker sequences of thin- entire outcrop length without perceptible
bedded turbidites and shaly channel-margin and changes in thickness or grain size and little
interchannel deposits. The slope facies associa- organization of beds into thickening- or thin-
tion consists of thick sequences of hemipelagic ning-upward megasequences, although locally
mudstone with thin-bedded siltstone and mud- outer-fan lobe and fan-fringe deposits can be
stone turbidites and local olistostromes and distinguished.
slide blocks of shallow-marine sandstone. On the Shumagin and Sanak Islands, Moore
Between Chichagof Island and Yakutat, the (1973a) reported the presence of 35% massive
Chugach flysch terrane consists largely of argil- beds of sandstone, 35% thin-bedded turbidites,
lite and greywacke with smaller amounts of 30% intermediate bed types, and less than
conglomerate and mafic volcanic rocks (Brew 0.5% conglomerate and tuff. Conglomerates
& Morrell 1979, table 15; Plafker & Campbell contain clasts as large as 20 cm and are de-
1979). Although Brew & Morrell (1979) sug- scribed as pebbly mudstone olistostromes.
gested that these deposits consisted mostly of Slumps are not uncommon and one as thick as
outer-fan facies, J. Decker (pers. comm. 1980) 135 m is reported; directions of slumping are
has interpreted these deposits as an inner-fan sub-parallel to the palaeocurrent flow direc-
channel and interchannel facies association. tions. Although Moore (1973a) did not inter-
The incorporation of conglomeratic units within pret facies associations, his descriptions suggest
thick sequences of argillite and thin-bedded the presence of slope and fan or basin-plain
 165 ° 160 ° 155 ° 150 ° 145 ° 140 ° 135 °
/ I / / I ~1<1
~ ~ ~ I_LJLJJ I f'~
~f'7"~'d -~'z

60 °
L AS
Chugach t e r r a n e

(k~ Slope
Inner-fan channel

M i d d l e - f a n channel
f"t.
O u t e r - f a n lobe

~ Basin plain

? ".. BP MF
. . .. ',0':::~=, CHICHAGOF I

56 c
• " 0 I00 200 3 0 0 KM IF1 2-
%'.. •
I I I I ._.

/ 1 I I
FIG. 3. Generalized distribution of facies associations from the Chugach terrane. Abbreviations (from east to west); IF, inner fan" MF, middle fan;
FF, fan fringe; BP, basin plain.

I',J
220 T. H. Nilsen & G. G. Zuffa

facies associations in these islands. No major samples of the Chugach flysch terrane from the
channelling was noted, so that the bulk of the Yakutat area are primarily arkose (G. R. Wink-
deposits may represent outer-fan or basin-plain ler, pers. comm. 1980), further suggesting that
deposits. these rocks may form a separate terrane from
In summary, preliminary analyses of facies the Chugach flysch terrane. In the Chichagof
associations in the Chugach flysch terrane sug- Island area, Chugach flysch terrane sandstone
gest the presence coevally of an inboard slope samples are primarily volcanogenic lithic aren-
facies association, locally difficult to distinguish ite, probably derived chiefly from a magmatic
from the adjacent older m61ange terrane be- arc provenance (J. Decker, pers. comm. 1980).
cause of its extensive synsedimentary deforma- In order to examine the lateral variability in
tion, and an outboard inner-fan to basin-plain composition of the Chugach fiysch terrane,
sequence that appears, on the basis of pub- Zuffa et al. (1980) studied the rock-fragment
lished interpretations, to be most proximal to petrography of about 150 samples of medium-
the SE and distal to the SW (Fig. 3). The clearly to coarse-grained sandstone from Baranof Is-
developed down-axis changes in facies associa- land in SE Alaska to Sanak Island in SW
tions of the trench-fill deposits suggest fill by a Alaska. The rock fragments of about 60 sam-
radiating fan system characterized by a distri- ples were divided into 26 categories, and 100 or
butary system rather than by a single large 200 grains were counted on each thin section.
channel depositing a wedge, such as that de- The resulting data indicate that two main
scribed by Schweller & Kulm (1978). The rela- source types supplied sediment to the Chugach
tions suggest penecontemporaneous slope de- flysch terrane.
position adjacent to a large, elongate, outbuild- The first source area shed argillite, siltstone,
ing fan system in a narrow trough, most prob- volcanic wacke, chert, limestone, basalt, basal-
ably a trench. Sediments appear to have been tic andesite, and andesitic felsite, in varying
derived chiefly from a point source at the SE amounts in different areas. This source area
end of the trough. Although not fully demon- probably consisted of an active volcanic arc,
strated to date by facies studies, the longitu- older accreted terranes, and an emergent or
dinal system of trench filling was undoubtedly submerged subduction complex. The second
supplemented by lateral infilling from canyons source area shed quartz, plagioclase feldspar,
that cut across the inboard trench slope to andesitic felsite, microlitic rock fragments, and
produce mixed sediments. potassium feldspar. This source area was prob-
ably a dissected arc terrane.
A ternary plot of polycrystalline quartz lithic
Petrography fragments, volcanic and metavolcanic lithic
fragments, and sedimentary and meta-
Burk (1965) inferred that the western part of sedimentary lithic fragments clearly indicates
the 'slate and greywacke' terrane, later in- an arc-orogen source field for the Chugach
cluded in the Chugach terrane, was derived flysch terrane (Fig. 4A). The plot also indi-
chiefly from a volcanic source area. From study cates an increasing degree of dissection of the
of sandstone of Shumagin and Sanak Islands, arc-orogen eastward. The petrographic changes
Moore (1973a) concluded that the Chugach suggest lateral as well as longitudinal input of
flysch terrane in those areas consisted primarily sediments to the trench. Sandstone from the
of litharenite derived chiefly from andesitic Yakutat area contains the highest amount of
volcanic rocks, with minor contributions from K-feldspar, which together with abundant
plutonic, sedimentary, and metamorphic rocks. plutonite fragments (Fig. 4B), is indicative of a
Connelly (1978) reported that samples of the fairly distinct magmatic arc provenance and
Kodiak Formation from the NW edge of the supports the inference that these rocks either
outcrop belt on Kodiak Island were more feld- form a separate terrane or were derived from a
spathic than samples of the Chugach flysch completely different source area.
terrane to the west. Winkler (in Nilsen & Conglomerate clasts in the Chugach flysch
Moore 1979) determined that the composition terrane include granitic rocks, chert, limestone,
of Kodiak Formation sandstone samples from quartzite, greywacke, argillite and silicic volca-
other parts of Kodiak Island is similar to that nic rocks in SE Alaska (Brew & Morrell 1979);
obtained by Moore (1973a) farther west. granitic rocks, chert, greenstone, limestone,
Mitchell (1979, 1980) reported for the Valdez argillite and greywacke in southern Alaska
Group south of Anchorage a litharenitic sand- (Tysdal & Case 1977); and chert, volcanic
stone composition, 60% volcanic, 30% meta- rocks, greywacke, mudstone and granitic rocks
morphic, and 10% plutonic types. Sandstone in SW Alaska (Moore 1973a). Thus, the con-
 The Chugach Terrane, a Cretaceous trench-fill deposit 221

s,,,,,.~ ~oc,u,~ w PM
QP • so.o~
/~ 0 $humogin A
/ \ + Kodiok / \
A / \ . S,~dovio / \ la
'-" / \ II Seword / \ L,,
/ \ e Voldez / \
/ \ ,r Chichagof / \

Lv S

FIG.4. Ternary diagrams showing rock-fragment petrography of Chugach terrane.

(A) Plot of polycrystalline quartz (Qp), volcanic and metavolcanic lithic fragments (Lv), and
sedimentary and metasedimentary lithic fragments (Ls). Dashed line defines arc-orogen source
field as proposed by Dickinson & Suczek (1979).
(B) Plot of all rock-fragment varieties: polycrystalline quartz-plutonic-gneiss-schist-phyllite (PM),
felsite-microlitic grains---ophitic grains--vitrophyric grains--undetermined cherty grains--
hypabyssal volcanic (V), and sandstone--siltstone--limestone--argillite---chert--shale--slate
(S). The three poles in a continental margin-arc-trench geometry may indicate magmatic arc
roots (PM), volcanic-arc (V), and subduction complex (S) sources. Dashed line around Yakutat
area samples indicates greater influence of magmatic arc source in this part of Chugach terrane.
Dashed line with arrow indicates from left to right eastward increasing dissection of magmatic
arc and increasing amounts of sedimentary detritus.

glomerates reflect the same two major sources belt parallel to the outcrop trend of the Chu-
as the sandstone petrography. gach flysch terrane and contain abundant shal-
low-marine and non-marine sequences, and (4)
granitic intrusions that form a major sub-
Discussion and conclusions parallel magmatic arc. The complete arc-trench
system is at present bounded by major faults,
The deformation history, petrography, mostly right-lateral strike-slip faults of Ceno-
palaeocurrents, facies associations, and region- zoic age.
al stratigraphic and structural relations suggest The style of folding, faulting, and meta-
that the turbidites of the Upper Cretaceous morphism is typical of deposits emplaced into a
Chugach flysch terrane represent a trench-fill subduction zone. The filling of the trench was
deposit, inclusions of pillow basalt, sheeted by two separate processes, longitudinal and
dykes, serpentinized dunite, and chert suggest lateral. The major palaeocurrent trends and
that the turbidites were deposited on ocean facies associations indicate an axial east-to-west
floor and were subsequently offscraped on the change from proximal to distal turbidite sedi-
upper plate of the subduction zone. Regional mentation, suggesting the development of very
evidence for the trench depositional setting is elongate and highly channelized deep-sea fans
indicated by the presence progressively to the within the narrow trench (Fig. 5). The principal
north of coeval: (1) slope deposits with south- source areas for these axial deposits were at the
trending submarine canyons and trench-slope eastern end of the arc-trench system. Lateral
basins, (2) remnants of a m~lange zone sepa- infilling from the north is suggested by the
rated from the Chugach flysch terrane by a presence of south-directed palaeocurrents and a
landward-dipping major fault zone, (3) south- slope facies associations along the northern
transported forearc basin deposits that form a edge of the outcrop belt (Fig. 6). The at least
 bO
bO
t-J

165 ° 160 ° 155 ° 150 ° 145 o 140 ° 135 °

/ I I I I

i-- i--I <::

Chugach t e r r a i n A L A S K A
60 °

, k ¢'-'~
v f •

BERING SEA 6
c~

0
0

GUL F OF ALASKA
"•o
'~ ",, ; le °° ,
o
. .,,..."" ...'2,~,
56 °
~~
. ~ ~ e . ~ ... o
. • o ,oo 200 3OOKM
. . ~ " t I L ,I

I / I I I I L

FIG. 5. Generalized palaeocurrent directions from the Chugach terrane. Longer arrows indicate longitudinal flow, shorter arrows lateral flow.
 165 ° 160 ° 155 ° 150 o 145 ° 140 ° 135 °
/ / / I I I I

m/,,,a.
F"
A L A S K A

60 °
"7
~ L !

--.
I
tl I
Submarine canyon ¢5

56 °

Proto-Pacific ocean floor "7

(KULA PLATE)

0 I00 200 3 0 0 KM
I I I I

,/x

,e..,.
MOUNTAINS F O R E A R C BASIN TRENCH SLOPE FAN C H A N N E L S FAN LOBES BASIN PLAIN
(magmatic arc) (arc-trench gap) (inner trench wall)

FIG. 6. Palaeogeographic map of Chugach terrane showing magmatic-arc source area, forearc basin, trench slope, trench and proto-Pacific ocean
floor. Large arrows show approximate direction of seafloor movement. Small arrows show principal directions of sediment transport. Incised
areas on trench slope and forearc basin indicate submarine canyons, some of which transport sediment to the trench floor. The Late Cretaceous
t~
trench is shown in its present position without palinspastic restoration to a more southerly position, as indicated by palaeomagnetic data, or t'~
straightening, as indicated by structural studies of the southern Alaska orocline.
224 T. H. N i l s e n & G. G. Z u f f a

partly coeval Chugach m61ange terrane, Eocene, pre-middle Miocene accretion to Alas-
although severely deformed and metamorph- ka. On i t s - n o r t h w a r d or possibly north-
osed, could in part represent deformed olisto- eastward trek, the eastern end of the block
stromes emplaced by large-scale submarine impinged on the west coast of North America,
landsliding on the trench slope. resulting in oroclinal bending of the plate dur-
The trench formed in response to north- ing its subsequent attachment to Alaska. Con-
directed subduction within the framework of tinued drag along the eastern part of the block
present distribution of the trench-fill, forearc resulted in slivering of the terrane along numer-
basin, and magmatic arc. It is clear that the ous Cenozoic strike-slip faults, especially in SE
Chugach terrane and its related Late Cre- Alaska. The present Aleutian trench and arc
taceous m61ange terrane, forearc basin and system was superimposed on the Chugach block
magmatic arc constitute a coherent larger ter- in Cenozoic time; this arc system contains vol-
rane that moved northward as a single unit. canic rocks as old as Eocene, yielding an over-
Oceanic plate reconstructions for the Pacific lap in the timing for arrival of the Chugach
suggest that the Kula plate was moving north- block and initiation of the new arc system.
ward and being subducted beneath Alaska The Chugach terrane and related belts, which
during the Late Cretaceous. However, palaeo- might be called the Chugach mesoplate, form
magnetic data from both Jurassic and Eocene part of a series of upper Mesozoic subduction
rocks in southern Alaska indicate that these complexes along the west coast of North Amer-
units formed at latitudes considerably south of ica (Jones et al. 1978). The term "mesoplate' is
the present 56-60° N (Stone & Packer 1976, used to indicate a larger plate constructed of
1979). If valid, these data require that the previously accreted smaller plates generally re-
Chugach terrane and its associated belts form a ferred to as 'microplates'. The terrane has
large, mostly continental plate that accreted to clearly been post-depositionally transported
southern Alaska by middle Miocene time, prior northward, and formed a major continental
to deposition of the Yakataga Formation. The plate within the proto-Pacific Ocean, much as
accretion must have occurred long after the the Salinian-Peninsula-Baja California plate
Late Cretaceous subduction zone had ceased to west of the San Andreas fault will in the future.
be active. It probably formed the much-debated source
The Chugach terrane is oroclinally bent for the Aleutian abyssal plain turbidites (Stew-
about an axis extending through the Prince art 1976). A speculative source for the Chugach
William Sound area. The time of bending is not mesoplate may have been the Pacific Northwest
known with certainty, and it remains possible of the contemporaneous United States, where
that the Chugach terrane originally filled an Palaeocene rocks are missing from the coastal
arcuate trench. The terrane in SE Alaska may area and the extensive north-trending late
have been partly a zone of right-lateral slip Mesozoic Franciscan assemblage and Great
rather than orthogonal subduction, as is the Valley sequence arc complex of California is
case at present. Palaeomagnetic data are re- abruptly truncated (Nilsen & McKee 1979).
quired from both limbs of the Chugach flysch Our thinking about the tectonic evolution of
terrane, SE and SW Alaska, to test for post- western North America must account for large-
depositional rotation of the limbs. scale translation of continental mesoplates,
Within the framework of available data, it which in the present case includes an outstand-
appears most reasonable that the Chugach ing example of a trench-fill sequence.
flysch terrane and its associated belts formed a
ACKNOWLEDGMENTS: We thank J. Decker, J. C.
large NE-SW-trending continental fragment Moore, and G. R. Winkler for helpful reviews of this
along the southern margin of which the Kula paper. Zuffa's work on rock-fragment petrography
plate was subducting during the Late Cre- was made possible by the contributions of thin sec-
taceous. Subsequent northward migration of tions by a large number of geologists, who are
the large block resulted in its probable post- individually acknowledged in Zuffa et al. (1980).

References
BEIKMAN, H. M. 1974a. Preliminary geologic map of east quadrant of Alaska. U.S. geol. Surv. Misc.
the southwest quadrant of Alaska. U.S. geol. Field Studies Map MF-612, scale 1:1,000,000.
Surv. Misc. Field Studies Map MF-611, scale -- 1975a. Preliminary geologic map of southeast-
1 : 1,000,000. ern Alaska. U.S. geol. Surv. Misc. Field Studies
1974b. Preliminary geologic map of the south- Map MF-673, scale 1:1,000,000.
 The Chugach Terrane, a Cretaceous trench-fill deposit 225

1975b. Preliminary geologic map of the Alaska H1LLHOUSE,J. W. & GROMME,C. S. 1977. Paleomagnetic
Peninsula and Aleutians: U.S. geol. Surv. Misc. poles from sheeted dikes and pillow basalt of the
Field Studies Map MF-674, scale 1:1,000,000. Valdez(?) and Orca Groups, southern Alaska.
BERG, H. C., JONES, D. L. & CONEY, P. J. 1978. Map LOS, 58, 1127.
showing pre-Cenozoic tectonostratigraphic ter- HUDSON, T. 1979a. Plutonism and regional geology in
ranes of southeastern Alaska and adjacent areas. southern Alaska. U.S. Circ. geol. Surv. 8 0 4 - B ,
Open-File Rep. U.S. geol. Surv. 78-1085. B86-8.
& RICHTER,D. H. 1972. Gravina-Nutzotin belt- 1979b. Mesozoicplutonicbeltsofsouthern Alaska.
tectonic significance of an upper Mesozoic Geology, 7, 230--4.
sedimentary and volcanic sequence in southern 1979c. Calc-alkaline plutonism along the Pacific
and southeastern Alaska. Prof. Pap. U.S. geol. rim of southern Alaska. Open-File Rep. U.S. geol.
Surv. 800-D, D1-24. Surv. 79-953, 31 pp.
BREW, D. A., JOHNSON, B. R., FORD, A. B. & - - & PLAEKER, G. 1979. Paleogene anatexis along
MORRELL, R. P. 1978. Intrusive rocks of the the Gulf of Alaska margin. Geology, 7, 573-7.
Fairweather Range, Glacier Bay National Monu- - - & PETERMAN, Z. E. 1977. Paleogene
ment, Alaska. Circ. U.S. geol. Surv. 772-B, anatexis in flyschoid rocks along the Gulf of Alaska
B88-90. margin. Abstr. Prog. geol. Soc. Am. 9, 439.
& MORRELL, R. P. 1979. Correlation of the JONES, D. L., BLAKE, M. C., BAILEY, E. H. &
Sitka Graywacke, unnamed rocks in the Fair- MCLAUGHLIN, R. J. 1978. Distribution and charac-
weather Range, and Valdez Group, southeastern ter of upper Mesozoic subduction complexes along
Alaska. Circ. U.S. geol. Surv. 804-B, B123-5. the west coast of North America. Tectonophysics,
BUDNIK, R. T. 1974a. The geologic history of the 47, 207-22.
Valdez Group, Kenai Peninsula, Alaska--deposi- - - & CLARK, S. H. B. 1973. Upper Cretaceous
tion and deformation at the Late Cretaceous (Maestrichtian) fossils from the Kenai-Chugach
consumptive plate margin. Thesis, PhD, Univ. Mountains, Kodiak and Shumagin Islands, south-
California at Los Angeles, 139 pp. ern Alaska. J. Res. U.S. geol. Surv. 1, 125-36.
- - 1974b. Deposition and deformation along an - - & DETrERMAN,R. L. 1966. Cretaceousstratigraphy
Upper Cretaceous consumptive plate margin, of the Kamishak Hills, Alaska Peninsula. Prof. Pap.
Kenai Peninsula, Alaska. Abstr. Prog. geol. Soc. U.S. geol. Surv. 500-D, D53-8.
Am. 6, 150. - - & S I L B E R L I N G , N. J. 1979. Mesozoic stratigraphy-
BURK, C. A. 1965. Geology of the Alaska Peninsula-- -the key to tectonic analysis of southern and central
island arc and continental margin. Mere. geol. Alaska. Open-File Rep. U.S. geol. Surv. 79-1200, 37
Soc. Am. 99, 250 pp. pP-
CLARK, S. H. B. 1972. Reconnaissance bedrock KARL, S., DECKER, J. & JONES, D. L. 1979. Early
geologic map of the Chugach Mountains near Cretaceous radiolarians from the McHugh Com-
Anchorage, Alaska. U.S. geol. Surv. Misc. Field plex, south-central Alaska. Circ. U.S. geol. Surv.
Studies Map MF-350, scale 1:250,000. 804-B, B88-90.
1973. The McHugh Complex of south-central LONEY, R. A., BREW, D. A., MUFFLER, L. J. P. &
Alaska. Bull. U.S. geol. Surv. 1372-D, DI-11. POMEROY, J. S. 1975. Reconnaissance geology of
CONNELLY, W. 1978. Uyak Complex, Kodiak Islands, Chichagof, Baranof, and Kruzof Islands, southeast-
Alaska: a Cretaceous subduction complex. Bull. ern Alaska. Prof. Pap. U.S. geol. Surv. 792, 105 pp.
geol. Soc. Am. 89, 755-69. MACKEVEIT,E. M., JR & PLAFKER,G. 1974. The Border
COOPER, A. K. & MARLOW, M. S. 1979. Geologic and Ranges fault in south-central Alaska. J. Res. U.S.
geophysical cruise across the outer Bering con- geol. Surv. 2, 323-9.
tinental margin. Circ. U.S. geol. Surv. 804-B, MCLEAN, H. 1979. Pribilof segment of the Bering Sea
132-4. continental margin: a reinterpretation of Upper
CSEJTEY, B., JR 1979. Regional significance of tecto- Cretaceous dredge samples. Geology, 7, 307-10.
nics of the Talkeetna Mountains, south-central MAGOON, L. B., EGBERT, R. M. & PETERING, G. 1978.
Alaska. Circ. U.S. geol. Surv. 804-B, B90-2. Upper Jurassic and Cretaceous rocks of the
DECKER, J., NILSEN, T. H. & KARL, S. 1979. Turbi- Kamishak Hills-Douglas River area, lower Cook
dite facies of the Sitka Graywacke, southeastern Inlet. Circ. U.S. geol. Surv. 772-B, B57-9.
Alaska. Circ. U.S. geol. Surv. 804-B, B125-9. --, GRIESBACK, F. B. & EGBERT, R. M. 1980.
DE~ERMAN, R. L. 1978. Interpretation of deposition- Nonmarine Upper Cretaceous rocks, Cook Inlet,
al environments in the Chignik Formation, Alas- Alaska. Bull. Am. Assoc. Petrol. Geol. 64, 1259-66.
ka Peninsula. Circ. U.S. geol. Surv. 772-B, B62-3. MANCINI, E. A., DEETER,T. M. & WINGATE,F. H. 1978.
DlCKINSOrq, W. R. & SUCZEK,C. R. 1979. Plate tectonics Upper Cretaceous arc trench gap sedimentation on
and sandstone composition. Bull. Am. Assoc. the Alaska Peninsula. Geology, 6, 437-9.
Petrol. Geol. 63, 2164-82. MITCHELL, P. A. 1979. Geology of the Hope-Sunrise
GRANTZ, A. 1964. Stratigraphic reconnaissance of the (gold) mining district, north-central Kenai Peninsu-
Matanuska Formation in the Matanuska Valley, la, Alaska. Thesis, MS, Univ. Stanford, 123 pp.
Alaska. Bull. U.S. geol. Surv. 1181-1, I1-33. -- 1980. Genesis of the Valdez Group south of
HILL, M. D. &MORRIS,J. D. 1977. Near-trenchplutonism Turnagain Arm, Alaska. Abstr. Prog. geol. Soc.
in southwestern Alaska. Abstr. Progr. geol. Soc. Am. 12, 142.
Am. 9, 436-7. MONGER, J. W . H . , SOUTHEY, J. G . & GABRIELSE, I--I.
226 T. H. N i l s e n & G. G. Z u f f a

1972. Evolution of the Canadian Cordillera. Saint Elias Mountains and Alexander Archipela-
Can. J. Earth Sci. 8, 259-78. go. Circ. U.S. geol. Surv. 722, 14-6.
MOORE, G. W. 1969. New formations on Kodiak and , JONES, D. L. & PESSAGNO, E. A., JR. 1977. A
adjacent islands, Alaska. Bull. U.S. geol. Surv. Cretaceous accretionary flysch and melange ter-
1274-A, A27-35. rane along the Gulf of Alaska margin. Circ. U.S.
-- & BOLM, J. B. 1977. Orientation of Late Cre- geol. Surv. 751-B, B41-3.
taceous and early Tertiary subduction, Kodiak REED, B. L. & LANPHERE, M. A. 1973. Alaska-
Islands, Alaska. Abstr. Prog. geol. Soc. Am. 9, Aleutian Range batholith: chronology, chemis-
1099-100. try, and relation to circum-Pacific plutonism.
MOORE, J. C. 1972. Uplifted trench sediments-- Bull. geol. Soc. Am. 84, 2583-610.
southwestern Alaska-Bering Shelf edge. Science, & -- 1974. Chemical variations across the
175, 1103-5. Alaska-Aleutian Range batholith. J, Res. U.S.
1973a. Cretaceous continental margin sedi- geol. Surv. 2, 343-52.
mentation, southwestern Alaska. Bull. geol. Soc. REID, M. & GLEE, J. B. 1980. Near trench volcanism,
Am. 84, 595-614. Kodiak Island, Alaska: Implications for ridge-
1973b. Complex deformation of Cretaceous trench encounter. Abstr. Prog. geol. Soc. Am.
trench deposits, southwestern Alaska. Bull. geol. 12, 148-9.
Soc. Am. 84, 2005-20. SCHOLL, D. W. & MARLOW, M. S. 1974. Sedimentary
1978. Orientation of underthrusting during sequences in modern Pacific trenches and the
latest Cretaceous and earliest Tertiary time, deformed circum-Pacific eugcosyncline. In:
Kodiak Islands, Alaska. Geology, 6, 209-13. D o n , R. H, & SttAVER, R. H. (eds). Modern and
-- & CONNELLY, W. 1979. Tectonic history of the Ancient Geosynclinal Sedimentation. Spec. Publ.
continental margin of southwestern Alaska, Late Soc. econ. Paleontol. Mineral. Tulsa, 19, 193-
Triassic to earliest Tertiary. In." SlSSON, A. (ed). 211.
The Relationship of Plate Tectonics to Alaskan SCnWEH,ER, W. J. & KULM, L. D. 1978. Depositional
Geology and Resources, H1-29. Alaska geol. patterns and channelized sedimentation in active
Soc. Symp. eastern Pacific trenches. In: SI'ANLEV, D. J. &
- & WHEELER, R. L. 1978. Structural fabric of a
- KELLING, G. (eds). Sedimentation in Submarine
melange, Kodiak Islands, Alaska. Am. J. Sci. Canyons, Fans and Trenches, 311-24. Dowden,
278, 739-65. Hutchinson & Ross, Stroudsburg, Pennsylvania.
McTrl, E. & R i c o LuccHI, F. 1972. Le torbiditi dell' STEWART, R. J. 1976. Turbidites of the Aleutian
Appennine settentrionale--introduzione all' Abyssal Plain--mineralogy, provenance, and
analisi di facies. Mere. Soc. geol. Ital. l l , 161-99. constraints for Cenozoic motion of the Pacific
- & -- - 1975. Turbidite facies and facies associa- plate. Bull. geol. Soc. Am. 87, 793-808.
tions. In: Examples of turbidite facies and facies STONE, D. B. & PACKER, D. R. 1976. Tectonic
associations from selected formations of the implications of Alaska Peninsula paleomagnetic
northern Apennines. Ninth Int. Congr. data. Tectonophysics, 37, 183-201.
Sedimentology, Field Trip Guidebook, A l l , 21- - - & 1979. Paleomagnetic data from the
36, Nice, France. Alaska Peninsula. Bull. geol. Soc. Am. 90, 545-
NILSEN, T. H. 1977. Turbidite facies and sedimenta- 60.
tion patterns. In: NILSEN, T. H. (ed). Late TVSDAL, R. G. & CASE, J. E. 1977. The McHugh
Mesozoic and Cenozoic Sedimentation and Tec- Complex in the Seward quadrangle, south-
tonics in California, Short Course, San Joaquin central Alaska. Circ. U.S. geol. Surv. 751-B,
geol. Soc. 39-52. B48-9.
- & BOUMA, A. H. 1977. Turbidite sedimentology
- , WINKLER, G. R. & CLARK, S. H. B.
and depositional framework of the Upper Cre- 1977. Sheeted dikes, gabbro, and pillow basalt in
taceous Kodiak Formation and related strati- flysch of coastal southern Alaska. Geology, 5,
graphic units, southern Alaska. Abstr. Prog. 377-83.
geol. Soc. Am. 9, 1115. & PLAFKER, G. 1978. Age and continuity of the
- & MCKEE, E. H. 1979. Paleogene paleogeogra-
- Valdez Group, southern Alaska. Bull. U.S. geol.
phy of the Western United States. Pacif. Coast Surv. 1457-A, A120-4.
Paleogeogr. Symp. Soc. econ. Paleontol. Miner- UNDERWOOD, M. B. & KARIG, D. E. 1980. Role of
al. 3, 257-76. submarine canyons in trench and trench-slope
- & MOORE, G. W. 1979. Reconnaissance study
- sedimentation. Geology, 8, 432-6.
of Upper Cretaceous to Miocene stratigraphic VON HUENE, R., MOORE, J. C. & MOORE, G. W. 1978.
units and sedimentary facies, Kodiak and adja- Cross section of Alaska Peninsula-Kodiak Is-
cent islands, Alaska. Prof. Pap. U.S. geol. Surv. land-Aleutian Trench. Map Chart Series, geol.
1093, 34 pp. Soc. Am. MC-28A.
PLAFKER, G. & CAMPBELL, R. B. 1979. The Border ZUFFA, G. G., NILSEN, T. H. & WINKLER, G. R.
Ranges fault in the Saint Elias Mountains. Circ. 1980. Rock-fragment petrography of the Upper
U.S. geol. Surv. 804-B, B102-4. Cretaceous Chugach terrane, southern Alaska.
, JONES, D. L., HUDSON, T. & BERG, H. C. Open-File Rep. U.S. geol. Surv. 80--713, 28 pp.
1976. The Border Ranges fault system in the
 The Chugach Terrane, a Cretaceous trench-fill deposit 227

TOR H. NILSEN, U.S. Geological Survey, Branch of Western Environmental

Geology, 345 Middlefield Road, Menlo Park, California 94025, U.S.A.
GIAN G. ZUFFA, Dipartimento di Scienze della Terra, Universita della Calabria,
87030 Castiglione Scalo (Cosenza), Italy.
 Structural evolution of coherent terranes in the Ghost Rocks
Formation, Kodiak Island, Alaska

Tim Byrne
SUMMARY: The Late Cretaceous to Palaeocene Ghost Rocks Formation, Kodiak
Island, Alaska, comprises two structural styles: multiply deformed m61ange and relatively
coherent terranes of turbidite deposits which together constitute a probable ancient
accretionary complex. The m61ange consists almost exclusively of orientated, elongate
sandstone boudins which lie in a foliated pelitic matrix. In the coherent terranes, structural
histories, rock types and hornfels metamorphism delineate two NE trending units: Unit A,
and a probably younger Unit B. Unit A can be subdivided into two gradational structural
belts, a seaward conjugate fold belt and a landward spaced cleavage belt. The conjugate
fold belt contains 'S' and 'Z' type kink folds that often intersect to form conjugate fold
sets. The folds of this fold belt are hornfelsed by a sill which intruded partially lithified
sediments; the fold belt therefore also formed when the sediments were partially lithified.
Analyses of the fold axial surfaces indicate a shortening direction of 319°, when later
folding is removed. Sill emplacement and hornfels metamorphism interrupted the
conjugate fold deformation and resulted in a change in style of deformation in Unit B to
broad, open to tight folds associated with a spaced cleavage. The cleavage strikes 046° and
dips approximately vertically, fanning divergently about fold axial surfaces. The approxi-
mately coaxial conjugate folding and folding with cleavage in unit A constitute D 1. Unit B
generally contains sub-horizontal beds except for a large-scale, overturned, thrusted fold,
which emplaced m61ange to the NW and above Unit B. Slickenlines, fold axes and poles to
bedding indicate this unit was shortened along a trend of 334°. This deformation is termed
D 2 and is correlated elsewhere in the Ghost Rocks Formation to thrust faults which place
m61ange to the NW and under the coherent terranes.
The structural histories of the coherent terranes in the Ghost Rocks Formations show
that accretionary prism deformation can involve coaxial deformations and the tectonic
deformation of partially lithified sediments. The structural histories also show that the
Palaeocene deformation axes in the Ghost Rocks Formation have not rotated out of their
original sub-horizontal orientations.

The structural evolution of accretionary ter- mildly metamorphosed and may record events
ranes presents an outstanding problem in con- associated with the later stages of Kula Plate
vergent margin geology. Structural models of subduction.
deformation and accretion at convergent mar-
gins based on marine geophysics (Seely et al.
1974; Karig & Sharman 1975), scale models Tectonic setting
(Cowan & Silting 1978; Seely 1977) and on-land
geological studies (Cowan 1974; Suppe 1973; The continental margin of SW Alaska has ex-
Moore 1973; Moore & Karig 1979) have been perienced repeated episodes of subduction-
enlightening but have been most successful in related magmatism and accretion since at least
showing the wide variety of unconstrained para- the early Mesozoic. While magmatism has been
meters. Moreover, neither geophysical profiles concentrated on the Alaskan Peninsula, and is
or scaled models are directly applicable to continuing today, accretion of deep-sea rocks
outcrop-scale problems. In order to understand has progressively broadened the continental
the geological consequences of plate converg- shelf to the SE. The Kodiak Islands constitute
ence and plate reorganizations, and to interpret an emergent portion of this continental shelf
ancient subduction complexes accurately, de- and expose NE-trending belts of deep-sea rocks
tailed structural studies and appropriate that dip N W and young progressively toward
kinematic interpretations of ancient accretion- the present day Aleutian trench (Fig. 1; Moore
ary terranes are essential. The Late Cretaceous & Connelly 1979; Moore 1969). The Ghost
to early Palaeocene Ghost Rocks Formation, Rocks Formation ( A r m e n t r o u t 1979a; Moore
Kodiak Islands, Alaska, provides an excellent 1969) comprises the next to youngest sequence
opportunity to resolve convergent margin pro- of deep-sea deposits in the Kodiak Islands and
cesses as it is well-exposed, young in age, only is in fault contact with older rocks to the N W

229
230 T. Byrne

JAP ~
t AREA

0 I0 50 km

× SCALE

~ALASKA
EXPLANATION
AGE SEDIMENTARYROCKS IGNEOUS ROCKS
MIOCENE ~ NARROWCAPE FM
EOCENE ~ SITKALIDAK FM
PALEOCENE ~ GRANITIC INTRUSIVE
ond ~] GHOSTROCKS FM. ROCKS
CRETACEOUS
CRETACEOUS I~] KODIAKFM.

/ /' \

Fro. 1. Generalized geological map of SE Kodiak Island, Alaska, after Moore (1967).

and younger rocks to the SE. While the Kodiak shales indicate turbidite deposition and the
margin may have lain south of its present occurrence of massive sandstones, pebbly mud-
position (Stone & Packer 1979) at the time stones and conglomerates suggest deposition in
when these rocks were deformed, geochemical channelized environments.
trends (Reed & Lanphere 1974) in the Alaska- Fossiliferous limestones are locally associated
Aleutian Range batholith clearly indicate plate with the igneous rocks in the Ghost Rocks
convergence was toward the NW during the Formation. The igneous rocks are clearly inter-
time of deposition and accretion of the Ghost bedded with the sedimentary rocks and include
Rocks Formation. pillow lava, pillow breccia, tuff, dykes and sills
of andesitic to basaltic composition (Reid &
Gill 1980; Hill 1978). The limestones typically
overlie or are interbedded with the volcanic
rocks and contain planktonic foraminifera of
The Ghost Rocks Formation
Palaeocene age at some localities, and of Late
The Ghost Rocks Formation consists of com- Cretaceous age at others.
plexly deformed sandstone and shale with inter- Silicic igneous rocks also intrude the Ghost
bedded volcanic rocks and scattered hypabyssal Rocks Formation. These consist of quartz diori-
intrusions. The sedimentary rocks are predomi- tic to tonalitic plutons that probably represent
nantly thin (1-5 cm) to thick (tens of metres) satellitic intrusions of the Kodiak batholith.
bedded, mostly massive sandstones with inter- The plutons are associated with zones of iron-
bedded shales of similar thicknesses. Pebbly staining and zeolitization, are chilled at their
mudstone and conglomerate also occur locally. margins, and have thermally altered the
Although tectonic deformation has severely sedimentary rocks of the Ghost Rocks Forma-
dismembered almost all original stratigraphy, tion. The pluton exposed NW of Jap Bay (Fig.
sedimentary structures in the sandstones and 2) is similar to other early Tertiary intrusive
 Structural evolution of coherent terranes in Alaska 231

rocks of the Ghost Rocks Formation and has included blocks are unresolvable at a mapping
been isotopically dated by the Mobil Oil Com- scale of 1:25 000. Sandstone and argillite are
pany. The ages of this pluton (62.6 + 0.6, 62.1 the dominant rock types but conglomerate,
+ 0.6: K-Ar, biotite, and 63 + 3: Rb-Sr limestone, greenstone and tuffaceous units also
isochron defined by biotite and plagioclase, occur, both interlayered in the matrix and as
Armentrout, pers. comm. 1978) are slightly inclusions. On a mesoscopic scale the foliation
older but not greatly different from the 57 to occurs only in the pelitic lithologies and distinc-
60 Ma ages determined from the main body of tively anastomoses around the more competent
the Kodiak batholith (Armentrout 1979b, and sandstone and greenstone lithologies, resulting
written comm. 1979) and probably record the in a 'phacoids in a matrix' appearance at most
time of pluton emplacement. Locally, the outcrops.
Ghost Rocks pluton is cross-cut by discrete The limestones in the m61ange terrane con-
zones of low grade alteration and cataclasis tain planktonic foraminifera of Late Cretaceous
suggesting that some deformation post-dates age (Poore, written comm. 1980; Rau, written
the emplacement of the pluton. comm. 1980; see section B-B' Fig. 3 for loca-
The Ghost Rocks Formation crops out in a tion) and are intimately associated with the
belt 160 km long and 15 km wide along the SE volcanic rocks, suggesting that the volcanic
side of the Kodiak Islands (Fig. 1). During rocks are of Late Cretaceous age, also.
three seasons of fieldwork, I examined most Detailed mapping of the m61ange (1:300) in
exposures of this formation. The conclusions the vicinity of Jap Bay shows that the steep,
presented here, however, are derived from a NW-dipping foliation crosscuts and locally
detailed study in the Jap Bay area and regional transposes an earlier foliation. The earlier folia-
studies of the bays and headlands to the NW tion is defined by the parallel alignment of
(Fig. 2). elongate sandstone boudins that plunge gently
SW. The intersection of the two foliations
locally produces a pencil cleavage that plunges
Jap Bay area gently to the SW, parallel to the sandstone
boudins.
Detailed mapping in the vicinities of Jap, The remaining area, designated as undiffer-
Kiavak and Kaiugnak Bays (Fig. 2) reveals the entiated (Fig. 2), was not mapped in detail but
presence of two dominant structural styles with- reconnaissance mapping shows it contains both
in the Ghost Rocks Formation. The structural m61ange (as described above) and relatively
styles are broadly differentiated as m61ange and more coherent, but still unresolvable, units of
coherent terranes, which occur as mappable interbedded sandstone and pelite with minor
belts that strike NE, parallel to the structural greenstone and conglomerate (section C-C'
grain of Kodiak Island. The coherent terranes Fig. 5). These relative coherent areas are litho-
are sedimentary deposits that may be analogous logically indistinguishable from unit A in the
to the slope basin deposits recognized in many coherent terranes.
modern convergent margins. As such, the de-
formation history of the coherent terranes re-
cord not only the kinematics of specific events Coherent terranes
in the history of the accretionary prism but
perhaps also events present in the more com- The structure of the coherent terranes contrasts
plexly deformed m61ange terranes. The fun- strikingly with that of the m61ange terranes.
damental contribution of this study is the con- The coherent terranes are characterized by
clusion that the deformation axes in the cohe- gently to tightly folded turbidite deposits.
rent terranes have not rotated from their origin- Pinch-and-swell structures and boudinage are
al sub-horizontal orientation since Palaeocene rare and individual sandstone beds can be
time. Systematic imbrication and underthrust- traced for tens of metres. Tectonic cleavage,
ing of younger more seaward deep-sea rocks is where present, penetrates both the pelitic and
not indicated. sandstone lithologies.
Critical to understanding the tectonic signifi-
cance of the coherent terranes is knowing their
M~lange terranes relationship to the underlying m61ange. Studies
of modern convergent margins suggest that
Areas designated as m61ange (Fig. 2) are char- relatively less-deformed slope sediments be-
acterized by strongly foliated and discontinuous come incorporated into more deformed,
argillite-sandstone units in which structures and accreted trench deposits through time, resulting
232 T. Byrne

A'~x'~ C'
H~V
KAIUGNAK BAY

ii,,r,-
/ ~ ' /
-" . .> ">.5.:..;% ..>-"
- " .-" I " :i ~
:
.~:i:i:i:i:i:i:i C

.- .- >~ " 5 .-I-" . ".~" .. i ..i . ~ i " ,f,:i!:i:irii!!!:ip>- \

~ ... --. -I-' - - -- .i. . -... .--1.::.:.:.:,:.:.:.:..

/..-.I. -..I / . - . i . ./ . .i i...::::;:.

...... - f ....-if...--,.. /. ... - - . j J ~ - - ~ KIAVAK \

.I.. • ..... " / . ~..'." .~- . ." "./'. -'. ~ J( BAY

" ..--".-"" "---./'~ I." .-i -I ..--~ .... . - ~ / IX.
.... ~...--f . . .
~i..-
- .. . . ....-
. ~ . d,..< - / ~"-"
: ~ ""- .-.-~ .......
y -->.... <. -. ;.......
,y. ;,.-"
%~j.;;; "~';~.
" I " ...i-" .;.i.... - ; -.-;~..- ~'..../--;
..~.--...:~--.-. / . I ...-I.. .. ~ - . - .. -- ... ......-'~ ..
"; .~.,:'-:::
" ~......--..~ .-I ..I .... ./.; .- . . .. -:>-.-~..-. ...I.: ~f::::::::::!:
~:,-........_.d-:~',-~,~~.......... >"-" .--"!../. - - ; i";":..../."';'"""~-.;~Y..?!:i:.::::i:i:i:i:i:!:....
:i:!:!:::::~.-'-/ - ....
===================================

,- /- ,-

.',)l'-'~I...
''''~ ~~..~.
.....
" " *""~-d,....
~..~.--"
--"
..... - .... ~'. I-'.. " - 7 .
... i.-.I . . ..~ ..#- . ~
I-H
....,, ,. - i.,..I..

) N
============================================J A P BAY B'

iJSiiiiJ!JiiJZ!iJiiijJfi!
v . . \l. iiiil
0 3
| I
KM

Fi6. 2.(A) Geological map and (B) cross-section of the Jap Bay area, generalized from coastline
(excellent exposure) and inland (poor exposure) traverses.

in a transition from more-deformed to less- studies (see below) indicate, however, that the
deformed sedimentary sequences (Lundberg & thrust faults are some of the latest structural
Moore 1981). In the Jap Bay area the contacts features in the Ghost Rocks Formation. Thus,
between the m61ange and coherent terranes, no direct evidence as to the primary rela-
where they are exposed, are either moderately tionship between the coherent and m61ange
dipping thrust faults or the relationships are terranes is available. To help resolve this prob-
ambiguous and the contacts could be deformed lem a detailed study of the structural history of
depositional or tectonic boundaries. Structural the m61ange terranes is in preparation so that it
 Structural evolution of coherent terranes in Alaska 233

B EXPLANATION

CONTACTS: ___...- .--- / PRIMARY ..-----_.. --" HIGH ANGLE FAULT

(dashed where
approximate)
t ll~ ~~ m ~ GRADATtONAL .~..~j~. ~ T H R U S T FAULT, T E E T H
ON UPPER PLATE

EOCENE ~ SITKALIDAK FM
PALEOCENE GHOST ROCKS FM.
end
CRETACEOUS ~ UNiT B ~ GRANITIC ROCKS
UNIT A
(CONJUGATE FOLDS) MAFIC IGNEOUS ROCKS

UNIT A ~ HYPABYSSAL
(SPACED CLEAVAGE ) ~'-H

MELANGE ~vVOLCANtC
UNDIFFERENTIATED
CRETACEOUS I~ KODIAK FM. 0 1.5 3
I- K'M I
cross- SECTION SCALE

~ ",. "(-- UNIT A "--),_ "+UNIT B l _, UNIT A I .,,

A ,,~, ~, A I A

~k~}~//]~//;~°~k\is,,'~14i~@/~/(I~t( ~g~g~(~(i~x\\~:xx~`4~*,~
V H V V I ' VH

FiG. 2.(B)

can be compared with the structural histories of

Unit A
the coherent terranes.
The coherent terranes can be subdivided into Unit A contains a wide variety of rock types
two stratigraphic units based on contrasting including conglomerate, pebbly mudstone,
lithologies and structural and metamorphic thick bedded sandstone (tens of metres thick),
histories; Unit A and a relatively younger Unit B. thin bedded sandstone (less than 10 cm thick),

8 /
STRIKE AND DIP OF BEDDING
...(" STRIKE AND DIP OF CLEAVAGE
83 , ~ X STRIKEAND DiP OPMELANGE . ,~"/i~7~..-f.;.
vv/ ~ FOLIATION ;T~'~ .rC~.~~C

B, ~'d.'

58 67 ~,t

( SPACED CLEAVAGE
B . . . . C.F. ~ fl'

11/I / I' i ~ I',P'--t41" i i I" iA.l /" .-'...~ ----~ ~ --- 1"#.-- ~l~l,n,i~'.qrl,;I
l;'' /,/ , ~,..../ .,/.," . ..-- . I....--f-~,..~--~ ~ ~-Z].~,i~)~d;~.~,l
i i , \."--"-'/i /- ..// / / / / .--~ ~-'~=~--G~ /l'T/l~.,~l/v,~j,,tl

FI6.3. Geological map and cross-sectionof Unit A. See Fig. 2 for location and symbols. Fold axes
were determined from more data than shown in maps. In cross-section long and short lines represent
apparent dip directions of bedding and cleavage, 'C' shows late Cretaceous fossil locality and 'CF'
shows extent of conjugate fold belt.
234 T. Byrne
massive argillite and interbedded extrusive and interaction of the sill with partially lithified
hypabyssal igneous rocks. Overall, the litholo- sediments. These field relations indicate that
gical character of Unit A is strikingly similar to hornfelsing apparently accentuated the dia-
that of the m61ange and undifferentiated ter- genetic processes that cause lithification. A
ranes, and the gross lithologies alone cannot be change in the degree of lithification of the
used to distinguish these terranes. Areas of sediments after sill emplacement is also re-
Unit A therefore may represent parts of a flected by a change in structural styles.
coherent terrane that were folded or thrust into Deformation styles within Unit A define two
the m61ange, relatively less deformed remnants mappable structural belts that grade into each
of an originally coherent stratigraphic sequ- other, a seaward belt characterized by conju-
ence, or both. gate folds, and a landward belt characterized by
Unit A appears to be early Palaeocene in a spaced cleavage (Figs 2 & 3). The structures
age. This unit is the most extensive coherent of the conjugate fold belt are hornfelsed by the
terrane exposed in the Jap Bay area and is andesite sill exposed at the mouth of Jap Bay.
lithologically and structurally similar to rocks Thus, this fold belt also is interpreted to have
that crop out locally along the entire length of developed when the sediments were partially
the Ghost Rocks Formation. These similar and lithified. Both the sill and the conjugate folds
probably correlative rocks contain Palaeocene are locally cross-cut by the cleavage of the
fossils at two localities (Nilsen & Moore 1979; spaced cleavage belt. Unit A therefore was
Poore, written comm. 1978), suggesting that deformed by two modes of deformation, conju-
Unit A may be Palaeocene in age also. Thus, if gate folding in partially lithified sediments and
Unit A is gradational with the m61ange, the spaced cleavage development in lithified sedi-
original sedimentary sequence extended from ments. Together, these structural belts docu-
Late Cretaceous to early Palaeocene. ment a progressive deformation in the Jap Bay
All the sedimentary rocks in Unit A have area that I have termed DI.
been mildly recrystallized by a hornfels meta-
morphism related to emplacement of the Conjugatefold deformation
hypabyssal rocks. The recrystallization is most The folds of the conjugate fold belt occur in a
intense 1 to 2 m below and above the andesitic wide variety of styles and forms. Typically, they
sill that crops out at the mouth of Jap Bay (Figs are disharmonically and/or polyclinally folded
2 & 3). Here, the pelites and sandstones are and locally sandstone layers are faulted,
very well indurated, and break conchoidally. At boudinaged or show pinch-and-swell structures
greater distances from this sill, a tectonic cleav- (Fig. 4). The mesoscopic scale disruption of the
age in the pelites cross-cuts the hornfels texture folded layers is consistent with the deformation
but the indurated texture indicative of recrystal- of relatively weak, partially lithified sediments
lization is still present in the pelites and sand- as indicated by the intrusive character of the
stones. In thin section, the recrystallization is andesite sill. The folds are not chaotic, howev-
recorded by fine-grained epidote and chlorite er, and generally f o r m ' S ' or 'Z' type asymmet-
porphyroblasts, patches of stilpnomelane, and ric fold sets which locally intersect to form
veins of prehnite and quartz. Near the contact, conjugate folds (Fig. 4A & B). Symmetrical
the epidote porphyroblasts are especially con- folds are also common (Fig. 4C).
spicuous in thin section because they occur in The direction of maximum shortening during
pelitic layers as large subhedral crystals. deformation of the conjugate fold belt can be
The andesitic sill appears to have intruded determined from either individual conjugate
partially lithified sediments. In most areas the folds or from the average orientation of the 'S'
upper and lower contacts between the sill and and 'Z' asymmetric fold sets (Gay & Weiss
the sedimentary rocks are sharp and regular. 1974; Tobisch & Fiske 1976). The individual
Locally, however, the contact consists of a conjugate folds yield a shortening direction that
several metre thick zone of angular to rounded is subparallel to bedding and trends 326 ° when
greenstone fragments in a pelitic matrix. The later folding is removed. The extension direc-
fragments typically have chilled margins and tion is orientated normal to bedding. The
vary from a few centimetres to several metres in shortening direction in these folds bisects either
size. Pelitic dykes usually less than 5 mm thick the acute (e.g. Fig. 4B) or obtuse (e.g. Fig. 4A)
also commonly penetrate both the greenstone angle formed between the fold axial surfaces,
fragments and the sill for several tens of cen- although the obtuse bisectrix is more common.
timetres. The fragmentation and dyking are The average orientation of the 'S' and 'Z' fold
interpreted as being results of explosive brec- sets yields a shortening direction that is sub-
ciation and steam-mud injections related to parallel to bedding and trends 319 °, when later
 Structural evolution o f coherent terranes in Alaska 235
7

B) _ -.--. _ .

t
cM

FIG. 4. Sketches from photographs of conjugate (A and B) and symmetrical (C) folds that formed in
partially lithified sediments in Unit A. Lines within sand layers indicate traces of bedding. Abundant
quartz filled fractures in sand layers are not shown. (D) Sketch of spaced cleavage cross-cutting axial
surface of asymmetric fold of conjugate fold belt. Bedding is shown right side up in all folds and fold
axes and cleavage surfaces are approximately normal to plane of sketches.

folding is removed. The extension direction is An important problem in interpreting the

normal to bedding. These results are in good tectonic significance of this fold belt is defining
agreement with the results from the individual its possible origin; did the folds form from
conjugate folds and indicates that both types of gravity-induced slumping or from tectonic pro-
data can be used to determine the direction of cesses? Although there is no direct way of
maximum shortening. These combined data answering this question, three field observa-
(Fig. 6A) show that on a regional scale the tions suggest the folds are of tectonic origin.
shortening direction trended sub-horizontal First, the northern contact of the conjugate fold
about 319 ° when the conjugate fold belt was belt is gradational. The asymmetric folds grade
deformed. In rotating the fold axial surfaces laterally into undeformed rocks of approxi-
around the later fold axes, I assume that bed- mately the same stratigraphic position. There is
ding was horizontal prior to conjugate folding. no discontinuity suggestive of a zone of slump
This appears to be a valid assumption because folds. Moreover, there is no evidence within the
the sediments were only partially lithified when belt to suggest that the folds were surficial
they were deformed and there is no evidence of features. Secondly, the development of the
an earlier deformation. This assumption is also folds post-dates both the compaction of the
supported by the consistency in the orientation sediments (i.e. post-dates 50% volume loss by
of the axial surfaces after they have been ro- dewatering) and the formation of calcareous
tated so that bedding is horizontal (Fig. 6A). concentrations. Thus, if the folds formed during
236 T. Byr?le
gravitational sliding, the slumped unit was of not geometrically analogous to many slaty
sufficient size to cut deeply into a compacted cleavages that parallel axial surfaces of folds
and partially lithified sequence. Finally, the and define the plane of maximum flattening
conjugate fold deformation is nearly coaxial (Wood 1974; Ramsay 1967), the average
with a later deformation that is clearly of tecto- orientation of the spaced divergent cleavage
nic origin. The simplest interpretation for this parallels related fold axial surfaces and prob-
belt is that it records the initial stages of a ably also defines the plane of maximum
tectonic deformation that produced, after sill shortening in the cleavage belt. The direction of
emplacement and lithification, broad folds maximum shortening indicated by the cleavage
associated with a spaced cleavage (see below). trends sub-horizontally 336° (Fig. 6B). The
direction of extension during cleavage develop-
Spaced cleavage deformation ment, as indicated by quartz fibres around
The spaced cleavage belt of Unit A is char- pyrite framboids and crystals, has a sub-vertical
acterized by broad (hundreds of metres) open orientation. These axes of the deformation are
to tight folds and a 5 mm spaced cleavage best coaxial with the directions of shortening and
developed in pelitic layers (Fig. 3). The cleav- extension determined from the conjugate fold
age typically fans divergently (Hobbs et al. belt.
1976) around the axial surfaces of related Thus, D1 is interpreted to have been a coa-
mesoscopic and macroscopic folds although xial, progressive deformation that started with
areas of axial planar cleavage are also locally the localized development of mesoscopic-scale
present. On a regional scale the cleavage has an conjugate folds in partially lithified sediments
average sub-vertical orientation, striking 046 ° (Fig. 7A). Sill emplacement apparently lithified
(Fig. 6B). Near the conjugate fold belt, the the sediments and interrupted this deformation,
spaced cleavage cross-cuts the axial surfaces of after which the mode of deformation consisted
the conjugate folds (Fig. 4D) indicating it post- of open to tight large-scale folds with an associ-
dates these structures. ated divergent spaced cleavage (Fig. 7B). Dur-
On a microscopic scale, the spaced cleavage ing both phases of Dt the shortening axes were
occurs in two gradational modes: (1) as local- sub-horizontal and trended about 318°. Signifi-
ized zones of concentrated, preferentially cantly, the shortening axis of the latest phase of
orientated phyllosilicates, and (2) as a pervasive this deformation is still horizontal.
but poorly developed fabric defined by the
parallel alignment of fine-grained phyllosili-
Unit B
cares. The phyllosilicates that define both
modes of cleavage are very fine-grained (less Unit B rock types and structural styles are
than a few microns) relative to the phyllosili- unique in the Jap Bay area and in the Ghost
cates that parallel bedding and occur as indi- Rocks Formation. The belt consists of medium
vidual grains and composite stringers. The bedded (10-40 cm thick beds) sandstone inter-
cleavage trace in the phyllosilicates parallels the bedded with similar thicknesses of shale. The
trace of the tectonic cleavage. Clusters and sandstones and shales are rhythmically bedded
trains of opaques also define both fabrics. The and sole markings and graded beds indicate
cleavage folia often cross-cut or abut against the turbidite deposition. Stratigraphic sections
conspicuously large phyllosilicates that parallel show no systematic cycles, however. Distinc-
bedding. This relationship and the strong dis- tively, this unit is not hornfelsed and contains
parity in size between bedding and cleavage only rare thick-bedded (greater than 1 m thick)
phyllosilicates indicate that the cleavage phyllo- sandstones; conglomerate, pebbly mudstone
silicates are not reorientated detrital grains. and igneous rocks are absent. Moreover, Unit
Instead, they have apparently grown at the B was apparently not deformed by the multiple
expense of the detrital micas through dissolu- deformations of D1. Thus, although no basal
tion-neocrystallization processes. Evidence for contacts were observed, the contrasting litholo-
dissolution of detrital grains is also shown by gies and structural and metamorphic histories
the sutured contacts between well-rounded suggest that Unit B was deposited after de-
chert pebbles in conglomeratic units. Not all of formation and metamorphism of Unit A.
the strain was accommodated through dissolu- Although Unit B has not been metamorph-
tion-neocrystallization, however, as many of osed, it has been thermally altered. Seven shale
the detrital quartz grains are also undulatory samples from this unit, examined by Mobil Oil
indicating internal deformation of the quartz Company, indicate a degree of organic matura-
grains. tion (average R0 = 2.15) 3 to 4 times higher
Although the spaced, divergent cleavage is than that recognized in the Eocene Sitkalidak
 Structural evolution of coherent terranes in Alaska 237

Formation (Moore & Allwardt 1981). Con- Rare SW dipping faults show evidence of NW
sidering the age and tectonic setting of the overthrusting of m61ange. A stereographic plot
Ghost Rocks Formation sediments, the high of slickenlines (Fig. 6D), commonly present on
reflectance values are explained most simply by both sets of fault surfaces, indicates a shorten-
the thermal alteration associated with the em- ing direction of 336 °. M61ange is also thrusted
placement of the early Palaeocene pluton ex- below Unit A across Jap Bay (Fig. 2) but no
posed to the NE of Jap Bay, suggesting that kinematic data were obtained from this fault.
Unit B is of earliest Palaeocene age. However, minor thrust faults occur within Unit
A in this area and show north-westerly dips and
Structural geology slickenline trends parallel to the shortening
Unit B is generally deformed into outcrop- direction recognized in Unit B.
scale asymmetric folds displaying landward or The timing of D2 deformation is ambiguous
seaward vergence. Bedding-parallel slicken- but landward-verging structures like those in
sides indicate that folding occurred through Unit B have been documented elsewhere in the
flexural slip and that the direction of slip was Kodiak Islands only in the Eocene Sitkalidak
normal to the fold axes. Stereographic plots of Formation. Moore & Allwardt (1981) show that
fold axes and poles to bedding indicate that the the Sitkalidak Formation on Sitkinak Island
direction of maximum shortening was orien- represents a thick, obducted trench fill sequ-
tated about 334 ° (Fig. 6C). This direction of ence that was accreted to the continental mar-
shortening is different from that of D1 and, gin in late Eocene time. Thus, the landward-
along with the contrasting tithologies and meta- verging structures of D2 in the Ghost Rocks
morphic histories, suggests that Unit B records Formation may be a local response within this
a separate deformation, D2, younger than D1. formation to obduction and accretion of the
On the west side of Jap Bay, D2 is repre- Sitkalidak Formation.
sented in part by a large-scale thrusted fold
which placed m61ange to the NW (landward)
and above Unit B (section D-D', Fig. 5). Structural synthesis and discussion
Smaller-scale landward-verging asymmetric
folds, unrelated to the larger fold, are also The coherent terranes in the Ghost Rocks
present. The north-eastern end of Unit B on Formation in the area of Jap Bay record two
this side of Jap Bay (not shown in Fig. 5) is major tectonic deformations, D I and D 2 (Fig.
poorly exposed but the limited exposures sug- 7). D 1 is recorded in Unit A and was initiated in
gest that this area is more complexly deformed partially lithified sediments. After sill emplace-
than the area of Fig. 5. ment, D~ progressed coaxially into a deforma-
On the east side of Jap Bay, beds of Unit B tion that involved lithified sediments. D2 caused
dip homoclinally 40-45 ° to the NW and are large-scale folding in Unit B and thrust m61ange
locally folded by small-scale asymmetric folds to the NW above and below Unit B and below
that consistently verge to the SE (seaward). No Unit A.
evidence of NW vergence and overthrusting of Critical to understanding the tectonic signifi-
m61ange was observed on the east side of Jap cance of the coherent terranes is knowing the
Bay. Thus, Jap Bay appears to d e f n e , and may tectonic setting in which they were deposited
be the topographic expression of, a north- and deformed: do the terranes represent
trending tear fault separating NW verging struc- accreted trench or slope basin deposits? Unit B
tures on the west from SE verging structures on is lithologically, metamorphically and structur-
the east. ally unique relative to other units in the Ghost
Elsewhere in the Jap Bay area, D2 is repre- Rocks Formation and appears to be the
sented by thrust faults (Fig. 7C). Near the youngest unit in this formation, suggesting that
mouth of Jap Bay, the contact between it is a slope basin deposit (Fig. 7B). A more
m61ange and Unit A is a moderately dipping seaward, trench environment is possible but
thrust fault which places m61ange to the NW unlikely because there is no evidence in the
and below Unit A (Figs 2 & 3). In the m61ange deformation histories of the coherent terranes
nearest the thrust contact (less than 0-15 m that indicates Unit B was thrust into its present
from the contact) subsidiary thrust faults perva- position.
sively cross-cut the m61ange fabric. The faults The structural history of Unit A suggests it
are sharp planar discontinuities generally was also deposited in a slope basin environment
spaced about 10 cm apart. Drag of the m61ange (Fig. 7A). Recent studies of modern conver-
fabric along the faults indicates underthrusting gent margins suggest that trench sediments
of m61ange on all of the NW dipping faults. become imbricated and tilted landward as they
238 T. Byrne

26 ~ ~ ~METERS

\20
" 0 60 .~

20

O /7/ D'
i {l~,i s/71/// I _z
/ ~-~-~ ~;.f ~ _----- . ~ - . ~ . - ~ ~ . ~ ~..~---_x. ~ - ..- ~__~ ~ ~ "><'~|

Fro. 5. Geological maps and cross-sections of undifferentiated terrain and Unit A (section C-C') and
Unit B (section D-D'). See Figs 2 and 3 for locations and symbols. Small-scale folds in D-D' are
exaggerated for clarity. 'C' in section C-C' shows coherent areas in undifferentiated terrane (see
text).

are deformed (Moore et al. 1979). In Unit A of from studies of modern convergent margins and
the Ghost Rocks Formation the sub-horizontal Unit A may represent the remnants of complex-
deformation axes that d e f o r m e d this unit are ly deformed trench deposits. A detailed study
still in their sub-horizontal orientation. This of the deformation history of the m61ange ter-
unit apparently was not imbricated and tilted ranes (in preparation) could resolve this prob-
landward during deformation as would be lem. In either case, the deformation history of
expected in a trench environment. Unit A is Unit A is significant because it indicates that
therefore most simply interpreted, as a slope uplift of the Ghost Rocks Formation occurred
basin deposit. Alternatively, trench environ- without imbrication and landward tilting of
ments may be more complicated than inferred younger more seaward deposits (Fig. 8).

Fro. 6. Stereographic plots and contour diagrams for D] (A and B) and D 2 (C and D). A and B
respectively show poles to axial surfaces in conjugate fold belt and poles to cleavage in spaced
cleavage belt. Closed circle and square indicate directions of maximum shortening and extension,
respectively. (C) Poles to bedding and fold axes (shown as open circles) in Unit B. (D) Slickenlines
(arrows) and poles to fault planes (+) of thrust faults subsidiary to large thrust fault that separates
Unit A from m61ange. SE-plunging slickenlines are on SE dipping fault planes.
 Structural evolution of coherent terranes in Alaska 239
N N

41 °
44 °

tl

per i%

A) CONJUGATE FOLDS N SPACED CLEAVAGE N

N N
26° j _ ~ 24 °

rl% I%

POLES TO N D) FAULT SLICKS AND N

PLANES f
BEDDIN

,t ,t

-4-

N=IOI N:I8 < I / N=32

240 T. Byrne
~'ozcqN/~ ##(C.

SE. ......... J \ ..... N~rV"

( CONJUGATE.~
A) FOLDS
UNIT A

DI
~1 SPACED CLEAVAGE

SILL
UNIT B
B)
A "'.- , ,
f .1
,A

C)

D2

FIG. 7. Interpretative deformation history of the coherent terranes in the Ghost Rocks Formation.
Unit A is inferred to grade into more complexly deformed units but not necessarily into m61ange
with depth (see text). (A) Sub-horizontal shortening in Unit A causes development of conjugate fold
belt in partially lithified sediments. Size and extent of conjugate folds are exaggerated for clarity. (B)
Unit A is intruded and hornfelsed by a shallow level sill which accentuated lithification. This resulted
in a change in style of deformation during continued sub-horizontal shortening. Broad folds with an
associated cleavage cross-cut the sill and the conjugate folds. (C) D 2 resulted in NW-directed
overthrusting and underthrusting in Units B and A, respectively. (NW-directed underthrusting also
occurred in Unit B but is not shown in this diagram.)

KODIAK
ISLAND
Arc Trench 4 I>

.-l>j-~<~-

LANDWARD TILT"I NG VERTICAL UPLIFT

(A) (B)
 Structural evolution of coherent terranes in Alaska 241

ACKNOWLEDGMENTS" This paper is part of a PhD provided comments on early drafts of this paper.
thesis completed at the University of California, John Armentrout and Mobil Oil Company graciously
Santa Cruz, and I greatly appreciate the stimulating provided the isotopic ages, the vitrinite reflectance
environment the people there provided. Mary Reid's measurements and first introduced me to the geology
dedicated assistance and encouragement both in and of Kodiak Island. National science Foundation grant
out of the field, and Casey Moore's inspiring enthu- E A R 77-07836, Mobil Oil Corporation, Exxon
siasm played significant roles in completing this pap- Company, Union Oil, Amoco Production, U.S. Geo-
er. Gene Gonzales still makes better thin sections logical Survey and the U.S. Coast Guard provided
than I, and it is greatly appreciated. Casey Moore, financial support.
Othmar Tobisch, Ed Beutner and George Moore

References
ARMENTROUT, J. 1979a. Cenozoic stratigraphy, MOORE, G. F. & KARIG, D. E. 1979. Structural
Kodiak Island archipelago, Alaska. Abstr. geol. geology of Nias Island, Indonesia: implications
Soc. Am. Progr. 11, 66. for subduction zone tectonics. Am. J. Sci. 280,
1979b. Paleocene plutonism in southern Alaska 193-223.
(abstr.). Proc. 14th Pacif. Sci. Congr. Khabar- MOORE, G. W. 1969. New Formations of Kodiak and
ovsk, U.S.S.R. adjacent islands, Alaska. Bull. U.S. geol. Surv.
BACHMAN, S. B. 1978. A Cretaceous and early Terti- 1274-A, 428-35.
ary subduction complex, Menodocino coast, MOORE, J. C. 1973. Complex deformation of Cre-
northern California. In: HOWELL, D. G. & taceous trench deposits, southwestern Alaska.
McDOUGALL, K. A. (eds). Mesozoic Bull. geol. Soc. Am. 84, 2005-20.
Paleogeography of the Western United States, -- & ALLWARDT, A. 1981. Progressive deforma-
419-29. Pacif. Sec., Soc. Econ. Paleontol. tion of a Tertiary trench slope, Kodiak Islands,
Mineral. Tulsa. Alaska. J. geophys. Res. (in press).
COWAN, D. S. 1974. Deformation and metamorphism -- & CONNELLY, W. 1979. Tectonic history of the
of the Franciscan Subduction Zone Complex continental margin of southwestern Alaska: Late
northwest of Pacheco Pass, California. Bull. Triassic to earliest Tertiary. In: The Relationship
geol. Soc. Am. 85, 1623-34. of Plate Tectonics to Alaska Geology and Re-
--& SILLING, R. M. 1978. A dynamic, scaled sources, H1-29, Alaska Geol. Soc., Anchorage,
model of accretion at trenches and its implica- Alaska.
tions for the tectonic evolution of subduction -- et al. 1979. Progressive accretion in the Middle
complexes. J. geophys. Res. 83, 5389-96. America Trench, southern Mexico. Nature, Lon-
GAY, N. & WEISS, L. 1974. The relationship between don, 281,638-42.
principle stress directions and the geometry of NILSEN, T. & MOORE, G. 1979. Reconnaissance study
kinks in foliated rocks. Tectonophysics, 21,287- of Upper Cretaceous to Miocene stratigraphic
300. units and sedimentary facies, Kodiak and adja-
HILL, M. D. 1978. Volcanic and plutonic rocks of the cent islands, Alaska. Prof. Pap. U.S. geol. Surv.
Kodiak-Shumagin Shelf, Alaska: subduction de- 1093.
posits and near-trench magmatism. Thesis, PhD, RAMSAY, J. G. 1967. Folding and Fracturing of Rocks.
Univ. California, Santa Cruz. McGraw-Hill, New York, 568 pp.
HoBas, B., MEANS, W. & WILLIAMS, P. 1976. An REED, B. L. & LANPnERE, M. A. 1974. Chemical
Outline of Structural Geology. Wiley, New York, variations across the Alaska-Aleutian Range
571 pp. batholith. J. Res. U.S. geol. Surv. 86, 819-29.
KARIG, D. E. & SHARMAN, G. F. Ill. 1975. Subduc- REID, M. R. & GILL, J. B. 1980. Near trench
tion and accretion in trenches. Bull. geol. Soc. volcanism, Kodiak Island, Alaska: implications
Am. 86, 377-89. for ridge-trench encounter. Abstr. Progr. geol.
LUNDBERG, N. & MOORE, J. 1981. Structural features Soc. Am. 12, 148.
of the Middle America trench slope off southern SEELY, D. R. 1977. The significance of landward
Mexico. In: MOORE, J. et al. (eds). Initial Rep. vergence and oblique structural trends on trench
Deep Sea drill. Proj. 66, (in press). inner slopes. In: TALWANI, M. & PITMAN, W. C.,
MOORE, G. 1967. Preliminary geologic map of Kodiak III (eds). Island Arcs, Deep Sea Trenches, and
Island and vicinity, Al~iska. Open File Rep. 271, Back-arc Basins, 187-98. Am. geophys. Union
U.S. geol. Surv. Reston, Virginia. M. Ewing Ser. 1.

FJ6. 8. Two models for uplift of the Ghost Rocks Formation. In both, opposing-arrows show
shortening directions inferred from conjugate fold and spaced cleavage belts. In imbricate thrust
model (A) (after von Huene 1979; Seely et al. 1974) sub-horizontal deformation axes formed in the
area of the lower slope or trench become tilted, landward as younger more seaward rocks are
accreted. (B) shows the model required by the structural history of Unit A. In this model
deformation axes remain in their sub-horizontal orientation as the prism rises vertically.
242 T. B y r n e

-- , VAIL, P. R. & WALTON, G. G. 1974. Trench Nevada, California. Bull. geol. Soc. Am. 87,
slope model. In: BURK, C. A. & DRAKE, C. L. 1411-20.
(eds). The Geology of Continental Margins, 249- VON HUENE, R. 1979. Structure of the outer conver-
60. Springer-Verlag, New York. gent margin off Kodiak Island, Alaska, from
STONE, D. B. • PACKER, D. R. 1979. Paleomagnetic multichannel seismic records. In: WATKINS, J.,
data from the Alaska Peninsula. Bull. geol. Soc. MONTADERT, L. & DICKERSON, W. (eds). Geolo-
Am. 90, 545-60. gical and Geophysical Investigations of Continen-
SUPPE, J. 1973. Geology of the Leech Lake Moun- tal Margins. Mere. Am. Assoc. Petrol. Geol. 29,
tain-Ball Mountain Region, California. Univ. 261-72.
Calif. Publ. Geol. Sci. 107, 82. WOOD, D. S. 1974. Current views of the development
ToI31SCH, O. & FISKE, R. 1976. Significance of conju- of slaty cleavage. Ann. Rev. Earth Sci. 2, 1-35.
gate folds and crenulations in the central Sierra

TIM BYRNE, Earth Sciences Board, University of California, Santa Cruz, California
95064, U.S.A.
 Sedimentation in the Sunda Trench and forearc region

Gregory F. Moore, Joseph R. Curray & Frans J. Emmel

SUMMARY: Sedimentation in the Sunda Trench and forearc region is dominated by the
transport of terrigenous detritus from orogenic and volcanic terranes around the margins
of the north-eastern Indian Ocean. Ouartzose Himalayan detritus is transported into the
Bay of Bengal and down the trench axis as far south as the Sunda Strait. Prior to
development of the outer-arc ridge, the majority of arc-derived sediment bypassed the
forearc basin and was deposited in the trench. Nearly all sediment derived from the arc
terranes of Java and Sumatra during the Neogene has been trapped in the forearc basin
and has not reached the trench.
Hemipelagic sedimentation dominates on the lower part of the inner trench slope.
Calcareous microfossils and volcanic ash are the dominant constituents of these hemipela-
gic deposits. Higher on the slope, terrigenous and hemipelagic sediments accumulate in
large trench-slope basins.
The Sunda forearc basin is a series of smaller individual basins separated from each
other by transverse highs, isolating the respective sedimentary sequences. Off north
Sumatra, quartzose detritus accumulates, whereas off Java, volcaniclastic sediments
predominate.
Long-distance transport of sediment down the trench axis, damming of sediments
behind the outer-arc ridge, and the segmented nature of the forearc basin leads to the
juxtaposition of sediment bodies with different provenances.

Sedimentation along convergent margins re- results serve to illustrate the complexity of
flects a complex interaction between tectonic sedimentary facies and spatial relationships that
and depositional processes. Sediments are de- develop in forearc regions.
posited seaward of the trench, in the trench This study is primarily based on the analysis
axis, on the inner trench slope, and in the of seismic reflection, 3.5 kHz and bathymetric
forearc basin. These depositional environments data collected by ships of the Scripps Institution
are continuously modified as the accretionary of Oceanography since 1968. Our data were
prism grows. Sediment provenance and disper- supplemented by seismic data collected by the
sal are also complicated. For example, ter- Lamont-Doherty Geological Observatory (pub-
rigenous detritus may be derived from a nearby lished by Hamilton 1979), the U.S. Naval
arc terrane, it may be transported long distances Oceanographic Office (Bowles et al. 1978), and
down the trench axis, or it may be carried into Union and Mobil Oil Companies (see Karig et
the trench on the subducting oceanic plate. al. 1980a). These geophysical data comprise a
Biogenic and volcanogenic detritus are also regional reconnaissance study of the entire Sun-
important components of convergent margin da Arc, as well as two detailed transect studies,
sediments. and include approximately 30 000 km of track
Convergent margin sedimentation is poorly (see Moore et al. 1980b for locations of most
understood, partly because of the paucity of tracks). Bottom samples from the trench and
studies of modern margins, but also because of lower trench slope are insufficient to define
the great variability among convergent margins sediment types and facies distributions precise-
around the world (Dickinson & Seely 1979; ly. The geophysical data, however, allow infer-
Underwood et al. 1980). The purpose of this ences to be made concerning depositional pro-
paper is to outline recent sedimentation pat- cesses. Data from several exploration wells on
terns and discuss the development of strati- the landward margin of the forearc basin off
graphic relationships along one convergent central and north Sumatra provide insights into
margin: the Sunda Arc of the north-eastern forearc basin development. Our sedimentolo-
Indian Ocean. In this paper, we define the gical studies of uplifted Tertiary trench and
depositional framework of the Sunda Trench trench slope strata exposed on Nias Island, west
and forearc region, outline the patterns of of Central Sumatra (Moore 1979; Moore et al.
sediment dispersal and provenance, and infer 1980a), provide additional data on forearc de-
the characteristics of the forearc deposits. Our positional processes.

245
246 G. F. M o o r e , J. R. C u r r a y & F. J. E m m e l

ped in the rapidly subsiding delta and on the

Regional setting
inner shelf, and relatively little sediment is
The Sunda Arc is continuous for more than transported to the fan (Moore et al. 1974).
5000 km from Burma in the NW to Sumba Off the Nicobar Islands, the Ninetyeast
Island in the SE (Figs 1 & 2). It is a continental- Ridge has recently impinged on the trench and
margin arc-trench system with a ridged forearc cut off the flow of turbidites to the Nicobar Fan
(Dickinson & Seely 1979) and marks the site of (Fig. 2; Curray & Moore 1974; McDonald
active subduction of the Indian (or Australian) 1977). Pelagic deposits are now accumulating
Plate beneath the Eurasian (or China) Plate. on the Nicobar Fan (Bowles et al. 1978). Prior
The arc is bounded on the NW by the to the mid-Pleistocene, however, sediments
continent-continent collision zone of the Hima- flowed down the Nicobar Fan as far south as the
layas and on the SE by the incipient continent- latitude of the Sunda Strait (Fig. 2). At DSDP
arc collision zone of the Banda Arc. Subduction site 211, west of the Sunda Strait (Fig. 2),
has apparently been occurring along the Sunda Upper Pliocene-Quaternary pelagic silicic ooze
Arc, at least intermittently, since the Permian overlies Pliocene silt turbidites. At the time of
(Katili 1973). The most recent episode of sub- deposition of the Pliocene turbidites, approx-
duction may have begun in late Oligocene time imately 200-300 km less convergence of the
(Karig et al. 1979; Hamilton 1979). Indian Plate and the Sunda Trench had occur-
Neogene sedimentation in the northern part red, and turbidites were transported directly to
of the Sunda Trench has been dominated by this part of the Nicobar Fan from Ganges-
detritus eroded from the Himalayan collision Brahmaputra sources (Curray & Moore 1974;
belt (Moore et al. 1974). Accretion of these McDonald 1977). These sediments are under-
thick sediments to the Sunda margin during the lain by brown clay of unknown age and Lower
Neogene has led to outbuilding of a wide Cretaceous nannofossil ooze ( v o n d e r Borch,
accretionary prism and upbuilding of a high Sclater et al. 1974).
outer-arc ridge (Curray & Moore 1974; Hamil- Off Java, mainly pelagic and hemipelagic
ton 1979; Karig et al. 1979, 1980b). sediments have been deposited on the Indian
Plate (Cook et al. 1978; Heirtzler et al. 1978).
The thickness of pelagic and hemipelagic sedi-
ments seaward of the trench varies from 200 to
400 m (unpublished data from S.I.O. cruise
Depositional environments
Rama, September 1980).
Deposition in the Sunda forearc region occurs
in four distinct environments: the Indian Plate, Sunda Trench
the trench, the trench inner slope (defined as
the area between the trench and the outer-arc West of the Andaman Islands, where the
ridge), and the forearc basin. The types of Bengal Fan is being subducted, the trench has
sediments deposited, and the depositional pro- little morphological expression and is less than
cesses that operate in each environment differ 3000 m deep (Fig. 4A). A more distinct morpho-
significantly as we demonstrate below. logical depression appears off north Sumatra,
where the thickness of Nicobar Fan sediment
decreases, the trench deepens to 5000 m, the
Indian Plate
trench floor narrows, the outer trench slope
Terrigenous, hemipelagic, and pelagic sedi- steepens, and a trench sediment wedge is de-
ments that are deposited on the Indian Plate veloped (Fig. 4C). The trench has relatively
seaward of the Sunda Trench are carried into steep inner and outer slopes south and south-
the trench by plate convergence. In the Bay of east of Java, and is greater than 6000 m deep
Bengal, deposition is dominated by transport of (Fig. 1). It has several isolated troughs that are
terrigenous detritus onto the Bengal Fan (Cur- greater than 6500 m deep and are separated by
ray & Moore 1971). Deposits in the Bay of bathymetric sills (Fig. 3). The width of the
Bengal are predominantly turbidites that are trench floor varies markedly south of Java, and
greater than 16 km thick west of Burma and locally, a trench sediment wedge is well-
thin markedly to the south (Fig. 3; Curray & developed (Fig. 4D).
Moore 1971, 1974). During low stands of sea- A major submarine channel (Fig. 4A,B)
level, great amounts of sediment are delivered transports sediment from the northern end of the
to the Bengal Fan. In contrast, during the trench to at least 10°N. The channel is not in the
present high stand of sea-level, most of the trench axis as in the eastern Aleutian Trench
sediment derived from the Himalayas is trap- (von Huene 1974), but is 10-50 km seaward of
 Sedimentation in the Sunda Trench 247

85 ° 90 ° 95 ° 100 °
, +++;++
++++++++++ +:+:,:~+:+:+:+~+~,;~3 0 °

~QUATERNARY and TERTIARY

VOLCA N I C S
QUATERNARY STRATA
TERTIARY STRATA
MESOZOIC STRATA
25 ° 25 ° PALEOZOIC and O L D E R STRATA
M E S O Z O I C and P A L E O Z O I C
I N T R U S I V E S and M E T A M O R P H I C S
4~ LOCATION of I L L U S T R A T E D
SEISMIC PROFILE

20' 20 °

15 ° 15 °

I0 o 0 °

:~-"'---:: .,.~

0°
.:',,,..--..... ~.,.,.. _.

+ ~+,.,,

115° 120°
5°

o )3°
J
10 ° 10"

15 ° E ~
85* 90 ° 95 ° 100 ° 105 ° 110 ° 115 ° 120 °

FIG. 1. Regional bathymetry of the north-east Indian Ocean (from Curray et al. 1981) and simplified
geological map of landmasses (from United Nations 1971; Hamilton 1979; and Page et al. 1979).
Locations of seismic profiles of Figs 4 and 5 are also indicated.
248 G. F. Moore, J. R. Curray & F. J. Emmel

:30`= I
\~"xT~ OUTER-ARC RIDGE
HIMALAYAS
( ~ 3 FORE-ARC BASIN
/
// '~""~ SUNDA TRENCH AXIS
25 =

" • "PATH- • DISPERSAL

SEDIMENT

k
i "'-.. PLEISTOCENE DISPERSAL
20 °
""- PATH

15 ° ,/I; I i1~ ',

7
I / ~ ~
# /,," ~,, /
r

I0 •

/
i
5e /

0°

/:

,o.F DSDP
811

15" l I I I I I I
8 5 `= 90`= 95`= I00 a 105 a I10`= 115° 120=

FIG. 2. Tectonic map showing major sediment dispersal paths to the Sunda forearc region. Solid
arrows indicate present dispersal paths; dashed arrows are dispersal paths that were active during the
Pleistocene low stands of sea-level.
 Sedimentation in t h e S u n d a Trench 249

I SUMATIIA I I J A Y A - - I EASTEliN SUNDA

t- ISLANDS
, - ,,:oh., .~o" . . .+~" o

- ~ IX 7\ / \ A ~/~ AA %~'~ ~ / O u t e r - O r C R,clge

E
_o // \ r ,n
7.. 5 VE'RTICYlI_ EXA66ER4TION ~\ l \ ~..

&
/ ""J \ i / "\ _
l
/t ", ~ ~ ~ ~ ~
8 _.I Fore- OFC Bosln /"
2 8osernent

_ o
3

8
#k ~0 ~, /" ~Oceon,c B0semen)
9
z a:
~0 ..as 7~
tll ~ i I I i [ I I I I I I I I I I I I I [ i I I I I I I I
o 1000 7000 3000 4000 5000
DISTANCE, kliomeires

FIG. 3. Depth profiles along the Sunda forearc region from south-west Burma to Sumba Island.
Lower diagram is an axial profile along the trench. Depth to oceanic basement based on seismic
reflection and refraction measurements (triangles are data points). Trench axis depths are from
echo-sounder crossings (circles). Arrows are locations of sills that may block the flow of turbidites.
Upper diagram is an axial profile along the forearc basin. Depth to forearc basin basement from
Curray et al. (1977), Hamilton (1979), Kieckhefer et al. (1980) and unpublished S.I.O. seismic
refraction data. Depth of forearc basin from echo-sounder crossings (circles). Depth of outer-arc
ridge measured from regional bathymetric map (Curray et al. 1981).

the base of the inner trench slope as in the having been transported from the Himalayas.
Peru-Chile Trench (von Huene 1974; Schweller The rough topography and variable width of the
& Kulm 1978). This channel, probably fed by a trench axis south of Java suggest that sediments
canyon from the Bangladesh shelf (Figs 1 & 2), are not transported along the trench axis in this
has been filled SW of Burma by a large area at present.
submarine slide and associated debris flows (see Sediments from Sumatra and Java are at
below). West of the A n d a m a n Islands, another present prevented from reaching the trench by
major channel from the Bengal Fan intersects the outer-arc ridge along most of the arc.
the trench at 12°N and trends southward (Fig. Arc-derived sediments generally are trapped in
2). the forearc basin (Hamilton 1979). An excep-
Transport of large amounts of sediment down tion is in the area of the Sunda Strait, where the
the trench axis is indicated by the broad trench Sumatran Fault cuts across the forearc region
floor west of Sumatra (Fig. 1) and the relatively (e.g. Hamilton 1979; Curray et al. 1981). Tur-
smooth axial gradient from Burma to the south bidity currents from south Sumatra and west
tip of Sumatra (Fig. 3). Himalayan detritus may Java apparently flow down the canyon formed
be transported as far south as the Sunda Strait along the fault and into the trench axis (Figs 1 &
approximately 3000 km from the head of the 2).
trench. Hamilton (1979) interprets quartzose, The morphotectonic setting of the Sunda Arc
micaceous, silty turbidites recovered from the was probably very different during the early
trench axis at 6°S (Anikouchine & Ling 1967) as and mid-Tertiary. Sumatran detritus is found in
250 G. F. M o o r e , J. R. Curray & F. J. E m m e l

tO

E
i---

(3
4 9 D

I
(M
I0

7 12

FIo. 4. Seismic reflection profiles across the Sunda Trench (locations shown in Fig. 1). (A) Profile
west of Andaman Islands shows thick sediments of Bengal Fan with channel west of trench axis. (B).
Profile west of Nicobar Islands shows wider channel with levees. Acoustic basement is the Ninetyeast
Ridge. (C) Profile off North Sumatra shows seaward-dipping trench wedge and onlapping
relationship of trench sediments on to Indian plate sediments. (D) Profile off Java shows oceanic
pelagic section onlapped by trench wedge with seaward-dipping surface.

uplifted upper Oligocene/lower Miocene trench aprons are widespread along the trench south of
deposits on Nias Island, suggesting that at 2°N (Bowles et al. 1978), suggesting that small-
least some arc-derived sediments were bypas- scale slumping and redeposition of accreted
sing the forearc basin and flowing into the material and surficial slope sediments through
trench (Moore 1979). Prior to the Miocene, the mass flows and turbidites is common. In areas
outer-arc ridge was probably much smaller of the trench off Java that are isolated from the
in size because thick fan sediments had not axial transport of turbidites, mass flow deposits
yet been accreted (Curray & Moore 1974; may form a substantial portion of the trench
McDonald 1977). The time of formation of a wedge. Small fans emanating from minor can-
morphological outer-arc ridge probably varied yons on the lower trench slope (Karig et al.
along the arc, occurring first in the north and 1980b) probably direct turbidity currents and
later in the SE. debris flows of slumped hemipelagic slope sedi-
Many seismic reflection and 3.5 kHz profiles ments and uplifted trench sediments to the
across the trench show a seaward-dipping sea- trench. The fans SW of Nias have a mounded,
floor at the base of the landward trench slope chaotic seismic character (Karig et al. 1980b)
(Fig. 4B,C). The echo character of this sedi- and extend across the trench axis, possibly
ment apron (type IIB of Damuth & Hayes blocking or disrupting the flow of turbidites
1977) indicates that it is formed of large down the trench axis (Fig. 3).
amounts of bedded sands and silts. Several A large sediment slide (over 2000 km 2) has
crossings of the apron show hummocky been identified in the trench off Burma (Moore
topography (echo type IIIC) characteristic of et al. 1976). This slide has transported a huge
debris flows or small slumps. These sediment volume (over 900 km 3) of sediment which may
 S e d i m e n t a t i o n in the S u n d a T r e n c h 251

have come from a westerly Pleistocene lobe of ridge west and north of Sumatra are supplying
the Irrawaddy River delta. Other slump masses detritus at present from the uplifted accretion-
tens of kilometres wide have not been identified ary prism to the inner trench slope. For exam-
elsewhere in the Sunda Trench (Hamilton 1979; ple, there are several large rivers that drain the
Moore et al. 1980b), suggesting that slumping of interior of Nias and discharge large amounts of
large-scale blocks (several kilometres in size) is sediment on to the trench slope. We identified
not an important mechanism of sediment trans- several canyons on the Nias shelf that lead into
port to the Sunda Trench. trench slope basins (unpublished data from
Expedition Rama, October 1980). The trans-
Trench inner slope port of accretionary prism material to the
trench slope has varied considerably along the
Sediments accumulate on the landward arc. It has probably been occurring since the
trench slope between the trench and the outer- Oligocene in Burma, whereas it is a fairly
arc ridge. The inner slope west and NW of recent phenomenon off Sumatra, because the
Sumatra is characterized by a ridge and islands of the outer-arc ridge have only been
trough morphology, and the width of the emergent for a few million years. Accretionary
troughs generally increases upslope (Moore & prism material is probably not being trans-
Karig 1976; Karig et al. 1980b). The basins are ported to the trench slope off Java, because the
continuous along strike for only a short distance outer-arc ridge is below sea-level and blanketed
(a few tens of kilometres), and the basin floors by hemipelagic deposits. Data from Nias
generally exhibit an irregular topography. (Moore et al. 1980a) suggest that redeposition
Many small canyons traverse short sections of of slumped hemipelagic sediments into the
the slope, providing a mode of transport of slope basins is important near the base of the
sediment into the basins (Karig et al. 1980b). inner trench slope. In the past, submarine
The lower trench slope off Java is steeper than canyons probably cut across the forearc basin
off Sumatra, and slope basins are not as well- and delivered arc-derived sediment from Java
developed. and Sumatra to the slope basins.
Slope basin sediments are thin near the base
of the slope, but are as thick as 2-3 km near the Forearc basin
outer-arc ridge. Over many of the ridges,
3.5 kHz profiles show evenly stratified reflec- Between the outer-arc ridge and the active
tors (echo type IB), indicating a uniform pela- arc terrane from the eastern Sunda Arc to north
gic or hemipelagic drape. The echo character of Sumatra is a series of forearc basins (Figs 1-3).
the basin sediments ranges from prolonged The basins are 50-100 km wide and 150-
(type IIA) to bedded (type IIB), suggesting that 500 km long. The water depth of the basins off
the strata are a mix of hemipelagic sediments Java is 3500-4000 m, but is only 500-2000 m off
and bedded silts and sands (turbidites). Sumatra. They are distinct depositional basins
Basins near the top of the slope have been that are separated from each other by trans-
filled, causing sediment to spill over into the verse structural highs. The sedimentary fill in
deeper basins (Fig. 5A). As small basins fill each basin is thus expected to be different in
with sediment they combine to form larger composition from adjacent basins because of
basins (Fig. 5A). Some basins near the top of the differences in source terranes along the arc
the slope contain up to 1.5 km of layered (see below). The sediment fill in the basins is up
sediments. The shallow reflectors are generally to 6 km thick (Fig. 3). The maximum sediment
continuous across the basins, but the deeper thickness in each basin underlies the present
reflectors are discontinuous (Karig et al. 1980b) deep basins (Hamilton 1979), and the basin
and many unconformities are visible (Fig. 5A). sediments onlap the structural highs, indicating
We believe that the discontinuous nature of the that the structural separation of the basins has
older sediments is due primarily to deposition persisted for a long period of time.
in smaller, discontinuous basins lower on the Sediments deposited in the forearc basins are
slope and secondarily to slumping. The uncon- derived primarily from the arc terranes at their
formities are due to tilting and folding contem- landward boundaries. Submarine canyons are
poraneous with sedimentation. common on the slope from the arc terrane into
Piston cores taken on the lower slope off Nias the forearc basins, and are apparently the main
(Karig et al. 1980b) indicate that hemipelagic dispersal paths for arc-derived sediments. Can-
sediments are presently accumulating at a high yons are more numerous along the Java margin
rate. Volcanic ash layers are common in the than along the Sumatra margin (Fig. 2). In
cores. The emergent islands of the outer-arc addition to the arc-derived terrigenous and
252 G. F. Moore, J. R. Curray & F. J. E m m e l

2
V

o ~
E :5
t--

0
B
0.5
0

I
Od

1.0

Fro. 5. (A) Seismic profile across trench-slope basin west of the Andaman Islands showing several
periods of deformation as indicated by many unconformities. Upper basin has been filled and
sediments are now spilling into lower basin. (B) Seismic profile across small fan in forearc basin
landward of Simalur Island showing characteristic concave-upward surface and chaotic internal
seismic facies. Location of profiles shown in Fig. 1.

volcaniclastic sediment, shallow-water carbon- that spread laterally across the basin floors.
ate debris is an important component of the The forearc terrane north of Sumatra is
forearc basin sediments. anomalous. There is at present no forearc basin
Surprisingly, we have identified only a few between the Nicobar Islands and Burma (Fig.
small submarine fan complexes in the forearc 1). A forearc basin probably existed prior to the
basin. This may be because the morphological Miocene, but it has been structurally disrupted
fans are smaller than the resolution of our by the opening of the A n d a m a n Sea (Curray et
seismic data. One fan, in the basin off Simalur al. 1981). The morphotectonic outer-arc ridge
Island (Figs 1 & 2), has a characteristic convex- becomes subaerial as the Indoburman Ranges
upward morphology and chaotic seismic charac- of Burma, and the forearc basin is the western
ter (Fig. 5B). Seismically identifiable buried fan trough of the Central Valley of Burma. The
and channel complexes are also uncommon. sediments in the basin are many kilometres
The flat floors of the basins and the reflection thick and probably have been derived and
configuration of the sedimentary fills suggest transported longitudinally within the trough
that deposition is mainly by turbidity currents from the Himalayan terrane.
 Sedimentation in t h e S u n d a Trench 253

Sediment provenance Sumatra, Palaeozoic schists, gneisses and gra-

nites, Mesozoic volcanic and volcaniclastic
Terrigenous detritus is at present the major con- rocks, and quartzose Tertiary strata are ex-
stituent of the sediments being deposited in the posed (Page et al. 1979). The Quaternary volca-
Sunda Trench and forearc region. Volumetri- noes of north and central Sumatra lie NE of
cally, the most important provenance is the the drainage divide, so recent volcanic detri-
uplifted basement terrane of the Himalayas. tus is transported mainly to the backarc region
The uplifted crystalline basement rocks are (Fig. 2). Thus, detritus delivered to the forearc
eroded rapidly and the detritus is carried by the region off north and central Sumatra reflects
confluent Ganges-Brahmaputra River system primarily a metamorphic and sedimentary
into the Bay of Bengal (Curray & Moore 1971), provenance. Such a provenance should contri-
where it is dispersed southward on the Bengal bute quartzose detritus with low ratios of pla-
deep-sea fan system (Curray & Moore 1974; gioclase to total feldspar (P/F) and low abund-
Curray et al. 1981). ance of volcanic lithic clasts (Fig. 5; Dickinson
Distal turbidites of late Miocene-Pleistocene 1970; Crook 1974; Schwab 1975; Potter 1978;
age were recovered at DSDP Site 211 west of Dickinson & Suczek 1979; Ingersoll & Suczek
Sumatra (yon der Borch, Sclater et al. 1974). 1979). Because there are several ophiolite
Detailed point counts of these turbidites bodies exposed in north Sumatra (Page et al.
(Ingersoll & Suczek 1979) document the petro- 1979), a small amount of ophiolitic detritus
logical characteristics of sediments derived would be expected to be delivered to the fore-
from a Himalayan provenance: they are arc basin.
micaceous and highly quartzose (quartz- In south Sumatra, Tertiary and Quaternary
feldspar-lithic modes average Q57 F28 L14), and volcanic rocks make up a large portion of the
lithic types are dominated by metamorphic exposures (Hamilton 1979; Fig. 1). Mesozoic
fragments (metamorphic-volcanic-sedimentary and Tertiary sedimentary rocks also contribute
lithic populations average Lms7 Lv4 Lsg). Other detritus to the forearc region. Sediments de-
petrological studies of Himalayan sediments rived from southern Sumatra are probably less
confirm their quartzose, metamorphic nature quartzose and more volcanic-rich than those of
(e.g. Raju 1967; Thompson 1974; Mallik 1976, north and central Sumatra.
1978). The range of modal compositions for Java is composed largely of Quaternary vol-
Himalayan detritus is shown in Fig. 6. canics and Tertiary sedimentary rocks (Hamil-
Terrigenous detritus eroded from the arc ton 1979; Fig. 1). Palaeogene strata are quart-
terrane of Sumatra is transported to the west zose, but Neogene strata are volcanic-rich
coast by rivers and delivered to the forearc (Hamilton 1979). Sediment derived from Java
region by submarine canyons. Although no should therefore be dominated by volcanic
quantitative point count information exists on lithics and plagioclase feldspar with subsidiary
the character of Sumatran detritus, inferences quartz. The islands east of Java are composed
can be made based on knowledge of the rocks mainly of volcanics and volcaniclastic sedi-
exposed on Sumatra. In north and central ments. Detritus derived from the eastern is-

Q Lm

- ; N SUM/~TRA

A PROVENANCE
J

///i i
/
FL "''-, L
Lv JJ// c .....
FIG. 6. Detrital modes of sands transported from various provenances to Sunda forearc region.
Himalayan modes are from detailed point counts (see text for references); Java and Sumatra modes
are inferred from the lithologies exposed in the source terranes (Page et al. 1979; Hamilton 1979) and
from studies in other arcs (Dickinson & Suczek 1979). (A) Quartz-feldspar-lithic diagram. (B)
Metamorphic-volcanic-sedimentary lithic fragment diagram.
254 G. F. Moore, J. R. Curray & F. J. Emmel
lands of the Sunda Arc is probably highly Andaman section
lithic-rich and low in quartz, is dominated by
volcanic lithic clasts and has high P/F ratios The oceanic crust entering the trench off the
(Fig. 6). Andaman Islands is inferred to be mid-
Cretaceous in age (e.g. Curray et al. 1981) and
is probably overlain by a thin Cretaceous pela-
gic section. Overlying the pelagics are Upper
Facies relationships Cretaceous-Palaeocene strata that probably
Due to differences in sediment transport to the represent a continental rise sediment wedge
different depositional environments of the fore- that developed off India prior to its collision
arc region and to changes in character of these with Asia (Curray & Moore 1974). These sedi-
environments along strike, sediments of diffe- ments are overlain by a sequence of ]Eocene to
rent facies and composition are deposited adja- mid-Miocene turbidites of the ancestral Bengal
cent to each other. These facies relationships Fan, which is in turn overlain by approximately
are best illustrated with cross-sections from the 1.5 km of Upper Miocene to Pliocene turbidites
Indian Plate, across the trench, trench inner of the Bengal Fan (Moore et al. 1974), including
slope, and forearc basin. However, because of large channel and levee complexes (Curray &
the variations along the arc, any cross-section is Moore 1974). Quaternary trench fill is essen-
valid only for describing one narrow transect. tially a part of the Bengal Fan, and consists
We will present here three schematic cross- dominantly of turbidites of Himalayan prove-
sections from transects inferred to be repre- nance.
sentative of large areas of the arc (Fig. 7). Young sediments deposited on the inner

H
~ 0

uM- P: " ~:7~ Q SAND

CORE LEGEND

PROVENANCE
-
LITHOLOGY
PeDb~yMud
_==r I \ I ~M --

0 -- - I I Pe(a~,c on~

-- Accrehooof'~

,~ __: "< -r I °...... ~....

AGES
Quote f no r,p Q u Upper
t ;~ ~ Phocene pl m M,cld~e
L/ ~-~7--zl Pi 0
MloceBe M
F~
I=~!I o,,qoc.~ o
o s 1:7:-31 M ~. . . . . E
-- L*S~-M PI I,\-'-:I Paleocene P

- -M 0 . . . .
I ,//,I ~ NIA$ - ~-o,

~ [::. 1 E . ~-- - " = -

=1,0 "
PI-Q

o~ ~M_ _ _ _ ~ - -

50 Kilometres
t0
V E : 5X

FIG. 7. Schematic depth sections across the Sunda forearc region off the Andaman Islands (top),
central Sumatra (middle) and central Java (bottom). Columns are inferred stratigraphic sequences
developed in various depositional settings. Structure is based on data presented by Curray et al.
(1977), Karig et al. (1979), and Moore et al. (1980b).
 Sedimentation in t h e S u n d a Trench 255

trench slope off the Andaman Islands are prob- tionary complex (Karig et al. 1980a). The
ably a mixture of terrigenous debris derived sedimentary fill in the basins records several
from the exposed ophiolitic and sedimentary periods of sedimentation and deformation.
terranes on the Andaman Islands (e.g. Karuna- Palaeogene strata fill a structural low formed
karan et al. 1968) and infrequently some terri- landward of a palaeo-outer-arc ridge. Uplift
genous detritus delivered to the trench slope and erosion of the entire margin occurred at the
by western distributaries of the Irrawaddy River end of the Oligocene (Karig et al. 1980b),
of Burma. Because the strata exposed on the producing a regional unconformity. During the
Andaman Islands are uplifted Bengal Fan de- Neogene, turbidite sedimentation has domin-
posits, the composition of the slope sediment ated. The forearc basin strata off Sumatra are
derived from the Andamans should be similar 4-6 km thick and are dominantly turbidites
to the sediment that has accumulated in the derived from Sumatra. They range in age from
trench with the addition of minor ophiolitic Palaeogene to Recent. In the central Sumatra
debris. Sedimentary rocks which may be older area, the terrigenous turbidites are quartzose
slope deposits exposed on the Andaman Islands with andesitic detritus present in varying
are hemipelagic and pelagic deposits interbed- amounts (Fig. 6).
ded with coarse strata that contain abundant
ophiolitic debris. Java section

Sumatra section Upper Cretaceous oceanic crust is at present

being subducted off Java (Heirtzler et al. 1978).
The age of the crust being subducted off It is overlain by a thin (300-400 m) section of
Sumatra is uncertain, but is inferred to be pelagic deposits that are Cretaceous to Pleis-
Eocene from marine magnetic data (Sclater & tocene in age. As off Sumatra, volcanic ash is
Fisher 1974). Oceanic crust is probably suc- an important component in the pelagic sedi-
ceeded by Eocene to Oligocene pelagic sedi- ments, and carbonate contents are less than
ments (Fig. 7), then by a 1 km thick sequence 10%. The pelagic section is overlain by Quater-
of Nicobar Fan turbidites which is overlain in nary trench deposits. Trench sediments south
turn by the trench deposits. Both the Nicobar and SE of Java are less than 1 km thick, and
Fan and trench turbidites reflect a Himalayan several isolated areas are nearly devoid of
provenance. Sediment cores from the Indian trench fill (Fig. 3). The trench off Java is
Plate contain a large component of terrestrial probably starved of terrigenous sand, so the
clay that has probably been transported from trench deposits are mostly hemipelagic silts and
India (Venkatarathnam & Biscaye 1973) and muds and redeposited material that was
considerable amounts of volcanic ash (Nink- accreted to the inner trench slope and subse-
ovich 1979). Carbonate comprises less than quently slumped back into the trench.
10% of the sediment because most of the Indian The inner trench slope is blanketed by a
Ocean floor is below the carbonate compensa- hemipelagic slope sediment section that is over
tion depth (Kolla et al. 1976). 1 km thick on the outer-arc ridge. Unlike the
Very thin slope sediments are deposited on slope deposits off Sumatra, those off Java prob-
top of accreted trench and fan sediments at the ably do not contain large amounts of ter-
base of the slope off Sumatra. Slope sediments rigenous material from the outer-arc ridge,
are dominantly hemipelagic clays and mud because the ridge is not subaerial along the Java
turbidites, because the flow of arc-derived sedi- section of the arc.
ment is cut off by the outer-arc ridge. Higher on The forearc basin deposits off Java are 4-
the slope, turbidites carrying terrigenous detri- 5 km thick and generally display similar sedi-
tus from the emergent islands on the outer-arc mentation patterns to those in the Sumatra
ridge are important. As inferred from Nias forearc basins. Compositionally, however, the
Island data (Moore et al. 1980a), a hypothetical Java deposits should differ markedly from those
section through a slope basin on the upper part off central and north Sumatra. The sediments
of the trench slope has thin hemipelagics at the probably are dominantly volcanic-rich turbi-
base overlain by a thicker sequence of ter- dites from Java, reflecting their arc provenance.
rigenous turbidites (Fig. 7).
The basement of the Sunda forearc basins is
believed to be either thinned continental crust Discussion and conclusions
(Kieckhefer et al. 1980), thickened oceanic
crust (Curray et al. 1977; Hamilton 1979), or an Sedimentation patterns in the Sunda forearc
old subsided and highly metamorphosed accre- region vary considerably both spatially and
256 G. F. M o o r e , J. R. C u r r a y & F. J. E m m e l

temporally. Transport of Himalayan detritus nally transported are deposited in the trench off
along the trench axis dominates in the north. As north Sumatra, while finer-grained sediments
the amount of sediment transported from the may be deposited in the adjacent forearc basin.
north decreases, hemipelagic sediments be- A similar juxtaposition of facies in Mesozoic
come more important components of the trench deposits of California (Franciscan Complex and
wedge. The isolated nature of the forearc basins Great Valley Series; Blake & Jones 1974) might
and the changes in provenance along the arc be the result of deposition in a similar tectonic
lead to marked spatial changes in composition setting (Hamilton 1978). Off south Sumatra,
of sediment that is deposited in the basins. At quartzose turbidites derived from the Hima-
the present time, most of the coarse sediment layas are deposited in the trench, whereas
derived from Java and Sumatra is trapped in the volcanic-rich turbidites derived from Sumatra
forearc basins. Prior to the Neogene, the outer- are deposited in the forearc basin, producing a
arc ridge was a much smaller feature, and juxtaposition of facies of markedly different
arc-derived detritus was able to bypass the compositions. Off Java, hemipelagic sediments
forearc basin and flow into the trench. Because dominate both in the trench and on the lower
the time of formation of the outer-arc ridge trench slope, so offscraped sediments might be
varied along the arc, this change in sedimenta- indistinguishable from the overlying slope sedi-
tion patterns also varied temporally along the ments.
arc. ACKNOWLEDGMENTS: Most of the S.I.O. data used in
A subsidiary effect of this change in tectonic this paper were collected with funds from the Office
regime was a corresponding change in sedi- of Naval Research as part of a large-scale study of
mentation rates in the trench and forearc tectonic and sedimentary processes in the north-
basins. Prior to uplift of the present outer-arc eastern Indian Ocean. Data acquisition in the central
ridge, sedimentation rates were relatively high Sumatra and Java areas and preparation of this paper
in the trench and relatively low in the forearc were supported by the IDOE/SEATAR Office of the
basin. After uplift of the ridge, sedimentation National Science Foundation under Grants OCE76-
rates must have decreased in the trench and 24101, OCE79-18185 and OCE79-20482. We thank
Dr Fred Hehuwat of the Indonesia National Institute
increased in the forearc basins. Until filling of of Geology and Mining for his efforts in coordinating
the basins is complete, this pattern should field studies of the Sunda Arc. Reviews by M. A.
continue. Arthur, S. B. Bachman, K. J. McMillen, J. C.
Coarse turbidites that have been longitudi- Moore, and T. H. Shipley improved the manuscript.

References
ANIKOUCHINE, W. A. & LING, H. 1967. Evidence for CURRAY,J. R., EMMEL,F. J., MOORE, D. G. & RAITr,
turbidite accumulation in trenches in the Indo- R. W. 1981. Structure, tectonics, and geological
Pacific region. Mar. Geol. 5, 141-54. history of the northeastern Indian Ocean. In:
BLAKE, M. C., JR & JONES, D. L. 1974. Origin of the NAIRN, A. E. M. & STEHLI, F. G. (eds). Ocean
Franciscan melanges in northern California. In: Basins and Margins: the Indian Ocean. (in
Dorr, R. H., JR & SHAVER,R. H. (eds). Modern press).
and Ancient Geosynclinal Sedimentation. Spec. & MOORE, D. G. 1971. Growth of the Bengal
Pubi. Soc. econ. Paleontol. Mineral. Tulsa, 19, Deep-Sea Fan and denudation in the Himalayas.
345-57. Bull. geol. Soc. Am. 82, 563-72.
BOWLES, F. A., RUDDIMAN, W. F. & JAHN, W. H. -- & -- 1974. Sedimentary and tectonic pro-
1978. Acoustic stratigraphy, structure, and de- cess in the Bengal Deep-Sea Fan and geosync-
positional history of the Nicobar Fan, eastern line. In: BURK, C. A. & DRAKE,C. L. (eds). The
Indian Ocean. Mar. Geol. 26, 269-88. Geology of Continental Margins, 617-27. Sprin-
COOK, P. J., VEEVERS, J. J., HEIRTZLER, J. R. & ger-Verlag, New York.
CAMERON,P. J. 1978. The sediments of the Argo , SHOR, G. G., JR, RAITT, R. W. & HENRY, M.
Abyssal Plain and adjacent areas, Atlantis II 1977. Seismic refraction and reflection studies
Cruise, Northeast Indian Ocean. BMR J. Aust. of crustal structure of the Eastern Sunda and
Geol. Geophys. 3, 113-24. Western Banda Arcs. J. geophys. Res. 82,
CROOK, K. A. W. 1974. Lithogenesis and geotecto- 2479-89.
nics: the significance of compositional variation DAMUTH, J. E. & HAYES,D. E. 1977. Echo character
in flysch arenites (graywackes). In: D o n , R. H., of the east Brazilian continental margin and its
JR & SHAVER,R. H. (eds). Modern and Ancient relationship to sedimentary processes. Mar.
Geosynclinal Sedimentation. Spec. Publ. Soc. Geol. 24, 73-95.
econ. Paleontol. Mineral. Tulsa, 19, 304-10. DICKINSON, W. R. 1970. Interpreting detrital modes
 Sedimentation in the S u n d a Trench 257

of graywacke and arkose. J. sediment. Petrol. 40, the western part, Bay of Bengal, Indian Ocean.
695-707. Mar. Geol. 22, 1-32.
DICKINSON, W. R. (~¢ SEELY, D. R. 1979. Structure --1978. Mineralogy of deep-sea sands of the
and stratigraphy of forearc regions. Bull. Am. Indian Ocean. Mar. Geol. 27, 161-76.
Assoc. Petrol. Geol. 63, 2-31. MOORE, D. G., CURRAY, J. R. & EMMEL, F. J. 1976.
--& SUCZEK, C. A. 1979. Plate tectonics and Large submarine slide (olistostrome) associated
sandstone composition. Bull. Am. Assoc. Petrol. with Sunda Arc subduction zone, northeast
Geol. 63, 2164-82. Indian Ocean. Mar. Geol. 21,211-26.
GRAHAM, S. A., DICKINSON, W. R. & INGERSOLL, R. , RAITI, R. W. & EMMEL, F. J. 1974.
V. 1975. Himalayan-Bengal model for flysch Stratigraphic-seismic section correlations and im-
dispersal in Appalachian-Ouachita system. Bull. plications to Bengal Fan history. In: VON DER
geol. Soc. Am. 86, 273-86. BORCn, C. C., SCLATER, J. G. et al. (eds). Initial
HAMILTON, W. 1978. Mesozoic tectonics of the west- Rep. Deep Sea drill. Pro]. 22, 403-12.
ern United States. In: HOWELL, D. G. & MOORE, G. F. 1979. Petrography of subduction zone
McDOUGALL, K. A. (eds). Soc. econ. Paleontol. sandstones from Nias Island, Indonesia. J, sedi-
Mineral. Pacif. Coast Paleogeogr. Syrup. 2, 33- ment. Petrol. 49, 71-84.
70. -- , BILLMAN, H. G., HEHANUSSA, P. E. & KARIG,
-- 1979. Tectonics of the Indonesian Region. Prof. D. E. 1980a. Sedimentology and paleo-
Pap. U.S. geol. Surv. 1078, 345 pp. bathymetry of Neogene trench-slope deposits,
HEIRTZLER, J. R., CAMERON, P. J., COOK, P. J., Nias Island, Indonesia. J. Geol. 88, 161-80.
POWELL, T., ROESER, H. A., SUKARDI, S. (~ -- , CURRAY,J. R., MOORE, D. G. & KARIG, D. E.
VEEVERS, J. J. 1978. The Argo Abyssal Plain. 1980b. Variations in geologic structure along the
Earth planet. Sci. Lett. 48, 21-31. Sunda Forearc, northeastern Indian Ocean. In:
HAYES, D. E. (ed.). The Tectonic and Geologic
INGERSOLL, R. V. ,~ SUCZEK, C. A. 1979. Petrology
Evolution of Southeast Asian Seas and Islands.
and provenance of Neogene sand from Nicobar
Geophys. Monogr. Washington, 23, 145-60.
and Bengal Fans, DSDP Sites 211 and 218. J. NINKOVlCH, D. 1979. Distribution, age, and chemical
sediment. Petrol. 49, 1217-28.
composition of tephra layers in deep-sea sedi-
KARIG, D. E., LAWRENCE, M. B., MOORE, G. F. & ments off Western Indonesia. J. Volcanol.
CURRAY, J. R. 1980a. Structural framework of geotherm. Res. 5, 67-86.
the forearc basin, NW Sumatra. J. geol. Soc. PAGE, B. G. N., BENNEI~, J. D., CAMERON, N. R.,
London, 137, 77-91. BRIDGE, D. McC., JEFFERY, D. H., KEATS, W. &
, SUPARKA, S., MOORE, G. F. & HEHANUSSA, P. THAIB, J. 1979. A review of the main structural
1979. Structure and Cenozoic evolution of the and magmatic features of northern Sumatra. J.
Sunda Arc in the central Sumatra region. Mere. geol. Soc. London, 136, 569-79.
Am. Assoc. Petrol. Geol. 29, 223-37. PoI"rER, P. E. 1978. Petrology and chemistry of
, MOORE, G. F., CURRAY, J. R. & LAWRENCE, modern big river sands. J. Geol. 86, 423-49.
M. B. 1980b. Morphology and shallow structure RAJU, A. T. R. 1967. Observations on the petro-
of the trench slope off Nias Island, Sunda Arc. graphy of Tertiary clastic sediments of the
In: HAYES, D. E. (ed.). The Tectonic and Geo- Himalayan foothills of North India. Bull. Oil nat.
logic Evolution of Southeast Asian Seas and Gas Comm. Delhi, 4, 5-16.
Islands. Geophys. Monogr. Washington, 23, SCHWAB, F. L. 1975. Framework mineralogy and
179-208. chemical composition of continental margin-type
KARUNAKARAN,C., RAY, K. K. & SAHA, S. S. 1968. sandstone. Geology, 3, 487-90.
Tertiary sedimentation in the Andaman-Nicobar SCLATER,J. G. (~ FISHER, R. L. 1974. Evolution of the
Geosyncline. J. geol. Soc. India, 9, 32-9. east-central Indian Ocean, with emphasis on the
KATILI, J. A. 1973. Geochronology of west Indonesia tectonic setting of the Ninetyeast Ridge. Bull.
and its implication on plate tectonics. Tectono- geol. Soc. Am. 85, 683-702.
physics, 19, 195-212. THOMPSON, R. W. 1974. Mineralogy of sands from the
Bengal and Nicobar Fans, Sites 218 and 211,
KIECKHEFER, R. M., SHOR, G. G., .JR, CURRAY,.J.R., eastern Indian Ocean. In: YON DER BORCH, C. C.,
SUGIARTA, W. & HEHUWAT, F. 1980. Seismic SCLATER, J. G. et al. (eds). Initial Rep. Deep Sea
refraction studies of the Sunda trench and fore- drill. Proj. 22, 711-3.
arc basin. J. geophys. Res. 85, 863-90. UNDERWOOD, M. B., BACHMAN, S. B. & SCHWELLER,
KOLLA, V., BE, A. W. H. & BISCAYE, P. E. 1976. W. J. 1980. Sedimentary processes and facies
Calcium carbonate distribution in the surface associations within trench and trench-slope set-
sediments of the Indian Ocean. J. geophys. Res. tings. In: FIELD, M. E., BOUMA, A. H. &
81, 2605-16. COLBURN, I. (eds). Quaternary Depositional En-
MCDONALD, J. M. 1977. Sediments and structure of vironments on the Pacific Continental Margin,
the Nicobar Fan, northeast Indian Ocean. Thesis, 211-29. Pacif. Sec. Soc. econ. Paleontol. Miner-
Ph.D. Univ. Calif., San Diego, 148 pp. (un- al. Tulsa.
publ.). UNITED NATIONS 1971. Geologic map of Asia and the
MALLIK, T. K. 1976. Shelf sediments of the Ganges Far East. Scale 1:5,000,000. 2nd ed.
delta with special emphasis on the mineralogy of VENKATARATHNAM,K. 1974. Mineralogical data from
258 G. F. Moore, J. R. Curray & F. J. Emmel
sites 211,212,213,214, and 215 of DSDP Leg 22 Res. Oceanogr. Abstr. 20, 727-38.
and origin of noncarbonate sediments in the VON DER BORCH, C. C., SCLATER,J. G. et al. 1974.
equatorial Indian Ocean. In: VON DER BORCH, C. Initial Rep. Deep Sea drill. Proj. 22, U.S. Govt
C., SCLATER,J. G. et al (eds). Initial Rep. Deep Printing Office, Washington, 890 pp.
Sea drill. Proj. 22, 489-501. U.S. Govt Printing VON HUENE, R. 1974. Modern trench sediments. In:
Office, Washington. BURK, C. A. & DRAKE, C. L. (eds). Geology of
-- & BISCAYE, P. 1973. Clay mineralogy and sedi- Continental Margins, 207-11. Springer-Verlag,
mentation in the eastern Indian Ocean. Deep Sea New York.

GREGORY F. MOORE, JOSEPH R. CURRAY, FRANS J. EMMEL, Geological Research

Division, Scripps Institution of Oceanography, La Jolla, California 92093,
U.S.A.
 Development of the North Island Subduction System,
New Zealand

Gerrit J. van der Lingen

SUMMARY: Subduction of the Pacific plate underneath the North Island of New
Zealand began near the beginning of the Neogene, when the Tonga-Kermadec Subduc-
tion System propagated southwards into the New Zealand continental crustal block,
together with the southwards migration of the relative pole of rotation between the Pacific
and Australian plates. Many geological and geophysical aspects of the North Island, both
on land and offshore, are in agreement with models of subduction systems from
elsewhere. Several NE-trending zones can be distinguished: Taupo Volcanic Zone
(volcanic arc), Axial Ranges (exposed part of arc massif), East Coast Depression (forearc
basin), Coastal Ranges (structural high of subduction complex), continental shelf and
slope (trough-inner-slope of subduction complex). A west-dipping Benioff zone underlies
the volcanic arc and arc massif. A negative gravity anomaly coincides with the surface
extension of the Benioff zone. However, no subduction trench exists below the negative
gravity anomaly. Sensu stricto, Fhe Hikurangi Trough, to the east of the subduction
complex, is not a subduction trench.
The forearc basin and subduction complex are riding passively on top of the Pacific plate
to the east of the Benioff zone. The forearc basin initially was shallow marine, but
subsequently was filled with coarse debris shed from the rising arc massif, and is now
above water. The subduction complex consists of highly deformed autochthonous
Cretaceous-Palaeogene strata, possibly deposited in a pre-subduction continental border-
land. Accreted oceanic sediments have not been recorded, but may be present in the lower
part of the subduction complex, or, to a large degree, may have been subducted. During
the formation of the subduction complex, small basins were created by elongate narrow
structural ridges parallel to the trough-inner-slope. Several are now exposed in the Coastal
Ranges structural high. Their sedimentary fills comprise sediment gravity flow deposits,
hemipelagic muds, and arc-derived ash. Similar sedimentary facies have been cored in
present-day offshore slope basins. Some sediment gravity flows reach the Hikurangi
Trough via submarine canyons, bypassing the slope.
A directional, morphological and geophysical break exists between the North Island
and Kermadec subduction systems, probably caused by the Vening Meinesz Fracture
Zone.
Early-upper Miocene rhyolitic tuff beds in trough-inner-slope flysch basins are the first
sign of volcanic arc activity. They cannot have come from the Quaternary Taupo Volcanic
Zone, but were probably derived from the upper Miocene Coromandel Arc further to the
NW. As subduction is oblique, strike-slip movement has moved the forearc region east of
the Axial Ranges from a position closer to the Coromandel Arc to its present situation.

The present-day boundary between the Pacific (Fig. 1; Walcott 1978, 1979; H a t h e r t o n 1978;
and Australian plates passes through the New Oliver & Coggon 1979).
Zealand continental crustal block. To the north The New Zealand crustal block once formed
of New Zealand the plate boundary follows the part of Gondwana. It started to separate from
structurally relatively simple Tonga-Kermadec the Australian-Antarctic block in the Campa-
Volcanic Arc system, and to the south the more nian (Christoffel & Falconer 1972; McDougall
complicated Macquarie Ridge system, which & van der Lingen 1974; Weissel & Hayes 1977).
joins the present-day spreading ridge between Spreading between New Zealand and Australia
the Australian-Pacific and Antarctic plates in a ceased c. 60 Ma ago, but spreading between
triple junction (Christoffel 1971; Christoffel & New Zealand and Antarctica is still continuing.
Falconer 1972; Hayes & Talwani 1972; Weissel Plate tectonic reconstructions based on
& Hayes 1972). Nowadays it is generally marine geophysical data go back to the Cre-
accepted that the plate boundary in the New taceous (Fig. 1; Pittman et al. 1968; Molnar et
Zealand area consists of two opposing subduc- al. 1975; Weissel et al. 1977; Barron & Harrison
tion systems, the North Island and Fiordland 1979). The further back in time, the greater the
ones, connected by the Alpine Fault system degree of uncertainty in the reconstructions.

259
260 G. J. van der L i n g e n

.%
J ~ I165 i ) h 90

-I .~bg-

g ~
SOUTH PlJI

,YH
2 ,., ~ BASIN

'+~J \ ~ j ) TUb .~ ~ ..,

I n" ,o ~ _

./ <<3~. 4s-

,
TA~MAN -- / ( • • ~ ~ ./
..., ., / )[> J> \ ~ P
BASIN~ / .#/, v / / / / I
t.i /'~ CAMPBELL ~ +- ~" ../I
~ - 7 _ ~ ', // ~ ~,,/ l
.oX << ./ A
-"-~ / ,/# ~ ~/"° ~ / / 1

t.\,_..,)\ l ~\ i,,I\ \/ ,~
\. ~, ~ / / 's~ \ \...2.v" t t'~
\ ~l ~ 1155 I I ~165 t " I ~ l / 1180 i 1190
1800-- Bathyme~ric contours (in m)
A - - ..L~... Subduction plate boundary
27 Magnetic anomaly
• Andesite volcano
:::::::::::::::::::::::::;::::H Fossil spreading ridge
. . . . . Fracture zone /k Submarine volcano

Piate boundary (~ Finite pole for last 21 Ma

Fro. 1. Present-day plate tectonics setting of New Zealand in the Southwest Pacific. The line A-B-C
connects three arbitrary points on the Pacific plate. A'-B'-C' is the position of this line 10 Ma ago,
and A"-B"-C" 21 Ma ago. VMFZ--Vening Meinesz Fracture Zone. Seafloor magnetic anomaly data
after Christoffel & Falconer (1972), Weissel & Hayes (1972, 1977), and Malahoff et al. (1981).
 Development of the North Island Subduction System 261
Trying to match the geology of New Zealand Island has been part of the forearc region of all
with the various 'geometric' reconstructions has six volcanic arcs, and that these arcs rotated
provided problems, many of which still await anticlockwise away from their position above
solution. However, for the last 38 Ma (since the North Island Benioff zone by a combination
magnetic anomaly 13), it has been possible to of NNE-trending dextral strike-slip movement
determine the positions of the relative poles of and sinistral displacement along the Vening
rotation between the Pacific and Australian Meinesz Fracture Zone (Fig. 1; van der Linden
plates with a reasonable degree of reliability 1967).
(Walcott 1978, 1979). During this period the This paper presents a review and discussion
pole migrated southwards, from a position near of current ideas on the North Island Subduction
the Bounty Trough to its present position at System. A brief description of its various geolo-
latitude - 6 2 . 0 ° + 1.8 ° , longitude 174.3°+1.8 ° gical zones and geophysical elements will first
(Chase 1978; see also Walcott 1979). The posi- be given, together with a discussion of their
tion of this migrating pole, close to the plate likely position and role in the subduction sys-
boundary through New Zealand, has resulted in tem, followed by a summary and discussion.
a very complex structural history of the New The volcanic arc terminology of Seely (1979) is
Zealand crustal block. The southwards pole- followed as closely as possible.
shift was accompanied by a gradual southward
propagation of subduction into the North
Island of New Zealand. This subduction is The North Island Subduction System
oblique to the main geophysical, structural, and
volcanic elements (Figs 1 & 2; Ballance & The forearc region of the North Island Subduc-
Reading 1980). The further southward the pole tion System is bounded to the west by a line of
migrated, the larger the compressional (subduc- andesite volcanoes of the Taupo Volcanic
tion) component and the smaller the strike-slip Zone, and to the east by the Hikurangi Trough
component of plate movement became. (Fig. 2). Its maximum width is 270 km. From
The volcanic arc characteristics of the North west to east the forearc region can be subdi-
Island were first described in detail by Dickin- vided into the following zones; Axial Ranges,
son & Hatherton (1967), Hatherton & Dickin- East Coast Depression, Coastal Ranges, and
son (1968), and Hatherton (1969, 1970a, b). the continental shelf and slope.
Since then the overall picture has been changed
in minor detail only. Taupo Volcanic Zone
Estimates of the time of initiation of sub-
duction underneath the North Island vary. Volcanic rocks of the Taupo Volcanic Zone
Ballance (1976) suggested 20 Ma, and Hay- are calc-alkaline, and consist of rhyolites, ande-
ward (1979) 22 Ma, both based on the first sites, dacites and basalts (Cole 1979--Dr Cole's
occurrence of island-arc type volcanics. paper appeared after completion of the manu-
Sameshima (1975) gives 19 Ma, calculated from script of my paper). Volumetrically, rhyolite
spreading rates and the maximum depth of the extrusives form the dominant component
Benioff zone. Walcott (1978), using similar, but (approximately 16 000 km3; Challis 1978).
more up-to-date geophysical data, calculated at Widespread ignimbrite sheets and thick ash
least 15 Ma. deposits have blanketed the surrounding area
A speculative reconstruction of the subduc- and partly filled a tectonic graben in Permian to
tion history of the North Island was made by Jurassic sedimentary rocks of the Torlesse
Ballance (1976), who distinguishes six succes- Supergroup (Figs 2 & 3). Rhyolitic volcanism
sive, geographically separate volcanic arcs; the commenced c.1 Ma ago. Andesite volcanism
Waitakere Arc (20-15 Ma), the Northland Arc began in the Late Quaternary, about 50 000 BP
(18-15 Ma), the Coromandel Arc (15-6 Ma), (Cole 1979); although volumetrically far less
the Whitianga Arc (6-3 Ma), the Tauranga Arc important (about 800 km3), it forms spectacular
(3-0.75 Ma) and the present-day Taupo Arc active volcanoes. A few dacite domes and some
(from 0.75 Ma, but others have suggested that isolated basalt flows are volumetrically insigni-
the Taupo Arc had started somewhat earlier, ficant.
e.g. Challis 1978; Cole 1979). The first five, The Taupo Volcanic Zone is situated approx-
now extinct, arcs are aligned in a NNW direc- imately 120-150 km above the Benioff zone
tion, whereas the Taupo Arc (or Taupo Volca- (Fig. 4). Based on geochemical evidence, Ewart
nic Zone) trends NNE, close to the direction of et al. (1977; see also Cole 1979) considered the
the Tonga-Kermadec Arc. Ballance (1976) sug- andesitic lavas of the Tonga-Kermadec-North
gested that the east coast area of the North Island volcanic arc to be derived from Benioff
262 G. J. van der Lingen

\
176 1 3----~ Section line iFig. 3)
Permian-Jurassic
/",<>% ', <,, ('" Torlesse Supergroup " )

I,',,~£ / o~ , Cretaceous- Paleogene

36-
(" Basement ")
I\ 'q""~/~
IX Id I~
11 I-C, Neogene flysch basins

~-3s
Neogene sediments

Quaternary sediments

~ Rhyolite

~.~...0."
- ~ Andesite

WHITE ISL. •

+ +

+ + + +-I-..i_/
:::: :,;,;,;L::::::: ;:
-38

i,,. I.
-..
.i ..."i,II'',;'......
i
• MT. EGMONT

• Andesite volcano
Fault ' i

* M~lange zone

..~ Shelf edge

' .~ N
.....
Foot of continental slope
......,~
( _ ..) Slope basin .- .. ¢j 42 ~
Ridge crest ¢/ ¢/~ / /
~ Submarine canyon
~ ::i ;II
- . ~ Seamount

% •

.." .. "
.
•
.

.
.
..
.

20 0 20 6O
I J
1
I

173
'1 ~.e Y 'i 179
 Development of the North Island Subduction System 263

zone melting, with extensive modification by emerge. The depression was never very deep,
low-pressure crystal fractionation processes. and shallow-marine sediments were deposited
They believe the rhyolites of the Taupo Volca- until the middle of the Pleistocene (Grant-
nic Zone to be the product of fusion of the Taylor 1978), after which it filled with abundant
crustal continental rocks caused by the magmas gravels from the rapidly rising ranges, and
rising from the Benioff zone. ceased to be marine. The basement of the
Another Quaternary andesitic volcanic area, depression is known only from a few drill holes
the Egmont Volcanic Chain, exists in the south- sunk in the Ruataniwha Plains to the SW of
west of the North Island, well outside the Hawke Bay (Fig. 2). They proved the existence
Taupo Volcanic Zone (Fig. 2; Challis 1978). It of Torlesse rocks below mid-Pliocene sediments
consists of three volcanoes, of which Mt (Grant-Taylor 1978).
Egmont has been active in historic time. These In the Eketahuna area (Fig. 2), mid-
andesite volcanoes have a higher K-content Pleistocene to Quaternary sediments uncon-
than those of the Taupo Volcanic Zone. This is formably overlie Cretaceous to lower Pliocene
in agreement with the theory that the K-content sediments (including Pliocene flysch sedi-
is directly related to the depth of the Benioff ments), which, in turn overlie Torlesse rocks
zone below the volcanoes (Hatherton & Dick- (Neef 1974). It may be that, structurally, the
inson 1968). The Benioff zone is situated be- Eketahuna area originally belonged to the
tween 150 and 250 km below Mt Egmont Coastal Ranges, later to be incorporated in the
(Adams & Ware 1977). East Coast Inland Depression.
No Quaternary basin sediments occur north
Axial Ranges and NW of Wairoa but may have been removed
by erosion, middle Miocene to Pliocene sedi-
The Axial Ranges consist of complexly de-
ments form a broad syncline, called the 'Wairoa
formed, well-indurated geosynclinal sediments
Basin' (Figs 2 & 3; Grindley 1960). To the south
of the Permian to Jurassic Torlesse Supergroup.
this syncline disappears underneath the Quarter-
During much of the Tertiary, the Axial
nary strata of the East Coast Inland Depression.
Ranges were submerged. Parts started to
Modern sedimentological data on the Wairoa
emerge at the beginning of the Pliocene. In
Basin are sparse, but Miocene sediments seem to
late Pliocene time some parts of the ranges
be mainly shelf sands and mudstones with minor
had acquired substantial relief (Grant-Taylor
algal limestones, grading into shallow-marine
1978).
Pliocene sediments (Grindley 1960). North of a
Apart from forming the Axial Ranges and
line due east through Gisborne, structures be-
the floor of the Taupo Volcanic Zone graben,
come very complicated. Tertiary strata have been
the Torlesse rocks also underlie at least part of
involved in several phases of d6collement tecto-
the East Coast Depression to the east of the
nics, as a result of uplift of areas to the NE (Ridd
Ranges (Beu et al. 1980). The eastern boundary
1964; Stoneley 1968).
of the Torlesse rocks is not known. In a volcanic arc context, the East Coast
In a volcanic arc context, the Torlesse rocks
Inland Depression probably represents a fore-
can be regarded as constituting the Arc Massif.
arc basin. The Wairoa Basin may represent an
East Coast Depression earlier phase of forearc basin development,
which gradually propagated southwards,
In between the Axial Ranges and the Coastal together with the southwards migrating subduc-
Ranges exists a structural depression (Figs 2 & tion system. I have adopted the term 'East
3). From Cook Strait to Wairoa, just north of Coast Depression' for both the East Coast
Hawke Bay, this depression is filled with Inland Depression and the Wairoa Basin.
Pliocene-Quaternary sediments. This part of
the depression is called the 'East Coast Inland Coastal ranges
Depression' by Beu et al. (1980). It started to
subside in mid-Pliocene time, when the Axial A series of small Neogene flysch basins,
and Coastal Ranges at either side began to bounded by narrow zones of highly deformed

Fro. 2. Geological map of the North Island Subduction System. ab---Akitio Basin; CP--Castlepoint;
E--Eketahuna; eb---Eketahuna Basin; EC--East Cape; G--Gisborne; mb---Makara Basin; MP--
Mahia Peninsula; N--Napier; RP--Ruataniwha Plains; tb--Tawhero Basin; W--Wairoa; wab---
Wairoa Basin; wb--Whareama Basin. Insert: zonations of the North Island Subduction System, and
locations of section lines (Fig. 3). Offshore data after Lewis (1980), and Katz & Wood (1981).
264 G. J. van der Lingen
EAST COAST DEPRESSION
TAUPO VOLCANIC ZONE AXIAL RANGES WAIROA BASIN
(Volcanic arc) _ , (arc massif).i ~ (forearc basin)
ko < ",--. -~ ~ _ - - --0

~ O
i
20
i
km

EAST COAST DEPRESSION COASTAL RANGES

AXIAL RANGES (forearc basin) (structural high)
(arc massif) MAKARABASIN (SB)

(~ 0L 10km

• Permian-Jurassic
COASTAL RANGES
(structural high) ~ Cretaceous-
TAWHERO BASIN (SB) WHAREAMABASIN Paleogene
om 'ss' 0
~ ' / > : / / . ; ~ Neogene
I"<7 / / /
0:///
0 10 km ~ Neogeneflysch
secs bB SE
L F-""~ SB
I Rhyolite

sees SE

~ - ~ SB

HIKURANGI
2 t__~/ ' '..........
\
/ TROUGH
I-~._ / \

s;,~:s
SB
. . . . ~ , k ~ SB HIKURANGI TROUGH CHATHAMRISE
3- \'~ ~ ' ~ ~ FF channel
"~-~_~-
,.-I-., ~'--._~~ ~ ~ ~~

7 ('~, / ~ _

FIG. 3. Sections across the North Island Subduction System. For locations, see insert of Fig. 2.
Section 1 after Grindley (1960): section 2 after van der Lingen & Pettinga (1980) and Grant-Taylor
(1978); section 3 after Johnston (1975); sections 3-6 are tracings of offshore seismic profiles recorded
by Mobil Oil Corp. Vertical scale is two-way travel time. SB--slope basin; SE--shelf edge;
FF--frontal fold.
 Development of the North Island Subduction System 265

"" A \ \ 4' B o ,oo oo.. I

,~. .:......... , i
Depth in km ~'" " • ~ l• O ~ -50
• 33-so ,., -. "~ • ,,~ o

• ,, ,oo .~½~iP-~ • •
.,o,_,,o \ f. --~':;" ,,r..._ % \ k ~. { ( \ ~ ' % - 3 / 2 " I

.
t.
• o •
dl°o
\
\
",7
\
/Mo
~,,, "<-t.
t
._ 4" • )- oI .... t

• • \ ~% iiitt
FiG. 4. Geophysical maos of the North Island Subduction System. (A) Deep earthquake loci. After
Walcott (1979). (B) Isostatic gravity anomalies and projected surface extension of the Benioff zone
(dashed line). After Hatherton & Syms (1975), and Reilly et al. (1977). C---cross-section D-E (Fig.
4A), showing the Benioff zone and the cluster of very deep earthquake foci. Projected earthquake
foci are from area between dashed lines in (A). After Adams & Ware (1977).

Cretaceous-Palaeogene rocks, is exposed in a sedimentation) pebbly mudstones (debris flow

coastal belt between Cape Palliser and Hawke deposits, derived from the thrust zones), and
Bay (Fig. 2). In recent years, one of the flysch rhyolitic tuff beds (as derived from the volcanic
basins, the Makara Basin, has been studied in arc) (Fig. 5). Microfauna indicate an outer-shelf
detail in the light of modern ideas about deep- to upper bathyal palaeodepth for the basin.
sea sedimentation processes and plate tectonics Descriptions from the literature and my re-
(van der Lingen & Pettinga 1980). Some of the connaissance fieldwork suggest that the other
other basins have been described in the context flysch basins have similar settings. Several more
of general mapping only: the Akitio Basin m61ange zones have been found (Figs 2 & 5)
(Lillie 1953), the Whareama and Tawhero and no doubt others exist elsewhere. A striking
Basins (Johnston 1975), and the Eketahuna feature everywhere in the Coastal Ranges is the
Basin (Neef 1974). pronounced unconformity between the highly
Based on tectonic setting and basin-fill litho- deformed Cretaceous-Palaeogene 'basement'
logies, van der Lingen & Pettinga (1980) inter- and the Neogene sediments (Fig. 3). This un-
preted the upper Miocene Makara Basin as a conformity is interpreted as marking the onset
basin which originated on the trough-inner- of subduction in this area.
slope (for reasons to be explained later, the The flysch of the Makara Basin does not
term 'trough' will be used instead of 'trench') of contain typical submarine-fan facies (Mutti &
a subduction complex, in accordance with the Ricci-Lucchi 1972), such as thickening- and
model proposed by Moore & Karig (1976). The thinning-upward, or coarsening- and fining-
basin measures about 20 by 30 km, and is upward sequences (Fig. 5; van der Lingen &
divided in half and bounded by narrow zones of Pettinga 1980). The same is true for flysch
highly deformed (in part m61ange) Cretaceous- strata observed in other flysch basins (Fig. 5).
Palaeogene rocks (so-called 'thrust zones'; Fig. This feature may be typical for trench-inner-
3). Thrust faults associated with these thrust slope basins in general, and could be a function
zones dip to the west. The basin sediments of the small size of the basins, in which sub-
consist of flysch strata (mass-flow deposits: marine fans cannot fully develop.
turbidites to grain flows, derived from up- North of Hawke Bay, relationships between
slope), hemipelagic mudstones (normal slope flysch basins and 'basement' are less clear,
 I'..J

!i~i~ii~i~i~i~i~i
~,~,

¢..,,.

0~

FIG. 5. Field photographs from the Coastal Ranges. (A) Flysch strata in the Makara Basin. (B) Flysch strata in the Whareama Basin. (C) Pebbly
mudstone (debris flow) intercalated in flysch strata of the Makara Basin. Pebbles are of Cretaceous-Tertiary lithologies. Diameter of lens cap is 55
mm. (D) M61ange zone exposed on shore platform, north of Castlepoint. Loose blocks represent a variety of Upper Cretaceous and Palaeogene
lithologies, left behind on the shore platform as erosional remnants.
 Development o f the North Island Subduction System 267

partly due to the earlier mentioned d6colle- piston and gravity cores and dredge samples
ment tectonics. But flysch sequences similar to (Lewis & Kohn 1973; Katz, 1975).
those in the Makara Basin are exposed on Morphologically, the slope consists of NE
Mahia Peninsula, and Cretaceous-Palaeogene aligned narrow, elongate ridges and basins, cut
m61ange zones, surrounded by d6collement perpendicularly by several submarine valleys
units, have been described north of Gisborne and canyons (Fig. 2; van der Linden 1968;
(Ridd 1964). On Fig. 2 I have therefore tenta- Lewis & Kohn 1973).
tively extended the boundary line between the Seismic profiles show that the ridges are
East Coast Depression and the Coastal Ranges actively growing structures and that sediments
north of Hawke Bay. are being ponded in small basins formed by the
By analogy with the Makara Basin, the other ridges (Fig. 3). Similar structures exist under-
Neogene flysch basins in the Coastal Ranges neath the planed shelf. Attitudes of faults are
are interpreted as (fossil) trough-inner-slope difficult to determine on the profiles, which has
basins as well. led to different structural interpretations. Lewis
Because of the intense deformation of the (1980) interprets the highs as imbricate thrust
Cretaceous-Palaeogene basement rocks, it is zones controlled by west-dipping faults, in
difficult to reconstruct the palaeogeography of accordance with generally accepted ideas about
the east coast area at the onset of subduction. A the development of subduction complexes
reconstruction has only been attempted for the (Seely 1979). Katz & Wood (1981), in contrast,
Lower Cretaceous, by Moore & Speden (1979). interpret the highs as being controlled by tecton-
But lithological characteristics of Upper Cre- ics caused by diapiric movements of overpress-
taceous and Palaeogene sedimentary rocks can ured clays; the influence of overpressured clays
provide some clues as to palaeoenvironments. has been suggested also for the Makran subduc-
All Cretaceous-Palaeogene sediments were de- tion complex (White 1979). Whatever the inter-
posited in a marine environment (Kingma pretation, the seismic profiles are very similar
1971). In Upper Cretaceous time the dominant to those from trench-inner-slopes of other sub-
lithologies were glauconitic sandstones, carbo- duction complexes (e.g. the Sunda Arc: Karig
naceous and siliceous mudstones, conglomer- et al. 1979, the Makran Arc: White 1979, and
ates, and flysch sediments. In Palaeocene- the Aleutian Arc: von Huene & Shor 1969 and
Eocene time sedimentation was generally slow, Grow 1973).
and the sediments fine-grained. Locally, res- In situ deformation of slope basement litholo-
tricted circulation conditions prevailed (e.g. the gies by diapiric movements seems well estab-
'Waipawa Black Shales'). Typical lithologies lished (Katz & Wood 1981) but compressional
are massive mudstones and bentonites. Depth deformation no doubt also plays an important
of deposition was probably shelf to upper bath- role. For instance, as in other subduction com-
yal. In the Oligocene there is a gradual change plexes (e.g. the Makran Arc, White 1979,
to more rapid, coarser-grained sedimentation, 1981), several profiles show a so-called 'frontal
accompanied by a greater variety of environ- fold', representing the first compressional fold
ments, probably shelf with localized deeper of trough (trench) sediments at the foot of the
basins. A possible modern analogue for this continental slope (Fig. 3).
environment could be the continental border- Slope basin and trough sediments consist of
land that exists offshore from California. hemipelagic muds, turbidites, and tephra
The area of the Coastal Ranges was uplifted layers. One dated ash layer (3400yr) was
to shallow-marine depth in the early Pliocene, encountered in several slope basins as well as in
forming a 'structural high' (Seely 1979). Barna- the Hikurangi Trough (Lewis & Kohn 1973). In
cle and coquina limestones were deposited un- basins high up on the slope, up to nine turbi-
conformably on this high and along the borders dites exist above the ash layer, while in the
of the East Coast Depression, which started to trough only two turbidites occur above the
form in this same period (Beu et al. 1980). The same ash layer. This suggests that most down-
Coastal Ranges finally emerged in the Pleis- slope-moving mass flows are being trapped in
tocene. slope basins, but that some reach the trough
(most probably via submarine canyons).
Continental shelf and slope The sediment types filling the slope basins
thus seem to be similar to those of the Neogene
The structure of the offshore area is known flysch basins in the Coastal Ranges, apart from
from seismic profiles (Figs 2 & 3; Lewis 1971, pebbly mudstones, which were not encountered
1980; Katz 1975; Katz & Wood 1981), and data in the short piston cores.
on lithologies come from a limited number of Dredge samples have shown that the struc-
268 G. J. van der L i n g e n

tural highs consist of older strata (Katz 1975; to the Hikurangi Trough (Adams & Ware
Lewis 1980). 1977).
The continental shelf and slope thus display A small cluster of very deep earthquakes was
many of the features typical of trench-inner- recorded, at about 600 km depth (Fig. 4;
slopes of subduction systems elsewhere. Adams & Ware 1977). It has been suggested
Together with the Coastal Ranges they form a that these earthquakes originated in a detached
subduction complex (Seely 1979). lithospheric slab (Barazangi et al. 1973; Christ-
offel & Calhaem 1973).
The Benioff zones of the North Island and
Hikurangi Trough
Kermadec systems are not continuous (Eiby
Morphologically, the Hikurangi Trough is 1977). In a transitional zone between 35 ° and
not a trench (Katz 1974). It only represents the 37 ° south latitude, the strike changes from
deepest area where the gently dipping abyssal N35.5E (North Island) to N18E (Kermadec). In
plain of the Pacific plate meets the foot of the that transitional zone is a gap in deep earth-
continental slope. In this configuration it has quake foci, where the Kermadec Benioff zone
acted as a valley along which sediment gravity shallows from 500 to only 200 km. In this gap
flows are being channelled. Most profiles across the earthquakes are also of smaller magnitude.
the trough show a shallow channel (Fig. 3). The
average water depth of the trough is 3000 m. Gravity anomalies
Although the North Island Subduction Sys-
tem is generally regarded as a continuation of Gravity anomaly maps of the North Island
the Tonga-Kermadec system, there is no bath- have been produced by Reiily et al. (1977). The
ymetric continuity between the Kermadec isostatic anomalies show a negative belt extend-
Trench and the Hikurangi Trough (Katz 1974; ing from East Cape, via Hawke Bay to Cook
Katz & Wood 1981). The area directly south of Strait (Fig. 4). Its maximum approximately
the termination of the Kermadec Trench is coincides with the surface intersection of the
characterized by several seamounts, probably Benioff zone. Its southern boundary roughly
consisting of oceanic igneous material (Fig. 2, coincides with the termination of the Benioff
Katz & Wood 1981). zone.
'Classical' subduction trenches (like the Ton- A positive gravity anomaly covers the Taupo
ga-Kermadec Trench) are characterized by two Volcanic Zone. A large part of the Coastal
associated geophysical features: the projected Ranges is also covered by a positive anomaly,
continuation of the Benioff zone cuts the sur- which extends eastwards over the Hikurangi
face near the trench, and a negative gravity Trough (Fig. 4).
anomaly is situated above the trench (Hather- There is no continuity with the negative
ton 1969). However, in the North Island this gravity anomaly belt of the Kermadec system.
geophysical association occurs about 200 km to Like the two Benioff zones, the two belts differ
the west of the Hikurangi Trough (Fig. 4). in strike (Hatherton 1970a). The North Island
Sensu stricto, the Hikurangi Trough therefore negative belt ends in a circular low with a
cannot be considered a subduction trench. minimum of - 1 3 0 regal, just north of East
Cape, and is separated from the Kermadec
system by a saddle of only - 2 5 mgal (Fig. 4;
Benioff zone
Hatherton & Syms 1975). This saddle lies in the
The Benioff zone underneath the North Is- transitional zone between the two Benioff
land has been described in detail by Hatherton zones. There is also a dextral offset between the
(1970a, b) and Adams & Ware (1977). A two anomaly belts.
well-defined seismic zone, dipping at about 50 ° The gravity anomalies thus have a spatial
to the west, has a maximum depth of 350 km at relationship with the Taupo Volcanic Zone and
its northern end and gradually shallows along the Benioff zone which is typical for subduction
strike to about 200 km at its southern end, systems.
where it terminates underneath the northern
part of the South Island (Fig. 4). This shallow-
ing and termination has been explained as being Discussion
a function of the position of the pole of rotation
in relation to the plate boundary (see Walcott That subduction of the Pacific plate underneath
1978). Earthquake foci shallower than 50 km the North Island takes place is now fairly well
show a diffuse pattern extending from the top established. Calculations derived from strict
of the Benioff zone (:~t about 50 km depth) geometric plate tectonic reconstructions suggest
 Development o f the North Island Subduction System 269

that subduction commenced near the beginning underlying the offshore slope basins and form-
of the Neogene. Subduction along the Tonga- ing the ridges. It therefore seems likely that the
Kermadec Arc began somewhat earlier, about subduction complex for the larger part is being
25 Ma ago, and it is thought that subduction formed by sediments from a pre-existing
propagated southwards from there into the borderland that, passively riding on the Pacific
New Zealand continental crustal block, follow- plate, is being squeezed together. The border-
ing a southward migration of the relative pole land area was probably much wider originally
of rotation. than the present-day subduction complex. This
Many geological and geophysical aspects of process, of course, can also be regarded as a
the North Island, both on land and offshore, form of accretion. It is possible that sediments
are in agreement with characteristics of subduc- deposited in the Hikurangi Trough and on the
tion systems in other parts of the world. Several Pacific Plate are being accreted to the lower
NE-trending geological zones can be inter- part of the subduction complex, adjacent to the
preted as representing specific elements of a trough. Alternatively, substantial amounts of
subduction system. Volcanic arc: the Taupo sediments could have been subducted, as has
Volcanic Zone; arc massif: rocks of the Torlesse been suggested for other subduction systems
Supergroup, forming the Axial Ranges, and (e.g. Moore et al. 1979). The necessary data are
underlying the Taupo Volcanic Zone and part lacking to solve this problem.
of the East Coast Depression; forearc basin: the The style of deformation of the subduction
East Coast Depression; subduction complex: complex is, however, similar to that of 'classic-
the Coastal Ranges (structural high), and con- al' subduction complexes. The 'accreting' rocks
tinental shelf and slope (trough-inner slope). A are being strongly deformed, often into
Benioff zone underlies the volcanic arc and arc m61ange zones, forming narrow, elongate
massif, and a negative gravity anomaly exists thrust ridges on the trough-inner-slope, creat-
near the projected surface extension of the ing small flysch basins in which sediments mov-
Benioff zone. ing downslope are being ponded. Some sedi-
In some aspects, however, the North Island ment gravity flows reach the Hikurangi Trough
Subduction System differs from subduction sys- via submarine canyons, bypassing the slope.
tems elsewhere. There is no subduction trench Most authors have considered the North Is-
coinciding with the negative gravity anomaly land and Kermadec subduction systems as
and the projected surface extension of the being one continuous system. Although they
Benioff zone. The forearc basin and subduction are closely related, there is a definite break
complex are situated to the east of a line at between the two. The Hikurangi Trough does
which the subducting plate starts its descent not form a continuation of the Kermadec
into the mantle. For a distance of over 200 kin, Trench, and their relative positions within the
the forearc basin and subduction complex are two systems is quite different. In a transitional
riding passively on top of the WSW-wards zone north of East Cape the strike direction
moving Pacific plate. Where the Pacific plate changes, and breaks exist between the Benioff
disappears underneath the subduction complex, zones and the negative gravity anomaly belts.
the continental slope and abyssal plain form the The gravity belts are also offset dextrally. As
Hikurangi Trough. In relation to the Benioff this transitional zone is situated along the direc-
zone and the negative gravity anomaly, the tion of the Vening Meinesz Fracture Zone, it
Hikurangi Trough is not a subduction trench, seems likely that the break between the two
but in relation to the subduction complex it systems is an expression of this fracture zone,
could be considered as such. This poses the dividing the Pacific plate into two parts.
problem where exactly to draw the plate bound- Authors have used two different sets of data
ary. to determine the onset of subduction below the
Although the Coastal Ranges, shelf and con- North Island, the depth of the Benioff zone
tinental slope are interpreted as a subduction (Sameshima 1975; Walcott 1978), and the first
complex, no evidence has been found on land occurrence of island arc-type volcanics
that oceanic sediments, scraped off the Pacific (Ballance 1976; Hayward 1979). The latter is
plate, have been incorporated into the subduc- the less reliable, as the relationship of the
tion complex. Lithological characteristics suggest volcanics with the present-day west-dipping Be-
that we are dealing with sediments originally nioff zone is purely speculative. The Waitakere
deposited in a continental borderland setting. Arc, for instance, could well be related to the
Continuity in structural style from the Coastal detached slab at 600 km depth. It is unlikely
Ranges to the continental shelf and slope sug- that this detached slab originally belonged to
gests that the same may be true for the rocks the west-dipping Benioff zone, as it is situated
270 G. J. v a n d e r L i n g e n

to the east of the lower end of the Benioff zone brings the subduction complex closer to the
(Fig. 4C). It could well be a remnant of an Coromandel arc back in time. Any rotational
originally east-dipping Benioff zone. This movement, however, of the early Neogene
would require a 'flip' in the subduction direc- Arcs or the forearc region, remains highly
tion, but similar flips have been suggested for speculative. Opening of the South Fiji Basin
segments further to the N W along the same cannot be invoked, as spreading in this basin
plate boundary, e.g. the New Hebrides and had already terminated before the end of the
Solomon Islands systems (Coleman 1975). This Oligocene (Watts et al. 1977; Malahoff et al.
hypothesis requires further testing. 1981).
A timing problem is presented by the rhyoli- According to Walcott (1978), the strike-slip
tic tuff beds in the Neogene flysch basins of the movement was mainly taken up by faults along
Coastal Ranges. It is assumed that these tufts the eastern side of the Axial Ranges. The
are derived from a volcanic arc. In which case forearc region to the east of the Axial Ranges
the Taupo Volcanic Zone cannot be the source, thus moved southwards during the Neogene,
as the first tuff beds recorded are of early- while being compressed due to the gradually
late Miocene age, whereas the rhyolitic erup- increasing compressional component of Pacific
tions of the Taupo Volcanic Zone did not start plate movement.
until the Pleistocene. Of the several Neogene
volcanic arcs proposed by Ballance (1976), the
Coromandel Arc was active when the tuff beds ACKNOWLEDGMENTS: The author benefited from dis-
were deposited. In a speculative reconstruction, cussions with Drs R. I. Walcott, F. J. Davey, I.
Ballance (1976) juxtaposes the Coastal Ranges Speden, and R. A. Wood. Messrs R. A. Cook and T.
opposite the Coromandel Arc in late Miocene Montague assisted with the interpretation of the
time. Recent fission-track dating of some Coro- seismic reflection profiles. Mr B. D. Field helped
mandel rhyolites gave a late Miocene age with the reconnaissance fieldwork. The figures were
(Rutherford 1978). But more work, especially drafted by Mr E. T. H. Annear. The photographs
geochemical, is required to test this hypothesis. were developed and printed by Mr A. Downing and
Miss Carol Hulse. Miss Christina Johnstone typed the
As subduction is oblique, a strike-slip compo- various versions of the manuscript. Drs J. D. Brad-
nent has to be taken into account in any spatial shaw, S. D. Weaver, R. I. Walcott, M. G. Laird and
reconstruction. Over the last 10 Ma, about H. R. Katz, and Mr T. Montague critically read the
190 km of dextral strike-slip parallel to the manuscript and made suggestions for its improve-
strike of the Benioff zone, has taken place. This ment.

References
ADAMS, R. D. & WARE, D. E. 1977. Subcrustal Series Map. No. 13. N.Z. geol. Surv., Dept Sci.
earthquakes beneath New Zealand; locations Industr. Res., Wellington.
determined with a laterally inhomogeneous CHALLIS, G. A. 1978. Upper Quaternary volcanism.
velocity model. N.Z. J. Geol. Geophys. 20, In: SUGGATE,R. P., STEVENS,G. R. & TE PUNGA,
59-83. M. T. (eds). The Geology of New Zealand, 2,
BALLANCE,P. F. 1976. Evolution of the Upper Ceno- 651-70.
zoic Magmatic Arc and plate boundary in north- CHASE, C. G. 1978. Plate kinematics: the Americas,
ern New Zealand. Earth planet. Sci. Lett. 28, East Africa and the rest of the world. Earth
356--70. planet. Sci. Lett. 37, 253-68.
BALLANCE, P. F. & READING, H. G. (eds). 1980 CHRISTOFFEL, D. A. 1971. Motion of the New
Sedimentation in Oblique-slip Mobile Zones. Zealand Alpine Fault deduced from the pattern
Spec. Publ. Int. Assoc. Sediment. 4. of sea-floor spreading. Bull. R. Soc. N.Z. 9,
BARAZANGI, M., ISACKS, B. L., OLIVER,J., DUBOIS,J. 25-30.
PASCAL, G. 1973. Descent of lithosphere -- & CALHAEM, I. M. 1973. Upper mantle viscosity
beneath New Hebrides, Tonga-Fiji, and New determined from Stokes' law. Nature (phys. Sci.)
Zealand: evidence for detached slabs. Nature. 243, 51-3.
London, 242, 98-101. -- & FALCONER, R. K. H. 1972. Marine magnetic
BARRON, E. J. & HARRISON, C. G. A. 1979. Recon- measurements in the south-west Pacific Ocean
structions of the Campbell Plateau and the Lord and the identification of new tectonic features.
Howe Rise. Earth planet. Sci. Lett. 45, 87-92. In: HAYES, D. E. (ed). Antarctic Oceanology II,
BEU, A. G., GRANT-TAYLOR, T. L. & HORNIBROOK, the Australian-New Zealand Sector. Am. geo-
N. DE B. 1980. The Te Aute limestone facies, phys. Union. Ant. Res. Ser. 19, 197-209.
Poverty Bay to northern Wairarapa. Misc. COLE, J. W. 1979. Structure, petrology, and genesis
 Development of the North Island Subduction System 271
of Cenozoic volcanism, Taupo Volcanic Zone, Margins, Mem. Am. Assoc. Petrol. Geol. 29,
New Zealand--a review. N.Z. J. Geol. Geophys. 223-37.
22, 631-57. KATZ, H. R. 1974. Margins of the south-west Pacific.
COLEMAN,P. J. 1975. On island arcs. Earth Sci. Rev. In: BURK, C. A. & DRAKE, C. L. (eds). Geology
11, 47-80. of Continental Margins, 549-65. Springer-
DICKINSON, W. R. & HATHERTON, T. 1967. Andesite Verlag, Berlin.
volcanism and seismicity around the Pacific. Sci- - - 1975. Ariel Bank off Gisborne. An offshore late
ence, 157, 801-3. Cenozoic structure, and the problem of acoustic
Emv, G. A. 1977. The junction of the main New basement on the East Coast, North Island, New
Zealand and Kermadec seismic regions. Int. Zealand. N.Z. J. Geol. Geophys. 18, 93-107.
Syrup. Geodyn. SW Pacific, Noumea, New Cale- - - & WOOD, R. A. 1981. Submerged margin east
donia, 1976, 167-78. Technip, Paris. of the North Island, New Zealand, and its petro-
EWART, A., BROTHERS, R. N. & MATEEN, A. 1977. leum potential. CCOP/SOPAC, Techn. Bull. 3.
An outline of the geology and geochemistry, and KINGMA, J. T. 1971. Geology of Te Aute subdivision.
the possible petrogenetic evolution of the volca- Bull. N.Z. geol. Surv. 70, 173 pp.
nic rocks of the Tonga-Kermadec-New Zealand LEWIS, K. B. 1971. Growth rate of folds using tilted
island arc. J. Volcan. Geotherm. Res. 2, 205-50. wave-planed surfaces: coast and continental
GRANT-TAYLOR, T. L. 1978. The geolegy and struc- shelf, Hawkes' Bay, New Zealand. In: Recent
ture of the Ruataniwha Plains. In: SMALL et al. Crustal Movements. Bull. R. Soc. N.Z. 9, 225-
(eds). Geology and Erosion in the Ruahine 31.
Range. Rep. N.Z. geol. Surv. G20, 31-61. 1980. Quaternary sedimentation on the Hikur-
GRINDLEY, G. W. 1960. Sheet 8 Taupo, 1st ed. angi oblique-subduction and transform margin,
Geological Map of New Zealand 1:250 000. Dept New Zealand. In: BALLANCE, P. F. & READING,
Sci. Industr. Res., Wellington. H. G. (eds). Sedimentation in Oblique-slip
GRow, J. A. 1973. Crustal and upper mantle struc- Mobile Zones. Spec. Publ. Int. Assoc.
ture of the central Aleutian arc. Bull. geol. Soc. Sediment. 4, 171-89.
Am. 84, 2169-92. & KOHN, B. P. 1973. Ashes, turbidites, and
HATHERTON, T. 1969. The geophysical significance of rates of sedimentation on the continental slope
calcalkaline andesite in New Zealand. N . Z . J . off Hawke's Bay. N . Z . J . Geol. Geophys. 16,
Geol. Geophys. 12, 436-59. 439-54.
1970a. Upper mantle inhomogeneity beneath LILLIE, A. R. 1953. Geology of the Dannevirke
New Zealand: surface manifestations. J. geo- subdivision. Bull. N.Z. geol. Surv. 46, 156 pp.
phys. Res. 75, 269-84. McDOUGALL, I. & VAN DER LINGEN, G. J. 1974. Age
1970b. Gravity, seismicity, and tectonics of the of the rhyolites of the Lord Howe Rise and the
North Island, New Zealand. N . Z . J . Geol. Geo- evolution of the south-west Pacific Ocean. Earth
phys. 13, 126-44. planet. Sci. Left. 21, 117-26.
1978. Astride two plates--an account of current MALAHOFF, A., FEDEN, R. H. & FLEMING, H. S. 1981.
tectonics in New Zealand. Endeavour, 2, 180-5. Magnetic anomalies and tectonic fabric of mar-
& DICrINSON, W. R. 1968. Andesite volcanism ginal basins north of New Zealand. J. geophys.
and seismicity in New Zealand. J. geophys. Res. Res., in press.
73, 4615-19. MOLNAR, P., ATWATER,T., MAMMERICKX,J. & SMITH,
& SYMS, M. 1975. Junction of Kermadec and S. M. 1975. Magnetic anomalies, bathymetry,
Hikurangi negative gravity anomalies (note). and the tectonic evolution of the South Pacific
N.Z. J. Geol. Geophys. 18, 753-6. since the late Cretaceous. Geophys. J. R. astron.
HAVES, D. E. & TALWANI, M. 1972. Geophysical Soc. 40, 383-420.
investigation of the Macquarie Ridge Complex. MOORE, G. F. & KARIG, D. E. 1976. Development of
In: HAYES, D. E. (ed.). Antarctic Oceanology II, sedimentary basins on the lower trench slope.
the Australian-New Zealand Sector. Am. Geo- Geology, 4, 693-7.
phys. Union, Ant. Res. Ser. 19, 211-34. MOORE, J. C. et al. 1979. Off Mexico. Middle Amer-
HAVWARO, B. W. 1979. Eruptive history of early to ica Trench. Rept. Leg 66 IPOD. Geotimes, Sept.
mid Miocene Waitakere Volcanic Arc, and 1979.20-2.
palaeogeography of the Waitemata Basin, north- MOORE, P. R. & SPEDEN, I. 1979. Stratigraphy,
ern New Zealand. J. R. Soc. N.Z. 9, 297- structure and inferred environments of deposi-
320. tion of the Early Cretaceous sequence, eastern
JOHNSTON, M. R. 1975. Sheet N159 and part sheet Wairarapa, New Zealand. N.Z. J. Geol. Geo-
N158 Tinui-Awaitoitoi, ist ed. Geological Map phys. 22, 417-33.
of New Zealand 1:63 360. Dept Sci. Industr. MuTrI, E. & Rlcci-LuccnI, F. 1972. Le torbiditi
Res., Wellington. dell'Apennino settentrionale: introduzione
KARIG, D. E., SUPARKA, S., MOORE, G. F. & all'analisi de facies. Mere. Soc. geol. Ital. 11,
HEHANUSSA, P. E. 1979. Structure and Cenozoic 161-99.
evolution of the Sunda Arc in the central NEEF, G. 1974. Sheet N153 Eketahuna, 1st edn.
Sumatra region. In: WATKINS, J. S., MONTADERT, Geological Map of New Zealand 1:63 360. Dept
L. & DICKERSON, P. W. (eds). Geological and Sci. Industr. Res., Wellington.
Geophysical Investigations of Continental OLIVER, G. J. H. & COGGON, J. H. 1979. Crustal
272 G. J. van der L i n g e n

structure of Fiordland, New Zealand. Tecto- obliquely convergent plate boundary. In:
nophysics, 54, 253-92. BALLANCE, P. F. & READING, H. G. (eds).
PIrrMAN, III, W. C., HERRON, E. M. & HE1RTZLER,J. Sedimentation in Oblique-slip Mobile Zones.
R. 1968. Magnetic anomalies in the Pacific and Spec. Publ. Int. Assoc. Sediment. 4, 191-215.
sea-floor spreading. J. geophys. Res. 73, 2069- VON HUENE, R. & SHOR, JR. G. G. 1969. The struc-
85. ture and tectonic history of the eastern Aleutian
REILLY, W. I., WHITEFORD, C. M. & DOONE, A. 1977. trench. Bull. geol. Soc. Am. 80, 1889-902.
North Island, 1st ed. Gravity Map of New WALCO~, R. I. 1978. Present tectonics and Late
Zealand, 1:1 000 000, Isostatic Anomalies. Dept Cenozoic evolution of New Zealand. Geophys.
Sci. Industr. Res., Wellington. J. R. astron. Soc. 52, 137-64.
RIDD, M. F. 1964. Succession and structural inter- - 1979. Plate motion and shear strain rates in the
-

pretation of the Whangara-Waimata area, GiN- vicinity of the Southern Alps. In: WALCO~, R. I.
borne, N . Z . N . Z . J . Geol. Geophys. 7, 279-98. & CRESSWELL, M. M. (eds). The Origin of the
RUTHERFORD, N. F. 1978. Fission-track age and trace Southern Alps. Bull. R. Soc. N.Z. 18, 5-12.
element geochemistry of some Minden Rhyolite Wa'FrS, A. B., WEISSEL,J. K. & DAVEY,F. J. 1977.
obsidians. N.Z. J. Geol. Geophys. 21,443-8. Tectonic evolution of the South Fiji marginal
SAMESHIMA,Z. 1975. Silica indices of volcanoes in and basin. In: TALWANI, M. & PITMAN III, W. C.
around New Zealand with reference to volcanic (eds). Island Arcs, Deep Sea Trenches and Back-
zones in the North Island. N.Z.J. Geol. Geo- arc Basins. Am. Geophys. Union, M. Ewing Ser.
phys. 18, 523-39. l, 419-27.
SEELY, O. R. 1979. The evolution of structural highs WEISSEL, J. K. & HAYES, D. E. 1972. Magnetic
bordering major forearc basins. In: WATKINS, J. anomalies in the southeast Indian Ocean. In:
S., MONTADERT, L. • DICKERSON, P. W. (eds). HAYES, D. E. (ed). Antarctic Oceanology H: the
Geological and Geophysical Investigations of Australian-New Zealand Sector. Am. geophys.
Continental Margins. Mem. Am. Assoc. Petrol. Union, Ant. Res. Ser. 19, 165-96.
Geol. 29, 245-60. 1977. Evolution of the Tasman Sea reappraised.
STONELEY,R. 1968. A lower Tertiary d6collement on Earth planet. Sci. Lett. 36, 77-84.
the East Coast, North Island, N . Z . N . Z . J . Geol. - , HAYES, D. E. & HERRON, E. M. 1977. Plate
-

Geophys. II, 128-56. tectonics synthesis: the displacements between

VAN DER LINDEN, W. J. M. 1967. Structural rela- Austra|ia, New Zealand and Antarctica since the
tionships in the Tasman Sea and south-west late Cretaceous. Mar. Geol. 25, 231-77.
Pacific Ocean. N.Z.J. Geol. Geophys. 10, 1280-- WroTE, R. S. 1979. Deformation of the Makran
301. continental margin. In: FARAlt, A. & DE JONG, K.
-- 1968. Cook bathymetry, N.Z. Oceanogr. Map A. (eds). Geodynamics of Pakistan, 295-304.
Chart. Oceanic Series 1:1 000 000. Dept Sci. Geol. Surv. Pakistan, Quetta.
Industr. Res., Wellington. WroTE, R. S. 1981. Deformation of the Makran
VAN DER L|NGEN, G. J. & PETrlNGA, J. R. 1980. The accretionary sediment prism in the Gulf of Oman
Makara Basin: a Miocene slope-basin along the (north-west Indian Ocean) (this volume).
New Zealand sector of the Australian-Pacific

GERRIT J. VAN DER LINGEN, Sedimcntology Laboratory, New Zealand Geological

Survey, University of Canterbury, Christchurch, New Zealand.
 The Barbados Ridge Complex: tectonics of a mature forearc
system

G. K. Westbrook

SUMMARY: The Barbados Ridge Complex is the wide accretionary sediment pile
associated with the Lesser Antilles island arc. Its width is so great (>200 kin) that a trench
no longer exists on the oceanward side of it. This development is a product of the age of
the system (>50 Ma) and the thickness of sediment on the ocean floor (0.8 km in the
north, >4 km in the south). The northward decrease in elevation of the sediment pile and
the variation in the style of initial deformation at the leading edge of the pile are related to
the northward change in sediment thickness and type. The region in which deformation is
prevalent has a westward limit just west of the axis of the minimum negative Bouguer
gravity anomaly. In the south the axis of this minimum is coincident with the Barbados
Ridge (an outer 'sedimentary arc'). The deformed rocks of the accretionary pile are
overlain by later sediments, which show varying degrees of deformation, and often occupy
small basins.
In the northern part of the complex, the relief and structure of the accretionary pile are
complicated by ridges and troughs running east-west across the general trend of structures.
These appear to be related to variations in the relief of the subducted oceanic basement.

The Barbados Ridge Complex is a region with and magnetic anomalies (Westbrook 1975).
complicated bathymetry lying along the eastern A n o t h e r possible contributory cause of the
margin of the Lesser Antilles Island arc in the magnetic anomalies in the vicinity of Barbados
eastern Caribbean (Fig. 1). It takes its name may be diagenetic magnetite formed in sedi-
from the northerly trending Barbados Ridge ments from reduction of hydrated ferric oxides
which lies 150 km east of the arc. The island of by petroleum microseepage (Donovan et al.
Barbados is its most elevated part. The complex 1980); hydrocarbons are found in Barbados.
generally deepens towards the east, and most The overall thickness of the complex in-
bathymetric features have a northerly trend, creases towards the south, and broadly reflects
except between latitudes 13°30'N and 15°30'N a similar southward thickening of undeformed
where easterly trending ridges and troughs sediment on the Atlantic ocean floor. Along the
occur. The complex becomes broader and more eastern margin of the complex, the sediments of
elevated towards the south (Fig. 2). the Atlantic ocean floor are progressively de-
The considerable quantity of geophysical in- formed as they pass into the complex, eventual-
formation obtained in the region shows that the ly losing any structure that can be recognized
complex is composed of sedimentary rocks from those seismic reflection sections that are
above a linear depression in the igneous base- currently available. The deformed sediment is
ment (Fig. 3). The axis of this depression, overlain by more recent sediment which thick-
which reaches a depth of 20 km beneath Barba- ens towards the west, and is absent in the
dos, lies up to 50 km west of the eastern limit of easternmost part of the complex. North of 14°N
seismicity associated with the subduction zone this overlying layer shows gentle deformation.
beneath the Lesser Antilles, and is where the South of 14°N, the Tobago Trough, a well-
crystalline crust of the Atlantic Plate passes developed basin containing a thickness of at
beneath the crystalline crust of the Caribbean least 3 km of undeformed sediment, lies be-
Plate. The axis is strongly out of isostatic tween the Barbados Ridge and the volcanic
equilibrium, being depressed beneath its island arc. Similar, but smaller, basins lie to the
equilibrium position. The eastern part of the east of the Barbados Ridge. On the island of
complex is above its equilibrium position and Barbados, strongly folded and faulted Eocene
corresponds to the outer trench rise of most flysch (the Scotland Group) is overlain by pela-
island arcs which do not have such a broad gic sediments of upper Eocene to Miocene age
sediment complex. The existence of meta- (Saunders 1979). Pleistocene to recent coral
morphosed sediments beneath the Barbados rock covers most of the island. The trend of
Ridge (7-10 km beneath Barbados) is indicated deformational structures on Barbados is gener-
by seismic velocities of around 5.0 km sec -1 ally NE.

275
276 G. K. Westbrook

65 64 63 62 61 60 59 58 57 56 55 54

/ \.
20 o /.~.20
(-, • -" "C-__
i0 , ,~:

19 .... ~--/)(-s-( ~ 19

2 o
%

• , ~

17

16
- --..
....x
~ )..
J5 L \

14 ,~-"
• ;" ,'
\

y,--~
• :.- ~,, ",/ ,' . , : : : • ,..:..,:.
12 .~ e

--~" '- ~ ~ % .-~-~ 7,',',',~

I0 iiii~iiiIiiii,iiii "~~-~ "',,

-.:"ii. : ...i.)i . . i . . . ~ '~."~.~

8 " " "' ~
65 64 63 62 61 60 59 58 57 56 55 54

Fi6. 1. Eastern Caribbean showing: bathymetry from U.S.G.S. Geologic-Tectonic Map of the
Caribbean Open File 75-146, & Kearey et al. (1975) (isobaths at 1 km intervals); lines of profiles 1-7
of Fig. 2; the eastern limit of the Barbados Ridge Complex (solid black line); the region of the initial
slope of the complex (close stipple); the western limit of deformation (dashed line); the Barbados
Ridge (open stipple); the line of section for Fig. 3 (y); the lines of the seismic reflection sections of
Fig. 5 (g) and Fig. 6 (a,b,c,d,e,f).
 The B a r b a d o s R i d g e C o m p l e x 277

I reported by Chase & Bunce (1969), who drew

an analogy with the sandbox experiment of
~,~ ..,,~i,i~ ~E 11.5:~ Hubbert (1951) to explain the mechanism of
thrusting and folding. Westbrook et al. (1973)
2 subsequently suggested that the deformation
was brought about by successive thrusts ema-
nating from a master d6collement close to the
sediment/basement interface accompanied by
shearing parallel to bedding planes, and pro-
ducing a reverse stratigraphy of thrust slices in
the complex which is characteristic of models
published for other forearc regions (Seely et al.
1974; Karig & Sharman 1975). This was in-
corporated later into an evolutionary model for
the whole of the complex by Westbrook (1975).
The style of deformation varies considerably
along the eastern margin (Peter & Westbrook
1976; Mascle et al. 1977; Biju-Duval et al.
1978). In the south it takes the form of gently
asymmetric east-facing folds (amplitude
0.5 km; wavelength 8 or 9 km) riding on thrusts
,Jo 2~o 3~o 0~0 5~o dipping westward at 20 ° which probably have a
km listric form at depth, presumably becoming
parallel to the basement as they run into a
Fro. 2. Bathymetric profiles across the Bar-
bados Ridge Complex showing the varia- major d6collement (Fig. 6d,e,f). North of
tion in its width and shape. The left edge of 15°30'N the deformation of sediment is so
each profile ends at the active volcanic intense and/or chaotic that no structure can be
island arc. The arrows mark the limits of recognized in it from seismic reflection sections.
the region in which deformation can be This deformed material, however, overlies un-
seen on seismic reflection sections. The deformed bedded sediment on a plane of dis-
positions of the profiles are shown on continuity (probably a d6collement) which lies
Fig. 1. at shallow depth within the undeformed
sedimentary section (Fig. 6a,b). A much
Eastern margin smaller part of the ocean sediments is involved
in the initial deformation (a layer 0.2 km thick
Between 17°30'N and 12°N the distance of the as opposed to 1.5-2 km in the south). The
eastern edge of the complex from the Lesser plane of discontinuity extends at least 30 km
Antilles increases from 200 to 450 km. Similarly beneath the deformed sediment, and how it
the distance between the axis of the Bouguer does so without being deformed does pose
gravity anomaly minimum (lying over the line some problems (Peter & Westbrook 1976; Mas-
of subduction of the Atlantic crystalline crust) cle et al. 1977). If the deformed material was
beneath the Caribbean crystalline crust) and essentially a submarine slide formed by the
the eastern edge increases from 95 to 265 km. gravitational collapse of the frontal slope of the
The increase in width of the complex is gradual, complex it would explain the lack of disruption
except at 15°N where the edge runs WNW for beneath the discontinuity and the gentler gra-
about 90 km because of the influence of a ridge dient of the frontal slope in the region where
in the oceanic basement. South of this change in this discontinuity exists. However, reflection
strike of the edge of the complex, the margin is profiles run by Lamont-Doherty (Ewing et al.
typically in the form of a slope rising from the 1974) and N O A A (Peter & Westbrook 1976
Atlantic ocean floor by about 1.5 km in 50 km and unpublished data) show that this discon-
then flattening off or becoming less steep. tinuity exists over such a wide area (between
North of the change in strike, the slope is more 15°30'N and 16°40'N) that a submarine slide,
gentle, 0.5 km in 50 km, until the Puerto Rico although possible, is an unlikely explanation.
Trench is reached, where it becomes steeper Given that the plane of discontinuity is likely to
again (Fig. 2). be a d6collement, then the presence of a parti-
The deformation of the sedimentary horizons cularly weak horizon above the d6collement,
of the Atlantic ocean floor at the eastern mar- probably with a more competent (more lithified)
gin of the Barbados Ridge Complex was first sequence of sediments lying beneath it, is re-
 OO

AVES RIDGE GRENADA TROUGH ST. V I N C E N T TOBAGO TROUGH BARBADOS

(extinct arc) (back arc basin'} (active arc~ (fore arc has.n] (fore arc ricl~e~,

undeformed sediment
Consolidated diorite or on ocean floor
voloanlos $ atn crustal edifice of island a r c ~x~ dacite
. pluton
'\\ tavas & undelorrr~d
i sed,ment small tore-arc basin trench-slope break (mainly turbidites)
~ ~ sedimentH . 7 . ..............:.......... ,,oy,ocl ....

L_._~,~_ _ ~ = ............ .~:i~UlnlllllllilmJlllllllllllllllllmrmlmnlll -~.:-:.:. . : .

"~~ " :::.%- .. : ::: :::::;: ~ ',,, // /
, /
# j ,~, decollement / ,/
eastward verging ' '
~~EEE!E~':}~-~ anomalously thick chamlaer .... old basement • metamorphosed thrusts and folds /" 'oceanic layer 2
Depth '~.~;=:=~ ~,!-i]]!!::i~E:' ocean crust _ . and deformed sediments in deformed sediment /
I~m oceanic layer 3
lower crust of island arc oceanic J~"f- tgneous
Moho layer 3 & basic & ultra basic cumulates ~?; Moho
intrus,ves

ascending magma ~ = transition of

and volatdes ~ abb o to ecloglte

I i L I i ] L i , i J I
100 200 300 400 500 600 700

Km east of long 64°W

FIG. 3. Interpretative cross-section of the Lesser Antilles and the Barbados Ridge Complex; from Westbrook (1975) and Boynton et al. (1979).
Position of section is shown in Fig. 1 (line y).
 The Barbados Ridge Complex 279

quired to enable it to persist so far without sion of westward directed thrusts (Westbrook
disruption. 1975; Mascle et al. 1981). Almost without ex-
There is a general correlation between the ception, the youngest sedimentary horizons are
thickness of sediment on the Atlantic Ocean affected by this deformation.
floor and style of deformation. The thinner the
sediment the more intense is the deformation,
because the stresses are distributed in a smaller Southern margin
volume of material. This feature of the Barba- In the south the Barbados Ridge Complex
dos Ridge Complex has been used by Moore abuts the continental margin of South America.
(1979) in developing a model for deformation in Although much of it has been explored for
forearc sediments, but thickness is not the only petroleum, the structure of the region where
factor. Lithological variation also plays a part, the two provinces merge is not well known.
as discussed above for the region north of East of Trinidad, the deformed margin of the
15°30'N where the development of a d6colle- Barbados Ridge Complex passes into the unde-
ment high in the sedimentary column results in formed passive margin of eastern South Amer-
only one-quarter of the sediment being in- ica. The area of transition is characterized by
volved in the deformation at the front of the diapiric structures, which can be seen on reflec-
sediment pile. The sediments in the north are tion profiles (Lowrie & Escowitz 1969; Ewing et
further from continental sources than those in al. 1974; Mascle et al. 1981). Whether these
the south (Bunce et al. 1971; Peter & West- result from salt as on parts of the African
brook 1976) and likely to have a higher clay Margin or from overpressured shales, such as
content. Consequently they may deform in a far those producing mud volcanoes on Trinidad
less competent fashion, especially if pore water (Arnold & Macready 1956) is not known, but
is trapped in them, causing them to become the latter seems more likely.
overpresssured. In the west, as mentioned above, the forearc
basin beneath the Tobago Trough continues
Western margin south-westward until it pinches out between the
metamorphic rocks of the Venezuelan Coast
The sediments of the forearc region wedge out Range and the south-westward extension of the
against the Lesser Antilles island arc. Up to volcanic arc.
120 km from the active arc the sediments are In the region of Tobago there is a sharp drop
completely undeformed except for some occa- in bathymetry from the continental shelf down
sional slumping. South of 14°N this zone of to the Barbados Ridge and the continuity of
undeformed sediments exists in the Tobago structure from the Barbados Ridge into the
Trough which can be followed round to the shelf is very unclear, although some authors
south-west passing beneath the Venezuelan claim it exists (Weeks et al. 1971). Seismic
continental margin as far as the island of Mar- refraction and reflection show that rocks with a
garita (Feo-Codecido 1977). The thickness of high seismic velocity occur close to the surface
undeformed sediment exceeds 6 km to the east at the southern end of the Barbados Ridge
of Grenada. North of 14° the undeformed sedi- (Ewing et al. 1957; Edgar et al. 1971); these
ment does not occupy a bathymetric basin could be the Cretaceous metamorphic rocks
except opposite Dominica where the so-called that crop out on Tobago and Trinidad. These
Lesser Antilles Trench lies at the foot of the rocks deepen towards the north. The trend of
slope down from the arc. Much of the sediment structures in Trinidad suggest that even if there
comes from the island arc, but a proportion of is no direct continuity of structures between the
clay and silt size material comes from the Barbados Ridge Complex and the South Amer-
Amazon and Orinoco (Keller et al. 1972). The ican margin, then they have been produced in a
sediments become deformed as they pass east- similar tectonic situation. There is no very
ward into the Barbados Ridge Complex. The obvious evidence for any significant transcur-
limit of deformation lies just to the west of the rent faulting along the northern margin of the
axis of the negative Bouguer gravity anomaly present continental shelf.
(Fig. 4). In the east of the Tobago Trough the
deformation has been accompanied by con-
siderable uplift to form the Barbados Ridge. Variation in structure along the
The deformation is in the form of gentle west- Barbados Ridge Complex
ward facing folds, although on the western
flank of the Barbados Ridge there is the im- The overall elevation of the complex decreases
plication that they are the near surface expres- towards the north and the Barbados Ridge does
280 G. K. Westbrook

61 6O 59 58 57 56
62 I !
18
I I I
~b ~118
Q

17
17

B.R.

16 16

15
15
,.,
" °.,o,

a
\

14
L
.+ 14
i
]

<b
13 13

,I~, ,1 I~ ]'.I"." '

b ,, 7

lYI~ 12
12

11

...... .-.j ~iii p

. 2
I,\

10 I \\'~I I I I II10
62 61 60 59 58 57 56
F m 4. Lateral variations in the Barbados Ridge Complex. Northward facing slopes ('steps') are
stippled. Ridges in the oceanic basement are shown in close diagonal hatching where they form
bathymetric features and in open hatching where they are known from seismic profiles to exist
beneath the ocean floor. Hatching is not continued where the ridges extend beneath the Barbados
Ridge Complex, as these areas are already stippled. The eastern margin of the complex is shown by
the solid line with triangles on the overthrusting side. The large dots mark the axis of the negative
Bouguer gravity anomaly. B.R.--Barracuda Ridge; T.R.--Tiburon Rise; G--Guadeloupe; D - -
Dominica; M--Martinique; L--St Lucia; B--Barbados; V--St Vincent; T.T.--Tobago Trough;
Gr--Grenada; To--Tobago; Tr--Trinidad.
 The Barbados Ridge Complex 281

150 ~00 250 3oo 350 4oo 450

km

60.5"

FIo. 5. Line drawing of seismic reflection section across the Barbados Ridge Complex (g in Fig. 1)
from Westbrook (1975).

6 -ii!i!iii!i!i!iiii!i!iiiiiii!i!i!iii!!:i:i:i:!::::::::.:.:.... .... -

7--

"-~iiiiiiiiiiiiliiiiiii!iiii:iiiiiiiiiii!!iiiii!iiiiiiiiiiiiii iiiiiiiiiW-
~ ~ i ~ ! i ! ! ~ [ ~ ! ~ i ~ i i I N ~ I ~ I ~ i i ~ [ ~ ! ~ -
l I

a c
z - ~ - 6 -.--':-Y'~

t _ _ J

d
I-- 3--

a _..~, _ e

_" ,~ ,.'?~,'/;',~,_3G~
,v . ~ . ~'.~ :
7-- -- -- 7-- " ">~---==,'.
_j__
~- L
%-
~- - - --

8--i.-.. I_ 8-1 ~-

f deformed
sediments
6-- ,

oceanic
8- ......... basement
9-

I0--I i_ i I
I ~ j
10 km scale b a r

FIG. 6. Line drawings of seismic reflection sections across the eastern margin of the Barbados Ridge
Complex. (a) and (d) from Peter & Westbrook (1976), (b) and (f) from Biju-Duval et al. (1978), (c)
and (e) from Chase & Bunce (1969). The zones which are stippled on sections (a), (b) and (d)
strongly reflect seismic energy, but the pattern of reflections lacks coherency. Pecked lines show the
interpreted positions of thrusts.
282 G. K. Westbrook
not continue further north than 14°50'N (Fig. angle of slope of the ridge and the angle
2). This appears to be principally related to the between the strike of the ridge and the direction
northward decrease in thickness of the ocean of convergence.
floor sediments (>4 km at 12°N; 1 km at 16°N). A significant aspect of this process is that
This northward decrease in elevation is not structures with a trend oblique to the main
uniform however (Fig. 4). South of 13°N the structural trend are produced without any
elevation of the complex is fairly uniform at change in the overall geotectonic setting. No
around 2000 m, east of the Barbados Ridge, it north-south regional stresses are required other
then descends at 13°N and 13°50'N to 3500 m, than the reaction to the stresses produced at the
the biggest drop coming at 13°50'N. It descends ridges. Between 13°50'N and 15°20'N the com-
again at 15°50'N to around 5000 m and at plex is dominated by WNW to west trending
17°10'N to 5600 m. These 'steps' which all have ridges and troughs superimposed upon the
a WNW trend are associated with ridges in the northerly trending features characteristic of the
oceanic basement (Peter & Westbrook 1976). southern part of the complex. The northern-
The northernmost of these is the Barracuda most ridge is that associated with the Tiburon
Ridge which extends at least 100 km beneath Rise. The sediment in this ridge shows no clear
the complex (Schubert 1974), the next south is reflecting horizons by comparison with the
the Tiburon Rise at 15°20'N. Both these ridges ridges and troughs south of it, implying recent
have a parallel trend (WNW) and run beneath deformation by the oblique convergence of the
the 'steps' in the complex, which have the same Tiburon rise (Peter & Westbrook 1976, line
trend. At 13°55'N the ridge is not expressed as a M). The ridge east of Martinique may also be
bathymetric feature on the seafloor east of the associated with a basement ridge, but there is
arc, but its presence is shown on seismic reflec- no clear geophysical evidence to support its
tion profiles and by a linear positive gravity existence, except that the NW flank of this
anomaly (Westbrook 1975). ridge is the site of a positive gravity anomaly
The ridges can influence the complex in which locally modifies the form of the main
various way. (Fig. 7). They control the thick- negative gravity anomaly. The blanket of unde-
ness of sediment on the ocean floor and conse- formed sediment which overlies the tectonized
quently provide lateral control on the amount rocks of the complex in this region of cross-
of accreted sediment in the complex. When the trends is displaced vertically by topography
ridges first encounter the complex, their effect (Peter & Westbrook 1976). The troughs are not
must be to uplift material above them. filled by more recent sediment and the dis-
After some time this initial effect will be placed sediment layer shows no variations in
reversed by the accretion of thicker sediment thickness related to topography, implying fairly
from the troughs flanking the ridges. With recent movement. By contrast, in the area
turbidite sediments coming from the south, south of 13°30'N, the blanket of overlying
sediment thickness south of each ridge will tend sediment is undisturbed and the troughs are
to be greater than to the north, and this would sites of recent sedimentation.
lead to a northward decrease in accreted sedi- The buried basement ridge at 13°55'N is
ment across each ridge. probably responsible for the relatively high
Another important effect is produced by the elevation of Barbados and the portion of the
difference between the trend of the ridge and Barbados Ridge just north of it, compared with
the direction of convergence of the ridge with the southern part of the ridge. It is possible that
the complex. As the ridge moves obliquely into it has been a significant feature since the late
the complex the sediment will tend to be swept Eocene producing the southward dipping
to one side by its 'plough' action. Earthquake palaeoslope on which the Oceanic Formation of
first motions and the orientation of Caribbean Barbados was deposited (Lohmann 1973).
plate boundaries indicate that the direction of
convergence between the Caribbean and the
Atlantic is easterly (Jordan 1975; Tomblin,
pets. comm.). The trend of the ridges is 290 °, Development of the Barbados Ridge
which give a difference in angle of 20 °. This Complex
produces a southward sweeping action which
builds up material to the south of the ridge The Barbados Ridge Complex is the widest
producing compressional structures semi- example of an accretionary sediment pile
parallel to the ridge (Fig. 8). The cross-stress associated with an island arc. This implies that
produced by the ridge comes from the reaction some features of accretionary sediment piles,
normal to its surface, which depends on the which may only be nascent in many arcs, should
 a b

~ " -~:'~ / - ~ ~-,uk.~ ~

-. ~ - ~ ; " \, - -~_~__ . :..~-:

¢ d
r~

•~ - ~ . ~.j~-"~ .~ .~ ~.,~,~. ,

~ • - '" i :=.:"

Fx6.7. The influence of a ridge in oceanic basement upon a forearc complex:

(a) Ridge approaching the edge of the forearc complex• The direction of convergence is the same as the strike of the ridge.
(b) After coming beneath the edge of the complex the ridge locally uplifts the part of the complex above its crest.
(c) After further subduction the effect of the relief of the ridge is outweighed by the greater contribution of sediment to the complex from either
side of the ridge which builds the complex further out than over the ridge, where little sediment is added•
(d) If there is a greater thickness of sediment on one side of the ridge than the other, because of its acting as a barrier to turbidites, then one side
of the complex will grow out further than the other, leading to a change in strike of the front across the ridge. t~
L~
284 G. K. W e s t b r o o k

0 b
I ~ ,

t t

y ~ -X
"~;::"~"~- - y,

!, E

I'ii
' I I

It ;I

Ft6.8. The effect of a ridge in oceanic basement converging obliquely with a forearc complex:
(a) Ridge approaching the edge of the complex in a direction oblique to that of the strike of the
ridge.
(b) Shortly after passing into the complex the ridge begins to distort the form of the leading edge as
the point of entry of the ridge migrates along it.
(c) After further subduction, the southward sweeping action of the ridge has built up material on its
southern side. There should also be some depletion on the northern side, but this would not produce
such an obvious change in strike in the leading edge of the complex.
(d) Diagram showing the main elements of the situation. Y marks the end of the ridge. Y' marks the
point at which the crest of the ridge passes under the leading edge of the complex. This has migrated
along the leading edge, from X, the point at which the ridge entered the complex. Z is the zone in
which most of the compressional structures produced by the ridge exist.

be fully developed in the Barbados Ridge Com- exists in the southern part of the complex, and
plex. O n e of these is the trench-slope break, the even in the region where it exists the topogra-
break of slope at the top of the inner trench- phy and structure of the area between it and the
wall which is very often the highest point of an front of the complex is so varied that it is
accretionary sediment pile. The outer conceptually difficult to identify the whole of it
sedimentary ridges of arcs such as the Java- as the inner trench wall. The 'trench-slope
Sumatra, have also been described as the break' of the Barbados Ridge complex lies
trench-slope break (Karig et al. 1980). The 50-60 km west of its eastern edge (Figs 1 & 2).
Barbados Ridge, however, is clearly not the It is interpreted as representing the maximum
equivalent of a trench-slope break, since it only elevation attained by material involved in the
 The Barbados Ridge Complex 285

initial accretion process before gravitational Nicobar Islands and again 50 km from the
forces nullify the upward movement of material trench (Weeks et al. 1967). The Nicobar Islands
caused by the horizontal convergence. The and Nias are the equivalents of Barbados. In
opposition of gravitational body forces to tecto- the eastern Aleutians where the major break of
nically induced forces allows the complex to slope occurs 50-70 km from the trench, the
build forward rather than be pushed upward position of Barbados is analagous with that of
and backward. On a large scale the deformation Kodiak Island 100 km from the break of slope.
of the pile can be considered as a viscous or The presence of a large forearc basin between
plastic process and therefore the position of the Kodiak Island and the break of slope does not
trench slope break and the angle of the trench detract from the comparison. Forearc basins
slope will be a function of gravitational body occur between the Barbados Ridge and the
forces, tectonically induced stresses and the eastern edge of the complex. Although in Dick-
rheology of the rocks. This last is the most inson & Seely's (1979) classification of forearcs,
difficult to assess, because it varies with litho- the Barbados Ridge Complex is cited as an
logy, and the amount of deformation and meta- example of a narrow ridged forearc, it should be
morphism. Following from this, one can specu- classed as a broad ridged forearc. It is implicit
late that variations in the angle of the trench in the evolutionary model of Westbrook (1975)
slope on the eastern margin of the complex are that narrow ridged forearcs evolve into broad
related to lithological variations in the material ridged forearcs.
being accreted. Between the Tiburon Rise and It is common for geological models of forearc
the Barracuda Ridge where only the topmost ridges (e.g. Dickinson & Seely 1979; Seely
sediment has been stripped away from the 1979) to include rises of oceanic crust beneath
undeformed sedimentary column and intensely the ridges. The geophysical evidence is strongly
deformed, the angle of the trench slope is less against the existence of a rise of oceanic crust
steep than it is further south. beneath the Barbados Ridge (Westbrook 1975;
The Barbados Ridge has been formed by the Boynton et al. 1979), and is ambivalent for its
uplift produced by the horizontal compressional existence beneath the forearc ridge of the Java
forces in the complex. These arise principally Sumatra arc system (Curray et al. 1977; Karig et
from shear stresses applied to the base of the al. 1980; Kieckhefer et al. 1980).
sediment pile by the subducting ocean crust
which are opposed by the reaction of the island
Barbados and history of the complex
arc and consequently achieve a maximum value
immediately above the line along which the Barbados, the geology of which is succinctly
ocean crust of the Atlantic is subducted beneath summarized by Saunders (1979) is an important
the crystalline crust of the Caribbean. A wide source of information on the nature of the rocks
accretionary pile is subject to more horizontal comprising the Barbados Ridge Complex, but is
compression than a narrow one, because shear also a source of problems. The Scotland Forma-
stress has been applied over a greater surface tion (Pudsey & Reading 1981) comprises a
area. Also, a purely frictional model for impart- group of sediments of early Eocene (perhaps
ing the shear stress at the base of the accretion- Palaeocene) to middle Eocene age which can
ary complex is probably inadequate, as some be generally described by flysch. Current opin-
shear stress will be transmitted by viscous shear ion is that these rocks were deposited in deep
through deforming sediment. This time- water. Pudsey & Reading propose that they
dependent element in the deformational pro- were deposited as a deep sea fan in the former
cess implies that an increase in subduction rate trench of the Lesser Antilles arc. While there is
would increase the amount of compressional a strong likelihood of this, it is also possible that
stress acting on the accretionary complex. the fan was formed on the ocean floor and
engulfed later by the subduction complex. The
Comparison with other arcs very small amount of material in the formation
derived from a volcanic source lends some
If the Barbados Ridge Complex is taken as a support to the latter view.
model, then which other forearc systems pos- The Scotland Formation is quite strongly
sess a sedimentary ridge that is a separate deformed. Most of this deformation is tectonic,
feature from the trench-slope break? Off but some synsedimentary gravity-driven de-
Sumatra a major break of slope lies 30 km from formation has been recognized. Exploration
Nias Island and 50 km from the trench (Karig et wells show that there is tectonic repetition of
al. 1980), and further north along the same arc the Scotland Formation, at least four times to a
system, the break of slope is 80 km from the depth of at least 4.5 km (Baadsgaard 1960).
286 G. K. Westbrook
The strike of this deformation is NE, rather for the history of these movements is shown in
than the northerly trend expected from general Fig. 9.
consideration of the shape of the arc and the The great volume of material in the Barbados
direction of subduction. Herrera & Spence Ridge Complex is undoubtedly a consequence
(1964) suggested that this trend was the result of the proximity of the Lesser Antilles to
of emplacement of the Scotland Formation as sources of terrigenous sediment from the South
gravity slides from the south. The difficulty with American continent. How much was directly
this explanation is that after sliding into the deposited in a trench in front of the arc earlier
trench the material underwent relatively little in its history and how much was deposited on
deformation. The possibility that the deforma- the ocean floor can only be conjectured. A
tion reflects a southerly component in the direc- considerable proportion, however, must be
tion of subduction since the Cretaceous is sup- continental slope and rise deposits from the
ported by some plate tectonic reconstructions northern margin of South America, incorpo-
of the area (Malfait & Dinkleman 1972; Ladd rated as the Lesser Antilles moved eastward
1976) for the period from Late Cretaceous until along the margin (Fig. 10) (Chase & Bunce
late Eocene. Subsequent tectonic rotation of 1969). How far the Lesser Antilles have mi-
the small area of Barbados is a possible ex- grated past South America is undetermined,
planation which could be tested palaeomagneti- but if one follows Jordan (1975) this could be
cally. The Oceanic Formation (middle Eocene virtually the whole length of the Caribbean.
to Oligocene) and the Bissex Hill Formation (This is illustrated by Pudsey & Reading 1981.)
(early Miocene) which overlie the Scotland The tectonic history of northern Venezuela
Formation are also deformed, although by no suggests that it might not have moved as far as
means to the same extent as the Scotlands. this, however (Bell 1972; Maresch 1974). In the
They show the continuing influence of tecton- region currently occupied by the Barbados
ism, comparatively far from the leading edge of Ridge Complex, there must once have existed a
the complex. thick sedimentary cone built in front of a proto-
Reconstructions of plate motions and the Orinoco River system. The collision of the
stratigraphy of Trinidad suggest that Jurassic complex with this cone must have had a signifi-
and Cretaceous rocks should have been present cant influence on the development of the com-
on the ocean floor during the early history of plex. This collision could have begun at any
the arc, but they are not present in Barbados as time between the end of the Eocene and
might be expected from many models of sub- the middle of the Miocene, depending on the
duction complexes. Maybe these older sedi- model chosen for the development of the east-
ments were not scraped off, but taken further ern Caribbean (Fig. 9f).
down the subduction zone in the initial stages of
its formation, or exist further west in the com-
plex. Maturity of the Barbados Ridge
Barbados has undergone considerable verti- Complex
cal movement since the late Eocene when the
Oceanic Formation began to be deposited on Can the Barbados Ridge Complex be consi-
the Scotlands at a depth of 3500 m (Saunders dered to be a mature member of a sequence of
1979). In the early Miocene during the deposi- forearc evolution? Listed below are variables
tion of the Bissex Hill Formation the water which can influence the development of forearc
depth can have been as little as 300 m (Steineck complexes.
& Murtha 1981). In the middle Miocene dur- (1) Age of the subduction zone.
ing deposition of the Conset Marl the water (2) Rate of subduction.
depth was between 1000 and 1500 m (Steineck (3) Thickness of sediment:
& Murtha 1981). The seismic stratigraphy of (a) on ocean plate to be subducted,
the Tobago Trough suggests that Barbados has (b) in trench at the foot of the forearc.
risen about 3 km relative to the trough's centre (4) Lithologies of sediments on (a) and in (b).
since the Miocene. The vertical displacements (5) Shape of the basement surface.
are controlled by the subduction process, which (6) Sedimentation directly on to the forearc
over a long term would tend to increase the complex.
elevation of Barbados. On a shorter term a It is implicit in most models of accretionary
slowing of subduction would result in a rise of complexes that there will be growth but will this
the complex as the negative isostatic imbalance growth always proceed in the same way?
was reduced, whereas an increase in rate would Looked at in simple terms the volume of sedi-
initially depress the surface. A possible scenario ment in the complex should be the product of
 The Barbados Ridge Complex
volcanic are turbidites ®

pelagic sediments
/ ~
®
- - .

LATE ~ m e d sediment

©
--~.~.\:/.... :.:~: :.:,>~,>~.,.-~--:~
EARLY MIOCENE
~ isostatic uplift

~r downwarp

Q
LATE EOCENE? -'~ l
cone of Proto-Orinoco ?

EARLY MIOCENE?

I ~ I A I ~ I I I
0 I00 200 300 400 km

Fl6. 9. Evolution of Barbados Ridge Complex.

(a) Early Eocene: youthful stage in the evolution of the forearc which had a trench possibly with
direct deposition of Scotland Formation sediments into it.
(b) Late Eocene: region in which Barbados now is, no longer in active accretionary zone and
receiving pelagic sedimentation of Oceanic Formation. Elevated position keeps it free of terrigenous
sediments. The forearc may also have been far from a source of continentally derived material at this
time. This stage of development began in the mid-Eocene.
(c) Early Miocene: a hiatus in subduction has allowed the central part of the forearc complex to rise
by isostatic rebound. During this stage the Bissex Hill Formation of Barbados was laid down.
(d) Late Miocene: following the recommencement of subduction in the middle Miocene the central
part of the forearc has become depressed by the reimposition of the negative isostatic imbalance.
Some of the Conset Marl of Barbados was probably laid down at the beginning of this stage, as the
water began deepening.
(e) Present: continued accretion and consequent widening of the forearc have imposed more stress
on the centre, and caused it to rise forming the Barbados Ridge and Tobago Trough. Present rate of
uplift is about 400 m Ma -l.
(f) End Eocene---early Miocene?: at some time during the period, the Barbados Ridge Complex
encountered what was probably quite a thick cone of sediment produced by some proto-Orinoco
delta. This must have produced a rapid widening of the complex.
Note: The hiatus in subduction referred to in (c) and (d) may have been a considerable slowing
rather than a cessation.
288 G. K. Westbrook

VOLCANIC ~
A~C ----4 ~",. \

..... ~ ';;;' ., ,. . . . . , . . . . ..

+ + .:

-P + + + +
L

500 K rn +
+

F1G. 10. Migration of the Lesser Antilles island are eastward along the northern margin of South
America during the Tertiary. The continental slope deposits of the South American margin become
assimilated in the forearc complex. Eventually the sediment cone of the proto-Orinoco becomes
incorporated in the complex. The front of the complex is shown by a solid line with triangle. The line
of subduction of crystalline ocean crust beneath crystalline crust of Caribbean Plate is shown by a
dashed line.

variables (1), (2) and (3), plus (6), but other position of the outer structural high, from that
effects complicate this. Dewatering and com- proposed by Karig (1974b) in which sediment
paction of accreted sediment will reduce its accretes in a depression above the oceanic
volume. If the mean density of the Barbados basement and the trench slope break migrates
Ridge Complex is taken to be 2.35 Mg m -3 and outward. Evidence for Seely's model exists
the mean density of sediment before addition is from forearcs associated with continental mar-
1.9 Mg m -3, then the volume of sediment in the gins, e.g. Middle America, but not for intra-
complex is 0.66 of its original volume. Only part oceanic arcs. The Barbados Ridge Complex
of the sedimentary section on the subducted does not follow Seely's model.
oceanic lithosphere may be accreted in the Given the variety of situations in which arcs
forearc complex, the remainder being taken occur and the episodic nature of geological
down further into the crust and mantle. This is history it is obvious that the development of a
probably strongly dependent on lithology; ma- universal model for the evolution of forearcs
ture lithified rocks of biogenic origin being far which will predict what will occur under any
more likely to remain attached to the oceanic circumstances during the evolution of forearcs
lithosphere than overlying terrigenous sedi- may be an intractable problem. However, in a
ments which were probably only added to the situation where there is a plentiful supply of
section close to the subduction zone, under terrigenous sediment and a moderate rate of
most circumstances (Moore 1975). In some subduction, growth of the forearc complex is a
situations such as Peru-Chile it is possible that well-established consequence of continued sub-
the forearc complex has been 'eroded' at its duction, it is reasonable to suppose that the
base by the subducting lithosphere (Karig Barbados Ridge Complex is a mature forearc
1974a). As well as governing the rate at which complex, representing a state towards which
sediment enters the complex, subduction rate is forearcs will evolve, given sufficient time.
also important in controlling stress imparted to
a forearc complex, and consequently affecting
its shape and strain within it.
The shape of the basement must have a Conclusions
strong influence particularly on the early de-
velopment of the complex. The upthrusting of a (1) The Barbados Ridge Complex is an
slice of oceanic crust beneath which accreted accretionary forearc complex of broad
sediment is 'understuffed' (Seely 1979) pro- ridged type. Its width and thickness are a
duces quite a different evolution, with a fixed consequence of its long history (50 Ma)
 The Barbados Ridge Complex 289

and the thickness of terrigenous sediment fine significant deformation to the upper
on the ocean floor. part of the section.
(2) The changes in elevation along the com- (5) Increase in the width of the complex, and
plex are broadly related to variation in also perhaps, an increase in subduction
sediment thickness on the ocean floor, rate, has increased the horizontal stress
and are locally controlled by ridges in the in the complex. This has produced the
oceanic basement. uplift of the Barbados Ridge and the
(3) Oceanic basement ridges influence the westward verging structures on its west-
position of the leading edge of the com- ern flank.
plex and its thickness by laterally con- (6) Isostatic response to variations in sub-
trolIing the amount of sediment accreted, duction rate has superimposed changes in
and by deforming accreted material when the elevation of the Barbados Ridge
they pass into the complex at an angle upon the long-term uplift produced tec-
oblique to their strike. tonically.
(4) Style of deformation and accretion at the
leading edge of the complex are depen- ACKNOWLEDGMENTS: Some of the ideas concerning
dent upon sediment thickness and litho- basement ridges in this paper were originally dis-
cussed by George Peter and myself some years ago,
logy. Thicker sediment gives a more although not presented in our paper of 1976. I am
open style of deformation. More compe- grateful to Carol Pudsey and Harold Reading for
tent lithologies in the lower part of the a preprint of their paper on Barbados. Harold Read-
sedimentary section can restrict the ing and Casey Moore gave helpful reviews of this
amount of sediment accreted, and con- paper.

References
ARNOLD, R. & MACREADY, G. A. 1956. Island- DICKINSON, W. R. & SEELY, D. R. 1979. Structure
forming mud volcanoes in Trinidad, British West and stratigraphy of forearc regions. Bull. Am.
Indies. Bull. Am. Assoc. Petrol. Geol. 40, 2749- Assoc. Petrol. Geol. 63, 2-31.
58. DONOVAN, T. J., RANDAL, L. F. & ROBERTS, A. A.
BAADSGAARD,P. H. 1960. Barbados, W.I. Explora- 1980. Aeromagnetic detection of diagenetic
tion results 1950-1958. Int. Geol. Congr. Rep. magnetite over oil fields. Bull. Am. Assoc. Pet-
21st Session, 18, 21-7. rol. Geol. 63, 245-8.
BELL, J. S. 1972. Geotectonic evolution of the south- EDGAR, N. T., EWING, J. I. & HENNION, J. 1971.
ern Caribbean area. In: Studies in Earth and Seismic refraction and reflection in the Carib-
Space Sciences (Hess Vol). Mere. geol. Soc. Am. bean Sea. Bull. Am. Assoc. Petrol. Geol. 55,
132, 367-86. 838-70.
BIJU-DUVAL, B., MASCLE, A., MONTADERT, L. & EWING, J., EWING, M., WINDISCH, C. • ATHEN, T.
WANNESON,J. 1978. Seismic investigations in the 1974. Underway Marine Geophysical Data in the
Colombia, Venezuela and Grenada basins, and North Atlantic June 1961-January 1971. Parts E
on the Barbados Ridge for future IPOD drilling. and F: Seismic Reflection Profiles. Lamont-
Geol. Mijnbouw , 57, 105-16. Doherty Survey of the World Ocean (Talwani,
BOYNTON,C. H., WESTBROOK,G. K., BOTT, M. H. P. M. (ed)).
& LONG, R. E. I979. A seismic refraction inves- EWING, J. I., OFFICER, C. B., JOHNSON, H. R. &
tigation of crustal structure beneath the Lesser EDWARDS, R. D. 1957. Geophysical investiga-
Antilles island arc. Geophys. J.R. astron. Soc. tions in the eastern Caribbean: Trinidad shelf,
58, 371-93. Tobago Trough, Barbados Ridge, Atlantic
BUNCE, E. T., PHILLIPS, J. D., CHASE, R. L. Ocean. Bull. geol. Soc Am. 68, 897-912.
BOWlN, C. O. 1971. The Lesser Antilles Arc and FEO-CODECIDO, G. 1977. Un esbozo geologico de la
the eastern margin of the Caribbean Sea. In: plataforma continental Margarita-Tobago (A
MAXWELL,A. E. (ed). The Sea, 4, part II, Wiley, geological outline of the Margarita-Tobago con-
New York. tinental shelf). Bol. Geol (Venezuela), 3, 1923-
CHASE, R. L. & BUNCE, E. T. 1969. Underthrusting 46.
of the eastern margin of the Antilles by the floor HERRERA, R. C. & SPENCE, J. 1964. The geology and
of the western North Atlantic Ocean, and the oil prospects of the Scotland Group of sediments
origin of the Barbados Ridge. J. geophys. Res. at Barbados. Sinclair and B.P. Explorations Inc.
74, 1413-20. (unpubl. report).
CURRAY,J. R., SHOR, G. G., RAITr, R. W. & HENRY, HUBBERT, M. K. 1951. Mechanical basis for certain
M. 1977. Seismic refraction and reflection stu- familiar geologic structures. Bull. geol. Soc. Am.
dies of crustal structure of the eastern Sunda and 62, 355-72.
western Banda arcs. J. geophys. Res. 82, 2479- JORDAN, T. H. 1975. The present-day motions of the
89. Caribbean plate. J. geophys. Res. 80, 4433-9.
290 G. K. W e s t b r o o k

KARIG, D. E. 1974a. Tectonic erosion at trenches. MASCLE, A., LAJAT, D. & NELY, G. 1981. Sediment
Earth planet. Sci. Lett. 21,209-12. deformation linked to subduction and to argillo-
1974b. Evolution of arc systems in the Western kinesis in the southern Barbados Ridge from
Pacific. Ann. Rev. Earth planet. Sci. 2, 51-75. multichannel seismic surveys. Trans. 4th Latin
, LAWRENCE,M. B., MOORE, G. F. & CURRAY,J. American geol. Congr. (in press).
R, 1980. Structural framework of the forearc MOORE, J. C. 1975. Selective subduction. Geology, 3,
basin, NW Sumatra. J. geol. Soc. London, 137, 530-2.
77-91. 1979. Variation in strain and strain rate during
-- & SHARMAN, G. F. III. 1975. Subduction and underthrusting of trench deposits. Geology, 7,
accretion in trenches. Bull. geol. Soc. Am. 86, 185-8.
377-89. PETER, G. & WESTBROOK, G. K. 1976. Tectonics of
KEAREY, P., PETER, G. & WESTBROOK, G. K. 1975. southwestern North Atlantic and Barbados
Geophysical maps of the eastern Caribbean, J. Ridge Complex. Bull. Am. Assoc. Petrol. Geol.
geol. Soc. London, 131,311-21. 60, 1078-106.
KELLER, G. H., LAMBERT, D. N., BENNEIT, R. U. & PUDSEY, C. J. & READING, H. G. 1981. Sedimentol-
RUCKER, J. B. 1972. Mass physical properties of ogy and structure of the Scotland Group, Barba-
Tobago Trough sediments. Trans. Sixth Carib- dos (this volume).
bean Geol. Conf. 4(15-8. SAUNDERS, J. B. 1979. Field Trip I(J), Barbados.
KIECKttEt-TR, R. M., SHOR, G. G. JR. CURRAY, J. R., Field Guide: 4th Latin American Geol. Congr.
SUGIARrA, W. & HEHUWART, F. 1980. Seismic 1979, 71-88.
refraction studies of the Sunda Trench and fore- SCtIUBERT, C. 1974. Seafloor tectonics east of the
arc basin. J. geophys. Res. 85, 863-89. northern Lesser Antilles. Abstr. Seventh Carib-
LADD, J. W. 1975. Relative motion of South America bean Geol. Conf., 62-3.
with respect to North America and Caribbean SI-EI.Y, D. R. 1979. The evolution of structural highs
Tectonics. Bull. geol. Soc. Am. 87, 969-76. bordering major forearc basins. In: WArKINS,
LE PnctioN, X. & ANGELIER, J. 1979. The Hellenic arc J. S., MONrADERI, L. & DI('KERSON, P. W. (cds).
and trench system: a key to the neotectonic Geological and Geophysical Investigations of Con-
evolution of the Eastern Mediterranean area. tinental Margins. Mere. Am. Assoc. Petrol. Geol.
Tectonophysics, 60, 1-42. 29, 245-60.
LEwm, J. & ROBINSON, E. 1976. A revised stratigra- • VAn., P. R. & WALTON, G. G., 1974. Trench
phy of the Lesser Antilles. Trans. Seventh Carib- slope model. In: BURK, C. A. & DRAKf-, C. L.
bean Geol. Conf. 339-44. (eds). The Geology of Continental Margins, 249-
LOHMANN, G. P. 1973. Palaeo-oceanographic inter- 60. Springer-Verlag, New York.
pretation of the Oceanic Formation (Eocene- STEINECK, P. L. & MURTttA, G. 1981. Foraminiferal
Oligocene), Barbados, West Indies, Thesis, PhD, palaeobathymetry of Miocene rocks, Barbados,
Brown Univ. (unpubl.). Lesser Antilles. Trans. 4th Latin American geol.
LOWRIE, A. & ESCOWITZ, E. (eds) 1969. Kane 9. Congr. (in press).
Global ocean floor analysis and research data W)N HUENE, R. E. 1979. Structure of the outer
series, 1, U.S. Naval Oceanographic Office. convergent margin of Kodiak Island, Alaska,
MALFAIT, B. T. & DINKLEMAN, M. G. 1972. Circum- from multichannel seismic records. Mere. Am.
Caribbean tectonic and igneous activity and the Assoc. Petrol. Geol. 29, 261-72.
evolution of the Caribbean plate. Bull. geol. Soc. WFEKS, L. A., HARBISON, R. N. & PETER, G. 1967.
Am. 83, 251-72. Island arc system in the Andaman Sea. Bull.
MARESCH, W. V. 1974. Plate tectonics origin of the Am. Assoc. Petrol Geol. 51, 1803-15.
Caribbean Mountain system of northern South -- , LA'IqIMORE, R. K. & HARBISON, R. N. 1971.
America: discussion and proposal. Bull. geol. Structural relations among Lesser Antilles,
Soc. Am. 85, 669-82. Venezuela, and Trinidad-Tobago. Bull. Am.
MARLOW, M. S., GARRISON, L. E., MARTIN, R. G., Assoc. Petrol. Geol. 55, 1741-52.
TRUMBUEE, J. V. A. & COOPER, A. K. 1974. WESTBROOK, G. K. 1975. The structure of the crust
Tectonic transition zone in the northeastern and upper mantle in the region of Barbados and
Caribbean. J. Res. U.S. geol. Surv. 2, 289-302. the Lesser Antilles. Geophys. J.R. astron. Soc.
MASCEE, A., BIJU-DUVAL, B., LETOUZEY, J., 43, 201-42.
MONTADERT, L. & RAVENNE, C. 1977. Sediments 1976. The structure and evolution of the Barba-
and their deformations in active margins of differ- dos Ridge. Trans. Seventh Caribbean geol. Conf.
ent geological settings. Int. Syrup. Geodynamics 1974, 321-8.
in South-west Pacific Noumea (New Caledonia) , Borr, M. H. P. & PEACOCK, J. H. 1973. The
27 August-2 September 1976, 327-44. Technip, Lesser Antilles subduction zone in the region of
Paris. Barbados. Nature (phys. Sci. ) 244, 18-20.

G. K. WESTBROOK, Department of Geological Sciences, University of Durham,

Science Laboratories, South Road, Durham DH1 3LE, England.
 Sedimentology and structure of the Scotland Group, Barbados

C. J. Pudsey & H. G. Reading

SUMMARY: The lower to middle Eocene Scotland Group is about 1700 m thick at
outcrop and can be divided into: (A) The Lower Scotland Formation consisting of (1) the
Walkers Member (500 m): graded turbidite sands and clays, arranged in sand-dominated
and clay-dominated packets; (2) the Morgan Lewis Member (200 m): almost all clay, with
two sand intervals; (B) the Upper Scotland Formation consisting of (3) the Murphys
Member (about 350 m): a thickening and coarsening upward sequence including slumped
beds; (4) the Chalky Mount Member (0-250 m): coarse and conglomeratic sands with
large slide blocks and minor interbedded muds; (5) the Mount All Member (about 500 m):
coarse and fine turbidite sands and clays, much disturbed and displaying no overall
sequence. The group was deposited by mass-flow processes in a deep marine environment,
probably in a deep-sea trench or a submarine fan.
Current flow was from the SW and mineral composition supports derivation from South
American metamorphic rocks. Clay mineralogy shows an upward trend from mainly
smectite to mainly kaolinite.
An ENE structural trend is present throughout the island and is anomalous in the
present plate-tectonic setting adjacent to a north-south trending island arc. Deformation
increases in complexity up the succession; this may be due to sedimentary slumping, partly a
result of increasing disturbance by rising mud diapirs. This locally culminated in extrusion of
hydrocarbon-rich mud--the overlying Joes River Formation. The Barbados Ridge probably
comprises clastic sediments deposited both in a trench and on the Atlantic Ocean floor.

The island of Barbados in the eastern Carib- Barbados Ridge is underlain by a thick (20 km
bean is the only exposed part of a major beneath Barbados) slice of low-velocity, acous-
north-south structure known as the Barbados tically opaque material before oceanic crust is
Ridge (Fig. 1). The Ridge is a bathymetric reached (Westbrook 1975). The presence of
feature extending 400 km northwards from faint reflectors dipping shallowly west along the
Tobago, about 100 km wide at Barbados and eastern margin of the ridge led Chase & Bunce
rising 4-5 km from the Atlantic Ocean floor to (1969) to propose an origin by accumulation of
the east (Westbrook 1981). Its highest part crumpled and thrust-faulted sedimentary mate-
coincides with linear negative free-air and rial in the southern part of the Antilles trench
Bouguer gravity anomalies (Kearey et al. 1975; (sic). This idea of an accretionary prism has
Bowin 1976) and a small magnetic anomaly been elaborated by Westbrook (1975, 1981),
(Westbrook 1975). The Ridge also marks the Bowin (1976), Dickinson & Seely (1979) and
eastern limit of seismicity associated with a Speed (1979) among others; such an origin is
west-dipping Benioff zone (Sykes & Ewing certainly to be expected from the position of the
1965), i.e. the eastern boundary of the 'Carib- ridge just east of the Lesser Antilles island arc.
bean plate' (Molnar & Sykes 1969). Fault-plane Alternatively the ridge has been interpreted as
solutions for earthquakes along the east and an in situ northward extension of the continen-
north margins of this plate imply that it is today tal rise of S America (Senn 1947; Meyerhoff &
moving east with respect to North America and Meyerhoff 1972), and as a huge gravity slide
the central Atlantic: the rate of convergence of into a trench from the west (Daviess 1971).
the Caribbean and Atlantic plates is ½ to 2 cm Most of Barbados is covered by the Coral
yr -~ depending on the method of calculation Rock of Pleistocene to Recent age. An inlier on
(Molnar & Sykes 1969). Caribbean plate mo- the NE coast, the Scotland District (Fig. 2)
tion with respect to South America is less well exposes a highly deformed succession of ter-
known but appears to include a compressional rigenous clastic sediments (Scotland Group)
component: a number of small blocks may be overlain by much less deformed pelagic sedi-
behaving as microplates in a broad zone of ments (Oceanic Group). In the south of the
deformation along the north coast of Colombia inlier the Joes River Formation mudflow out-
and Venezuela (Jordan 1975; Bowin 1976). crops between these two groups. Foraminifera
Seismic reflection and refraction profiles, and molluscs from the Scotland Group yield an
combined with gravity data, show that the age of lower to middle Eocene and possibly

291
292 C. J. P u d s e y & H. G. Reading

C H A R T OF THE EAST CARIBBEAN SEA

km 500
S u b m a r i n e c o n t o u r s in k i l o m e t r e s

L "~ . . . . . . , \ ) Io ,/ ' ; 'co'

' ~" - .... < ...... :2 "'D /~tu/ ', ,

615 W 6f , ~

FIG. 1. Location map of Barbados. Bathymetry from Hurley (1966): Edgar & Saunders et al. (1973);
Westbrook (1975); Ladd & Watkins (1978).

include upper Eocene forms (Trechmann 1925; In the Scotland District the Tertiary rocks
Senn 1940, 1947; Cizancourt 1948; Caudri occur in a series of ENE-trending fault-
1972). The Joes River Formation contains bounded blocks (Fig. 2). The major faults are
blocks of Palaeocene limestone (Caudri 1972) not exposed and their nature can only be infer-
but its matrix has not been accurately dated. red. The occurrence of a few fault slices of
The Oceanic Group is largely Oligocene but its steeply dipping Oceanics and their presence in
base could be as old as middle Eocene (see boreholes which penetrate 4½ km of stacked
Lohmann 1974 for summary). It has been Oceanic/Scotland sediments led Speed (1979)
argued (e.g. Baadsgaard 1959; Speed 1979) that to propose that the Scotland District was a pile
the Oceanic and Scotland Groups are, in part, of totally unrelated fault slices. Sedimentary
time-equivalents, separated everywhere by a sequences and structural styles vary between
tectonic contact. We maintain that the contact blocks, but the five lithostratigraphic members
is unconformable in at least a few places (Jukes- of Senn (1940) have been mapped and corre-
Brown & Harrison 1891, p. 271, appendix) lated from block to block (Fig. 3). Pudsey
though the unconformity may have been a (1979) has presented a table of systematic stra-
plane of movement during subsequent de- tigraphy.
formation.
 Sedimentology of the Scotland Group, Barbados 293

GEOLOGY OF THE SCOTLAND DISTRICT, BARBADOS.

AYS COVE

CORAL ROCK

~I
AN LEWIS [~ OCEANIC GROUP
JOES RIVER FORMATION

::i:. ~ UPPER SCOTLAND FORMATION

LOWER SCOTLAND FORMATION

i ii!iiiii!i!: 0 km 5

; :?:!!::!!::!!::!!::!i::!!:~:.2.~::-
::::: ~ %

;KEETE'S B/I
"~ RAGGED

FIG. 2. Simplified solid geology of the Scotland district. Coral Rock, Oceanic and Joes River
outcrops from Baadsgaard (1959) and Saunders (1965). Marls of Bissex Hill etc., included within
Oceanic Group. Inset map shows location on Barbados.

Sedimentology beds are parallel-laminated, often with ripple

Lower Scotland Formation cross-laminated or convolute tops. Thinner
sands are rippled or convolute throughout. The
Walkers Member few unlaminated sands exhibit dish structures in
The member consists of alternating graded their middle portions. The sole markings and
sands and laminated clays, in turn arranged in ripples define a very consistent palaeocurrent
packets dominated by sand and packets domin- direction from the SW and west.
ated by clay. The tops of sands grade into the overlying
Sand bed thickness ranges continuously from clay over a thickness of ½to 1 cm. The clays are
a few centimetres to 2 m. Most sands over dark grey to brown, thickly laminated, non-silty
10 cm thick are medium or coarse-grained, and ferruginous. Clay units are a few cen-
smoothly graded to fine (Fig. 6A); there are timetres to 2 m thick, rarely up to 5 m; they
occasional granular basal lags. Fine, poorly generally contain very thin graded silts and very
graded sand constitutes most of the thinner fine sands, indistinctly parallel or ripple cross-
beds. The bases of sands are sharp (rarely laminated. Beds of all thicknesses are laterally
erosive) and, where exposed, about half the continuous across most outcrops.
beds bear sole marks: flutes, grooves, load On a scale of tens of metres vertically, there
casts, trails and prod marks. Most sands are are quite distinct sandy and clayey packets of
pale yellowish-grey to white, well-sorted, quart- beds, sandy packets containing about 50% sand
zose and non-micaceous. Internally, thicker and clayey packets 10% or less. Sand-
294 C. J. Pudsey & H. G. Reading

CORALROCK cernible is that the clayey packets contain

slightly fewer sands towards the top.
*l~![ OCEANIC GP.
Morgan Lewis Member
._.L~.~._....._.....~... ~ --~" " - ' - This member consists almost entirely of clay
Joes River Fm. with very thin graded silts and fine sands,
similar to the clay packets in the Walkers
Member (Fig. 6B). The clays are dark purplish-
grey to brown, thickly laminated and ferrugi-
?
Mount All Mb. s nous to the point of containing stringers of
e ovoid iron concretions. A few of the silt beds
0 are completely cemented by iron oxy-
Upper T hydroxides ('clay-ironstones') and trails are
Scotland t well-preserved on their bases. Scattered fine
Chalky Mount Mb. Fm. sands up to 12 cm thick bear sole marks indicat-
lkm A ing current flow from the SW. Isolated coarse
N sand beds from 10 cm to 1 m thick occur spor-
Murphys Mb. D adically: they have stepped erosive bases and
some contain clay flakes and fine sand clasts.
I; Two laterally impersistent sand intervals are
Morgan Lewis Mb. present in the type area, each about 25 m thick.
R The lower (and better exposed) contains c. 50%
Lower 0 sand in massive and parallel-laminated beds up
Scotland
Fm. u to 2 m thick. The sands are fine-grained, mod-
Walkers Mb. p erately sorted, poorly graded and slightly car-
bonaceous. A few sole marks indicate current
flow from the NW.

Upper Scotland Formation

Fi¢~. 3. Stratigraphic terminology used in
this paper, with approximate thicknesses Murphys Member
and a diagrammatic guide to lithology. The The base of the member is a thickening and
thicknesses of the Mount All Member and coarsening upward sequence, the sand:shale
the Joes River Formation have not been ratio increasing from 1:20 to 1:3 in a thickness
accurately determined, on account of the of 10-15 m. The bulk of the Murphys Member
complex structure of the former and the is quite variable both laterally and vertically,
lack of bedding in the latter. The Oceanics and does not show the good packeting de-
seem to be much thicker in some subsur- veloped in the Lower Scotland Formation.
face wells (Baadsgaard 1959), but this may
be due to tectonic repetition. Conformable Overall it becomes sandier upwards, sand:shale
contacts have been documented between ratio being 2:1 at the top; the modal grain size
each pair of members except for part of the remains constant at fine sand. In most mea-
Mount All Member. In particular the Mor- sured sections half the sands belong to a distinc-
gan Lewis-Murphys transition has been tive laminated carbonaceous facies, variously
mapped for several kilometres. coloured red, grey, yellowish or black due to
combinations of carbon and iron oxides. Both
dominated packets in the type area of the carbonaceous and non-carbonaceous fine sands
Walkers Member (for location see Fig. 9) may are moderately sorted, micaceous and rather
be correlated over several hundred metres (Fig. silty, usually poorly graded though the bases of
4) with no apparent interconnection between beds are sharper than their tops. Erosion is
them. virtually absent. There are a few medium and
Throughout the member there is no overall coarse graded sands but hardly any sole marks,
change in grain size of the sands, in bed thick- except for sporadic load casts and trails. Sand
ness or packet thickness. The only trend dis- bed thickness, commonly 10-40 cm, ranges up

FI~. 4. Measured sections in type area of the Walkers Member (for location see Fig. 9), showing
lateral correlations between sandy packets of beds. Part of two sections are enlarged to demonstrate
bed-by-bed correlation. Arrows indicate current flow directions.
Sedimentology of the Scotland Group, Barbados 295

@
b

°iI ,lOOm

@ 50

500m
296 C. J. Pudsey & H. G. Reading
to 2 m. The thick beds are commonly compo- conglomerates up to 20 m thick, separated by
site, consisting of several parallel or ripple carbonaceous silty clays. Westwards the sands
cross-laminated units stacked together and become thinner and finer and there is much
varying slightly in composition or type of ripple cross-lamination. The most westerly sec-
lamination. Palaeocurrents, almost all from tions comprise clays containing large isolated
ripple cross-lamination, are again consistently blocks of very coarse white sands as well as a
from the SW. few coarse turbidite sand beds, fine rippled
The clays are much siltier than those in the sand being unimportant except near the base.
Lower Scotland Formation and there are no The clays are contorted beneath some of the
clay-ironstones. Very thin silt and fine sand sand blocks.
interbeds are carbonaceous and ripple or In contrast to the view of Senn (1940) the
trough cross-laminated. The clay is almost Chalky Mount Member does not thicken con-
black and pyritic when fresh, weathering brow- sistently to the SE. The few palaeocurrent
nish grey and growing gypsum crystals on the measurements (Fig. 5) show that both coarse
surface. In the northern Scotland district a few and fine sands were deposited by currents flow-
very thin beds of silty cream-coloured chalk ing from the SW and west. Chalky Mount out-
were found--the same lithology as parts of the crops exhibit many slump features, particularly
Oceanic Group. synsedimentary normal faulting with a down-
Two conglomerates occur in the Murphys throw to the north. Trace fossils are rare, but
Member; the better exposed is 3-4 m thick and there are stellate impressions at the top of the
mappable for 1½ km along strike. Small and Chalky Mount Summit Beds, and two beds
medium pebbles are dominantly of well- containing vertical burrows in other localities.
rounded hard micritic limestone and clay- Conglomerates are prominent in the eastern
ironstone, with minor quartzite, shale and vein part of the type area: large rounded clay and
quartz, and locally numerous shell fragments fine sand intraformational clasts are most com-
(bivalves, gastropods and scaphopods) in a fine mon, followed by micritic limestone, iron con-
sand matrix. cretions, quartzite and vein quartz (see Speed
A principal feature of this member is synse- 1979). Baadsgaard (1959) reports an Upper
dimentary deformation, including normal fault- Cretaceous age for a limestone pebble.
ing (Fig. 6C), slumped packets of beds a few In the SE of Barbados there is no Oceanic
metres thick, disruption of sand beds by clay Group between the Coral Rock and the local
injection, and olisthostrome units up to 10 m Chalky Mount Member, and the coarse and
thick consisting of rafts of coarse sand up to 4 m granular sandstones are cemented by calcite.
x 30 cm in a red shaly matrix. Measurement of
slump fold and fault orientations yielded no Mount All Member
consistent palaeoslope direction. The member has such a complex structure
that it is not easy to derive a coherent sediment-
Chalky Mount Member ary picture: the thickness estimate of 500 m
This member is distinguished by the very is very tentative. Above the Chalky Mount
coarse grain size, massive nature and lateral Member immediately north of the Chalky
discontinuity of the sediments: it shows great Mount type area, the Mount All Member be-
variation between outcrops and only the main gins with 75 m of thickly laminated brown and
section will be considered in detail here. In the grey clays with very thin graded silts and no iron
type area (Fig. 9) the member has been divided concretions. Above these clays and at Mount
into the Chalky Mount Summit Beds (lowest All, it consists of thin to thick, fine to coarse
100 m) and the Chalky Mount North Beds sands interbedded with red-brown silty clays
(metre 100 to top) (Pudsey 1979). The former (Fig. 6E). The interbedding is usually random
consist of thick, amalgamated, coarse to granu- though packets of thin and thick sands are
lar clayey white sands. Some beds are well- well-developed in one or two places. The sands
graded (Fig. 6D) and a very few have single sets have sharp or erosive bases, gradational tops,
of tabular cross-bedding at the top. Silty mud and are quite commonly cemented by calcite or
interbeds are thin and impersistent because of silica. Sorting varies from moderately good to
widespread scouring. There is no upward poor. The fine sands are micaceous and carbo-
change in bed thickness or grain size and the naceous (though less so than the Murphys
unit has a sharp top. sands) and internally parallel or convolute-
In the eastern part of the type area the laminated: the coarse sands are quartzose, fer-
Chalky Mount North Beds (Fig. 5) consist of ruginous, clayey and well to poorly graded.
very distinct graded packets of coarse sands and Some contain dish structures and others water
 Sedimentology of the Scotland Group, Barbados 297

W
0 500m

li" o lOOm &

©
C

]
O
I@ Q
G

Fro. 5. Measured sections 3a to 3f at Chalky Mount (for location see Fig. 9), showing lateral vari-
ation in the Chalky Mount North Beds. Datum at base is top of 100 m thick, sheet sand. There is a
marker bed containing huge concretions at metre 40 in sections a, b and c. Note extreme discontinuity of
sand bodies in sections e, f and g. Units of sand rafts near base of sections e and g are similar to olistho-
stromes described in Murphys Member. Sections are overlain by thick clay unit at base of Mount All
Member.

escape features. Overall s a n d : s h a l e ratio is mainly turbiditic; it is also probably a deep-

about 2:1. Sole structures are uncommon and water deposit. Foraminifera show the whole
include flutes modified by load casting, and Scotland Group to be m a r i n e . Because of its
frondescent marks. Few current measurements very coarse grain size and the presence of a
were made, on account of the difficulty of shallow water benthic fauna, the Chalky Mount
unfolding the steeply plunging structures, but Member has often been interpreted as a shallow
current flow seems to have been from the marine deposit (e.g. T r e c h m a n n 1925; Baads-
southern quadrant. gaard 1960). We believe that the member is
At least some, and probably most, of the entirely a deep-water mass-flow deposit for the
complex folding in this member can be ascribed following reasons: (1) position within a turbi-
to slumping very shortly after sedimentation dite succession; (2) presence of grading and
(see below under 'Structure'). No consistent poor sorting of the coarse sands; (3) absence of
palaeoslope direction has been determined. wave ripples and of evidence for emergence,
The top of the M e m b e r is not seen and there and almost total lack of cross-bedding in the
are no interbeds of Oceanic lithology. sands; (4) shallow-water fossils occur only in
conglomerates and most of the shells are frag-
mented: shells were resedimented along with
Interpretation pebbles; (5) extreme rarity of trace fossils of
The Lower Scotland Formation is entirely any kind and only two instances of vertical
turbiditic and hemipelagic and we assume it to burrowing in all the sections examined (see
have been deposited in deep water. The U p p e r below).
Scotland Formation is more variable, but still The .Scotland Group on Barbados is part of
298 C. J. Pudsey & H. G. Reading

FIG. 6(A) Walkers Member, type area. Parallel-laminated medium-grained sand bed with
fine-grained top showing water-escape structures, overlain by thinner parallel and ripple
cross-lamin ated fi ne-grained sand bed. Slightly erosive base of sand is unusual. Scale: hammer
32 cm.

(B) Morgan Lewis Member, type area. Monotonous clays with very thin graded slits (lighter)
and one clay-ironstone (darker, at top). Scale: 1 m of tape extended, lens cap 5.5 cm.
 Sedimentology o f the Scotland Group, Barbados 299

the Barbados Ridge and any interpretation The three remaining possibilities of a deep-
must take into account the size, shape and sea trench, an abyssal cone and a deep-sea fan
position of the Ridge as well as the sedimentary must all be considered likely (see 'Regional
features of the miniscule outcrop area. Thick significance'). Certain features of the sub-
piles of deep-sea clastic sediments may accumu- marine fan model (e.g. Walker 1978; N o r m a r k
late in the following environments (Rupke 1978) are consistent with the Scotland G r o u p
1978): sequences (Fig. 7). For instance the non-
1 Open oceans erosive, laterally continuous, rather regularly
2 Deep-sea trenches spaced, sand packets of the Walkers M e m b e r
Basin plains 3 Enclosed seas may represent fan lobes (cf. Mutti et al. 1976).
4 Borderland basins During Morgan Lewis deposition sand supply
5 Miscellaneous small became finer and more intermittent: the
basins Murphys M e m b e r is the result of renewed
6 Abyssal cones progradation, the source now supplying more
7 Deep-sea fans carbonaceous and finer sand. The u n c o m m o n
Turbidite fans 8 Short-headed delta front and thin chalk beds are more likely to be
fans carbonate turbidites shed from an adjacent
9 Continental rise fans swell than to represent pelagic background
sedimentation. If the latter were so, one would
Some of these may be eliminated in the case of
expect many more such thin beds, or the clay
the Scotland Group. should have an appreciable carbonate content.
(1) is unlikely because of the thickness (nearly The Chalky Mount Member is interpreted as
2 km cf. a few hundred metres in modern channel deposits though the small size of the
examples) and the lenticularity within the outcrop area makes it difficult to distinguish
Upper Scotland Formation. between a series of small distributary channels,
(3) is not possible because there are no land and one major channel containing a complex
areas enclosing the Ridge. network of smaller braided channels. Both
(4), (5) and (8) involve relatively small sedi- associations have been described from the mid-
ment bodies. fan regions of modern submarine fans. The
(9) can probably be eliminated because an sheet-like non-erosive Chalky Mount Summit
essential characteristic of such accumula- beds perhaps represent a proximal fan lobe or
tions is extensive reworking by indigenous channel mouth bar. The rare cross-bedding
bottom currents. The Scotland Group sedi- could have been formed by the tails of turbidity
ments show very little evidence for rework- currents or by down-channel storm-driven cur-
ing. rents (Shepard et al. 1979). The slide blocks

(c) Synsedimentary normal faulting in fine laminated sands of Murphys Member. Dark
laminae are concentrations of carbonaceous material. Fault movement of the base of the bed
occurred before deposition of the top was completed, as uppermost lamination is undisturbed.
Scale: lens cap 5.5 cm.
(o) Chalky Mount Member, type area. Very thick, well-graded pebbly sands. Younging
direction to right. Scale: hammer 32 cm.
(E) Mount All member, type area. Thin to thick fine and coarse-grained sandsinterbedded with
silty clays. Younging direction to left. Scale: hammer 37 cm.

(F) Joes River Formation, Joes River. Blocks of sand in structureless mud matrix. The largest
block is of very coarse sand graded to medium and its younging direction is to the right. Scale:
hammer 37 cm.
(~) Syncline in clays of Morgan Lewis Member. Note simplicityof fold style and general lack of
parasitic folds even in this incompetent lithology. Scale: figure 1.6 m.

(H) Folding in lower part of Mount All member. Overall younging direction is upwards and to
the right. An almost upright anticline at left is followed by a slightly faulted syncline with
parasitic minor folds, and at right a recumbent anticline. Folding took place at a stage when the
sands were sufficientlylithifiedneither to thin nor thicken around the hinges of folds. Scale: hill
is about 40 m high.
300 C. J. Pudsey & H. G. Reading

/ f
.,.. .... .. o ........ .--..,
•
marked strain shadows and a few are polycrys-
/ . . . . . . . . .. . ' . \ talline, but about 30% have uniform extinction
and some approximate to the hexagonal form of
~¢ "[3~'"!'"'" '" outer "fan
';'" .... X
• fl quartz• The latter are probably of volcanic
origin (Todd & Folk 1957), the former plutonic
or metamorphic (Blatt et al. 1972). Feldspar is
/" mo0, ..... ." / uncommon and heavily altered: we did not find
widespread microcline as reported by Velbel
[ bar If£L.< i :] (1980). The heavy minerals in the sands were
I , /~ erosion of old .i, derived from a metamorphic terrane (Matley
I ch,n°e,./ 1932; Hedberg, in Senn 1940).
I fill / / The two external sediment source areas were

/I jj/ the metamorphics on the north coast of South

America and the volcanics of the Lesser An-
tilles arc, with the former predominating. This
is in agreement with the conclusions of Velbel
(1980). A local source is postulated for the
/ / I i l conglomerates since they contain no metamor-
/ phic pebbles.

FIG. 7. Some features of the submarine fan Clay mineralogy

model applicable to the Scotland Group.
Small square represents position of Barba- The clay mineralogy of 37 samples was deter-
dos (relative to the model) during Chalky mined using X-ray diffractometry and the re-
Mount deposition. sults are summarized in Fig. 8. The mineralogy

probably moved only a short way downslope

following erosion of a previously deposited A \
channel fill (Fig. 7). \ J Joes River
/' \ ~ Motirlf All
After the hiatus in sand supply represented / / \ / \ • rho,kyM.....
by the thick clays at the base of the Mount All \ [ ] Murphys
• '~ o Morgan Lewis
Member, progradation was renewed. The very , / ~ ",\ Z --, • wor,e,.~
Overall bend
thick laminated fine sands, seen both in the
Mount All and Murphys Members, are clearly
not the products of single classical turbidity
currents. They may have been laid down by
longer-lasting and gently pulsating river-
generated turbidity currents (Heezen et al. ,D- / k <-, '"-.-.-. " X
1964). The lack of organisation of the Mount if,, \
All Member as a whole may be a consequence
of the very frequent disturbances which trig- ~:.i,,;;'" . . . . "'. ,_,'
gered slumping: stable conditions were never
KO Sm
maintained for long enough for a consistent
sediment dispersal pattern to develop. FIG. 8. Clay mineralogy of the Scotland
Group. Ka = kaolinite, I1 = illite,
Petrography and provenance Sm = smectite. Minor components (chlo-
rite and vermiculite) not plotted.
The composition of the conglomerates has
been described above. In thin section, Scotland of each member of the Scotland Group is more
Group sandstones are poorly to moderately or less distinct. The Walkers and Chalky Mount
well sorted with a high matrix content. Most clays are the most readily identifiable, the
grains are angular, though large quartz grains former by their high smectite content (up to
are well-rounded. A number of quartz grains 75%) and the latter by the virtual absence of
show corroded edges particularly where the smectite and abundance of well-crystalline
rock is locally calcite-cemented. Composi- kaolinite. The Morgan Lewis clays plot in a
tionally, the sands belong to the category of very narrow field, all having Ka:I1 close to 3:2.
quartz wackes of Dott (1964). Of the large The Murphys and Mount All both show a
(>1 mm) quartz grains, a majority show rather wide range in clay composition. The
 Sedimentology o f the Scotland Group, Barbados 301

general trend is a decrease in smectite and sandy substrates. Many were shallow burrow-
increase in kaolinite during Scotland sedi- ers. There is a marked lack of bored and
mentation, accompanied by an increased illite encrusted shells. This may indicate burial by
contribution in the middle (see inset in Fig. 8). sand soon after death, or frequent slumping and
Only one sample of Joes River clay was analy- turbidity flow of the sediment into water too
sed: it bears little relation to the Scotland clays, deep for any epifauna to live.
being much richer in illite. The source area for a small part of Scotland
We believe that most of these clays are sedimentation was a shallow sandy marine
turbiditic rather than hemipelagic because they shelf, well-oxygenated and with an abundant
frequently have a gradational lower boundary food supply, supporting a very diverse benthic
with the c or d division of a turbidite sand, they molluscan fauna. Mixing of the pebbly debris
are not bioturbated, planktonic microfossils are with shell debris probably took place during
extremely rare (Senn 1940; Baadsgaard 1959), resedimentation, as there are no byssally
and interbedded clay bands of two or more attached molluscs.
distinct colours are not seen (cf. Rupke & No palaeoecological studies have been
Stanley 1974; Hesse 1975). undertaken on the larger foraminifera of the
The obvious source for the smectite is the Scotland Group.
volcanic Lesser Antilles arc, and for the kaoli-
nite the deeply weathered tropical soils of Trace fossils
northern South America. Kaolinite flocculates Most traces are post-depositional trails on the
very easily in sea-water and is in that sense a bases of thin turbidite sand beds. Those in the
proximal clay mineral; it is therefore not sur- Lower Scotland Formation include Palaeodic-
prising to find it in clays interbedded with tyon, Spiroraphe, Scolicia and other grazing
proximal turbidites. The smectitic Lower Scot- trails typical of quiet deep-water environments
land clays may be hemipelagic, or more prob- (Seilacher 1978). A small form of Chondrites
ably redeposited ashfalls. Most of the samples occurs in clay-ironstones. In the Upper Scot-
contain a trace of chlorite, three containing land Formation there are two isolated beds
10% or more: this may be derived from meta- containing Tisoa-like vertical pipes, and stellate
morphics. The effects of diagenesis can prob- traces cf. Haentzschelinia occur on the tops and
ably be neglected since the clays are quite bases of a few coarse sands, in addition to
young and have never been deeply buried; even nondescript sinuous grazing trails beneath Mur-
smectite, which is particularly unstable, re- phys and Mount All Member sands. Crimes
mains in large quantities. (1977) has described a similar association from
The progressive decline in smectite and in- an Eocene deep-sea fan in Spain: he postulates
crease in kaolinite may be attributed to either that 'shallow-water' trace-makers may live in
or both of (a) lessening of volcanic activity the mid-fan region as turbidity currents periodi-
during Scotland Group deposition, (b) swamp- cally bring in oxygen and food from the shelf
ing of the volcanic contribution as the conti- (see also Corbo 1979). Thus the trace fossil
nent-derived sediments prograded into the evidence is consistent with the deep marine
basin. interpretation.

Palaeoecology The Joes River Formation

Shelly fauna of the Chalky Mount Member This curious unit presents stratigraphical
The diverse molluscan fauna of a few con- problems: it has not been accurately dated and
glomerate beds was described and illustrated by at outcrop it is not clear to what extent it
Trechmann (1925) who dated the assemblage as overlies, interfingers with, or is faulted against
middle/upper Eocene. During the present the Scotland Group. It is a m61ange sensu lato
study small collections were made from Mur- containing blocks of fine to coarse sand of all
phys and Chalky Mount conglomerates, and sizes up to tens of metres long in a massive,
Trechmann's collection of nearly 100 species in often sheared, gritty mud matrix. The whole
the British Museum was re-examined briefly. formation is impregnated with heavy black oil.
The fauna consists largely of worn and broken Senn (1940) was able to detect a faint stratifica-
specimens, and the evidence for redeposition in tion and estimate a maximum thickness of
deep water is now substantial (see above). All 400 m.
the specimens are fully marine. The taxa repre- The sand inclusions are matrix-supported,
sented are all thought to have lived in shallow vary in shape from equant blocks to elongate
water (<100 m) deep) and are characteristic of rafts (Fig. 6F) and occur in random orienta-
302 C. J. Pudsey & H. G. Reading
tions. Some are well-graded and all are similar and their nature and exact positions can only be
to Scotland Group sands, from which they inferred. Their attitudes vary from low to high
could have been derived. There are also small angle (Speed 1979) and dips are to both north
(few centimetres) angular blocks of Lower and south. There is outcrop evidence for low-
Scotland-type clay-ironstones, some of which angle north-dipping reverse faults in the north-
appear slickensided. Senn (1940) mentions ern part of the area.
blocks of sandy foramineral limestone of
Palaeocene age, which do not outcrop elswhere
on the island. Herrera & Spence (1964) in a
detailed account, describe a locality with very
large sand blocks up to 300 m long and parallel
to the regional strike, in a contorted but de-
finitely bedded mud matrix.
The formation has been variously interpreted
as a fault breccia (Baadsgaard 1959), as a
submarine slide deposit (Hess 1938; Daviess
1971; Herrera & Spence 1964) and as the
product of sedimentary volcanism (Senn 1940).
Baadsgaard's theory is rejected because of the
thickness and outcrop width of the Joes River
Formation (Fig. 2) and because there are major
faults in the northern Scotland district where it
is not present. Submarine sliding is more likely
but does not explain the oil content, which is
much higher than in the exposed Scotland
Group. Lithologically the Joes River Formation
strongly resembles mud volcano deposits de-
scribed from Trinidad by Higgins & Saunders
(1967, 1974) and Arnold & Macready (1956)
where the mud matrix is often sheared during FIG. 9. Structural map of the Scotland
extrusion and large sand blocks can be detached District showing fold axes and bedding
from the walls of the vent and brought to the attitudes. Encircled letters are localities. W
surface. Generation of methane during bacter- = Walkers Member type area; CM =
ial decay of organic matter assists mud injec- Chalky Mount type area; MA = Mount All
type area; CW = Cattlewash, where a large
tion by rendering it more fluid (Hedberg 1974) faulted slice, now enclosed within the Joes
and the extruded mud is often very hydrocar- River Formation, exposes flat-lying
bon-rich. We therefore concur with Senn's in- Oceanic Group unconformable on gently
terpretation, which is supported by the contrast dipping Upper Scotland Formation.
in mineralogy between Joes River and Scotland
clays (see above); one would expect more illite
in clays brought up from greater depth. In general, the structure simplifies with in-
The mobile mud would of course, be subject creasing stratigraphic depth, this trend being
to downslope movement after extrusion, just as modified by variations in competence of the
the Joes River Formation is much disturbed by rocks. There is a progression upwards from
land-slipping at the present day. The formation simple, laterally persistent, gently east-plunging
is highly significant in interpreting the structure folds in the Walkers Member to very complex,
of the Scotland district (see below). steeply plunging structures in the Mount All
type area. In the Lower Scotland Formation
there are open to tight, upright to overturned,
Structure angular folds with wavelengths of tens to hun-
dreds of metres (Fig. 6G & 10). The axial traces
The ENE structural trend is constant not only of the folds undulate slightly between ENE and
at outcrop (Figs 2 & 9), but in the subsurface NE and the poles to bedding show some scatter
throughout the island. This trend is anomalous about a NNW-SSE girdle (Fig. 10): the initially
in the present-day tectonic setting adjacent to a almost planar axial surfaces have been refolded
north-south trending island arc. Major strike by later folds with near-vertical axes and steep
faults separate discrete slices in the Scotland NW-striking axial planes. This is similar to the
district. These faults are not normally exposed results of Speed's work on the Chalky Mount
 Sedimentology of the Scotland Group, Barbados 303

Walkers Mbr. type ares / o o:0o

• o J • • - - c ~

/o -

I
• • o I

:o .~, o.
/ /
I

,on.

8•~ • 7

o ,-,p • /

~ "~N .6
• .
• "" ~
\ ~ Uo~
Polos to bedding

. o8 °%
.~. %++° 00
\ o oooo oz~o o @ 9"

~ Ooo o o~' ~ /
-.,- o • o° ~ y
Poles to bedding -~__~ ~ .~_J N=110

FIG 10. Structure of the Walkers type area.

\
\
\
• /
Block diagram is schematic only: grading \
indicates way-up of beds. The stereogram
J
is a Lambert equal-area projection. J/ N: 19
Spots = beds right way up, open Mososcopic fold axes
circles = inverted beds.
FIG. 11. Structure of the Mount All type
area. Stereograms are Lambert equal-area
area (Speed 1979). Towards the top of the projections. In the upper diagram open
stratigraphic succession, folds become more circles = inverted beds. The only mesosco-
pic folds plotted are those where the
complex, with parasitic and disharmonic minor plunge could be measured directly at out-
structures (Fig. 6H), though the intensity of crop.
deformation (amount of shortening) probably
does not vary much. The Mount All area shows
widely scattered poles to bedding (Fig. 11) and Oceanic Group is virtually undeformed except
mesoscopic fold axes plunging to the north, for faulting and tilting of blocks, so the main
NE, SE, south and SW. There is very little folding took place before the middle of the
inversion of beds, but few fold axial traces are upper Eocene. At the top of the Scotland
mappable for more than 100 m, and it is possi- Group it is not possible or useful to separate
ble that there is considerable faulting in this 'tectonic' from 'sedimentary' deformation.
heavily forested area. Evidently local slopes were generated on the
There is no cleavage in the Scotland Group seafloor during and after Scotland deposition.
and all deformation took place at a very high Possible causes include: (a) bank collapse into
structural level. Much was synsedimentary, channels or sliding of levee material towards
slumping frequently occurring while the sedi- the interchannel area (Mutti 1977); (b) slump-
ments were poorly lithified. The overlying ing down the front of the aggrading sediment
304 C. J. Pudsey & H. G. Reading

pile, or down the frontal slope of an advancing Group sediments may have been deposited in a
accretionary wedge; (c) mud diaprism. Local deep-sea trench, an abyssal cone fed by a broad
slopes generated by the upward movement of delta front, or a deep-sea fan fed by a sub-
overpressured mud are a common cause of marine canyon. Although tectonic accretion
slumping on deltas (Roberts et al. 1976). may have led to the construction of the large
Although progradational slumping and bank Barbados Ridge from sediment deposited en-
collapse may have been responsible for some tirely within a narrow trench, it is more likely
small-scale faulting and the sand blocks in the that sediments originally deposited on the
Chalky Mount Member, there is no evidence Atlantic Ocean floor were also incorporated
that the superficial folds are related to channel (Fig. 12). The occurrence of packets of sand
margins and the random fold patterns conflict and clay, plus channel sequences, indicates that
with the regularity of the palaeoflow indicators. there was at least sometimes an organized
We therefore think that mud diapirism is the sediment dispersal system, i.e. some kind of
most important factor. As distal, organic-rich cone or fan: though this may of course have
muds become buried by an advancing sediment been a very elongate 'fan' within a trench,
pile, their impermeability quickly results in rather than an unconfined one on open ocean
overpressuring by water and hydrocarbon gases floor. The consistent palaeocurrent directions
(lledberg 1974; Chapman 1973, 1974). Diapirs support an elongate fan. Without knowing
form at a spacing related to the thickness of the more about the thicker, more proximal end of
stratigraphic units involved, to the viscosity the Barbados Ridge, it is impossible to say
contrast and to any basement structures. They whether the source was a submarine canyon or
may be circular, or linear parallel to the deposi- a broad delta-front.
tional strike (Chapman 1974). Slopes are In Fig. 12 the east margin of the Caribbean
formed in varying directions around their flanks plate has been drawn west of its present posi-
and slumping and faulting occur as the diapirs tion with respect to South America. How far
continue to rise. In some circumstances they west is not known: it is clearly inappropriate to
may break through to the surface. We believe extrapolate the present convergence rate back
that increasing upward disturbance by rising 50 or 60 Ma, and the Caribbean and South
mud diapirs locally culminated in extrusion on American plates may have been coupled
to the seafloor of hydrocarbon-rich m u d - - t h e together at some time during the Tertiary (Jor-
Joes River Formation. dan 1975). J. F. Dewey (pers. comm.) has
Diapiric structures have been identified on calculated, from magnetic lineations in the
seismic reflection profiles in the SE part of the Atlantic, that 50 Ma ago North and South
Barbados Ridge complex, and are indicated on America were converging very slowly, with the
the USGS Geologic-Tectonic map of the Carib- Caribbean plate squeezed sideways between
bean. Although the Scotland district is too them. Very oblique convergence along a com-
small, and now too severely faulted, to be able plex fault system between the Caribbean and
to delineate individual diapirs and their associ- South America might have resulted in uplift
ated radial and concentric structures, the and sediment supply to the newly initiated
present distribution of Joes River Formation (Malfait & Dinkelman 1972) Lesser Antilles
outcrops (Fig. 2) suggests they were elongated Trench to the north. Either the now inactive
parallel to strike. Aves Ridge, or the first volcanic islands of the
Lesser Antilles chain, could have supplied the
volcanic quartz detritus and the smectitic clays.
Regional significance The prodigious change in sedimentary facies
from Scotland Group to Oceanic Group could
In terms of plate-tectonic theory the Barbados be due to one or more of: (1) cessation of uplift
Ridge should be an accretionary prism of de- of source as convergence between North and
formed and thrust-faulted trench a n d - o c e a n South America changed to sideways motion
floor sediment, analogous to the Mentawai (J. F. Dewey, pers. comm.); (2) strike-slip
Islands (Karig et al. 1980). However, there are movement of depositional area eastwards away
major inconsistencies between Barbados and from the delta source (cf. Howell et al. 1980);
the accretionary prism model: the structural (3) development during strike-slip movement of
trend is at a high angle to that of the arc, there deep trenches such as the Venezuela Trench
being no outcrop evidence for major north- (Ladd & Watkins 1979) which trapped ter-
south structures; and pelagic sediments overlie rigenous clastics coming from the south; (4)
turbidites unconformably. uplift of the sediment pile during accretion of
As pointed out above (p. 299) the Scotland further slices of ocean-floor sediment under-
 Sedimentology of the Scotland Group, Barbados 305

SOUTH AMERICA

• , i j
. , I s i t

( ....... , ' /' ".',",".'TRENCH, y ~ / ~ / ¢ , i ' ',1/ '~l /) I / 11 /

/ /' / '. . . . . SEDIMEN S// ~ ~ - - - ~ ' ' ' ' " ....
/ / , / ' i • :'. • , . [v/ / ~ ~\~'~'.£.o CARIBBEAN~

/~, ,' .-.-:,.......!.:.:.'-;~: i...., - X ~ \ ~ z ~ 7 ~ ; ~ ,, .. ~ - - ~ o/

/ --_-....~.i::L~i:.:i:<;..i:.:.-?.,::::7."-:~\~,,~/; ~,C.~ g - ~ ~ -/
/ ---- /. ! "...:::" ..:;~:-- ~ \ \ \ ~ - - - d ~ / , / ~ / ~\~ - l/s, ~ f q , ,

/SUBMARINE FA OR ..... :: .-/~-----~v~JJk~_-~- --- ~ -------i /

FIG. 12. Block diagram of the eastern Caribbean, reconstruction for the middle Eocene (about 50 Ma
ago). View is to south towards the north coast of South America. Caribbean plate moving east with
respect to South America (black arrow). Both trench sediments and ocean-floor fan sediments pile
up to construct Barbados Ridge. L.A.T. = Lesser Antilles Trench.

neath it (Moore & Karig 1976; Dickinson & ridge from the Mid-Atlantic Ridge to Central
Seely 1979). Turbidity currents then went America during part of Cretaceous time. This
round the sides and only pelagic sediments plate boundary had become inactive by the
could accumulate on top (Westbrook 1974). mid-Eocene but probably remained as a topo-
This is consistent with the gradual shallowing graphic feature on the seafloor. Westbrook
during Oceanic Group deposition reported by (1981) gives a full discussion. Subduction in-
Lohmann (1974) and with continuing uplift volving any such east-west irregularity intro-
through the Pleistocene and Recent, witnessed duces many complications not predicted by a
by Coral Rock deposition in a series of terraces. simple two-dimensional subduction-accretion
Transverse trends occur in many active island model.
arcs (summary in Ranneft 1979). At the end of ACKNOWLEDGMENTS:Field work was financed by the
the Barbados Ridge, an east-west cross-warp U.K. Ministry of Overseas Development and the
separates the calc-alkaline volcanics of the Ministry of Trade, Barbados. C. J. Pudsey acknow-
Guadelupe-St Lucia sector of the arc from the ledges a N.E.R.C. research studentship.
predominantly basaltic Grenada-St Vincent The manuscript benefited by discussions with J. F.
Dewey, D. Karig, J. B. Saunders and G. K. West-
sector (Rea 1981). This feature and the E N E
brook.
structures on Barbados may be related to John Cooper, British Museum (NH) helped with the
oceanic transform faults in the downgoing shells and W. J. Kennedy with trace fossils.
Atlantic slab (Peter & Westbrook 1976). One We thank R. Holland, P. Jackson, G. Collins and D.
Caribbean reconstruction (J. F. Dewey, pers. Relton for assistance and O.D.M. for permission to
comm.) shows an east-west trending spreading publish.
306 C. J. Pudsey & H. G. Reading

References
ARNOLD, R. & MACREADY, G. A. 1956. Island- stone of Cretaceous flysch sections of the East
forming mud volcano in Trinidad, British West Alps and other basins. Sedimentology, 22, 387-
Indies. Bull. Am. Assoc. Petrol. Geol. 40, 2748- 416.
58. HIGGINS, G. E. & SAUNDERS, J. B. 1967. Report on
BAADSGAARD,P. H. 1959. Barbados geological review 1964 Chatham mud island, Erin Bay, Trinidad,
and evaluation. Unpubl. report, Gulf Oil Co., West Indies. Bull. Am. Assoc. Petrol. Geol. 51,
132 pp. 55-64.
BLATT, H., MIDDLETON, G. V. & MURRAY, R. C. - - & -- 1974. Mud volcanoes--their nature and
1972. Origin of Sedimentary Rocks. Prentice- origin. Verl. naturforsch. Ges. Basel, 84, 101-52.
Hall, Englewood Cliffs, New Jersey, 634 pp. HOWELL, D. G., CROUCH, J. K., GREENE, H. G.,
BOWIN, C. O. 1976. Caribbean gravity field and plate MCCULLOCH, D. S. & VEDDER, J. G. 1980. Basin
tectonics. Spec. Pap. geol. Soc. Am. 169, 79 pp. development along the late Mesozoic and Caino-
CAUDal, C. M. B. 1972. The larger foraminifera of zoic California margin: a plate tectonic margin of
the Scotland District of Barbados. Eclog. geol. subduction, oblique subduction and transform
Helv. 65, 221-34. tectonics. In: BALLANCE, P. F. & READING, H.
CHAPMAN, R. E. 1973. Petroleum Geology: a Concise G. (eds). Sedimentation in Oblique-slip Mobile
Study. Elsevier, Amsterdam, 304 pp. Zones. Spec. Publ. int. Assoc. Sediment. 4,
1974. Clay diapirism and overthrust faulting. 43-62.
Bull. geol. Soc. Am. 85, 1597-1602. HURLEY, R. J. 1966. Geological studies of the West
CHASE, R. L. & BUNCE, E. T. 1969. Underthrusting Indies. Can. geol. Surv. Pap. 66-15, 139-50.
of the eastern margin of the Antilles by the floor JORDAN, T. H. 1975. The present-day motions of the
of the western North Atlantic Ocean, and the Caribbean plate. J. geophys. Res. 80, 4433-9.
origin of the Barbados Ridge. J. geophys. Res. JUKES-BROWNE, A. J. & HARRISON, J. B. 1891. The
74, 1413-30. geology of Barbados. Part l--the Coral-rocks of
D~ CIZANCOURT, M. 1948. Nummulitcs de i'ile de Barbados and other West-Indian islands. Q. J.
Barbade. MOrn. Soc. g~ol. Fr. 57, Tome 27, Fasc. geol. Soc. London, 47, 197-250.
1. KARIG, D. E., LAWRENCE, M. B., MOORE, G. F. &
CORBO, S. 1979. Vertical distribution of trace fossils in CURRAY,J. R. 1980. Structural framework of the
a turbidite sequence, Upper Devonian, New fore-arc basin, NW Sumatra. J. geol. Soc. Lon-
York State. Palaeogeogr. Palaeoclimatol. don, 137, 77-91.
Palaeoecol. 28, 81-101. KEAREY, P., PETER, G. & WESTBROOK, G. K. 1975.
CRIMES, T. P. 1977. Trace fossils of an Eocene Geophysical maps of the eastern Caribbean, J.
deep-sea sand fan, north Spain. In: CRIMES, T. P. geol. Soc. London, 131,311-21.
& HARPER, J. C. (eds). Trace Fossils 2. Spec. LADD, J. W. & WATKINS, J. S. 1979. Tectonic de-
issue Geol. J. 9, 71-90. velopment of trench-arc complexes on the north-
DAVIESS, S. N. 1971. Barbados: a major submarine ern and southern margins of the Venezuela
gravity slide. Bulll geol. Soc. Am. 82, 2593-601. Basin. In: WATKINS, J. S., MON~ADERT, L. &
DICKINSON, W. R. & SEELY, D. R. 1979. Structure DICKERSON, P. W. (eds). Mere. Am. Assoc.
and stratigraphy of forearc regions. Bull. Am. Petrol. Geol. 29, 363-71.
Assoc. Petrol. Geol. 63, 2-31. LOHMANN, G. P. 1974. PaleD-oceanography of the
DOTF, R. H., JR 1964. Wacke, greywacke and mat- Oceanic Formation, Barbados. Thesis, PhD,
rix--what approach to immature sandstone clas- Univ. Brown.
sification? J. sediment. Petrol. 34, 625-32. MALFAIT, B. T. & DINKELMAN, M. G. 1972. Circum-
EDGAR, N. T., SAUNDERS, J. B. et al. 1973. Initial Caribbean tectonic and igneous activity and the
Rep. Deep Sea drill. Proj. 15, U.S. Govt Printing evolution of the Caribbean plate. Bull. geol. Soc.
Office, Washington, 1137 pp. Am. 83, 251-71.
HEDBERG, H. D. 1974. Relation of methane genera- MATLEY, C. A. 1932. The old basement of Barbados:
tion to undercompacted shales, shale diapirs and with some remarks on Barbadian geology. Geol.
mud volcanoes. Bull. Am. Assoc. Petrol. Geol. Mag. 69, 366-73.
58, 661-73. MEVERHOFF, A. A. & MEYERrtOFF, H. A. 1972. Con-
HEEZEN, B. C., MENZIES, R. J., SCHNEIDER, E. D., tinental drift, iv: the Caribbean "plate". J. Geol.
EWING, W, M. & GRANELLI, N. C. L. 1964. 80, 34-60.
Congo Submarine Canyon. Bull. Am. Assoc. MOLNAR, P. & SYKES, L. R. 1969. Tectonics of the
Petrol. Geol. 48, 1125-49. Caribbean and Middle America regions from
HERRERA, R. C. & SPENCE, J. 1964. The geology and focal mechanisms and seismicity. Bull. geol. Soc.
oil prospects of the Scotland Group of sediments Am. 80, 1639-84.
at Barbados. Unpubl. report, Sinclair and B. P. MOORE, G. F. & KARIG, D. E. 1976. Development of
Explorations Ltd. sedimentary basins on the lower trench slope.
HEss, H. H. 1938. Gravity anomalies and island arc Geology, 4, 693-7.
structure with particular reference to the West MurrI, E. 1977. Distinctive thin-bedded turbidite
Indies. Proc. Am. philos. Soc. 79, 71-96. facies and related depositional environments in
HESSE, R. 1975. Turbiditic and non-turbiditic m u d - the Eocene Hecho Group (South-central
 Sedimentology of the Scotland Group, Barbados 307
Pyrenees, Spain). Sedimentology, 24, 107-31. Texas: record of Appalachian tectonism during
, NILSEN, T. H. & RlccI Luccm, F. 1976. Outer Eocene. Bull. Am. Assoc. Petrol. Geol. 41, 2545-
fan depositional lobes of the Laga Formation 66.
(Upper Miocene and Lower Pliocene), east- TRECHMANN, C. T. 1925. The Scotland Beds of Bar-
central Italy. In: SXANLEY,D. J. & KELUNG, G. bados. Geol. Mag. 62, 481-504.
(eds). Sedimentation in Submarine Canyons, VELBEL, M. A. 1980. Petrography of subduction zone
Fans and Trenches, 210-23. Dowden, Hutch- sandstones--discussion. J. sediment. Petrol. 50,
inson & Ross, Stroudsburg, Pennsylvania. 303-4.
NORMARK, W. R. 1978. Fan valleys, channels and WALKER, R. G. 1978. Deep-water sandstone facies
depositional lobes on modern submarine fans: and ancient submarine fans: models for explora-
characters for recognition of sandy turbidite en- tion for stratigraphic traps. Bull. A m . Assoc.
vironments. Bull. Am. Assoc. Petrol. Geol. 62, Petrol. Geol. 62, 932-66.
912-31. WESTBROOK,G. K. 1974. The structure and evolution
PETER, G. & WESTBROOK,G. K. 1976. Tectonics of of the Barbados Ridge. 7th Caribbean Geol.
the south-western North Atlantic and Barbados Congr. 321-8.
Ridge complex. Bull. Am. Assoc. Petrol. Geol. 1975. The structure of the crust and upper
60, 1078-106. mantle in the region of Barbados and the Lesser
PUDSEY, C. J. 1979. Stratigraphy and sedimentology Antilles. Geophys. J. R. astron. Soc. 43, 201-42.
of the Scotland Group, Barbados. Trans. 4th. 1981. Barbados ridge complex: tectonics of a
Latin American Geol. Congr. (in press). mature fore-arc system (this volume).
RANNEFT, T. S. M. 1979. Segmentation of island arcs
C. J. PUDSEY& H. G. READING, Department of
and application to petroleum geology. J. Petrol.
Geology and Mineralogy, Oxford OX! 3PR,
Geol. 1, 35-53.
U.K.
REA, W. J. 1981. The Lesser Antilles. In." THORPE,
R. S. (ed). Orogenic Andesites, 167-85. Wiley,
New York. Appendix
ROBERTS, H. H., CRATSLEY, D. W. & WHELAN, T.
1976. Stability of Mississippi delta sediments as The base of the Oceanic Group
evaluated by analysis of structural features in
sediment borings. Offshore Tech. Conf. Paper B i o s t r a t i g r a p h i c e v i d e n c e is not yet c l e a r
No. OTC 2425, 14 pp. e n o u g h to c o n f i r m or disprove any age o v e r l a p
RUPKE, N. A. 1978. Deep clastic seas. In: READING, b e t w e e n the S c o t l a n d a n d O c e a n i c G r o u p s . W e
H. G. (ed.). Sedimentary Environments and believe that the O c e a n i c / S c o t l a n d c o n t a c t is at
Facies, pp. 372-415. Blackwell Scientific Publica- least partly u n c o n f o r m a b l e b e c a u s e :
tions, Oxford.
& STANLEY,D. J. 1974. Distinctive properties of
turbiditic and hemipelagic mud layers in the WEST IJ ~ - - ~ , EAST
_~ora/ro'ck~--~/
Algero-Balaeric Basin, western Mediterranean
Sea. Smithson. Contrib. Earth Sci. 13, 40 pp. A. Oceanics -"'~~/"}'~
SAUNDERS,J. B. 1965. Field trip guide. Barbados.4th .f/ J z ././~"?~.
Caribbean Geol. Congress, Trinidad, 443-9.
SEtLACHER, A. 1978. Use of trace fossil assemblages
for recognising depositional environments. In: B.
BASAN, P. B. (ed.). Trace Fossil Concepts. Short
Course Soc. econ. Paleontol. Mineral. Tulsa, 5, Cattlewash Member ( L o h m a n n 1974)
167-81. of O c e a n i c Group
SENN, A. 1940. Paleogene of Barbados and its bear-
ing on history and structure of Antillean-
Debris f l o w - m u d with d i s o r g a n i s e d ,
Caribbean region. Bull. Am. Assoc. Petrol.
m a t r i x s u p p o r t e d pebbles & rafts of
Geol. 24, 1546-610. 5m
fine sand. Presence of v e r t i c a l
1947. Die Geologie der Insel Barbados B.W.I. burrows shows this is not a thrust
(Kleine Antillen) und die Morphogenese der breccia.
umliegende marinen Grossformen. Eclog. geol.
Helv. 40, 199-222. Upper Scotland Formation, p r o b a b l y
SHEPARD, F. P., MARSHALL,N. F., McLOUGHLIN, P. L........... Murphys Member
A. & SULLWA~, G. G. 1979. Currents in Sub-
marine Canyons and other Sea Valleys. Studies in FI6 13 (A) Double unconformity Coral
Geology, 8, Am. Assoc. Petrol. Geol. Tulsa, rock/Oceanics/Scotland sandstones at
Oklahoma, 173 pp. Skeete's Bay; from Jukes-Browne & Harri-
SPEED, R. C. 1979. New views on the geology of son (1891).
Barbados. Trans. 4th Latin American Geol. (B) Measured section at base of Oceanic
Congr. (in press). Group, Cattlewash. Section is horizontal
SYKES, L. R. & EWING, M. 1965. The seismicity of the and right way up. Murphys Member con-
Caribbean region. J. geophys. Res. 70, 5065-74. sists of interbedded sand and clay: Cattle-
TODD, T. W. & FOLK, R. L. 1957. Basal Claiborne of wash Member consists of muddy cherts.
308 C. J. Pudsey & H. G. Reading
(a) Jukes-Brown & Harrison (1891, p. 217) Fig. 13B). This is measured in a large block
figure a section at Skeete's Bay showing the within the Joes River Formation (Fig. 9) but
double unconformity (Fig. 13A). Unfortunately subsequent fragmentation and incorporation of
this locality is now entirely overgrown. These the block into a mudflow does not invalidate
authors were certainly able to recognize a tecto- the point that the base of the Oceanic Group
nic contact between the Oceanic and Scotland was at least partly unconformable on the Scot-
Groups in other places. land Group. In many places the unconformity
(b) A new locality at Cattlewash (Barbados has been a plane of movement during deforma-
grid. ref. 67346185) shows the short section in tion.
Subduction and tectonics on the continental margin off northern
Spain: observations with the submersible Cyana

Jacques-Andr~ Malod, Gilbert Boillot, Raymond Capdevila, Pierre-Alain

Dupeuble, Claude Lepvrier, Georges Mascle, Carla Miiller & Josette
Taugourdeau-Lantz
SUMMARY: Short-term latest Mesozoic-early Cenozoic convergence between the
Iberian and European plates transformed the passive north Spanish margin into an active
margin with southward subduction of the oceanic crust of the Bay of Biscay. This margin
became passive again when subduction ceased in late Eocene or Oligocene time. Nine
dives with the submersible Cyana allowed us to sample the rocks outcropping on the Le
Danois Bank slope. A Mesozoic sequence includes shelf limestones of the Dogger to the
Lower Cretaceous, and Lower Cretaceous detrital sediments (sandstones, conglomerates,
pelites); it overlies a typical Palaeozoic Asturian basement.
The main objective of the diving campaign was to study the structure of these strata in
order to elucidate deformation of the continental wall of an active margin during its early
stages of evolution, before the formation of an accretionary prism. The occurrence in
Lower Cretaceous pelites of a weak schistosity in a reversed fold limb gives evidence of
compressive structures, probably northward-verging overfolds. Additional evidence of
compressive tectonics is the dislocation of quartz grains and dolomite crystals in Lower
Cretaceous sandstones and Jurassic dolomites. These observations suggest that the
continental slope in the Le Danois Bank area has a predominantly imbricate structure.
Deformation and uplift of the northern Spanish margin appear to be related to the early
stages of subduction of the European plate under Iberia.

The northern and southern continental margins profiles and rock samples have been collected
of the Bay of Biscay are very different. The in the area of the Le Danois Bank, a marginal
continental slope off Brittany is in general plateau located on the northern Spanish margin
gently inclined and 50-100 km wide, whereas (Fig. 1). The Le Danois Bank was a part of the
the northern Spanish one is narrow (25 km) and deep margin before the Pyrennean orogeny and
steep (Brenot & Berthois 1962; Laughton et al. culminates now at a depth of 450 m. Boillot et
1975). A deep sediment-filled trough characte- al. (1979) proposed that the impressively steep
rized by a large negative gravity anomaly lies at slope of this bank is the result of an imbricate
the base of the northern Spanish margin (Fig. structure, a hypothesis which we examine fur-
1). According to Sibuet & Le Pichon (1971) and ther in this paper.
Le Pichon et al. (1971) this is a fossil trench We used the submersible Cyana during July
associated with an Eocene subduction zone. 1979 to address three problems on the Le
Convergence between the Iberian and Euro- Danois Bank:
pean plates, during the latest Mesozoic and the the early Mesozoic stratigraphy,
early Cenozoic, and the resulting southward the nature of the basement,
subduction of the oceanic crust of the Bay of the style of deformation.
Biscay, have transformed the passive northern We surveyed two geological cross-sections of
Spanish margin into an active one. Short term the northern slope of the bank between 3000
subduction at this time is put forward to explain and 500 m depth during nine dives. Fifty-four
the deformation and uplift of the rocks on the samples of rocks (most of them taken in situ),
northern Spanish continental shelf (Boillot et numerous photographs and video-records were
al. 1971, 1973; L a m b o y & Dupeuble 1975; collected (Malod et al. 1980; Capdevila et al.
Boillot & Capdevila 1977; Boillot et al. 1979). 1981). We concentrate here on the structural
The margin became passive again when the observations.
convergence and the subduction ceased, owing
to collision between the Iberian and European
plates along the Pyrenees. The margin assumed Mesozoic stratigraphy and
its current physiography after further phases of palaeogeography
normal faulting, transcurrent faulting and subsi-
dence. A continuous survey of the bedrock is not
Much new data including seismic reflection possible, because of local soft mantling sedi-

309
310 J . - A . M a l o d et al.

LE DANOIS B,e

FRANCE

200

PANISH TRENCH

A S T U RI A -"--~.~,...~c~~_~ --/- '

5"
I I I I I

FIo. 1. Location of the Cyana dives on the Le Danois Bank.

The divcs are numbered from I to 9. Insert' location of the chart with rcspect to the Bay of Biscay
and the fossilized north Spanish trench.

ments. However, rock outcrops are numerous in age from Portlandian to earliest Cretaceous
and we can draw an accurate geological cross- (perhaps Valanginian). Turbidite-like Hauteri-
section for each dive, plotting lithologies by vian sandstones rich in plants debris follow, and
means of our numerous samples. We found are capped by Barremian pollen-rich pelites.
three stratigraphic units in different depth These pelites and a previous sample of Aptian
zones. These were located respectively during age (H.76 DR.09 in Boillot et al. 1979) dredged
dives 1, 2, 3, 4, 9, dives 5, 6, 7 and dive 8, and at the same level on the slope are comparable in
are summarized in Fig. 2. lithology to black shales sampled in Deep Sea
Submarine cliffs associated with Upper Juras- Drilling Project holes 398 (Sibuet & Ryan et al.
sic outcrops show clearly on a bathymetric chart 1979) and 400, 402 (Montadert & Roberts et al.
of the Le Danois Bank slope (Malod & Vanney 1979).
1981). These cliffs are remarkably parallel, The intermediate unit (II in Fig. 2) occurs
extending several kilometres along the slope between 2000 and 1500 m depth in the west,
at nearly uniform depths. The observed and between 2000 and 600 m in the east. This
sedimentary units are therefore probably of unit consists of azoic dolomites succeeded by
good lateral continuity. Despite the existence of thick pollen-rich Aptian micaceous sandstones.
faults, oblique to the general east-west orienta- In the upper part of the unit one or more levels
tion of the slope, we consider that the succes- of conglomerate indicate coarse detrital sedi-
sion of sedimentary units in Fig. 2 may be mentation of unknown but probably Aptian to
representative for all the studied area. Albian age. Because of their stratigraphical
The lower unit (I in Fig. 2) occurs between position and because Portlandian to earliest
3000 and 2000 m depth. Kimmeridgian to Port- Cretaceous limestones of unit I are slightly
landian limestones in decimetre-scale beds, dolomitized, we suppose that dolomites in unit
sometimes with marly interbeds, occur at the II are of Late Jurassic to earliest Cretaceous age.
base. They are succeeded by massive lime- The upper unit (III in Fig. 2) occurs from
stones, commonly slightly dolomitized, ranging 1500 m depth to the top of the bank in the west
 S u b d u c t i o n o n the c o n t i n e n t a l m a r g i n o f f n o r t h e r n S p a i n 311

IIETERS
......
UNITS (I)
I ~
II
9 3 4E4,8E6,9E 3,9E 5,9E9
II1 ,
2
7
1500
3
_ _ I ,,
¢- --'=-.~, 3ES,4E3 [ [
i i
- i i

BARREMIAN

HAUTERIVIAN ..... 7E I. 7E2 ,, .

IERRIAS IAN VALANG IN IAN

6E5
8E4,8E5
7E7,/E8
5E2 6E6
, , q.:}
K I . ~ R IDG[AI~ 6E4
6E3
OXFORDIAI~
to BAJOCIAN 6E2

~?6EI,9EI0

i 5
,° / • ,, ......... ,2

FIG. 2. The three stratigraphic units of the Le Danois Bank.

1: Cambrian quartzite; 2: limestone in decimetre-scale beds; 3: limestone with interbedded marl;
4: massive limestone; 5: dolomite; 6: pelite; 7: sandstone (turbiditic); 8: micaceous sandstone; 9:
conglomerate; 10: early Tertiary reefal limestone; 11: Aquitanian limestone and Pliocene marl; 12:
Quaternary mud.
Sample locations, with numbers, are shown. (1) After Montadert et al. (1979); (2) partly after
Boillot et al. (1979).

and only on the top of the bank in the east. This red in Hauterivian time, probably as a result of
unit begins with in situ pre-Mesozoic basement, the initial rifting of the margin. We assume that
here consisting of a two-mica quartzite of prob- during the Cretaceous opening of the Bay of
able Cambrian age. Above it massive strata of Biscay the evolution of the north and south
undetermined nature crop out. Limestones margins were very similar. On this basis, the
(Bajocian to Oxfordian) and dolomites com- profile OC.412 recorded by Montadert et al.
plete the succession observed with the sub- (1979) on the Armorican margin has been
mersible. Samples cored and dredged on the taken as a model for the structure of the north
top of the bank show that the upper part of this Spanish margin (Fig. 3A, in which we have
unit consists partly of grey marls of Early Cre- removed the post-Aptian layers to draw the
tm:eous age (Boillot et al. 1979). Aptian reconstruction). The Early Cretaceous
Late Jurassic and earliest Cretaceous de- sediments are located in narrow basins between
posits indicate platform-type sedimentation and tilted blocks of the stretched continental crust.
are similar in all three stratigraphic units. Later In this way the post-Hauterivian facies contrasts
Early Cretaceous facies are more detrital and between the three stratigraphic units are inter-
contemporary deposits differ within the three preted as being the result of sedimentation in
stratigraphic units. The transformation occur- different basins.
312 J . - A . M a l o d et al.

B Present Time
0 kmL, C~ ( ~ ) - ~ I I I Le Danois Bank
" (i early Cretaceous
Jurassic
pre-Mesozoic basement

o| 5km
]
lO

A Aptian Time
III
~ Le Danois Bank
okmN L.~ ~ IS

1
7a t NiiiiN iiiiiiiiii!ii ! i ii!iiiiil !i iill
....................................,:-:::~~:-iiiii!i~:i!ili!i!iii!iii!iiiiiiiiiiiiiiiiiiiilililili!!iii
I

FIG. 3. A sketch of the geological evolution of the studied area.

Stratigraphic sequences l-Ill as in Fig. 2.
(A) Aptian time: the cross-section is based on a profile of the Armorican margin (OC.412 of
Montadert et aI. 1979). On this profile, the post-Aptian layers have been removed and the thickness
of the Jurassic is based on diving data. a and b are the positions of future overthrusts.
(B) Present time: the slope of the Le Danois Bank is characterized by an imbricate structure, a, b
and c represent overthrusts.

Structural observations
No obvious folds have been observed in any sedimentary sequences up-slope (Fig. 2). It may
dives. Notwithstanding the narrow field of view be caused by normal faulting or by thrust
from the submersible, it appears to us that the faulting (Fig. 3). The schistosity and the micro-
outcropping layers mainly form a homoclinal scopic breakage of crystals in some rocks are
sequence dipping gently to the south. However, strong arguments for the second hypothesis.
we have observed local N, E and SE dips. Our Moreover, the close proximity of the three
structural observations may be summarized as different sedimentary units and the facies con-
follows: trasts between them (Fig. 2) argue against
(1) An outcrop of Barremian pelites shows normal faulting. They may be more easily ex-
two cleavages (Fig. 4). The more inclined one is plained by superposition in an imbricate struc-
parallel to bedding; the other, an apparently ture (Fig. 3B). The overthrusts may have taken
horizontal fabric, is interpreted as a weak schis- place either at the base of the Mesozoic or in a
tosity. It suggests that the outcrop belongs to plastic level of the pre-Mesozoic basement.
the reversed limb of a fold, perhaps associated (4) Normal faults (Fig. 5) and probable
with a thrust fault. strike-slip faults, also occur. They explain the
(2) At a microscopic scale, some rocks show limited repetition of strata with the same facies,
other evidence of a compressional tectonic but not, in our opinion, the repetition of units
event. Dolomite crystals in the Jurassic and where the Lower Cretaceous has a different
quartz grains in the Early Cretaceous sand- lithology.
stones are often broken and crushed. How can we explain the coexistence of
(3) The most important observation is prob- extensional and compressive structures? The
ably the tectonic repetition of the Mesozoic structural evolution of the margin has been
 Subduction on the continental margin off northern Spain 313

FI6.4. Outcrop of Barremian pelites showing two cleavages.

The more inclined fabric is bedding-parallel; the other apparently horizontal fabric is probably a
weak schistosity (dive 1, 17h09, 2299 m depth, scale bar: about 30 cm).

described by Boillot et al. (1973) and Lamboy &

Dupeuble (1975), using a detailed survey of the
Asturian and Galician continental shelves. The
deformation of this part of the margin, situated
60 km south of Le Danois Bank, is polyphase.
During the latest Cretaceous and the early
Eocene, compression induced folding, uplift
and erosion by subaerial processes. At the end
of the Oligocene, after the Palaeogene trans-
gression, extensional (horst and graben) struc-
tures appeared. Finally NE and NW strike-slip
faults cut the margin.
On the Le Danois Bank, the Eocene hiatus
does not permit us to work out such a precise
history. Aquitanian neritic limestones lie on a
flat erosion surface on top of the bank, resting
unconformably on deformed Mesozoic strata.
We ascribe the main compressive deformation
on the slope of the Le Danois Bank to the latest
Cretaceous-early Eocene tectonic phase and
the normal or transcurrent faults to later Oli-
gocene and Miocene tectonic phases.

Conclusions
FI~. 5. Normal fault slickenside with vertic- The north Spanish margin has been shortened,
al striae in Tithonian limestones (dive 7, deformed and uplifted during the latest Cre-
13h15, 2355 m depth, scale bar: about taceous and early Eocene, coincident with con-
20 cm). vergence between the Iberian and European
314 J.-A. Malod et al.
I
EUROPEAN PLATE - .mmr4--IBERIAN PLATE
I
! /-~ Pyrenees Le Danois

, ,~;',i,:..,::.;..-".~i~";::,.~T".".,.... ~" ' """ '::::~'::':::'"?:":':~'%~"

CONTINENTAL CRUST OCEANIC CRUST ~'~\

z / MANTLE "-- .~. \\

FIG. 6. Reconstruction showing how Le Danois Bank, near the junction of the European and Iberian
plates, acquired an imbricate structure during the latest Cretaceous-early Eocene by contempor-
aneous subduction of the oceanic crust of the Bay of Biscay under Spain.

plates (Fig. 6). The Cretaceous stable margin buoyancy of the young lithosphere (Molnar &
was telescoped by thrusting resulting from Atwater 1978). The rather intense deformation
southward subduction of the oceanic litho- of the stable north Spanish margin is probably a
sphere of the Bay of Biscay under Spain. The consequence of this 'forced subduction'.
resulting imbricate structure is completely dif-
ferent from that of an accretionary prism, in- ACKNOWLEDGMENTS: We particularly acknowledge
volving deformed pieces of the old stable mar- the Spanish Government who authorized the diving
gin, not trench and ocean floor material. campaign. We are grateful to all the people and
This subduction seems anomalous in several organizations who made the cruise and the dives
possible: the Commander and crew of the R.V.
respects: Nadir; Commander Caillart, responsible for the
(a) The lack o f calc-alkaline volcanism. This Cyana team, and Y. Morel, in charge of logistic
may be explained by the small amount of operations. We especially acknowledge Messrs
oceanic lithosphere involved in the subduction Kientzy, Leroux, Arnoux and Nivaggioli, pilots and
(less than 130-150 km according to Boillot & co-pilots of the Cyana. The following organizations
Capdevila 1977). supported the diving campaign and subsequent analy-
(b) The importance o f the deformation o f the sis of the data: Centre National pour l'Exploitation
margin, despite the brevity o f the subduction. At des Oc6ans, Universit6 P. & M. Curie de Paris,
the beginning of subduction ( - 6 5 Ma) the old- Universit6 de Rennes, Universit6 de Rouen, Centre
National de la Recherche Scientifique and Comit6
est oceanic lithosphere of the Bay of Biscay was d'Etudes P6troli6res Marines.
45 Ma old, according to Montadert et al.
(1979) (seafloor spreading under the Bay of
Biscay began at - 1 1 0 Ma). Consequently re- Contribution no. 128 of the 'Groupe d'Etude de la
sistance to the initiation of subduction may Marge Continentale' (University of Paris and
have been great, because of the relative C.N.R.S., ERA 605).

References
BOILLOT, G. & CAPDEVILA,R. 1977. The Pyrenees: 1971. Structure et histoire g6ologique de la
subduction and collision? Earth planet. Sci. Lett. marge continentale au Nord de l'Espagne (entre
35, 151-60. 4° et 9°W). In: Histoire structurale du Golfe de
-- , DUPEUBLE, P. A., HENNEQUIN-MARCHAND,I., Gascogne. V. 6.1-52. Technip, Paris.
LAMBOY, M. & LEPRETRE, J. P. 1973. Carte & MALOD, J. A. 1979. Subduction and tectonics
g6ologique du plateau continental nord-espagnol on the continental margin off northern Spain.
entre le Canyon de Capbreton et le Canyon Mar. Geol. 32, 53--70.
d'Aviles. Bull. Soc. geol. Fr. XV, 367-91. BRENOT, R. & BERTHOIS, L. 1962. Bathym6trie du
-- , LAMBOY,M., D'OzouVILLE, L. & SIBUET,J. C. secteur atlantique du banc Porcupine (Ouest de
 S u b d u c t i o n o n the c o n t i n e n t a l m a r g i n o f f n o r t h e r n S p a i n 315

l'Irlande) au cap Finist6re (Espagne). Rev. Tray. de la pente du banc Le Danois (marge continen-
Inst. POches Mar. 26, 219-46. tale nord-espagnole) d'apr~s un relev6 bathym6t-
CAPDEVILA, R., BOILLOT, G., LEPVRIER, C., MALOD, rique au sondeur multifaisceaux et des observa-
J. A. d~ MASCLE, G. 1981. Les formations cristal- tions en submersible Cyana. Ann. Inst.
lines du Banc Le Danois. C.r. Acad. Sci. Paris. Oc~anogr. 56, 73-83.
LAMBOY, M. & DUPEUBLE, P. A. 1975. Carte g6ologi- MOLNAR, P. & ATWATER, T. 1978. Interarc spreading
que du plateau continental nord-ouest espagnol and cordilleran tectonics as alternates related to
entre le Canyon d'Aviles et la fronti~re portu- the age of subducted oceanic lithosphere. Earth
gaise. Bull. Soc. geol. Fr. XVII, 442-61. planet. Sci. Lett. 41,330-40.
LAUGHTON, A. S., ROBERTS, D. G. & GRAVES, R. MONTADERT, L., ROBERIS, D. G. el al. 1979a. Initial
1975. Bathymetry of the northeast Atlantic: Mid- Rep. Deep Sea drill. Proj. 48, U.S. Govt Printing
Atlantic Ridge to northwest Europe. Deep Sea Office. Washington.
Res. 22, 791-810. , DE CHARPAL, O., ROBERTS, D. G., GUENNOC,
LE PICHON, X., BONNIN, J., FRANCHETEAU, J. & P. & SIBUET, J. C. 1979b. Rifting and subsidence
SmUET, J. C. 1971. Une hypoth6se d'6volution processes. In: TALWAN~, M., HAy, W. & RVAN,
tectonique du golfe de Gascogne. In: Histoire W. B. F. (eds). Deep Drilling Results in the
Structurale du Golfe de Gascogne, VI. 11.1-44. Atlantic Ocean: continental margins and paleo-
Technip, Paris. environment. M. Ewing Ser. 3, 154-86.
MALOD, J. A., BOILLOT, G., CAPDEVILA, R., SmUET, J. C. & LE PICHON, X. 1971. Structure
DUPEUBEE, P. A., LEPVRIER, C., MASCLE, G., gravim~trique du Golfe de Gascogne et le loss6
MULLER, C. (~ TAUGOURDEAU-LANTZ, J. 1980. marginal nord-espagnol. In: Histoire Structurale
Plong6es en submersible au Sud du Golfe de du Golfe de Gascogne, VI. 9.1-18. Technip,
Gascogne: stratigraphie et structure de la pente Paris.
du Banc Le Danois. C.r. Somm. Soc. geol. Ft. 3, --, RYAN, W. B. F. et al. 1979. Initial Rep. Deep
73-6. Sea drilling Proj. 47, U.S. Govt Printing Office,
- - (;~ VANNEY, J. R. 1980. Etude morphologique Washington.

J. A. MALOD, G. BOILLOT, C. LEPVRIER, G. MASCLE /~ J. TAUGOURDEAU-LANTZ,

Universit6 P. & M. Curie, 4 place Jussieu, 75230 Paris Cedex 05, France.
R. CAPDEVILA,Universit~ de Rennes, 35042 Rennes Cedex, France.
P. A. DUPEUBLE, Universit~ de Rouen, 76130 Mont-Saint-Aignan, France.
C' MOLLER, BEICIP, 232, avenue Napol6on Bonaparte, 92500 Rueii-Malmaison,
France.
 Subduction in the Hellenic Trench: probable role of a thick
evaporitic layer based on Seabeam and submersible studies

X. Le Pichon, P. Huchon, J, Angelier, N. Lyb~ris, J. Boulin, D. Bureau,

J. P. Cadet, J. Dercourt, G. Gla~on, H. Got, D. Karig, J. Mascle, L. E. Ricou
& F. Thiebault
SUMMARY: A Seabeam survey of the Hellenic Trench in 1978 was followed by a
submersible survey in 1980. Fifteen dives were completed between 1500 and 3000 m
covering 58 km on the bottom and obtaining 48 samples. Thus, the Hellenic Trench is
probably now the most intensively studied trench as previous work included seismic
reflection, dredging, coring, bottom photography, side-scan sonar and drilling. Because of
the high sedimentation rate and the relatively shallow depth, we did not expect outcrops
except along faults and in canyons. Active tectonics with numerous faults and folds have
been observed. But we have also discovered vertical cliffs with hard rock exposure over a
depth range of 1000-1500 m. In this paper, we discuss the nature of the hard rock outcrops
observed both on the inner wall and on two hills that we interpret as diapirs. We conclude
that at least part of these hard rocks are probably Messinian evaporites and concentrate on
the tectonic role of this evaporitic layer in the Hellenic subduction. It is likely that the
evaporites do not subduct but tend to pile up in the trench, to form a floating evaporitic
basin.

We call Aegea the landmass which includes Hellenic Trench, which is approximately per-
southern Greece, the A e g e a n sea and Western pendicular to the direction of subduction.
Anatolia. Aegea is overriding the Mediterra- Three dives were made in zone 3, in the Strabo
nean seafloor toward the south-west along the Trench, one of two parallel trenches forming the
Hellenic Trench (McKenzie 1978; Le Pichon & S W - N E Levantine branch. This branch makes
Angelier 1979; Angelier 1979; Le Pichon & an angle of only 20 ° with the direction of
Angelier 1981). As a contribution to the H E A T subduction and is thus a nearly pure transform
program (Hellenic Arc and Trench: Le Pichon portion of trench. These areas had previously
& Hs/i 1979), we have surveyed parts of the been mapped using Seabeam on board J e a n
Hellenic Trench using multi-beam echo- C h a r c o t (Le Pichon et al. 1979a; Lyberis et al.
sounding (Seabeam) in 1978 (Le Pichon et al. 1981) and had been the object of numerous
1979a) and then the C y a n a SP-3000 submersible sea-surface cruises (e.g. Got et al. 1977; Le
in 1979. Fifteen dives were conducted between Quellec et al. 1980).
water depths of 1500 and 3000 m (Le Pichon et The principal aim of these systematic detailed
al. 1979b, 1980; Huchon et al. 1981); the three studies was to characterize the strain pattern
diving zones are located on Fig. 1. A single dive over the whole length of the Hellenic subduc-
was made in zone 1. Eleven dives were made in tion zone, using an interpretation of the fine-
zone 2 in the N W - S E Ionian branch of the scale bathymetry, as given by Seabeam, and the
submersible observations of tectonics on the
seafloor. Le Pichon et al. (1981) have shown
that the resulting superficial strain pattern is
different over the inner wall, where it is purely
extensional (01 vertical) and over the main part
of the trench floor area where it is purely
compressional (~r3 vertical). Thus the trench
floor is the locus of a major discontinuity in the
strain pattern. The A e g e a n extensional zone
(Aubouin & Dercourt 1965; Angelier 1979)
extends over the whole of the Hellenic arc at
the foot of the inner wall. The trench floor, on
the other hand, belongs to a compressional
FIG. 1. Location map. 1,2,3: diving zones. zone which may extend seaward to a large part
a,b,c: locations of Figs 3, 5 & 6 respective- of the Mediterranean ridge (Le Pichon & Ange-
lier 1981). The deformation in the trench floor

319
320 L e P i c h o n et al.

, o

1 o

I
A
'7,,.
B C ;~I" z~"
/
/ f 2~" 2]"

Fro. 2. Stress-strain relationship along the Hellenic subduction zone after Lc Pichon et al. (1981).
1,2,3 and 4: areas mapped with Seabeam. Open arrows: horizontal projection of the Africa-Hellenic
arc slip vector according to Le Pichon & Angelier (1979). Black arrows: direction of the axis of
maximum shortening o 1 based on Seabeam analysis. Dashed arrows: direction of o I based on the
analysis of tectonic structures observed during submersible dives. A,B,C: theoretical relation
between stress and relative motion (schematic).

area is compatible with subduction of the H s i & Ryan 1974; Got et al. 1977). However,
Mediterranean seafloor toward the north-east, our observations indicate that tectonic features
ol being orientated 030 ° (perpendicular to are frequent within the unconsolidated to semi-
the trench) along the Ionian branch and N 0 ° consolidated sedimentary cover (Le Pichon et
(oblique to the trench) along the Levantine al. 1981) but also that hard rock outcrops exist
branch (Fig. 2). over large portions of the trench system. Speci-
In this paper, we wish to emphasize a unique fically, extensive hard rock outcrops have been
characteristic of the Hellenic subduction zone, observed in two areas: in the Matapan site (area
which is the presence of a thick evaporitic layer, 1 in Fig. 1) and in the Poseidon site (area 2 in
and to discuss observations which seem to Fig. 1).
suggest that the evaporites are tectonically
thickened below the trench. They consequently
tend to act as a decoupling layer between the
The Matapan site
P l i o - Q u a t e r n a r y sedimentary cover and the The only dive in the Matapan area was made
deeper more compact pre-Messinian sediments. on the upper portion of the inner wall, about
2000 m above the 4600 m deep trench floor,
which is too deep for Cyana. Fig. 3 shows the
Observations of consolidated geological section along the dive. The inner
s e d i m e n t a r y outcrops wall consists of a series of steps with sub-
vertical cliffs facing the trench floor, each a few
Prior to our submersible survey, it was feared metres high. Three of these cliffs are several
that all tectonic activity might be hidden by the tens of metres up to about 100 m high and have
high sedimentation rate (about 10 cm/1000 yr: been interpreted as major fault scarps. Prob-
 Subduction in the Hellenic Trench 321
-1000
WNW ESE

Innerwall
ooze NN2|
~ R .... t .... ' 1500

San0,tone

Conglomerate

ooze NN21
~S
__,,~¢_~- - - -
2000

ch,or,t,c \ ~ i - _
sandstone rl~lnrit "r ~ ~ ~ --
! Oligo .... tO S a n~'s'to'ne ~ - . . Oiigo - rrliocene

2500

.......................

P{iO - q u a t e r n a r y ?
0 500 m
3000
m

FIG. 3. Geological section across a part of the inner wall of the Matapan Trench (site 1; see Fig. 1).

able slickensides were observed within the

The Poseidon site
massive rocks which outcrop on the cliffs; they
are compatible with pure normal faulting. Two Poseidon Trench (2 in Fig. 1) is much shal-
samples collected at the foot of this cliff consist lower than Matapan Trench, in part because of
of sandstones containing plant debris. One of the presence of a complex of circular hills rising
them has been dated by the nannoflora (Muller, from the trench floor. The largest one we called
pers. comm. 1980) as late Oligocene to early Pollux (P in Fig. 4); it is 5 km in diameter and
Miocene. Their composition suggests that they rises 1000 m above the 3000 m deep trench floor.
result from the erosion of rocks similar to those A smaller hill, named Castor, which occurs
which outcrop in the Peloponnesus, immediate- immediately to the SE, is only 500 m high (C in
ly to the north of site 1 (Chamley, pers. comm. Fig. 4). The hills are asymmetrical, the steeper
1980). side facing seaward. They divide the trench
Although we have no indications as to the floor into two portions of unequal width, the
nature and composition of the lower inner wall narrowest portion to the south. Actually, Cas-
in the Matapan area, the upper part of the inner tor very nearly encroaches upon the outer rise,
wall is most probably the result of extensional leaving only a narrow corridor, a few metres
block faulting within an Oligo-Miocene detrital wide. The Seabeam survey has demonstrated
sedimentary series. that such hills are a common feature of the

N 17-
outer wall Diapirs Innerwall
P
" C --

FIG. 4. Three-dimensional diagram of the diving zone of site 2. The observer looks to the NW from
the surface of the sea, 30 km away from the centre of the diagram. Vertical exaggeration: x2. The
thick line indicates the axis of the trench. C: Castor hill. P: Pollux hill. 5,6: position of sections 5 and
6.
322 Le Pichon e t al.

Hellenic Trench (Le Pichon et al. 1979a; Lyberis deep. Finally, the inner wall has an average
et al. 1981), and seismic reflection suggests that slope between 25 ° and 45 ° and rises to less than
they are diapiric (Le Quellec et al. 1980). 1500 m.
Because the depth at the axis of the trench is Extensive hard rock outcrops have been
less than 3000 m, we were able to investigate observed both on the flanks of Castor and
two complete sections from the outer rise to the Pollux hills and on the inner wall. These out-
inner wall, one crossing Pollux hill and the crops occur on the faces of sub-vertical cliffs a
other crossing Castor hill. few metres to a few tens of metres high, sepa-
Detailed portions of the Castor section are rated by portions of 25°-45 ° slopes covered by
shown in Figs 5 & 6 and general interpretation a P t e r o p o d - F o r a m i n i f e r ooze of Holocene or
is shown in Fig. 7. The section is about 10 km uppermost Pleistocene age (see Fig. 5).
long. From south to north, the outer rise on the A c o m m o n characteristic of these outcrops is
seaward side has a maximum slope of 7 ° and that they are affected by erosion resulting from
progressively d e e p e n s from 2300 m to about dissolution. However, the intensity of erosion is
3000 m at the axis of the narrow corridor. The extremely variable from thin micro-grooves to
straight steep southern flank of Castor has a deep grooves (Fig. 8). In quite a few cases, the
slope of about 40 °. The northern flank has a dissolution has been sufficiently intense to iso-
semi-circular outline and an average slope of late columns or pillars a few centimetres to a
10°-20 ° . The northern portion of the trench few tens of centimetres in diameter (Fig. 9). In
floor is about 3 km wide and 2600-2800 m

SW NE
-- 1500

Recent ooze Innerwall

Ooze NN21
MaSSive rocks ~ ~
with caverns and Holocene ooze ~ ' " 2000
clissoJutlon forms Ca v ~ I : ' , : .......

Ill I I

ooze NN2| ~ i " ' ~/i'"

pt e r~ ~ , ' , ' , "
25,00
~ . ~ . . . , ~,:.v,,'...-:" ,'-',': ' " ;i

0 bOOm
I 3000
m

FIG. 5. Geological section across the lower part of the inner wall of the Poseidon trench (site 2; see
Fig. 1).

S Outerwall N
V
2600-
o -3-- . ~ _ Diapir
~ - Iz ~ o o 0 o -- - ~ ' - ~ II
\. . . . ~--~<__ :. --_:= ~_ _ o o o _ ---~ ~t// Trench
2700
o o

2800

^^^?^^?.^.~.&~ -o f o \ ~ : ? ~
A A ^ ^ ^ .A A ^ A A " ' ~ A A A ~ /~ ^ -~ .... ~ . --^. "~ ~--
2900
m - 0 100 200m RF - ~'A ~'?
Me ? I - F - ~ - --~ Me?

FIG. 6. Geological section across the outer wall to the base of Castor diapir, after Le Pichon et al.
(1981) (site 2, see Fig. 1). PI-Q: Plio-Quaternary, with overlying crust. Me: Messinian. O: ooze. F:
fold. RF: reverse fault. NF: normal fault. I: base of diapir. II: zone of overthrusting. III: zone of
intense shortening. IV: zone of limited extension. V: zone with little tectonic activity. NN20 is the
calcareous nannoplankton biozone recognized in the crusts.
 Subduction in the Hellenic Trench 323

SSW
NNE
Hellenic trench
Outerwall Innerwall ,/~ IOOO
~/ Diapir ~ Ill ~

^ A "~~---~'~-~" . ~ ~ ~ ~ ,, ~ '~ / ~ L 30o0

A A ^ 4 /~ A A /~ ~ A 4 / ~ m
A A A 4 A A
A A A A
A /1 A 4

Plio- quaternary A

I ~ Messinian
0 I 2kin
1 I i

F1G. 7. Interpretative geological section across the Hellenic Trench, from the outer wall to the inner
wall (site 2; see Fig. 1).

FIo. 8. Sub-vertical cliff, deeply affected by dissolution, at the foot of Castor diapir, site 2,w2956 m.
Note that gouges are less developed at the foot of the cliff, indicating uplift of the diapir with respect
to the talus.

The four numbers at the bottom left of each photograph (Figs 8-19) are, from top to bottom:
hour (e.g. 4400 = 14 h 40.0. Fig. 8);
heading of the submersible in degrees. This makes an angle of 20 ° to the right with the axis of the
camera (e.g. 0038 = 038 ° (submersible) = 018 ° (camera);
uncorrected depth in metres (e.g. 2781 = 2956 m after conversion of pressure to water depth;
altitude in decimetres with regard to the bottom (e.g. 0043 =4.3 m).
The scale of the photographs is variable, from 2 to 3 m in width for the foreground. For example, in
Fig. 10, the pillars are 1 m high.
324 Le Pichon e t al.

Ft(;. 9. Intersection of two fault scarps trending 110° (on the right) and 040 ° (on the left) on the inner
wall, site 2,--2310 m. The dissolution is here sufficiently intense to isolate columns or pillars (see
text).

Fi6. 10. Cavern with pillars supporting the sloping roof formed by indurated slope formation. Inner
wall, site 2,--2370 m.
 Subduction in the Hellenic Trench 325

one extreme case, the erosion has carved a The roof of the cavern is actually an indu-
cavern about 3 m deep and 1 m high, with rated slope formation. Such slope formations
numerous pillars supporting the sloping roof have been observed frequently (Figs 11 & 12).
(Fig. 10). They abut against the sub-vertical scarps and

FiG. 11. Drapes formed by indurated slope formation, hanging from the vertical cliff. Inner wall, site
2,--2311 m.

FIG. 12. Dissolved drapes with festoons and holes. In the foreground, the mechanical arm of Cyana.
Base of Castor diapir, site 2,--2937 m.
326 L e Pichon et al.
have a slope of 25°-45 °, similar to the slope of sub-vertical scarps or the indurated slope forma-
the sedimentary talus, forming drapes which are tion, except for a small piece of oolitic dolomi-
obviously affected by dissolution (as shown tic limestone, which was deposited in very
by the common occurrence of circular holes a shallow water and has been subjected to sub-
few centimetres in diameter). Some of these aerial erosion several times (Purser, pers. comm.
holes are still complete while others have been 1980). We have not been able to date this
eaten away by erosion (Figs 10, 11 & 12). We sample. More surprising, perhaps, has been our
suggest that these features resulted from a inability to sample fallen debris on the
process of induration in the uppermost layer of sedimentary talus at the base of the cliffs; but
the sedimentary talus. When the talus slumped such talus is lithified early by interstitial slope
they were left hanging in mid-water, which sediment and is equally resistant to sampling.
explains the occasional occurrence of a series of The other type of evidence we have is the
several such drapes at 1 m intervals. observation of a scarp a few metres high and a
An interesting point is that a fossil fauna of few tens of metres long at the base of Pollux hill
sponges and scleractinian corals, still in life which has a thin and wavy stratification (Fig.
position, covers the hard rock surfaces above 13). The thin beds stand out in relief because
2200 m depth. Dating gave an age of 18,200 + they are less affected by dissolution (Fig. 14). In
500 yr for one of the scleractinian corals (Delib- one location, the horizontal stratification is
rias, pers. comm. 1980). It is thus probable that interrupted by a pocket of breccias which has
this fauna is part of the cold water Mediterra- apparently filled a synsedimentary hole pro-
nean fauna which disappeared at the end of the duced by dissolution (Fig. 15).
last glaciation (Zibrowius, 1980). Some indu- Finally a polygenetic conglomerate contain-
rated slope formations include sponges and ing pebbles of Upper Miocene to Messinian
consequently may be less than 20,000 yr old. pelagic limestone has been sampled at the foot
of Pollux hill, on the top of a talus accumula-
Unfortunately, the evidence we have on the tion. The matrix is of Lower Pliocene age.
nature of the rocks observed is meagre. Despite However, although it is unlikely that the peb-
repeated attempts at sampling, we were unable bles have been transported a long distance,
to break and retrieve any sample of either the their exact origin is unknown.

"...F,~ .,' ,¢2.

FIG. 13. Formation with wavy thin bedding. Foot of Castor diapir, site 2,--2940 m.
 Subduction in the Hellenic Trench 327

FIG. 14. Thin beds standing out in relief by differential dissolution. Castor diapir, site 2,--2942 m.

FIG. 15. Pocket of breccias interrupting the horizontal bedding. Castor diapir, site 2,--2812 m.
328 L e P i c h o n et al.

Discussion of the lithology of Tectonics related to Castor and Pollux

consolidated sedimentary outcrops hills
Although we cannot yet reach firm conclusions As noted previously, the sub-vertical cliffs
concerning the lithology of these hard rock observed on the flanks of Castor and Pollux
outcrops, the available evidence leads to the probably correspond to fault scarps. However,
following tentative deductions. in contrast to the faults of the inner wall, their
There is a striking difference between the trend changes continuously and follows the
Oligo--Miocene sandstones observed on the up- general shape of the circular hills. The scarps at
per part of the inner wall in the Matapan area the base of the hills commonly show a zone
and the hard rock formations in the Poseidon about 1 m high which has been only recently
area. In the Matapan area, we did not find exposed to water and consequently is not cov-
evidence for intensive erosion by dissolution. In ered by the ubiquitous thin dark concretionary
the Poseidon area there is strong evidence for film which covers older rock exposures. At
major erosion by dissolution, although the in- other places, the base of the scarp over a similar
tensity of erosion varies widely from one place height is much less affected by dissolution than
to another. We conclude that the lithology the zone immediately above (Fig. 8). Both facts
observed in the Poseidon area includes a large indicate upward motion of the hill with respect
variety of interstratified rocks, ranging from to the adjacent sedimentary fill and suggest that
highly soluble ones to those which are barely this motion is geologically recent.
affected by dissolution. On the flank of Pollux, between 2700
The erosional forms observed are always and 2200 m, a possible tectono-sedimentary
vertical which indicates that the dissolution m61ange has been observed. It consists of
process has been controlled by gravity and that imbricated marly plates mixed with large con-
there has been no major tectonic tilting since. glomeratic blocks. The sedimentary formation
Furthermore, the scarps of the inner wall are observed on the summit of Pollux appears to be
cut by 030°-trending corridors which most conglomeratic. This may further indicate up-
probably correspond to normal faults, perhaps ward movement of the hills. In this interpreta-
with a small strike-slip component (Le Pichon et tion, Castor and Pollux are active diapirs which
al. 1981). The walls of these corridors are lift, occasionally on their flanks and perhaps on
affected by the same dissolution as the main their summits, portions of the sedimentary fill
scarps (see Fig. 9). Finally, we noted that the of the trench (which apparently includes a
slope formations, which may be quite recent, tectono-sedimentary m61ange).
are strongly affected by dissolution. This admit- In this respect, we need to discuss briefly the
tedly circumstantial evidence suggests that the tectonics of the outer rise, to the south of
process of erosion by dissolution is geologically Castor and Pollux, investigated in detail during
recent and perhaps still active in deep water. A three different dives and described by Huchon
subaerial erosion could date from the Messinian et al. (1981) and Le Pichon et al. (1981). A
desiccation stage (Hsti & Ryan 1974), but this cross-section extending south of Castor across
old age is quite difficult to reconcile with the the narrow southern corridor and the lower
above evidence. A much younger age would outer rise is shown in Fig. 6. It is representative
imply an implausibly rapid subsidence rate. of an east-west zone at least 1 km wide and
The stratification observed in Figs 13, 14 & indicates active shortening of the upper and
15 strongly resembles the type of thin bedding semi-consolidated sedimentary layer, ol being
observed in evaporitic rocks, either gypsum or close to north-south (Figs 16-19).
halite. The small piece of shallow water dolomi- Ritsema (1974) has given a reverse fault
tic limestone was sampled a few tens of metres solution for a 6.4 magnitude earthquake (De-
above these beds. It thus seems quite reason- cember 17 1952) with an epicentre in this
able to hypothesize that the hard rock forma- general area. This evidence confirms that the
tions observed in the Poseidon area belong, at zone is active and that the activity is compress-
least in part, to the Messinian evaporitic series ional.
and include beds of gypsum or even halite as The amount of shortening measured is about
well as dolomitic limestone, which would ex- 150 m; the actual amount may be somewhat
plain the variable intensity of erosion by dis- larger depending on the amount of overthrust-
solution. ing of the outer rise over the trench floor (see
Fig. 6), but probably does not exceed 500 m.
The tectonic observations are fairly precise
because the deformation is clearly inscribed in
 Subduction in the Hellenic Trench 329

F~6. 16. NNE-facing overthrust of marls and crusts of the outer wall on the recent ooze of the axis of
the trench. South of Castor diapir, site 2,--2985 m.

FIG. 17. Small folds with broken axis in a crust of biozone NN20, near the overthrust of Fig. 16, site
2,--2866 m.
330 Le Pichon e t al.

• ,lr,
' ,,

.. ~:r~, ,:. '

FI(;. 18. Small fold with brokcn axis (dctail), site 2,--2865 m.

FIG. 19. Tectono-sedimentary m61ange exposed on a fault scarp. South of Pollux diapir, site
2,--2683 m.
 Subduction in the H e l l e n i c T r e n c h 331

the indurated crust which covers the semi- gic cover immediately south of the trench. Le
consolidated sediments in this area. This crust is Pichon & Angelier (1979) have estimated the
the result of lithification of the calcareous rate of subduction as about 3-4 cm yr -1 in this
hemipelagic sediment from the top downward area which would represent 150-200 km since
(Allouc, pers. comm. 1980). The age of the the Messinian. The amount of evaporites shown
sediments forming the crust is 200-400 000 yr in Fig. 7 thus corresponds to less than 1 Ma of
(biozone NN20; nannoplankton determined by subduction if none of them are subducted.
C. Miiller). The age of the deformation is However, this reasoning assumes that the thick-
unfortunately unknown but is much more re- ness of evaporites was uniform, which is doubt-
cent than formation of the crust because broken ful. In any case, it is clear that most of the
portions of crust are not covered by the dark sediments, possibly including part of the evapo-
oxidized film. rites, are subducted and that the present situa-
A similar but less intensive shortening was tion is not a steady-state one.
observed on the outer rise 10 km to the north It is likely that a sizeable portion of the
along the Pollux section. Here, in addition, deformed sedimentary cover as well as the
semi-consolidated marls exposed over several diapiric hills will be later incorporated into the
metres on a fault scarp have the structure of a inner wall as suggested in Fig. 7. We are unable
sedimentary m61ange (Fig. 19). We conclude to estimate this amount. The tectonic con-
that the upper sedimentary cover of the outer figuration of the layers underlying the trench
rise is actively shortened and that a tectono- may thus be extremely complex and may have
sedimentary m61ange is probably the result of little resemblance to the schematic representa-
this shortening. The intensity of shortening is tion of Fig. 7.
greater on the Castor section, where the outer Finally, we shall briefly discuss the results
rise impinges on the diapir, than on the Pollux obtained in the Strabo trench. We noted pre-
section, which is separated from the diapir by a viously that the motion there is nearly parallel
portion of trench 1 km wide. Consequently, it is to the trench which, in detail, can be inter-
probable that the upward movement of the preted as a series of Riedel shears within
diapirs blocks the progression of at least the a general left-lateral strike slip zone with a
upper part of the sedimentary layer toward the limited component of subduction (Lyberis et al.
north and is the cause of the localization of the 1981; Huchon et al. 1981). No hard rock out-
shortening in this area. crops have been observed in the immediate
vicinity of the floor of the trench although large
exotic boulders suggest that hard rock outcrops
Tectonic role of the evaporites in the exist further upslope. This has been confirmed
subduction zone by dredging (Ariane 1979).
On the other hand, Biju-Duval et al. (1978)
G l o m a r Challenger drilling in the Mediterra- showed that the Strabo trench is the southern
nean has demonstrated that the Messinian limit to a large and thick evaporitic basin which
Stage is associated with a sedimentary evapori- is limited to the north by the Pliny trench (see
tic complex both in the Western and Eastern Fig. 2). This trapping of an evaporitic basin
basins (Hsfi & Ryan 1974; Hs0 et al. 1978). between two trenches suggests that the evapori-
Seismic reflection investigations confirm that tic basin was built by tectonic thickening in
evaporites, including halite beds, exist in the front of Pliny trench and that Strabo trench is
Poseidon area below a few hundred metres of the result of a recent migration of the subduc-
Plio-Quaternary pelagic sediments (Biju-Duval tion zone. As a result the thickness of evapo-
et al. 1978). It is thus difficult to avoid the rites entering the Strabo trench is now much
conclusion that the two diapirs Castor and less (see Biju-Duval et al. 1978). In addition,
Pollux are the product of tectonic thickening of the actual rate of subduction is low because of
the evaporites below the trench, with local ex- the mostly transform nature of the motion.
trusions. This would explain the asymmetry of Consequently, the tectonic style of the Strabo
the diapirs, with the steep side facing seaward. trench is quite different from that of the
We noted previously that the Seabeam survey Poseidon trench.
suggests that such diapirs are common within
the Hellenic trench (Lyberis et al. 1981).
Conclusion
Fig. 7 shows a hypothetical model of the
Poseidon trench. At the level of Castor, we Detailed Seabeam and submersible studies
show a thickness of 1 km for the Messinian have enabled us not only to decipher the re-
series and 300 m for the Plio-Quaternary pela- gional strain pattern but also to discover and
332 Le Pichon et al.
map extensive hard rock outcrops which had Seismic reflection has shown that only a few
not previously been detected from sea-surface hundred metres of pelagic Plio--Quaternary
studies. sediments cover the Messinian formation sea-
A remarkable character of these outcrops in ward of the trench. Consequently we suggest
the Poseidon area, but not in the investigated that the Messinian evaporites are tectonically
part of the Matapan area, is the presence of thickened below the trench and may occasional-
variable but often very intense erosion by dis- ly extrude through the thin sedimentary cover.
solution. This fact, coupled with the similarity The Hellenic trench may thus be a subduction
in stratification of some outcrops with gypsum zone where the tectonic style is dominated by
or halite beds, leads us to suggest that the the presence of a thick evaporitic layer which is
Messinian evaporitic complex is outcropping in difficult to subduct.
the Poseidon area.

References
ANGELIER J. 1979. Ndotectonique de l'arc dgden, of the Hellenic trench system. Earth planet, sci.
Thesis, Soci6t6 geologique du Nord, Publication Lett. 44, 441-50.
no 3,417 pp. ,--, BOULIN, J., BUREAU, D., CADET, J. P.,
ARIANE 1979. R6sultats de dragages sur la bordure CttAPEI,, A., DERCOURT, J., GI,A~ON, G., GOT,
externe de Farc hell6nique (M6diterran6e orien- H., KARIG, D., LYBERIS, N., MASCLE, J., RICOU,
tale). Mar Geol. 32, 291-310. L. E. & THII~aAULT, F. 198(I. Importance des
AUBOUIN, J. & DERCOURT,J. 1965. Sur la g6ologie de formations attribu6es au Messinien dans les fos-
I'Eg6e: regard sur la Cr~te (Grace). Bull. Soc. s6s de subduction hell6niques: observations par
geol. Fr. VII, 78%821. submersible. C. r. Acad. Sci. Paris, 290D, 5.
BtJu-DuvAl., B., LETOUZEY, J. & MONTADERT, L. , , --, DERCOURT,J., GLA~ON,
1978. Structure and evolution of the Mediterra- t3., GOT, HI, KARl(;, D., LYBERIS, N., MASCEE,
nean basins. In: Hs0, K. J. et al. (eds). Initial J., RICOU, L. E. & THI~BAUI:r, F. 1979b. Tecto-
Rep. Deep Sea drill. Proj. 42, 951-84. U.S. Govt nique active dans le loss6 de subduction hell6ni-
Printing Office, Washington. que: observations par submersible. C. r. Acad.
GOT, H., STANLEY, D. J. & SOREL, D. 1977. North- Sci. Paris, 289D, 1225.
western Hellenic arc: concurrent sedimentation --, HUCttON, P., ANGELIER, J., LYBERIS, N.,
and deformation in a compressive setting. Mar. BOULIN, J., BUREAU, D., CADEr, J. P., DER-
Geol. 24, 21-36. COUR't, J., GLA(~ON, G., GOT, H., KARIG, D.,
Hs0, K. J., MONTADERT, L., BERNOUILLI, D., CnA, MASCLE, J., RICOU, L. E. & THII~BAULT,F. 1981.
M. B., ERICKSON, A., GARRISON, R. C., KIDD, Active tectonics in the Hellenic trench. Oceanol.
R. B., MELIERES, F., MULLER, C. & WRIGHT, R. Acta, SP4, 273-81.
1978. History of the Mediterranean salinity cri- LE PICHON, X. & HSU, K. J. 1979. HEAT, Episodes,
sis. In." Hs0, K. J., MONTADERT, L. et al. (eds). !, 3--6.
Initial Rep. Deep Sea drill. Proj. 42, 1{153-78. LE QUELLEC, P., MASCLE, J., GOT, H. & VITTORI, J.
U.S. Govt Printing Office, Washington. 1980. Seismic structure of southwestern Pelopon-
& Ryan, W. B. F. 1974. Deep sea drilling in the nesus continental margin. Bull. Am. Assoc. Pet-
Hellenic trench. Geol. Soc. Greece, 10, 81-9. rol. Geol., 64, 242-63.
HUCHON, P., ANGELIER, J., LE PICHON, X., LYBERIS, LVBf~RiS, N., ANGELIER, J., LE PICHON, X. & RENARD,
N. & RlCOU, L. E. 1981, Les structures tectoni- V. 1981. Interpr6tation d'un foss6 de subduction
ques observdes en plong6e dans les losses hell6- partir des lev6s bathymdtriques au sondeur
niqucs C. r. somrn. Soc. geol. France, in multifaisceaux 'Seabeam': I'exemple du loss6
press. hell6nique. C. r. somm. Soc. geol. France,
LE PICHON, X. & ANGELIER, J. 1979. The Hellenic arc MCKENZlE, D. P. 1978. Active tectonics of the
and trench system: a key to the neotectonic Alpide-Himalayan belt: the Aegean Sea and
evolution of the eastern Mediterranean area. surrounding regions (tectonics of Aegean re-
Tectonophysics, 60, 1-42. gion). Geophys. J. R. astron. Soc. 5 5 , 21%54.
-- & -- 1981. The Aegean Sea, Proc. R. Soc. Rn'SEMA, A. R. 1974. The earthquake mechanisms of
London, in press. the Balkan region. R. Netherl. Meteorol. Inst.,
• AUBOUIN, J., LYBER1S, N., MONTL S., De Bilt, Sci. Rep. 74--4, 36 pp.
RENARD, V., GOT, H., Hs0, K., MART, Y., ZIBROWtUS, H. 1980. Thanatocoenose pleistocene
MASCI.I.:, J., MAVrHEWS, D., MITROPOULOS, D., profonde ~ spongiaires et Scl6ractiniaires dans la
"FsovI.IAS, P. & CnRo.~lS, G. 1979a. From sub- fosse hellOnique. X X V I I e Congr~s CIESM, Cag-
duction to transform motion: a Seabeam survey liari, October 9-18.
 Subduction in the Hellenic Trench 333

X. LE PICHON, P. HUCHON, J, ANGELIER, N. LYB~.RIS, J. BOULIN & D. BUREAU,

D6partement de G6otectonique, Universit6 Pierre et Marie Curie, 75230 Paris
Cedex 05, France.
J. P. CADET, D6partement des Sciences de la Terre, Universit6 d'Orleans, 45045
Orl6ans, France.
J. DERCOURT,Laboratoire de Stratigraphie, Universit6 Pierre et Marie Curie, 75230
Paris Cedex 05, France.
G. GLA~ON, D6partement de G6otectonique, Universit6 Pierre et Marie Curie,
75230 Paris Cedex 05, France.
H. GOT, Centre de Recherches en S6dimentologie, Universit6 de Perpignan, 66000
Perpignan, France.
D. KARIG, Department of Geological Sciences, Corneli University, New York
14853, U.S.A.
J. MASCLE,Laboratoire de G6ologie dynamique, Universit6 et Marie Curie, 75230
Paris Cedex 05, France.
L. E. RIcoc, Laboratoire de G6ologie historique, Universit6 Paris Sud, 91000
Orsay, France.
F. THX~BAULT, Laboratoire de G6ologie structurale, Universit6 de Lille, 59000
Lille, France.
Detailed tectonic trends on the central part of the Hellenic Outer
Ridge and in the Hellenic Trench System

N. H. Kenyon, R. H. Belderson & A. H. Stride

SUMMARY: Extensive new sonograph coverage in the Eastern Mediterranean allows a

re-assessment of the complex relief outside the Hellenic Arc. The trenches to the SE and
SW of Crete are similar in their plan view appearance, each having predominantly sinuous
or curved structures that are interpreted as the surface expression of major thrusts.
Thrusting, possibly imbricate, is also suggested on the central, shallowest part of the
Hellenic Outer Ridge where interlocking convex-outward, curved structures occur, along
with other faults and folds. These are thought to relate to d6collement-type displacement
within or beneath evaporite sediments. The structural style deduced from the morphology
is consistent with a continuing outward radial push of the Hellenic Arc to the SW, S and
SE, rather than with a migration to the SW and strike-slip on the SE side of the arc as
proposed by other workers.

The region considered here includes the central RRS Discovery surveys of 1977 and
portion of the Hellenic Trench System and 1979
Hellenic Outer Ridge in the vicinity of Crete
(Figs 1-3). The nature and significance of these The two cruises of RRS D i s c o v e r y in 1977 and
two major relief features are still incompletely 1979 doubled the area of ground previously
understood. The Outer Ridge may have been examined in the Eastern Mediterranean by
initiated during the late Miocene as it seems means of Gloria, the long-range side-scan sonar
that the Messinian evaporates thin onto the (Fig. 1). The new coverage included large parts
'Upper Plateau' region (Biju-Duval et al. 1978). of the Hellenic Trench System as well as an
Because the Outer Ridge evolved mainly dur- extensive survey of the central portion of the
ing the Plio--Quaternary it is probably related Hellenic Outer Ridge. Both cruises made use of
largely to the latest phase in the outward the improved two-way looking Mark 2 Gloria
growth of the tectonically more disturbed Hel- 'fish', which in this instance viewed a 30 km
lenic Trench System. wide swathe of seafloor. Single-channel sub-
The present consensus is that the relief of the bottom profiling was also used, and narrow-
Outer Ridge is largely tectonic in origin, beam echo-sounding provided more realistic
although there is disagreement about the profiles of the surface of the floor than can be
mechanisms involved in its development. Major obtained from conventional echo-sounding.
thrusting has been proposed within the Hellenic In analysing the sonographs it must be borne
Outer Ridge (Finetti 1976; Biju-Duval et al. in mind that linear features oriented parallel to
1978) while Kenyon & Belderson (1977) prop- the ship's track show up most clearly and small
osed a regional d6collement mechanism to ex- linear features at right angles to the track may
plain the folding and faulting along the whole not show up at all. Thus to be sure of detecting
length of the Outer Ridge. In addition, there all structural trends it is necessary to examine
are areas of postulated submarine salt karst the whole region on two sets of courses at an
within the tectonised ground (Belderson et al. angle to one another. Inspection of Fig. 1 will
1978). show that, although the majority of courses are
The new data consist of greatly increased roughly parallel to the Hellenic Arc, crossing
sonograph coverage over both the Outer Ridge tracks are spread throughout the region and so
and the Hellenic Trench System, together with most morphological trends should have been
additional narrow-beam echo-sounder profiles recognised.
and sub-bottom profiler data. These are used to
re-assess our earlier conclusions (Stride et al.
1977) as well as to make a broad comparison Salt karst
with the data from the multi-beam echo-
sounder 'Seabeam' (Le Pichon et al. 1979). We The existence of salt karst of submarine origin
contest the proposal by Le Pichon et al. that on the Hellenic Outer Ridge was postulated by
compression has been dominant to the SW of Belderson et al. (1978) to account for the
Crete while strike-slip motion has been domi- presence of sub-circular to irregular patches of
nant to the SE of Crete. rough ground that have a conspicuous appear-

335
336 N. H. Kenyon, R. H. Belderson & A. H. Stride
1

t~
GREECE
i

L- 5A
"2"

° ~

°-~

f--

FtG. I. Gloria long-range side scan sonar coverage in the Eastern Mediterranean up to 1979, the
approximate limits of the Scabcam surveys (boxes) and the location of Scripps Institution of
Oceanography dccp-tow sites (stars). The insct figure shows the location and direction of view of the
sonar beam for sonographs in Figs 4 & 5.

ance on 'fixed gain' s o n o g r a p h s . T h e m o r e c o n f i n e d to the i n n e r portion of the Hellenic

recent side-scan s o n a r c o v e r a g e has r e v e a l e d a O u t e r Ridge (Fig. 2). West of C r e t e two groups
few m o r e of the distinctive sub-circular patches of these sub-circular salt karst bodies are con-
of r o u g h g r o u n d , i n t e r p r e t e d as dissolution n e c t e d by p r o m i n e n t linear faults, or salt
surfaces a b o v e diapirs. Such r o u g h floors are 'dykes" ( B e l d e r s o n et al. 1978). A detailed

===========================
:':':':':':':':':':':':':':':', ~ l~ .:, I ~' ~ ~
~I ~~ ~.~.~i

:i.................... ::~i:::.
_3~o:~i~i!~i~i~ii~i::~!i~!i!::~:.. ~ ~. ".~

FiG. 2. Sub-circular to irregular shaped areas of rough floor (mostly pits) attributed to salt karst are
confined to a belt on the internal portion of the Hellenic Outer Ridge, north of the limit shown.
1
(Hellenic Outer Ridge shaded).
 \\ I,, # ~(~$ 4",.., .---I>-
r- .< y~'
F" ~o < ~ ~E4N ~,r.. /'7
r" "l,,,, v"
1~KARPATHOS
%
".bk'~ _ { /--'~~ A Rz.C ,..

• .~,...~= _Z-..~..h. -; /--~ ~i~\

~ ~\¢,t',' '~ .~-, ~ . . ~ ~

, ~ . ~ ~ , ,_~.

,, ),)2' " ---- --

X:
-34 ° i ,~t , ~--~_-~-~"

50i<m

.__..
20 ° 22" 24 ° ", 26 °
 Tectonic trends in the H e l l e n i c T r e n c h S y s t e m 337

echo-sounding survey was subsequently made longer straight ones (strike-slip faults). In many
of one of the sub-circular depressions and the cases interpretation of the features as folds,
associated lineament, in joint work with Scripps thrusts, strike-slip faults and normal faults is
Institution of Oceanography using their deep- inevitably rather subjective without making
tow instrumentation system. This confirmed the further more detailed observations of individual
sub-circular and rough floored nature of the features.
depression and indicated that the surface ex-
The Hellenic Trench system
pression of the lineament is a steep-walled
trough. It thus seems likely that a salt 'dyke' is The general form of the trenches outside the
associated with this trough. In contrast the Hellenic Arc has been evident on bathymetric
sub-circular bodies of salt karst in the south- maps for many years. The trenches are consi-
eastern part of Fig. 3 tend to be located within dered by most workers to indicate the lines of
relatively flat areas between bundles of what outcrop of major faults, probably thrusts. The
are probably folds (Belderson et al. 1978, fig. western trenches have, in general, a thicker
4). The presence of salt karst north of the line sediment fill than the eastern ones (Pliny and
on Fig. 2 may have resulted from the mobiliza- Strabo Trenches), so that details of their basal
tion of salt due to the higher degree of deforma- relief are somewhat obscured. Local surveys by
tion that has affected that ground. However, means of Seabeam (Fig. 1) have recently pro-
any enhanced deformation resulting from the vided detailed bathymetry of parts of the tren-
presence of salt cannot account for the struc- ches (Le Pichon et al. 1979). However, the
tures on the Upper Plateau, as the salt is simplified interpretations of Le Pichon et al.
thinnest beneath that ground (Biju-Duval et al. (1979) based on these contour maps for the
1978, fig. 8). Elsewhere south of the line the trenches SW and SE of Crete tend to impart a
Plio-Quaternary cover thickens towards both falsely angular appearance to the features that
the Messina and Herodotus abyssal plains. is not present on the data. This is due to the use
In the Hellenic Trench System there are of straight lines to represent what are, in real-
much larger, irregular-shaped areas which show ity, often curved features. In addition, some
up strongly on fixed gain sonographs (Fig. 3). relief on the contoured maps is ignored.
Many of these areas are expected to be rough- A detailed, yet much more regional, view of
surfaced outcrops of Messinian evaporites that the trend of these trenches (Fig. 3) is provided
have been exposed by faulting or slumping of by the sonographs, samples of which are shown
overlying Plio-Quaternary sediments, but other in Fig. 4. The most notable aspect of the shape
areas may only indicate steep slopes. Indeed, of the trenches is that they have a marked
an outcrop of evaporites has been observed tendency to be sinuous in plan view, rather than
recently from a submersible operating in Sea- angular. The supposed dominance of minor
beam survey area 2 (Fig. 1) of the Hellenic offsets in the trench axes that have been pro-
Trench System (Le Pichon et al. 1981). posed by Le Pichon et al. (1979), on a basis of
simplifications of the relief shown by Seabeam
surveys, are mostly unconfirmed by the sono-
Structural trends graphs. Just south of the outermost trench (the
The total amount of morphological information Strabo Trench) there is an interconnected
about structural trends provided by sonographs group of curved structures, with the convex
is now so great (Fig. 3) that it compares in detail sides facing mainly outwards (Fig. 4C). These
with maps of lineations produced from aerial features are interpreted as newer thrusts that
photography on the adjacent land. For exam- are developing external to the main Strabo
ple, Foose (1977) has mapped structural linea- Trench.
ments from Landsat images of the Tyrrhenian In general, the sonographs indicate a similar-
and Aegean Basins and distinguished between ity in the morphology of the narrow, sediment-
short, straight lineaments (normal faults), free trenches to SW and SE of Crete (Fig. 4A,
longer wavy-shaped ones (thrusts) and much B). This argues against the basic difference

FIG. 3. The main structural trends in the central portion of the Hellenic Trench System and Hellenic
Outer Ridge, derived only from sonographs. Faults and folds are not differentiated. Black patches
are areas of postulated salt karst. Some of the patches shown as salt karst on the walls of the trenches
may, in reality, merely indicate steep slopes. The generalized inner and outer limits of the Hellenic
Outer Ridge are shown by thick dashed lines. The inner dashed line also corresponds with the
Hellenic Trench System. The inset figure gives the locations of names referred to in the text.
 Profile

A
<3

lOkm A

Profile

A
 Tectonic trends in the H e l l e n i c Trench S y s t e m 339

between the two areas postulated by Le Pichon verse to the Outer Ridge. By chance, the only
& Angelier (1979), who proposed a thrust Seabeam coverage of the Outer Ridge reported
motion for those located SW and west of Crete on by Le Pichon et al. (1979, fig. 3) included a
and a predominant strike-slip motion along the part of this area and gave hints of its anomalous
Pliny and Strabo Trenches located SE of Crete. trends.
The high ground behind the trenches is Tectonic cross-trends are particularly well-
blocky, with the main structural trends made developed between Crete and Cyrenaica, with a
evident on bathymetric charts. The sonographs tendency to decrease in importance outwards.
show, in addition, many relatively short, The arrangement of these may be interpreted as
straight scarp-like features, both to the east and implying a conjugate shear system developed in
west of Crete, but trends are neither so consis- relation to an outwards radial push (Fig. 6).
tent nor so numerous as the structural trends on Between Crete and Cyrenaica there are adja-
the Hellenic O u t e r Ridge. They are considered cent areas of flat and intensely deformed
to be normal faults, which is consistent with ground. The large flat northern part of the
earlier interpretations of normal faulting both Hellenic Outer Ridge, to the SW of Crete
here and on Crete (e.g. Angelier 1978, fig. 5). ('Lower Plateau') is intersected in its western
In one area to the west of Crete there are half by a substantial ridge of deformed ground
several more sharply defined straight faults with (Stride et al. 1977, fig. 4) perhaps underlain by a
only minor throw; these are presumably the thrust. South-east and SW of the Lower Plateau
youngest ones recognizable at the scale of the there are some much narrower bands of re-
sonographs. latively flat floor that alternate with probable
fold bundles, possibly underlain by thrusts.
Hellenic Outer Ridge South of the Lower Plateau on the shoalest part
of the Outer Ridge ( ' U p p e r Plateau') there are
The new data re-emphasize the presence of a numerous tectonic trends with relief of up to
dominant lengthwise tectonic trend that is 130 m. These are frequently curved and more
traceable along virtually the whole of the Helle- widely spaced than elsewhere on the O u t e r
nic Outer Ridge (Figs 3 & 6), including some Ridge (Fig. 5a); they are interpreted as prob-
areas where there was previously little or no able outcrops of thrusts.
sonograph cover. The features comprising this
trend are known to consist of folds along the Morphological indications of
southern edge of the Outer Ridge, and else- compression between the Hellenic Arc
where are interpreted as folds, largely faulted
and the Hellenic Outer Ridge
and slumped along their strike (Stride et al.
1977). The localized kink in the lengthwise It is generally agreed from geological studies on
tectonic trend on the south edge of the Hellenic land that there has been outward migration of
Outer Ridge (Fig. 5B) is attributed to com- the Hellenic Arc since the Mesozoic. However,
pression of the Outer Ridge against a projecting a more recent assessment, while supporting
spur at the base of the African continental thrusting on the SW side of the arc suggests by
margin (as was previously suggested for similar first-motion studies of earthquakes (during the
relief located further west by Stride et al. 1977, short period of available observations) that
fig. 14). This is analogous, but on a much there is N E to SW strike-slip motion on the
smaller scale to the way the Outer Ridge wraps eastern side of that arc (Le Pichon & Angelier
around both sides of the foot of the Cyrenaican 1979). This proposed contrast between the east-
continental slope (Fig. 3). One notably anoma- ern and western sides of the arc needs to be
lous area lies just south of the Hellenic Trench examined in relation to the the structural trends
System to the west of Crete. Here a bundle of in the Hellenic Trench System and the adjacent
probable folds bends round to become trans- Outer Ridge.

FIG. 4. Sonographs (located on Fig. 1) showing the morphological similarity of the trenches (A) to
the SW of Crete (where there is supposed thrusting), and (B) to the SE of Crete (where there is
supposed predominant strike-slip motion according to Le Pichon & Angelier 1979). The narrow cleft
which defines each trench axis lies in shadow cast by the outer wall. Echo-sounder profiles show that
the outer wall is about 550 m high near the west end of (A), 300 m high near the centre of (B), and
330 m high near the east end of (B). Sonograph (C) shows overlapping curved and mainly convex
outward-facing structures interpreted as the early stages in the development of a new trench system
just south of the main Strabo Trench. Note that sonograph (C) is continuous with sonograph (B).
White arrows show the supposed motion. The black arrowheads point in the direction of view of the
sonar beam as indicated on Fig. 1.
340 N. H. K e n y o n , R. H. Belderson & A . H. Stride
The trench system outside Crete is notable ment. A seismic refraction line shot from Crete
for approximately NW-trending major struc- to Egypt across the Outer Ridge to the east of
tural lines in the west and approximately NE- the 'Upper Plateau' shows the crust beneath the
trending ones in the east. The partial sediment ridge to be 27 km thick, of which the sedi-
fill in the western examples obscures their mentary cover accounts for about 8 km, with
deeper relief, so that they contrast with the some evident uparching of the basement (J.
narrow cleft-like appearance of the eastern Makris, pers. comm.).
examples, where the deepest portions are still Thrusting involving even supposed Mesozoic
largely sediment-free. However, the almost sediments within parts of the Outer Ridge has
sediment-free trenches more immediately to already been interpreted on sub-bottom pro-
the SE and SW of Crete are remarkably similar files (Finetti 1976; Biju-Duval et al. 1978),
in appearance, as shown on Fig. 4. It is likely although such data are not entirely convincing
that there is progressive outward younging of on their own. We propose shallow thrusts which
trenches, from the Pliny Trench to the narrower we think may be related to a regional d6colle-
and less deep Strabo Trench and then to the ment or a series of d6collement horizons within
interconnected group of proposed thrusts im- or at the base of the Messinian evaporites
mediately to the south of the Strabo Trench, (Kenyon & Belderson 1977). However, we do
where the newest trench of all seems to be not exclude some deeper, higher angle thrusts
developing. At the same time the block of below the Upper Plateau. Messinian salt will
ground between the Pliny and Strabo trenches have contributed to the complexity of the de-
is probably being uplifted. formation.
On the Hellenic Outer Ridge, the two sets of In conclusion, the broad regional coverage
cross-trends which cut the Outer Ridge-parallel with sonographs does not lend support to the
folds (Fig. 6) are interpreted as strike-slip faults view that there is significant contrast between
related to a conjugate shear system which re- the shape of the bottoms of the trenches to the
sults from regional compression between the SE and SW of Crete, as has been suggested by
Hellenic Arc and the foot of the African con- Le Pichon & Angelier (1979) and Le Pichon et
tinental slope. The fact that such a pattern is al. (1979), but rather points to reasonable simi-
found both SW and SE of Crete implies that, larity in their morphology. The conflict be-
even if some strike-slip motion is now occurring tween the structural data presented and inter-
in the trenches SE of Crete, a substantial preted here and the first-motion studies (im-
amount of compression is also being transmit- plying transform motion to the SE of Crete)
ted across the trenches onto the Outer Ridge. might perhaps be resolved if the proposed
In any case the Outer Ridge and folds parallel transform motion is younger than the trench
to it are as well-developed SE of Crete as they and Outer Ridge relief. Certainly the first-
are to the SW of Crete and belie any long-term motion studies of this region are only available
difference in origin between the two sides. for some earthquakes of the past few years, and
Thrusts or incipient thrusts are considered to there is no reason to suppose that they should
lie beneath the bundles of folds separating be relevant for the very much longer period of
smooth ground in the south-eastern part of Fig. time needed to build present relief. Moreover,
3, while other well-developed thrusts (Fig. 6) Le Pichon & Angelier (1979, fig. 2 and table II)
may lie beneath the ridge that intersects the did not make use of three horizontal projec-
Lower Plateau, as well as beneath the southern tions of slip vectors which indicated thrusting
edge of the Hellenic Outer Ridge and beneath a normal to the trench system east of Crete
step located above it to the south of the Upper (south of Karpathos), which diverged by about
Plateau (Stride et al. 1977, figs 6 & 12). In 90 ° from others that were used preferentially by
addition, outcrops of thrusts, possibly in places them in the vicinity. Clearly it wilt be necessary
imbricate and overlapping, are suggested by the to obtain a much longer time series of first-
numerous curved, interlocking structural trends motions for the region before there can be any
on the 'Upper Plateau'. For example, the two certainty about whether the proposed trans-
thrusts in the western half of Fig. 5(B) corres- form motion east of Crete is typical of present
pond to the two that outcrop in profile in fig. 15 deformation in the region.
of Finetti (1976). Further east they merge into a
single thrust. The net effect of such thrusts
ACKNOWLEDGMENTS:Mr Kenyon thanks Professor F.
should be to cause some thickening of the N. Spiess and Dr W. B. F. Ryan for the opportunity
sediments in this, the shoalest, part of the to sail on R.V. Melville so as to explore in detaiL, with
Hellenic Outer Ridge. Alternatively, its eleva- the deep-tow system, some of the postulated salt
tion might be due to arching up of the base- karst revealed by Gloria sonographs.
 A
%

~,~°

Profile

lOkm

Profile
A spur

F~6.5. (A) Sonograph (located in Fig. 1) showing curved tectonic trends, with relief of up to about 130 m, on the shoalest part of the Hellenic Outer Ridge.
They are interpreted as the surface of two overlapping thrust sheets. White arrows show the supposed motion. These thrusts may be seen in depth on a
sub-bottom profile of Finetti (1976, fig. 15). The black arrowhead points in the direction of view of the sonar beam as indicated on Fig. 1. (B) Sonograph on
the south side of the Hellenic Outer Ridge (located in Fig. 1) showing ridge-parallel tectonic trends, with a general relief of about 40 m, wrapped around a
spur at the base of the African continental slope. White arrows show the supposed movement direction of two thrust sheets with curved fronts. The black
arrowhead points in the direction of view of the sonar beam as indicated on Fig. 1.
4~
 I'O

\\ . . . - - ' " ... ~. -~

V
"-0
% ~
_X "~,e,ra.sver~e ,~u, ,re~-

e~

. . . . . . .

L26: 22 ° - ~ 2 ~ ~ 26 °

FIG. 6. Simplified structural map showing the curved outcrops of some of the most probable thrusts observed on sonographs in the Hellenic
Trench System and on the Hellenic Outer Ridge, together with a summary of the Outer Ridge-parallel fold and fault trends and the conjugate
system of strike-slip fault trends. The principal stress trajectories will be normal to the ridge-parallel fold trends and bisect the angle between
the conjugate system of supposed strike-slip faults and lie parallel with the axes of bends in fold bundle trends. This is consistent with a scheme
of radial stress. The possible thrust bounding the southern margin of the Hellenic Outer Ridge is thought to represent the front of a
d~collement within the Outer Ridge. The continuity of this feature is based on additional bathymetric data.
 Tectonic trends in the Hellenic Trench System 343

References
ANGELIER, J. 1978. Tectonic evolution of the Hellenic nic and Cyprus Outer Ridges. In: BIJU-DUvAL,
Arc since the late Miocene. Tectonophysics, 49, B. & MONTADERT, L. (eds). Structural History of
23-36. the Mediterranean Basins, 233-40. Technip,
BELDERSON, R. H., KENYON, N. H. & STRIDE, A. H. Paris.
1978. Local submarine salt-karst formation on LE P~CHON,X. & ANGELIER, J. 1979. The Hellenic arc
the Hellenic Outer Ridge, eastern Mediterra- and trench system: a key to the neotectonic
nean. Geology, 6, 716-20. evolution of the Eastern Mediterranean. Tecto-
BIJU-DUVAL, B., LETOUZEY, J. & MONTADERT, L. nophysics, 60, 1-42.
1978. Variety of margins and deep basins in the - , AUBOUIN, J., LYBERIS,N., MONTI, S., RENARD,
-

Mediterranean. In: WATKINS, J. S. et al. (eds). V., GOT, H., Hs0, K., MART, Y., MASCLE, J.,
Geological and Geophysical Investigations of MATrHEWS, D., MITROPOULOS, D., TSOFLIAS, P.
Continental Margins. Mem. Am. Assoc. Petrol. t~ CHRONIS,G. 1979. From subduction to trans-
Geol. 29, 293-317. form motion: a Seabeam survey of the Hellenic
FINETrI, I. 1976. Mediterranean Ridge; a young sub- Trench System. Earth planet. Sci. Lett. 44, 441-
merged chain associated with the Hellenic Arc. 50.
Boll. Geofis. teor. appl. 19, 31-65. - et al. 1981. Subduction in the Hellenic trench:
-

FOOSE, R. M. 1977. Structural lineaments and tecto- probable role of a thick evaporite layer based on
nics of the Mediterranean Basin. In: BIJU- Seabeam and submersible studies (this volume).
DUVAL, B. & MONTADERT, L. (eds). Structural STRIDE,A. H., BELDERSON,R. H. & KENYON, N. H.
History of the Mediterranean Basins, 221-7. 1977. Evolving miogeanticlines of the East
Technip, Paris. Mediterranean (Hellenic, Calabrian & Cyprus
KENYON, N. H. & BELDERSON, R. H. 1977. Young Outer Ridges). Philos. Trans. R. Soc. London,
compressional structures of the Calabrian, Helle- 284A, 255-85.

N. H. KENYON,R. H. BELDERSON • A. H. STRIDE,Institute of Oceanographic

Sciences, Wormley, Godalming, Surrey, England.
 The structure of the Calabro-Sicilian Arc: result of a
post-orogenic intra-plate deformation

F. C. Wezel
SUMMARY: Morphotectonic and geological data show that this arc constitutes a
post-orogenic ring-like system concentric round the Marsili 'backarc' basin. Peripheral
concentric features comprise the 'Aeolian andesitic ring' and the arcuate troughs and
ridges characteristic of the 'forearc' region. The Hyblean horst-like plateau and the Strait
of Sicily represent the block-faulted foreland of the Sicilian segment of the Calabro-
Sicilian Arc. The present shape of the arc has largely resulted from Plio-Quaternary
vertical tectonic movements, superimposed on compressive structures caused by previous
phases of tectonic deformation. The effect can be visualized as an outward-moving
topographic wave producing inversion of tilting movements. This kinematic process
(termed 'krikogenesis') is postulated as being related to midplate domal upwarp, dilatation
and cauldron-like collapse of the Marsili block in the core of the arcuate system.
Volcanicity appears to be associated with various stages of this geological history. Rather
than originating by collision of two rigid plates, present-day morphological, volcanic and
seismic configurations seem to be due to interaction of intra-plate 'arches', activated by
local mantle movements.

Along with the Hellenic and Cyprus arc sys- 1979 and revised colour version 1980). The
tems, the Calabro-Sicilian Arc is considered by structures constitute a nearly complete circle or
many workers (e.g. Biju-Duval et al. 1978) to ellipse, but more commonly encompass a semi-
be a Mediterranean counterpart of Pacific is- circle or less (Fig. 1). These peculiar and pre-
land arcs. Subduction of Ionian Sea lithosphere viously undescribed arcuate structures surround
should be responsible for the calc-alkaline vol- composite crustal blocks that, according to their
canicity of the Aeolian Islands and for behind- diameter, I will name: local-blocks (up to 10' of
arc spreading which created the Tyrrhenian latitude or 18.5 km), minor-blocks (10'-1 ° or
Basin (backarc basin). However, a possible 18.5-111.1 kin), meso-blocks ( 1 ° 4 ° or 111.1-
difference with western Pacific systems arises 444.5 km) or 'small ocean basin'-blocks (40-7 °
from the fact that Mediterranean arcs must be or 444.5-777.8 km). Elsewhere, outside the
interpreted as the result of motion of small Mediterranean region, blocks with a diameter
plates, rather than of a general descent of greater than 7 ° (geo-blocks) are of the same size
African lithosphere beneath Europe (McKen- as certain 'lithospheric plates' (e.g. the Tibet
zie 1972; Alvarez et al. 1974). Plateau 20 degree-block can be compared in
The purpose of this paper is to show that the size with the Cocos plate).
recent tectonics and volcanism of the Calabro- Larger Tyrrhenian circular structures are of
Sicilian region more likely resulted from intra- fairly uniform size, having a roughly constant
plate deformation (here termed 'krikogenesis') diameter of between 1° to 2 °, averaging 150 km
instead of from the interaction and horizontal (Fig. 1). These nearly equidimensional meso-
convergence of two lithospheric plates. This blocks are subdivided morphologically into
different view arises from detailed land and minor component units.
marine studies, mainly of the Sicilian sector,
and from morphostructural analysis of the Tyr-
rhenian Sea floor. The 'backarc' region of the
Calabro-Sicilian Arc (Marsili Basin)
Morphotectonic configuration of the The arc constitutes a multi-ringed system con-
Tyrrhenian Sea floor sisting of a 'backarc' region, concentrically sur-
rounded by an inner ring of volcanic islands and
Structural and morphological observations led seamounts, and by an outer series of post-
to the identification, in different physiographic orogenic 'forearc' basins and ridges.
zones and depths, of several arcuate bathy- The core of the Calabro-Sicilian circular
metric structures on the Tyrrhenian Sea floor structure, the most impressive of the Tyrrhe-
(see morphological map of Savelli & Wezel nian rings, is represented by the Marsili Basin,

345
346 F. C. W e z e l
lO" 11-

.-. 2"O~, MAIN S T R U C T U R A L

LINEAMENTS
OF THE
Roma
o
TYRRHENIAN
BASIN

I '; "''
vi( L,.'.........O / ~

~[/ ';' . k~7"- ""

, ;........ 1~e../..'
.' " ' , .
i l~\t
.:' / i / , ': , , ~.s-i~.,I. ,. , . ~ :
- ".
,, ~,-"~:.,..;,~,: . ,,.

"~.~

f:i /

•' "~..-'-'~_.-:-L~,'/~.: ' ~",'~ ~ ~t :,.:.::.~-

12" 14" :~" le

LEGENO

A,._...=~--~ Normo[orrronscurrentfouhs{A=Moiorltructurolllnwametnts)
Ooshed where in~erred

A ,I,-L4--t_L~ Normal or tronscurrent (oufls proved by ~ o p ~ ond declivities

A~ Proboble I;strlc normo] |ouhs

Volconoes

5g- I1 M= K- Ar dating, Sources 111 MlrlniJl, 1975

iZl s , , ~1.~ sTz
Moin orcuate feotures 13] DSOP gl;~r, W~J_I1978

Position of DSDPdrilllngs

FIG. 1. Major tectonic lineaments in the Tyrrhenian Sea floor deduced from a morphological
analysis. Note that the constituent circular blocks ('krikogens') show a peripheral suture rim with a
width commensurate with size of the block. Circular fragmentation affected both submarine and
continental regions. C.G. = Campidano Graben.

1° in diameter. This nearly circular depression, Dredge samples include tholeiitic, alkaline
with a maximum depth of 3400 m, has in the (Del Monte 1972) and calc-alkaline basalts
centre the huge Marsili volcanic seamount, (Maccarrone 1970; Selli et al. 1977), perhaps
consisting of a ridge of extrusive rock, situated indicating different stages of igneous activity
on a major fault trending N17°E. The seamount (P. L. Rossi, pers. comm. 1980). According to
has 2900 m of relief, with its base lying at a Selli et al. (1977) the volcano displays in its
depth of about 4000 m according to seismic uppermost portion a calc-alkaline affinity
data (Selli et al. 1977). above a base with tholeiitic affinity. Radio-
 T h e s t r u c t u r e o f the C a l a b r o - S i c i l i a n A r c 347

metric data suggest a 0.1-0.2 Ma (late Pleis- horst-and-graben structures which probably
tocene) age for the more recent eruptives of control the eruptive activity (Iacobucci et al.
Marsili seamount, while the onset of volcanism 1977). Present-day tensional faulting is also
possibly took place in Pliocene times between 4 shown by numerous faults cutting the metamor-
and 2 Ma ago (Selli et al. 1977). phic basement and volcanic successions, and by
The Marsili Basin is characterized by signifi- geodetic measurements (Nappiet al. 1977). The
cant crustal attenuation (crust 16-18 km thick: 20 km thick intermediate continental crust in
Morelli et al. 1975), by extensional tectonics, the western (older and inactive) sector of the
and by the presence of an extremely strong arc includes several low-velocity layers, while in
local thermal field (maximum north-south dia- the eastern (active) sector the crust contains
meter of about 40') in the southern sector of only a single low-velocity layer at a depth of
Marsili volcano; here in the region of 39°N heat about 10 km (Morelli et al. 1975).
flow values are up to 11 H F U (Della Vedova & The annular volcanic ridge is limited on the
Pellis 1979). These geophysical data suggest the concave side by inward-sloping normal faults,
existence at a shallow crustal depth (about sometimes of listric type (see Fig. 2 and Savelli
18 kin) of hot mantle material (magma cham- & Wezel 1979), indicating tilting of stepped
ber?), especially beneath the seismically active blocks towards the 'forearc' region. The circu-
eastern half of the basin. Converging on this lar shape of the wall ridge and the evidence
area of abnormally elevated heat flow are from some seismic reflection profiles (e.g. pro-
several major transverse (radial) lineaments file MS 3 in Bacini Sedimentari 1980) for a
such as the Sangineto, Catanzaro and the Giar- vertical boundary fault suggest a spectacular
dini-Lipari lines (Fig. 2). In contrast, the west- huge cauldron-like collapse of the Marsili re-
ern aseismic half of the Basin has a Bouguer gion along ring fractures (Fig. 3). This may
gravity-high anomaly of +250 mGal, centred perhaps be connected with subsidence of mag-
over the deepest basinal area ('Marsili mas- ma in a magma chamber underlying the region.
con'). In summary, Aeolian andesitic volcanicity
mainly occurred in a region that seems to have
been subjected to extensional stress. Eruptive
The Aeolian calc-alkaline ring centres are chiefly located on the intersections
of crustal fractures. Volcanic features have
Alvarez et al. (1974, p. 312) were apparently resulted from long-active phases of magma
the first to draw attention to ' . . . the remark- ascent with formation of volcanic cones, fol-
ably circular pattern of the south-easternmost lowed by stages of volcano-tectonic collapse
Tyrrhenian, with the inner ring of the Aeolian possibly due to the emptying of a magma
Islands---M. Palinuro and the outer ring of the chamber.
southern Apennines-Calabria-northern Sicily In general it has been postulated that the
mountains concentric about the Marsili volcano change of composition of erupted lavas toward
• . .'. This peculiar circular feature was called the concave side of the arc may be related to the
'anello delle Eolie' by Selli (see Selli et al. 1977, thickness and the structure of the crustal base-
p. 228), who demonstrated the calcalkaline ment (e.g. stable, 'orogenized' or labile: see
nature of rocks from six volcanic seamounts Wezel 1977)• The Aeolian calc-alkaline erup-
belonging to the ridge (Figs 1 & 2). tive centres are presumably related to crustal
The start of the Aeolian volcanism shifted fractures cutting an 'orogenized' basement con-
eastward during the Quaternary. Lava com- sisting of folded flysch-type units, similar to
position evolves from calc-alkaline to shoshoni- those outcropping in the Sicilian mountain
tic suites (Barberi et al. 1974). Stromboli is still chain (see Bacini Sedimentari 1980)• Addi-
active with persistent moderate explosive activ- tionally, geochemical data indicate contamina-
ity. Though the available radiometric data indi- tion of some andesitic magmas (Lipari Island)
cate a Quaternary age for the older Aeolian with flysch-type sediments metamorphosed
activity (see Marinelli 1975), it is however under conditions of low-pressure/high-
possible that earlier volcanic activity took place temperature (Marinelli 1975, p. 210).
from the middle Pliocene, after sedimentation
of the Trubi Formation (Selli et al. 1977; Bacini
Sedimentari 1980). The 'forearc' region
A quantitative interpretation of magnetic and
gravity anomalies along the Aeolian arc axis The convex side of the arc system faces towards
shows the continental nature of the underlying the Calabro-Sicilian continental region. It con-
basement. Its morphology is characterized by sists essentially of troughs and ridges (Fig. 2)
348 F. C. Wezel

Z
<_
_1
O
w~,o
~o~

_1
0 .

j77=.,~

ti. • °

\
......... i:!i%i

::Y _

i::: : :?i:: ¸ "i

"::7..

,¢
y
ll,, y
y
y
Y

/ , L •

.. ~,
-/ ., ,,:,%

/ °.

J
'~'", i ~ ] ' *I,~ ~ ::"
'~
: ii!

_~_~~'~ "~..~y-':
•~.J./', g
- . ~:~:~ • ~

g
• ~ "~~ . . . ., "~
g ~
, ['~"~:
~ E

a ag Y,

\\/

Fro. 2. Simplified geological map of Calabro-Sicilian Arc showing the concentric regional distribu-
tion of selected key tectonic, volcanic and seismic features (derived from published data, mainly the
"Structural Model of Italy' and the 'Carta Tettonica d'Italia'). (A) Shows the outward lateral
migration, from middle Miocene to late Pleistocene, of compressive deformation, concomitantly
with a shifting of the foredeep axis. (B) Fault-plane solutions for earthquakes along the arc denoting
dip-slip faulting and left-lateral strike-slip motion. Note the lack of thrust events.

Fro. 3. Diagrammatic and grossly simplified cross-sections across Sicilian segment of the Calabro-
Sicilian Arc, showing its 'krikogenetic' evolution. (A) (early-middle Miocene); development, uplift
and erosion of volcanic arc; infilling of peripheral geosyncline with 'internal' preflysch and flysch
sediments, partially volcaniclastic. (B) (late Miocene): inversion of regional relief; outward growth
and subsidence of Marsili block; tectonization of preflysch and flysch sediments; Hyblean submarine
 The structure of the Calabro-Sicilian Arc 349

@ @ ®
a
a

i = _z

I
C~ ....a ii

~ r~

i
i;
- £

>- , ~

=i i " i i i i i i ! " i i i i i =-. ! " i i i i i i i " i i i" i

volcanic activity. (C) (Plio-Quaternary): caldera-like collapse of Marsili block; complementary
activity of Marsili volcano and Aeolian andesitic ring; development and filling of Cefalh Basin;
faulting and uplift of deformed orogenic sediments; outward migration of uplift axis and foredeep;
activation of foreland 'arch' and development of 'outer ridge'. (D) (Present): continued collapse of,
and sedimentation in, Marsili and Cefalfl basins; further uplift and erosion of mountain arc;
'subduction' of the Hyblean 'arch' ('krikogen').
350 F. C. W e z e l

that developed in outwards progression during The fragmentation of this surface occurred by
the Plio-Quaternary. tilting toward the Ionian side and by stepwise
downfaulting towards the Tyrrhenian side.
The Sicilian segment of the mountain arc (i.e.
The Cefalu and Paola 'outer-arc' troughs the Peloritani-Nebrodi-Madonie chain: Fig. 2)
was a horst structure bordered by mainly east-
Filled by over 1000 m of Plio-Quaternary
west orientated normal faults in Pleistocene
sediments (up to 1900 m in Sicily and 4000 m in
times (Di Geronimo et al. 1979). The magni-
Calabria) the Cefal0 and Paola basins repre-
tude of uplift is manifested spectacularly by the
sent major depositories for sediment accumula-
presence of the Peloritani of lower Pleistocene
tion along the present Tyrrhenian margin (Fig.
marine sediments at an altitude of more than
2). The morphology of the basement is com-
1000 m (see Bonfiglio 197(t).
plex, giving rise to several depocentres sepa-
The present-day structural configuration of
rated by structural highs. In the Cefal/~ Basin
the mountain arc results from intense Plio-
the basement consists (see Bacini Sedimentari
Quaternary vertical tectonic movements (tilting
1981)) of the same tectonic units (lower
and faulting). These are superimposed on pre-
Miocene flysch sediments) which constitute the
vious compressive structures (thrust and gliding
Sicilian mountain chain (cf. Ogniben 1960;
nappes: see Ogniben 1960, 1973) produced by
Wezel 1974). The undeformed Cefal/l Basin
preceding phases of orogenic deformation,
succession, lying unconformably on the tecto-
nized flysch strata starts with middle Miocene which migrated progressively towards the exter-
sediments of likely Tortonian age. In Tortonian nal (foreland) zones (see Fig. 2A). Ancient
and younger strata all the structures shown by thrust arcs internal to the trench, such as this
seismic reflection profiles are extensional and one, should not be confused with modern
indicate foundering of the Cefal/l Basin, with 'accretionary prisms' created by converging
distinct episodes of subsidence (Bacini Sedi- plates. Calabro-Sicilian arcuate chains owe
mentari 1980). In the initial stage a fault-con- their present structure not to lateral compres-
trolled subsidence was dominant, but later it sive stress but to a post-orogenic tensional
was replaced by regional subsidence lacking regime,
conspicuous fault control. The external (foreland) zones of the Cala-
Thus it appears that the upper Miocene and brian orogenic arc appear to extend to the east
beneath the Ionian seafloor and constitute a
Plio-Quaternary basin-fill deposits overlying
thrust pile ('accretionary prism'; see Biju-Duval
tectonically emplaced units are entirely post-
& Montadert 1977), much wider in its surface
orogenic and clearly indicate a tensional regime
expression than the Sicilian one. The thrust
following tectonic compression.
front, passing through Gela and Catania in
Sicily, might be expected to continue in the
The mountain arc Ionian Sea (Fig. 2) and to be traced southward
as much as about 36°35'N on the 17th meridian.
The arcuate mountain chains represent an
overall arch-shaped ('geanticline') ridge out-
cropping in Calabria and Sicily (Fig. 2). In The foredeep
Calabria the mountain arc comprises two uplift
axes separated by the intramontane Craft and The Catania-Gela trough represents the
Mesima valleys which are graben-like depress- emerged portion of the trench (Fig. 2). The
ions (Ogniben 1973). The western axis (i.e. the submarine trench below the Ionian Sea is not a
Tyrrhenian Coastal Chain and its southern ex- pronounced bathymetric feature but it is
tension, Fig. 2) was arched and normal-faulted marked by local and discontinuous deeps at
during Plio-Quaternary times. The total post- about 3000 m. The 'narrow incipient' Calabrian
orogenic vertical displacement along the steeply trench . . . 'has a flat floor underlain by up to
dipping fault surfaces amounts to more than 800 m of layered (and faulted) fill' (Belderson
500 m (Carrara & Zuffa 1976). The eastern et al. 1974).
axis, centred on the crystalline highlands of the The Catania-Gela foredeep contains only
Sila, Serre and Aspromonte (Fig. 2), was up- some hundred metres of Pleistocene marls and
lifted and tilted eastward during the Pleis- sands. Boreholes have shown the presence of
tocene. the 'Gela Nappe' in the foredeep inner wall.
Remnants of a lower Pleistocene erosional This gliding nappe was emplaced from the
peneplain can be traced from the Ionian coast north in lower Pleistocene times (see Ogniben
up to almost 2000 m (Ergenzinger et al. 1978). 1960; Di Geronimo et al. 1979).
 The structure of the Calabro-Sicilian Arc 351

The foreland region times a step-like shift of deformation occurred

every 2 or 3 Ma. As deformation progressed
In Sicily the foreland is represented by the toward the external zones (in the south), found-
Hyblean Plateau (carbonate platform), up to ering phases occurred in the internal (Tyrrhe-
986 m above sea-level (Fig. 2). On the whole nian) zones (Fig. 3). Tectonic movements can
it exhibits a horst-like structure (directed be visualized as a lateral migration of the
NE-SW), being dissected by normal faulting. 'positive uplift axis' or still better as an out-
The plateau shows a close correlation with a ward-moving topographic wave producing re-
gravity high which was considered by Vecchia versal in the direction of tilting movements.
(1954, p. 14) to correspond with a 'simatic Clastic material for the lower Miocene inter-
bulge'. Towards the Catania-Gela foredeep the nal (Sicilide) flysch sediments exposed in the
platform sequence includes several Plio- northern coastal zone or drowned beneath the
Quaternary volcanic units with an average present-day 'outer-arc' troughs, was un-
thickness of 500 m. Volcanic rocks consist doubtedly derived from a sourceland im-
largely of alkali-basalts with some interbedded mediately to the north (cf. Ogniben 1960;
tholeiitic lavas (cf. Marinelli 1975, p. 211). Wezel 1974). This probably corresponds to the
This foreland plateau is broadly analogous Marsili block, now sunk to depths of more than
(cf. Fig. 3) to the 'outer swell' or 'outer ridge', 3000 m beneath the sea. The presence of calc-
external to deep submarine trenches, inter- alkaline volcanic detritus in some internal flysch
preted in terms of plate tectonics as resulting sediments (e.g. Tusa Flysch) represents direct
from flexing in the lithosphere because of evidence for derivation from an uplifted andesi-
downward bending of the rigid, subducting tic volcanic island arc considered to lie to the
oceanic plate. A further analogy can be made north in early Miocene times (Wezel 1974) and
between recent Hyblean lavas and some which has now disappeared beneath the Tyr-
oceanic basalts (from areas of 'smooth base- rhenian Sea (Fig. 3A).
ment'), suggesting in the Hyblean case deriva- The beginning of the inversion of regional
tion from cracks and fissures along foundering morphological relief probably occurred during
continental blocks rather than from oceanic middle Miocene times, and was probably con-
crust. nected with an intense tectonic phase affecting
The submerged portion of the Sicilian car- the Sicilide flysch strata. Outward growth of the
bonate platform below the Sicily Channel is Marsili block has probably caused the closing of
characterized by Siculo-Tunisian graben-like the contiguous foredeep and the tectonization
troughs (up to 1700 m deep) resulting from of its flysch sediments (Fig. 3B). The pulsatory
Plio-Quaternary block-faulting, with contem- expansion of the block was subsequently fol-
poraneous volcanic activity (Pantelleria and lowed by its breaking up and collapse.
Linosa alkali-basalts).

Intra-plate krikogenetic fragmentation

Tectonic evolution of the arc region
From the available geological evidence it seems
I argue from the data presented that the Cala- likely that the Marsili block, the central core of
bro-Sicilian region constitutes a post-orogenic the Calabro-Sicilian arc system, was affected
multi-ringed system concentric round the early by positive and later by negative vertical
Marsili meso-block. Spatial concentricity is pro- tectonic movements which initiated the out-
duced by an arcuate series of swell-shaped ward-migrating topographic wave.
upheavals (horsts) and basin-shaped depress- This case history is used to propound a model
ions which become younger and more widely for the evolution of similar arc systems else-
spaced with increasing distance from the Marsili where. Empirical data suggest some distinct
cauldron-like structure. stages in the morphotectonic evolution of a
Field studies in Sicily indicate a southward circular block or 'krikogen'. The new term
migration of orogenic deformation (thrust and "krikogenesis', derived from the Greek words,
gliding nappes, olistostromes) from the lower ko{ko~, for ring and from 7~veat~, for forma-
Miocene to the lower Pleistocene (Fig. 2A), tion, is proposed to define the tectonic move-
concomitant with the formation and outward ments which create the arcuate (island and
shift of a foredeep axis with rate of progression mountain belt) structures so widespread on
in the early Miocene of about 2 to 10 km Ma-1 terrestrial (e.g. the Australasian island arcs)
(Wezel 1975). My as yet unpublished stratig- and planetary surfaces. The relief originating
raphical research suggests that in early Miocene from krikogenetic movements results primarily
352 F. C. Wezel
from extensional stress. Krikogens generally Pliocene updomed region (Locardi et al. 1977),
exhibit radial and concentric deformation pat- which continues southward a 'metallogenic-vol-
terns, with magmatic activity often accompany- canic' arch (in the southern Tuscan region)
ing the deformation. flanked by rifts. Intense volcanicity occurred
The envisaged time-sequence of geological during the culmination of the uplift phase.
events comprises the following stages (Fig. 3): Within the larger 'Pelagian Sea krikogen',
Stage 1 ('dome and trough' structure)--after the Sicilian Channel was an arch-like feature
the post-paroxysmal phase, krikogenesis de- which collapsed during the Pliocene (Burollet
velops a crustal mosaic of blocks (circular in & Winnock 1979). The isolation of the Pantel-
shape) of different size, bounded by arcuate leria, Linosa and Malta troughs occurred by
and asymmetric troughs. The width of the rifting in Quaternary times accompanied by
troughs is commensurate with the diameter of alkali basaltic volcanism. These rifted arch
blocks. Shallow seismic activity (less than basins are located within a region where a
40 km deep) is largely concentrated in trenches thinned crust is suggested by positive Bouguer
but also occurs in central elevated areas of anomalies (Vecchia 1954).
blocks. The pattern of faulting is radial and, Stage 3 ('cauldron-type collapse')--in re-
subordinately, concentric. At present this stage sponse to the vast effusion of lava the central
may be recognized in the Calabro-Sicilian Arc area of krikogens collapses. Surface subsidence
either in the foreland (e.g. the Hyblean was probably caused by sinking of the magma
Plateau) or in the 'forearc' (e.g. the Silan block column in the magma chamber. The sinking
in Calabria) continental regions. The presence crust produces volcano-tectonic depressions
of positive gravity anomalies corresponding to like that of the Marsili Basin. Mainly because of
such blocks indicates crustal thinning, possibly the coalescence of minor basins (Savelli &
above locally ascending mantle material. Wezel 1979), their size tends to increase with
Such an episode of development may have time, giving rise to 'small ocean basins' like the
occurred during Plio-Quaternary times in the Tyrrhenian Sea (see Fig. 1).
case of the 'Sardinian block', at present a giant In such basins the floor comprises a mosaic of
'semi-dome'. The central-eastern part of the collapsed blocks, at different stages of develop-
island was intermittently uplifted and tilted ment as manifested by their seismic activity.
westward. This led to phases of erosion and Deep earthquakes (hypocentral depths more
subsequent infilling of western peripheral than 200 km) are probably related to the verti-
troughs (e.g. the Campidano graben) with con- cal sinking of the older, more evolved, kri-
tinental elastics (Cocozza et al. 1974). Associ- kogens. In fact, fault-plane solutions for three
ated alkaline volcanic activity developed along Tyrrhenian deep-focus shocks (depth of 280-
fissures roughly contemporaneously. 300 km) show dip-slip faulting along vertical
Stage 2 ('axial collapse zone')---continued fault planes (see Fig. 2B and Riuscetti & Schick
arch-shaped uplift may lead to an axial collapse 1975). The complete lack of foci in the 40-
zone with graben structures. Tension across the 200 km depth interval, the absence of correla-
'geanticline' arch provides lines of weakness tion between shallow and deep shocks and the
exploited by active arc volcanism. At the overall lack of thrust events are arguments
periphery of larger blocks a submarine trench against the presence of a Benioff zone (Riuscet-
('geosynclinal trough') can originate. Turbidites ti & Schick 1975). Furthermore, the low level of
derived from the volcanic arc and from older seismicity of the Calabro-Sicilian Arc is con-
uplifted rocks fill the trench with flysch-type sistent with post-orogenic tectonic activity.
sediments. The vast caldera-like subsidence was prob-
This stage of development may have occur- ably either accompanied or immediately suc-
red during early Miocene times in the case of ceeded by the outward-moving topographic
the Marsili block (Fig. 3A) and in late Miocene wave, producing the lateral shift of the fore-
and Plio-Quaternary times for the 'Etruscan deep axis. The uplift of the mountain arcs com-
block' (Tuscan Archipelago, Umbro-Tuscan pensates the sinking of the Marsili block
and northern Latium regions: see northern part ('touche-de-piano' tectonics).
of Fig. 1). In the latter area block-faulting and In Plio-Quaternary times the foreland 'arch'
uplift of the arch were accompanied by the began to evolve (Fig. 3C). It is mainly repre-
emplacement of plutons, volcanism and metal- sented in the present Sicily Strait by grabens,
logenesis resulting from hydrothermal pro- banks and volcanic activity, indicative of a
cesses (cf. Marinelli 1975). Latium alkaline Pliocene phase of fracturing (stage 1) and a
potassic volcanism originated in an elongate Quaternary phase of intra-arch collapse (stage
collapse zone (rift valley-type) within a late 2). Thus, it appears that the present structural
 The structure of the Calabro-Sicilian Arc 353

configuration of the Sicilian 'forearc' region Quaternary volcanic activity. The thrusting of
may be considered equivalent to that of an the Sicilian orogenic system over the 'foreland'
'inter-arch zone' lying between intra-plate block is related to the rate of subsidence of the
arches at different stages of development. latter (i.e. to its relative stage of 'krikogenetic'
development). It seems difficult to account for
this intra-continental contact, manifested by the
Conclusions Catania-Gela foredeep, by appealing to an ac-
tive Benioff zone.
For the reasons illustrated in the previous sec- Also the underlying 'foreland' circular block
tions it seems most unlikely that the Calabro- ('krikogen') is an intra-plate feature whose
Sicilian Arc has resulted from descent of litho- tectonic history might be attributable to local
sphere along a Benioff zone. The concentric mantle movements.
and radial shape of the system in plan view 'Krikogenetic' movements are thought to be
suggests that the active tectonic force is vertical. induced by thermal regimes of mantle dilata-
Present-day morphological structure is entirely tional nuclei ('DIN' of Wezel 1977). Stage 1 is
post-orogenic and indicative of intracontinental related to the regional upwelling, stage 2 to
extensional tectonics. subsequent thermal expansion of the 'DIN',
The geological history of the area shows an while stage 3 manifests thermal contraction.
outward step-like growth of the orogenic sys- The fact that the whole Tyrrhenian floor
tem beginning from a central circular block (Fig. 1) and the Italian region appear to have
('krikogen'), presently forming the Marsili fractured in post-orogenic times in different
Basin and almost completely encircled by the circular blocks assembled like a mosaic, empha-
'Aeolian andesitic ring'. It is postulated that sizes the importance and the spreading of intra-
this block suffered 'krikogenetic' movements plate 'krikogenetic' deep-seated deformation in
(and igneous processes) consisting of Miocene shaping present-day geomorphological and
arching and rifting stages, followed in the structural configurations. I believe that this
Pliocene by large-scale cauldron collapse, poss- tectonic process is particularly efficient in de-
ibly triggering outward-moving 'touche-de- veloping the topographic features of the 'back-
piano' movements which structured the 'fore- arc' and the 'forearc' regions located within
arc' region. Foundering phases occurred in the previous Alpino-Mediterranean mountain
internal (hinterland) zones as deformation and belts.
migration of the foredeep axis progressed out-
wards in time toward the external (foreland)
ACKNOWLEDGMENTS" I should like to thank Drs J.
zones. Leggett, A. Robertson and A. Smith, for their many
Beginning in Messinian times the 'foreland' helpful suggestions and for revising the English. The
block (Sicily Channel) was also activated, ex- study of the Tyrrhenian Sea was part of an Italian
hibiting the early two stages of evolution (i.e. programme funded by the Consiglio Nazionale delle
arching and rifting), accompanied by Plio- Ricerche ('P.F. Oceanografia e Fondi Marini').

References
ALVAREZ, W., COCOZZA,T. & WEZEL, F. C. 1974. nean basins. Initial Rep. Deep Sea drill. Proj. 42,
Fragmentation of the Alpine orogenic belt by 951-84.
microplate dispersal. Nature, London, 248, 309- & MONTADERT, L. 1977. Introduction to the
14. structural history of the Mediterranean basins.
BAC1NI SEDIMENTARI1980. Dati geologici preliminari Int. Syrnp. Struct. Hist. Med. Basins, 1-12. Tech-
sul Bacino di CefalO (Mar Tirreno). Ateneo nip, Paris.
Permense, Acta Nat. 16, 3-18. BONFIGLIO, L. 1970. Facies biodetritica tardoplioceni-
BARBERI,F., INNOCENTI,F., FERRARA,G., KELLER,J. ca nei Peloritani a 1250 metri di altitudine. Boll.
& VILLARI, L. 1974. Evolution of Eolian arc Soc. geol. Ital. 89, 499-506.
volcanism (southern Tyrrhenian Sea). Earth BUROLLET, P. F. & WINNOCK, E. 1979. Essai de
planet. Sci. Lett. 21,269-76. synth6se. In: BUROLLET,P. F., CLAIREFOND,P. &
BELDERSON,R. H., KENYON,N. H. & STRIDE,A. H. WINNOCK, E. (eds). La Mer P~lagienne. G~ol.
1974. Calabrian Ridge, a newly discovered M~dit. 6, 321-7.
branch of the Mediterranean Ridge. Nature, CAFLISCH, L. & SCHMIDT DI FRIEDBERG, P. 1967.
London, 247, 453-4. L'evoluzione paleogeografica della Sicilia e sue
BIJU-DUVAL, B., LETOUZEY, J. • MONTADERT, L. relazioni con la tettonica e la naftogenesi. Mem.
1978. Structure and evolution of the Mediterra- Soc. geol. Ital. 6, 449-74.
354 F. C. Wezel
CARRARA,A. & ZUFFA, G. G. 1976. Alpine structures SQUYRES, C. H. (ed). Geology of Italy. Earth
in northwestern Calabria, Italy. Bull. geol. Soc. Sci. Soc. Libyan Arab Republic, Tripoli, 1,
Am. 87, 1229-46. 165-219.
COCOZZA, T.,JACOBACCI, A., NARDI, R. & SALVATORI, MORELLI, C., GIESE, P., CASSINIS, R., COLOMBI,B.,
I. 1974. Schema stratigrafico-strutturale del Mas- GUERRA, I., LUONGO, G., SCARASCIA, S. &
siccio Sardo-Corso e minerogenesi della Sardeg- SCHUTTE,K. G. 1975. Crustal structure of South-
ha. Mere. Soc. geol. Ital. 13, 85-186. ern Italy. A seismic refraction profile between
DELLA VEDOVA,B. & PELLIS, G. 1979. Risultati delle Puglia-Calabria-Sicily. Boll. Geol. Teor. Appl.
misure di flusso di calore eseguite nel Tirreno 17, 183-210.
sud-orientale. Atti Cony. Sci. Naz. P.F. NAPPI, G., BONASIA,V., FERRI, M., MONTAGNA,S. &
Oceanografia e Fondi Marini CNR (Roma 5-7 PINGUE, F. 1977. Primi risultati sulla sorveglianza
matzo 1979), 2, 693-712. del complesso vulcanico Lipari-Vulcano (Tirreno
DEE MONTE, M. 1972. I1 vulcanesimo del Mar Tirre- meridionale). Boll. Soc. geol. Ital. 95, 967-80.
no: nota preliminare sui vulcani Marsili e Pali- OGNIBEN, L. 1960. Nora illustrativa dello Schema
nuro. G. Geol. 38, 231-52. geologico della Sicilia nord-orientale. Riv. Min.
Dt GERONIMO, I., GHISEa-n, F., LENT1Ni, F. & Siciliana, I l, 183-212.
VEZZANI, L. 1979. Lineamenti neotettonici della 1973. Schema geologico della Calabria in base
Sicilia orientale. Mem. Soc. geol. Ital. 19, 543-9. ai dati odierni. Geologica Romana, 12, 243-585.
ERGENZINGER, P., GORLER, K., IBBEKEN, H., RIUSCETrl, M. & SCHICK, R. 1975. Earthquakes and
OBENAUF, P. & RUMOHR, J. 1978. Calabrian Arc tectonics in Southern Italy. Boll. Geol. Teor.
and Ionian Sea: vertical movements, erosional Appl. 17, 59-78.
and sedimentary balance. In: CLOSS, H., SAVELLI, D. & WEZEL, F. C. 1979. Morfologia e stile
ROEDER, D. & SCHMIDT, K. (eds). Alps, Apen- tettonico del Bacino Tirrenico. Atti Cony. Sci.
nines, Hellenides, 359-73. Schweizerbart'sche, Naz. P.F. Oceanografia e Fondi Marini CNR
Stuttgart. (Roma 5-7 matzo 1979), 2, 729-38.
IACOBUCCI, F., INCORONATO, A., RAPOLLA, A. & SFLLI, R., LUCCHINI,F., ROSSl, P. L., SAVELLI,C.
SCARASCIA, S. 1977. Basement structural trends DEE MONTE. 1977. Dati geologici, petrochi-
in the volcanic islands of Vuicano, Lipari and mici e radiometrici sui vulcani centro-tirrenici.
Salina (Eolian Islands, Southern Tyrrhenian J. Geol. 42, 221-46.
Sea) computed by aeromagnetic and gravimetric VECCHIA, O. 1954. Lineamenti geofisici e geologia
data. Boll. Geol. Teor. Appl. 20, 49-61. profonda nella Sicilia ed arec circostanti. Riv.
LOCARD1, E., LOMBARDI, G., FUNICIELLO, R. & Geof. Appl. 15, 3-34.
PAROrro, M. 1977. The main volcanic groups of WEZEL, F. C. 1974. Flysch successions and the tecto-
Latium (Italy): relations between structural nic evolution of Sicily during the Oligocene and
evolution and petrogenesis. Geol. Romana, 15, early Miocene. In: SQUYRES, C. H. (ed). Geolo-
279-300. gy of Italy. Earth Sci. Soc. Libyan Arab Repub-
MCKENZIE, D. 1972. Active tectonics of the lic, Tripoli, 2, 105-27.
Mediterranean region. Geophys. J. R. astron. 1975. Diachronism of depositional and di-
Soc. 30, 109-85. astrophic events. Nature, London, 253, 255-7.
MACCARRONE, E. 1970. Notizie petrografiche e pet- 1977. Widespread manifestations of Oligocene-
rochimiche sulle lave sottomarine del Seamount 4 Lower Miocene volcanism around western
(Tirreno Sud). Boll. Soc. geol. Ital. 89, 159-80. Mediterranean. Int. Syrup. Struct. Hist. Med.
MARINELLI, G. 1975. Magma evolution in Italy. ln: Basins, 287-301. Technip, Paris.

FORESE CARLO WEZEL. Istituto di Geologia, Universita degli Studi, 61029 Urbino,
Italy.
 Deformation of the Makran accretionary sediment prism in the
Gulf of Oman (north-west Indian Ocean)

Robert S. White

SUMMARY: The offshore Makran of Pakistan and Iran in the Gulf of Oman forms the
seaward margin of a folded and faulted accretionary sediment wedge which extends
several hundred kilometres inland across the onshore Makran. Sediment is being scraped
off the oceanic Arabian plate as it subducts shallowly northward beneath the Eurasian
plate. The submerged front of the accretionary prism has been traced along the entire
width of the Makran continental margin, from the Strait of Hormuz in the west to the
boundary with the Indian plate in the east. The structure of a 70 by 90 km portion of the
deformed sediment belt was mapped in detail by a grid of continuous seismic reflection
profiles. This survey demonstrates that the folds are laterally continuous and are well
lineated, that the deformed belt appears to be migrating southward at a rate of about
10 km every million years and that folding of the initially undisturbed abyssal plain
sediments occurs in the southernmost, or 'frontal' fold. Mapping of the three-dimensional
structure of the frontal fold shows that it systematically increases then decreases in height
along strike over a distance of about 80 km. Folding is initially restricted to the uppermost
2½ km of the sediment pile, and there is local evidence of small-scale shale diapirism. The
frontal fold is then incorporated into the accretionary prism by uplift along a thrust fault
which fattens out within the sedimentary section, possibly in overpressured shales. There
is little subsequent deformation of the uplifted material or of the sediment which infills the
basins between the fold ridges until some 70 km north of the frontal fold when further
rapid uplift occurs, eventually elevating the sediment above sea-level. I attribute this
second phase of uplift to major imbricate thrust faulting extending down to the oceanic
basement; most of the shortening and thickening of the sediment wedge occurs in this
region. The relatively simple pattern of uplifted folds and intervening basins in the frontal
70 km of the accretionary wedge results from a balance of compressive and normal
stresses on the thick sediment section. This simple deformational style of the seaward part
of the accretionary prism can be traced eastwards until it becomes chaotically disrupted by
the consumption of a basement ridge which protrudes to just beneath the seafloor.

The Makran continental margin of Pakistan and The M a k r a n accretionary wedge lies above
Iran in the Gulf of O m a n , north-west Indian the oceanic part of the Arabian plate which is
Ocean stretches from the Strait of Hormuz in subducting shallowly northwards beneath the
the west to near Karachi in the east and forms continental Eurasian plate at a convergence rate
the seaward edge of a large accretionary sedi- of about 50 mm y r - ' (see inset to Fig. 1). The
ment prism extending hundreds of kilometres easternmost boundary of the wedge is formed
inland. This paper examines the style of de- by an active right-lateral strike-slip fault system
formation in the offshore Makran where the extending along the Ornach Nal and C h a m a n
thick flat-lying sediments of the Gulf of O m a n faults (Abdel-Gawad 1971). The western end of
become folded and faulted in the initial stages the M a k r a n accretionary prism is at the Strait of
of their incorporation into the accretionary Hormuz where collision between the Arabian
prism. Previous work on R R S Shackleton leg and Eurasian plates changes to a continent-
3/75 in the central part of the Gulf of O m a n continent type (White & Ross 1979). On the
demonstrated that the M a k r a n continental mar- south-eastern side of the Gulf of O m a n the
gin comprises a series of fold ridges and in- Arabian plate is separated from the Indian
tervening basins (White & Klitgord 1976; White plate by the Murray Ridge (Barker 1966)
1979b); similar structures were found extending which, as our recently acquired seismic profiles
westwards to the Strait of Hormuz on R V indicate, is a tensional feature.
Atlantis II leg 96(13) (White & Ross 1979). I Preliminary results from a reversed seismic
report preliminary results from R R S Shackleton refraction line above the Gulf of O m a n abyssal
leg 1/80 in January 1980 which completes the plain show that normal oceanic crust lies be-
eastern end of the M a k r a n margin survey off neath the 6 to 7 km thick pile of flat or gently
the coast of Pakistan (Fig. 1). dipping sediments. Although there are no rec-

357
358 R . S. W h i t e

30*N

Suifon \~x~\

\ %..,._*J ~'~.
";f'/(~*'//\xiiTaf
"~ fan *
HAMUN-I
MASHKEL

\\-\

, 7-.:.,--.,,' - -~- -- ~-!.% f/

.................
;

,"'" 'i "'- '7~ ? Oetaded

\,

OMAN --~ ~ ::/" /'" ~"\'"",, ".... ~,._")k ' x \

"<'~%,x~__-~ .," OMAN \ ":, ~' \~, -
Eurastan Plat

-) % ....
\~- as. "
.%~ "(
Plate \
\\
\,
21" N4---
S6*E 67"E

FIo. 1. Map showing location of continuous seismic profiles (all with coincident gravimetric,
bathymetric and magnetic measurements), in the Gulf of Oman, north-west Indian Ocean. Dotted
tracks from RV Atlantis II leg 96(13), dashed tracks from RRS Shackleton leg 3/75 and
solid tracks from RRS Shackleton leg 1/80. Detailed survey tracks are shown expanded in Fig. 2.
Stars show location of volcanic centres, triangles the positions of active mud vents, solid areas
mark ophiolitic assemblages and 'coloured m61anges' and thin lines show the major faults on land.
Information compiled from Ahmed (1969), Barker (1966), Berberian (1976), Hudson et aL (1954),
Jacob & Quittmeyer (1979), Kazmi (1979), Nowroozi (1976) and Snead (1964). Inset shows general
setting of the area with the main plate boundaries.

ognizable seafloor spreading magnetic anoma- the measured heat flow values within the basin
ly lineations beneath the Gulf of O m a n by show little scatter the best limits that can be
which the crust can be dated, its age has been placed on the age of the crust if all possible
estimated by comparing the heat flow measured sources of error are considered are that it was
at a number of stations within the basin to the formed between 70 to 120 Ma ago (Hutchison
standard curves published by Parsons & Sclater et al., in press). This means that the crust is at
(1977). least as old as and may be older than the large
Only rather broad limits can be placed on the ophiolite sequences emplaced in Oman. Crustal
age of the oceanic crust because the sedimenta- creation during the Cretaceous quiet zone also
tion rates and history are poorly known and so explains the absence of seafloor-spreading mag-
the correction factor to convert from the netic lineations.
observed heat flow through the seafloor to the Accreted sediments have been uplifted and
heat flow through the buried basement is sub- exposed in the onshore Makran as thick,
ject to considerable uncertainty. Even though faulted flysch sequences (see e.g. Hunting 1960,
 Deformation o f sediment prism in the Gulf o f Oman 359

Arthurton et al. and McCall & Kidd 1981 for d. Gas hydrate~free gas reflection
discussions of the geology of the Makran). Accumulations of free gas, which have very
Between four to five hundred kilometres north low impedance to seismic waves and conse-
of the coast is a chain of Cenozoic volcanic and quently give rise to large-amplitude reflections
plutonic rocks of andesitic to rhyolitic composi- are common within the uppermost 700 m of
tion (Fig. 1) which may represent a volcanic arc sediment beneath both the abyssal plain and the
(Farhoudi & Karig 1977). There are large off- continental margin of the Gulf of Oman (there
sets in the volcanic chain such as to the east of is a good example at the south end of profile C,
the Jaz Murian depression at about 60°E, and Fig. 3, marked 'g'). The compelling evidence
Dykstra & Birnie (1979) have suggested that that these anomalously strong reflectors are
the subduction zone may be segmented. How- generated by free gas is reported by White
ever, we have found no significant offsets any- (1977b). Furthermore, the free gas is not found
where in the offshore Makran structures. There in stratigraphic or structural traps but forms a
is no well-developed Benioff zone, although the bottom simulating reflector (BSR) by accumu-
seismicity appears to deepen somewhat to the lating beneath an impermeable sediment plus
north in a way consistent with a shallowly gas hydrate layer which is stable in the topmost
subducting plate (Jacob & Quittmeyer 1979). few hundred metres of sediment. White's
The Makran subduction zone exhibits no (1979a) evidence that the BSR's are caused by
topographic trench, and the accretionary wedge free gas trapped beneath gas hydrate has been
is unusually wide; the very thick sediment pile confirmed in several ways: the measured
on the subducting plate is probably responsible temperature gradient above a prominent gas
for these abnormal features. reservoir (Hutchison et al., in press) agrees with
the temperature gradient predicted from the
Interpretation of seismic reflection profiles depth and the pressure and temperate (P-T)
conditions of the free gas to hydrate phase
The continuous seismic reflection profiles change, the marked increase in the seismic
illustrated in this paper are unprocessed except velocity of hydrated sediments found in labora-
for the application of linear time variable gain tory experiments (Stoll & Bryan 1979) has been
triggered by the seafloor return and of band observed in situ, and many more examples of
pass filtering from 5 to 70 Hz. They exhibit a BSR's mirroring the P-T conditions of the
number of features which we briefly review to hydrate phase change have been recorded on
prevent possible misinterpretation. our recently acquired profiles. This confirma-
tion that the BSR's are caused by trapped gas is
a. Bubble pulse important for our purposes beause it means that
A single 2.6 litre (160 cubic inch) airgun fired some of the strongest reflectors on the seismic
once every 12 sec was used as a sound source. It profiles bear no relation to the structure but
generates a quarter second long wavetrain simply record particular P-T conditions, and
which causes multiple returns from every inter- indeed frequently cross-cut major structural
face and obscures other arrivals within this boundaries (several such reflectors are marked
period. Higher resolution of the structure im- 'g' on Figs 3 & 9).
mediately beneath the seafloor was achieved by
making a separate record of the signal after Detailed survey across fold belt
high pass filtering at 100 Hz to remove the
dominant bubble pulse frequency. A detailed grid survey of a 70 x 90 km area
extending from the Gulf of Oman abyssal plain
b. Water multiple onto the Makran continental margin (Fig. 1)
The prominent return at twice the travel time shows that the sequence of ridges and interven-
to the seafloor (marked 'm' on Fig. 3) is an ing basins is strongly lineated parallel to the
artefact generated by an extra bounce in the coast. The survey consists of 11 profiles perpen-
water layer. dicular (A through K, Fig. 2), and 10 profiles
parallel (lines 1 to 10, Fig. 2), to the strike
c. Migration of the fold belt. Gravimetric, bathymetric and
The illustrated profiles are unmigrated time magnetic measurements were made in addition
sections; to turn them into true depth sections to a continuous seismic reflection profile along
they must be migrated and the normal inci- each line. The profiles were navigated using a
dence travel times multiplied by the seismic combination of moored radar transponder,
wave velocity in the sediment. This has been satellite and land radar fixes, which were in-
done for the frontal fold (Fig. 10). terpolated using measurements of the speed
360 R. S. W h i t e

25 through the water from an electro-magnetic

H IJK log.
Profile C (Fig. 3) from the western part of the
detailed survey exhibits the typical sequence of
ridges and intervening basins seen along the
2
entire Makran continental margin (see White &
3 Klitgord 1976 and White & Ross 1979 for
4 profiles from further west off the Iranian coast).
5 Gently dipping sediments beneath the abyssal
plain in the south are folded to a maximum
amplitude of about 400 m in the 'frontal fold'
(at 24 ° 10' N on Fig. 3), then uplifted by fault-
ing to a height of about 1¼ km above the abyssal
plain. Basins between the uplifted folds subse-
quently become filled by sediment, the amount
of infill increasing towards the coast until even-
tually the crests of the folds are completely
10 buried. The sediment is derived partly from
24 reworking of the material of the fold ridges
62 63 although the absence of slump accumulations at
the base of the ridges suggests that they are well
FIG. 2. Location map showing positions of consolidated and that most of the basinal infill
grid of continuous seismic reflection pro- comes from the surrounding land masses of
files in the detailed survey area over the
Arabia and the Makran. Most of the adjacent
offshore Makran accretionary belt. Gra-
vimetric, bathymetric and magnetic land is desert and much of the material is
measurements were also made along these wind-borne (Stewart et al. 1965). In the lower
tracks. part of the fold belt the ridges form effective
barriers to downslope sediment movement and

m 9

U • >L.

5.%~

2/'d/+0' LATITUDE NORTH 2Z+* I0'

PROFILE C
i i
10 k m

FIG. 3. Continuous seismic reflection profile along line C perpendicular to the strike of the fold belt
(see Fig. 2 for location). Arrowheads mark intersection with profile 5 (Fig. 5), and profile 9 (Fig. 9).
'm' marks water multiple and 'g' denotes prominent returns caused by free gas trapped beneath gas
hydrate. The profiling system comprised a single 2.6 litre (160 cubic inch) airgun fired once every
12 sec and a Geom6chanique hydrophone streamer. Linear time variable gain triggered by the
seafloor return and 5-70 Hz band pass filtering has been applied to the signal. Vertical exaggeration
at seafloor is approximately 7:1. Heavy vertical lines are noise caused by interference.
 Deformation of sediment prism in the Gulf of Oman 361

24'=SO' LATITUDE NORTH 2t~' 10'pROFILE 1

10 k m

FIG. 4. Continuous seismic reflection profile along line I perpendicular to the strike of the fold belt
(see Fig. 2 for location). Arrowhead marks intersection with profile 9 (Fig. 9). Vertical exaggeration
at seafloor is approximately 7:1. Vertical lines are one-hourly marks superimposed on the record.

the seafloor is often at different levels in adja- faulting of the basinal sediments until a major
cent basins. On profile I (Fig. 4), the southern- region of uplift near the coast is reached.
most basin at 24 ° 19' N is at a higher level than The bathymetric map of the detailed survey
the next landward basin at 24 ° 22' N, presum- area illustrated in Fig. 6 demonstrates the
ably as a result of recent uplift of the outermost linearity of both the scarp generated by the
ridge. uplifted ridge behind the frontal fold (at appro-
Profiles across the inter-ridge basins (e.g. ximately 24 ° 13' N), and of the ridges and
Figs 3 & 4) show the infilling sediment to be basins lying behind this. In the northernmost
little disturbed and in particular there is no part of the survey (north of 24 ° 47' N), is a
evidence of buckling. I infer that in the south- steep slope rising rapidly towards sea-level;
ernmost 60 km of the fold belt there is no major unlike the lower part of the margin this is cut by
shortening other than in the frontal fold, deep transverse valleys which feed terrigenous
although as suggested below further major de- sediment on to the upper part of the fold belt.
formation and faulting commences approx- Individual ridges can be traced along strike
imately 60 to 70 km landward of the frontal across the entire detailed survey area as can be
fold. Horizons within the basinal sediments dip seen in the bathymetric profiles of the north-
gently landward and the dips increase with south survey lines illustrated in Fig. 7. Some of
depth; this structure may be attributed partly to the ridges die out in the vicinity of profile G,
the unconformable deposition of sediment on giving in detail a different sequence of ridges
the landward side of the crest of the frontal fold and basins in the west of the area (e.g. profile
as it develops (discussed further below), and C, Fig. 3), than in the east (e.g. profile I, Fig.
partly to tilting of the sediment ponds as the 4). In Fig. 8 the north-south bathymetric pro-
frontal fold is uplifted by a thrust fault which files are shown superimposed to emphasize the
flattens with depth. Landward tilts of up to ½° in degree of similarity of shapes and depths of the
the seafloor of the basins are common (see, e.g. ridges and intervening basins. The lower part of
Fig. 4 and White & Klitgord 1976), so if we the figure shows the heights of all the ridge
assume that the infilling sediment was depo- crests where they are not buried by secondary
sited horizontally this indicates that recent post- sediment. After the initial uplift of the outer-
depositional tilting has occurred. most ridge to an elevation of 1¼ km above the
Profiles parallel to the strike along the inter- abyssal plain the ridge crests remain at roughly
ridge basins in the frontal 60 km of the margin the same height, increasing in elevation by an
exhibit remarkably little disturbance of the average of less than 1:100 (about 0.6°), until
infilling sediment (e.g. profile 5, Fig. 5). There rapid uplift occurs some 60 to 70 km landward
is no evidence of folding, of diapirism or of of the frontal fold.
362 R. S. White

C
2- " ! . . . .

TWO-WAY .~, . -t:~7-T.":~-- : ~ :. ~ " ~ : " v , ::"7~;:: .". ~"~-L ~ i : ;T.:~

TIME ". 7...- :::-.:: :[. £.},:i, t : . . t :, '.,;::i~.i:: :z-. ~ .~..,--i&~ - ~ . ~ ' 2 ' ~ 2 ~ ; : : - : ~ 2 ; ~ i !

62"20' LONGITUDE EAST 62"30'

PROFILE 5

Fro. 5. Part of continuous seismic reflection profile along line 5 parallel to the strike of the fold belt
(see Fig. 2 for location) showing flat-lying sediments. Arrowhead marks intersection with line C (Fig.
3). Vertical exaggeration at seafloor is approximately 7:1. Vertical lines are one-hourly marks
superimposed on the record.

The frontal fold

The frontal fold is sufficiently gently folded 400 m by profile F. Near profile G the fold
to allow us to investigate its structure by tracing becomes much broader and subsequently de-
reflectors into the fold from beneath the abyssal creases in amplitude to 240 m by profile K in
plain. This is not possible with the ridges futher the east (e.g. Fig. 7).
on to the Makran continental margin because b. The fold is asymmetric in shape with the
they are too highly deformed. I have chosen steepest limb facing seaward. The seaward side
three prominent horizons (marked by arrow- often exhibits a crumpled surface (e.g. Figs 3 &:
heads on Fig. 9) which can be seen on every 11), which may be tectonic in origin but may
crossing, and correlated them between profiles also be caused in part by sediment slumping off
using two tie lines orientated along strike; they the seaward slope.
are profile 10 which runs approximately east- c. Sediment accumulates in the depression
west across the abyssal plain sediments just between the crest of the frontal fold and the
south of the fold, and profile 9 (illustrated in adjacent uplifted ridge. Much of the infilling
Fig. 9), which is positioned just north of the sediment is probably derived from the steep
crest of the frontal fold. A very strong reflec- seaward-facing scarp of the elevated ridge. As
tion from the gas hydrate/free gas boundary the frontal fold develops the secondary sedi-
also occurs at 4.7 to 4.8 sec near the crest of the ment is deposited with increasing angle of un-
fold and can be traced continuously for more conformity against the landward limb of the
than 40 km ('g' on Fig. 9). Although some fold. The ponded sediment accentuates the
reflectors are visible from deeper into the skewness of the seafloor expression of the
sedimentary section (e.g. on Fig. 3), we have frontal fold (e.g. Fig. 10).
not mapped them because they are insufficient- d. With the exception of the uppermost
ly clear to correlate between profiles using the sedimentary unit which is about ½ km thick the
tie lines. The time sections were migrated into top 21 km of sediment is deformed in parallel
true depth sections (Fig. 10) using sediment folds (Fig. 10). Only the topmost ½ km thick
velocities calculated from variable angle sono- sediment unit thins appreciably as it is traced
buoy profiles in the Gulf of O m a n (reported by from beneath the abyssal plain into the frontal
White 1977a and Closs et al. 1969). The main fold. The fold at depth is generally broader than
features of the frontal fold are: the seafloor ridge.
e. Deep refectors within the sedimentary
a. The amplitude of the fold varies along column, where they can with confidence be
strike from 120 m on profile A in the west to traced from the abyssal plain into the frontal
 Deformation of sediment prism in the Gulf of Oman 363
i

2z,° 5C

2/+0 00'
62°10 ' 62'50'
FIG. 6. Bathymetric map of the detailed survey area over the offshore Makran accretionary belt
Contours in hundreds of metres, corrected for variations in the speed of sound in the water. Shading
is in 600 m intervals. Track control shown by light lines; ' - ' and ' + ' mark bathymetric highs and
lows, respectively. Note the well-lineated sequence of fold peaks and intervening basins which is
particularly clear on the lower part of the slope where it is not masked by overlying sediment.
Cross-cutting channels are visible on the upper portion of the slope.
364 R. S. White
2,o 5o, Latitude North 24Ooo,
i i i I

flDepfh
2 (kin)
t

K
FIG. 7. Bathymetric profiles from detailed survey lines projected on to north-south axis shown along
top of figure. See Figs ! and 2 for location of profiles. Vertical exaggeration is 7.7: 1. Note the
increase in height of the frontal fold over profiles A to G, then the change in width and decrease in
height over profiles H to K.

24o50,, , Latitude, North, , 2/+0,000

Oeplh
(kin)

.===v.2-& -z ~ _ _ ~_
-_ - _ _.._---____=_.__ ~..= =.~_...,_

FIG. 8. Top two diagrams show stacked bathymetric profiles from detailed survey lines projected
on to north-south axis (as in Fig. 7). Bottom diagram shows depths of crests of ridges in the detailed
survey where they are not buried beneath sediment. Vertical exaggeration is 7.7:1.
 Deformation o f s e d i m e n t p r i s m in the G u l f o f O m a n 365

[3 C I

..... ,4' : . . . . ~\ - I
/+.

5
TWO-WAY
TIME
(SECS)
" 7L;2. , ::'

•
"(i

~.: '7.':."!+'5 :.:.~: 7 -

62" 20' 62" 50'

LONGITUDE EAST
PROFILE 9

FIG. 9. Continuous seismic reflection profile along line 9 orientated parallel to the strike of the fold
belt approximately 1 km north of the crest of the frontal fold (see Figs 2 and 6 for location).
Arrowheads mark intersection with profiles B (Fig. 11), C (Fig. 3) and I (Fig. 4). Note the inferred
shale diapir marked 'd' also seen on profile B (Fig. 11), the strong bottom simulating reflector
marked 'g' caused by free gas trapped beneath a gas hydrate layer and the lateral continuity of
reflectors. The reflectors marked a, b, c are those used to construct the depth sections illustrated in
Fig. 10. Some of the deep returns are side echoes from the adjacent ridge. Vertical exaggeration at
seafloor is approximately 15:1. Vertical lines are one-hourly marks.

fold (which is generally only where the ampli- the processes involved in transferring the thick
tude of the fold is small), appear to dip beneath undeformed sediment pile from the subducting
the overlying 2½ km of folded material without Arabian plate to the folded and faulted accre-
themselves becoming buckled (see, e.g. Figs 3, tionary wedge of the Makran using the observa-
11 and profiles from further west in White tional constraints. In so doing, two points are
1977a). worth bearing in mind: firstly that, apart from
f. On some profiles the beds are upturned the frontal fold where we can see reflectors
where they abut against the adjacent uplifted from deep within the sedimentary section, we
ridge (e.g. profile B, Fig. 11). only have details of the surficial deformation
from which to infer the deep structure; secondly
that the entire offshore belt is controlled ulti-
Mechanisms of fold formation and emplacement
mately by gravity forces acting on the subduct-
In the previous two sections I have described ing plate and the overlying sediments.
the observed features of the offshore Makran One of the main questions concerns the
fold belt and have alluded to the likely emplace- extent to which diapirism rather than crustal
ment mechanisms. I now look in more detail at shortening controls the formation of the charac-
366 R. S. W h i t e

A B c
--....
p',,
.~ .-:;,

~"~___
,,. ,.,,

E F G H

/iii
"'...._.__

/"

I K

6.

0 16 20

FIG. 10. Migrated depth sections of the upper portion of the frontal fold in the detailed survey area
(see Fig. 2 for location). The horizons shown were correlated between profiles using profiles 9 (Fig.
9) and 10 as tie lines. Dotted area on profile B marks a shale diapir (see Fig. 11). Depths (vertical)
and distances projected perpendicular to the strike (horizontal), are in kilometres, with north on the
left of each section and distance 0 km at the foot of the scarp marking the seaward edge of the
adjacent uplifted ridge. Vertical exaggeration is 10:1.

teristic ridge and basin morphology of the Other, smaller-scale evidence that diapiric clay
offshore Makran. Salt diapirism cannot be re- is present in the Makran is given by the wide-
sponsible for the structures because no large spread occurrence of active mud volcanoes
outcrops of salt are found anywhere on land in along the coast (see Fig. 1 and Ahmed 1969,
the Makran, nor do the reflection and refrac- Snead 1964), which are frequently triggered by
tion profiles at sea show any evidence of the earthquakes and may be accompanied by out-
distinctive reflectors and seismic velocities pro- gassing (Sondhi 1947).
duced by salt. Clay diapirism is, however, much We have found only one example of a prob-
more likely in the thick, rapidly accumulated able clay diapir in all our reflection profiles
flysch deposits of the Gulf of Oman. Chapman across the Makran continental margin; this has
(1974) has suggested that thick regressive se- formed within the frontal fold on profile B
quences present ideal conditions for the genera- (illustrated in Fig. 11), at a depth of about
tion of a folded and overthrust belt across a 250 m beneath the seafloor. The diapir extends
continental margin and Buffler et al. (1979) to-about 700 m depth, the southern edge being
have attributed part of the Mexican Ridge marked by a curved fault surface which is
province in the western Gulf of Mexico entirely concave landward. Deeper reflectors show an
to downslope overthrusting; the sequence of inflexion beneath the diapir which is probably
ridges and basins in the Mexican Ridges has an artefact caused by low seismic velocities in
some similarities in general size and appearance the over-pressured diapiric clays. Profile 9 (Fig.
to those on the Makran continental margin. 9) crosses the diapir perpendicularly and shows
 Deformation of sediment prism in the Gulf of Oman 367

that it is of limited extent, with a diameter of has been locally modified by the step topo-
about 2 -~
2 km. graphy of the adjacent fold belt (see, e.g. King
Despite the evidence for localised diapiric 1978).
clay intrusions I believe that diapirism is not (b) The frontal fold is next uplifited by a
primarily responsible for the formation of the thrust fault which is concave towards the land
fold ridges on the Makran margin. The diapir and flattens out within the lower sediments.
on profile B discussed above is much smaller The frontal fold must be uplifted by a fault
than, and lies above, the folding in the frontal rather than simply by continued folding because
fold. The larger-scale buckling of the frontal the maximum height the frontal fold attains
fold nowhere exhibits an increase in amplitude anywhere along the entire Makran margin is
with depth as would be expected if it were only 400 m (see, e.g. White & Klitgord 1976)
driven by a deep diapiric intrusion. Furth- whilst the next landward ridge is 1¼ km above
ermore, there is no indication that any of the the abyssal plain. It may be that the fold
elevated ridges to the north of the frontal fold reaches a peak amplitude of only 400 m be-
show any signs of the continued growth in cause the sediments involved have by then
amplitude that might be expected if they were become sufficiently consolidated by the folding
cored by diapiric clay. Considering the overall to support the thrust fault. The upturned beds
morphology of the Makran continental margin at the landward edge of some of the frontal
it is unlikely that it is generated by the type of folds may be caused by drag folding on the
down-slope overthrusting controlled by clay footwall beneath the thrust fault.
diapirism postulated by Chapman; the rather The thrust elevates the fold by a very con-
fiat 60 to 70 km wide bench terminated in the sistent 1~ km, and then movement along the
south by a 1¼ km scarp down to the frontal fold fault ceases and a new frontal fold starts to form
and the abyssal plain is atypical of deformation in the flat-lying abyssal plain sediments to the
driven by down-slope slumping, which general- south. There is thus a delicate balance between
ly dies out gradually towards the seaward edge. folding and faulting of the sediments in the
I suggest that as the undeformed abyssal frontal fold. The deformation reverts to folding
plain sediments are carried northwards they when the additional normal stress across the
undergo the following sequence of events: thrust fault imposed by the weight of the up-
lifted ridge exceeds about 2 × 107 N m -2 (i.e.
(a) First the upper part of the sediment pile 200 bars; the additional stress caused by the
is folded in the frontal fold. The layer im- displacement of 1¼ km of water by uplifted
mediately beneath the seafloor is thinned over sediment). As we shall see in the next section
the crest of the fold possibly because it is this balance, which depends on the relative
unconsolidated and thus is buckled with a smal- magnitudes of the vertical and horizontal stres-
ler amplitude than deeper lithified layers (Biot ses and on the strength of the thick pile of
1961) but perhaps mainly because lack of tur- sediments, is easily upset by basement peaks
biditic sedimentation on the elevated crest of projecting into the sedimentary section.
the fold preferentially thickens the topmost (c) Over the next 60 to 70 km towards the
layer over the adjacent abyssal plain. This also coast there is little further deformation other
causes the bathymetric expression of the folding than a small amount of additional tilting result-
to be narrower than at depth by burying the ing from minor movement on the thrust faults.
seaward edge of the fold. The observation that From the very small average bathymetric slope
only the upper 2½ km of sediment become of only 0.6 ° I infer that the subducting plate is
folded suggests that there is a d6collement zone slipping beneath the fold belt along weak hori-
about 2½ km beneath the seafloor which may lie zons within the sediments (Chapple 1978), in a
in a mud unit weakened by abnormally high similar manner to that reported from the frontal
pore pressures. Overpressured clays, which we 50 km of the Barbados accretionary wedge
suspect are present in the uppermost 700 m of (Westbrook 1975).
sediment from the evidence of the diapir on (d) To the north of this gentle slope the fold
profile B, may exist as deep as 3 km without belt becomes abruptly and rapidly uplifted by
becoming gravitationally unstable (Chapman further thrust faults. These faults probably ex-
1974); they may then assist folding by providing tend down to or even into the basement rocks.
interfacial lubrication (Biot 1961). The consis- Uplift rates on the coast are typically 1.5 to
tent asymmetry of the frontal fold may be 2.0 mm yr -1 increasing towards the east (Vita-
explained by its growth in a region where one of Finzi 1975, 1979; Page et al. 1979). Extensive
the principal axes of the stress ellipsoid is not seismicity accompanies the faulting; it was re-
vertical (Price 1967), because the stress field ported that the magnitude 8.3 earthquake of 28
368 R. S. White

9
V
d
4

. . . . ~~''.~'~''~''~':
~t~..~,"e':~'~ ...~ '........
" -_~ :.~',.,....-.,.
' ' ' ' "" -

'::..... ~'"'"::~......... ":'"':: .............. "~" " ,....... :;"~',~..~_..~i~.

" :'~,~. . ........
......... ....... ," " , : . i,,-.:. "., " ,, ',.,,>.
~ ~
< . ~,.~

• ... ' :..:~":"~"" ". - ' ' . ' " ....i'. ;. ':"~,,~.,"
~'~-":

' .' . ,..,.. '........

:,'.i'"'", :<' ":""" ' " "': % "~'~
4,,

-/. . . . . . :,,..j.",-.-,. ,.', .........

..,, ;,/,;:,.

TWO -WAY -:.:::?,/...i:'.i

.......
:.....,.'..:i".i-:~.: ":. ~!:.:....:;:.....,;;..i~:i.,i::
, 4' , ~''.'-""" ~. ", ' ...... ' " ' '" '~

. ' . .' ~ . '.....

. ,:,t.....i~.': :~,,:,~/,
TIME t .: , ....
....
~... :-.~:~.
~.": ~ i ~ ~ .,~.
(S EC S) ~~:~..~!:
,-~.~ ........ :: ...:.,...,.,,; ,
,.,.~'.
:y.%..,,' ~
+>.'.,
. , ..,i,.~'.,,,
:., ~',,
', ':'?

'.',::::,.:~.-:~.:~,:'~'~;J~. ., ,~ , . . . .

l
." ,'! ...,., ....... ;,~.........
"'15..............
.. '.....,
6" ~...::.~
......
~'""" ...... .........
..~..:..:~.., ....,.~~' ~,,,.,-',r
..... _.,,.,
....~, ,-,. .......!i~::.,,.
............ ~,:',/~,~
~ .i

: ' :~'" "I rl.......)'#:'"~; 'i :..... ~' " -;'";~")"";'~. : ".... " " ,.,,.
' ~ ,. . . . . . . . . . .... ,~'~.~
;;:;'"t ,j;::' ';...:.T" "" ''
~,,',
,,~,,.,.s~ ,I.
'""
,.
."':~'""
'..... i.;,;~,'.,~,~;~';;.;..,
~
I. ...........
i.........
:.......'..... ;:i,q,..
,,,.,.~:,
."'i",i.......
.., :~:'",,.,'" .' '....

J, " ' ' ,' I " " ' " ',~v~' "

r..:.;.. ( "" ¢~b...-"'~i'""~

~!'°h!''"::'~':''''
°.' '-'t',,,t,'"f~.
,-,~,...,,:~,~:.;..;,;:::~.-,,,~'o.
i'' '~',~:,i',,~,

. ~ '' - ' ' ' ' ". " " ~-"~",~' i.' '

7
i ,II,L, ,i i
i,i.~ i, I1.~1,,

;,.,.L!i' h !:|I'

0 10 20
DISTANCE (KMS)

PROFI LE B
 Deformation of sediment prism in the Gulf of Oman 369

November 1945 created a 2 m high fault scarp past 15 Ma derived by summing the thicknesses
(Page et al. 1979). The pattern of incised of formations now exposed on land (from
streams and wave-cut terraces suggests that A h m e d 1969), and 0.50 mm yr -I over the past
uplift occurs episodically along the Makran 0.02 Ma measured by radio-carbon dating a
coast. piston core in the western Gulf of O m a n abyssal
There is no evidence of any major basement plain (Stoffers & Ross 1979). These rates sug-
discontinuity beneath the abyssal plain south of gest that it takes about 0.4 Ma for the frontal
the fold belt at the longitude of profile G near fold to grow to its maximum height of 400 m. If
which the pattern of uplifted ridges changes we allow for some of the sedimentation being
within the detailed survey area. This of course aeolian rather than turbiditic, and thus depo-
does not mean that there is not a basement sited uniformly over both the fold and the
feature now buried beneath the accretionary abyssal plain, our estimate of the time taken for
fold belt which was responsible for those per- the fold to develop is proportionately
turbations in the ridge lineations. However, it is greater. If some of the reduction in thickness is
more likely that even given a thick sediment due to tectonic thinning, the time taken for the
pile of great lateral extent then individual ridges fold to develop could be less.
will be of limited length because the sediment The convergence rate of the A r a b i a n plate
section is insufficiently strong to transmit stres- beneath the Eurasian plate is about 50 mm yr -1.
ses an indefinite distance laterally. The detailed Only part of this is taken up by deformation and
survey area did not extend far enough along southward migration of the seaward edge of the
strike to follow an individual uplifted ridge over continental margin; much of the shortening
its entire length. However, an estimate of the of the sedimentary section occurs by internal
lateral extent of the frontal fold can be made deformation and thickening by imbricate
from 19 roughly equally spaced crossings over a thrusting commencing more than 60 km further
distance along strike of 80 km over which north.
interval the amplitude increases from about
10 m in the west to a maximum of 400 m and Eastward extent of the Makran fold belt
then decreases again to 240 m in the east. This
lateral extent of over 80 km gives a length to The offshore Makran fold belt was traced
width ratio for the frontal fold in excess of 8:1. eastwards towards the triple junction near
Although we cannot directly measure the Karachi between the Indian, Arabian and
rate of deformation, we can estimate it from Eurasian plates. Deformed ridge and basin
two different approaches, as below: topography is found along the entire continen-
(i) Geological mapping in the onshore Mak- tal margin, but it becomes increasingly irregular
ran (Ahmed 1969) indicates southward regres- as the Murray Ridge impinges on the M a k r a n
sion of the marine-fresh water sedimentary margin. Similarly, the simple open fold and
boundary by 250 km since the Oligocene (i.e. basin structure is perturbed in the west where
about 6 to 10 mm yr-~). If we assume that the the Makran margin is in collision with the
deformational front is moving southwards at opposing continental margin of O m a n (White
the same speed as the coastline, then a new fold & Ross 1979).
must be added to the edge of the accretionary On profile M (illustrated in Fig. 12), to the
wedge approximately once every million years. east of the detailed survey, is a large basement
(ii) Alternatively we can calculate the fold ridge only 10 km south of the fold belt. This
growth rate provided we know the sedimenta- ridge causes large gravity and magnetic anoma-
tion rate and we assume that the uppermost lies (Taylor 1968). The basement, where it can
sediment layer is thin over the frontal fold last be seen adjacent to the fold belt, is only
solely as a result of non-deposition on the about 2 -~2km beneath the seafloor, and it has
developing fold. The ratio of the amount by already caused the pattern of ridges and basins
which the uppermost layer is thinned to the to be much less simple than within the detailed
amplitude of the fold is remarkably constant at survey area (Figs 3, 4).
0.5 for all profiles across the frontal fold (Fig. As the basement ridge passes obliquely be-
10 and White 1977a). Our best estimates of the neath the Makran fold belt in the east of the
sedimentation rate are 0.54 mm yr -~ over the Gulf of O m a n it generates chaotic seafloor

Fro. 11. Continuous seismic reflection profile across frontal fold on line B (see Fig. 2 for location).
Note the small inferred shale diapir marked by 'd' cutting through the upper sediments. The shale
diapir is also seen on profile 9 (Fig. 9), which intersects at the position marked by an arrowhead.
Vertical exaggeration at seafloor is approximately 7:1.
370 R. S. White
0-

::::: :::===================== :,:

, =======
======================
=======
:::::::::::
::::::: ::::. ::::::::::::::::::::::::::::::::::::::
:a<~,.:~ ,:~:~,:~:~,:,~:~-::':~.~+::~;<%:..:::a::

2~," 10'
2/,"50 LATI TUDE NORTH
PROFILE M
i

10 km

Fro. 12. Continuous seismic reflection profile along line M approximately perpendicular to the strike
of the fold belt (see Fig. 1 for location). Vertical exaggeration at seafloor is approximately 7:1. Note
the prominent offshore basement peak near 24° 12' N.

%
¸:" i •
. ,
: I

:.;:~:. . . . . . . . . . . •
..:..::.. ..: .,. . ~... .. .. .
. .z . :~.:.3
) : : ~]]i

:: :..~:;,.,u,~ :::' ::: ::i.. ," ;.-..: ,: :3:: ~.:::~,:.

2/,," 50' LATITUDE NORTH 2/*°30'

PROFILE 0

Fro. 13. Continuous seismic reflection profile along line O near the eastern end of the fold belt (see
Fig. 1 for location). Vertical exaggeration at seafloor is approximately 7:1. The accretionary fold belt
has now consumed the basement ridge seen in Fig. 12, generating chaotic seafloor relief.

relief (e.g. profile O, Fig. 13). T h e sediments o p e n folding and thrusting of the thick s e d i m e n t
are still being d e f o r m e d and i n c o r p o r a t e d in the pile (stages 'a' t h r o u g h 'c' of the previous
accretionary w e d g e by thrust faults, but the section), have b e e n bypassed and the shallow
slope of the continental margin is here much b a s e m e n t causes i m m e d i a t e rapid uplift and
steeper than to the west and t h e faulting ex- imbricate thrusting such as is seen near the
tends into the b a s e m e n t . T h e initial stages of coast in the detailed survey area (stage 'd').
 D e f o r m a t i o n o f s e d i m e n t p r i s m in the G u l f o f O m a n 371

Conclusions achieved by large-scale imbricate thrust faulting

extending to the basement and accompanied by
The systematic, well lineated pattern of uplifted major seismicity.
folds and intervening basins at the seaward The simple and relatively gentle deformation
margin of the offshore Makran is the result of observed over the seaward part of the accre-
scraping a very thick pile of sediment off a tionary wedge will not be preserved in the
shallowly dipping oceanic plate at a convergent geological record because in the process of
margin. The oceanic crust of the subducting e m p l a c e m e n t above sea-level the sediment be-
plate is Cretaceous in age and the great width of comes further faulted and stacked up in thrust
the accretionary prism extending across the slices in a similar m a n n e r to that reported from
continental margin and into the Makran of many other subduction zones. The occurrence
Pakistan and Iran is the result of continued of large seamounts or basement ridges on the
subduction from the Late Cretaceous or early subducting plate also disrupts the o p e n fold
Tertiary until the present. However, the pat- ridge and basin structure, as is seen in the
tern of plate interactions in this region has not eastern part of the Makran continental margin.
always been as simple as it is now, as is attested
to by the extensive Late Cretaceous ophiolites
and m61anges emplaced on the adjacent con- ACKNOWLEDGMENTS: Many people have contributed
tinental margins of O m a n , Iran and Pakistan to this work by generously giving their time and their
expertise both at sea and in Cambridge. Although
and by the large strike-slip faults east of the Jaz there are too many to list by name, I am indebted to
Murian and Lut depressions. Nevertheless them all and wish to acknowledge their assistance. In
there are no m a j o r offsets in the present sea- particular, I am grateful to Dr K. E. Louden who
ward margin of the accretionary wedge. was Chief Scientist on RRS Shackleton leg 1/80,
Material from the flat-lying abyssal plain is to Dr D. H. Matthews who initiated my study of the
added to the seaward edge of the accretionary Gulf of Oman and to Drs G. C. P. King and D. P.
belt by the upper 2½ km first becoming folded McKenzie, amongst many others, for their comments
above a d6collement zone within the on the work. I thank the captain, officers and crew of
sedimentary section and then elevated by a the RRS Shackleton on leg 1/80 for navigating
us through blissfully calm seas. The Natural Environ-
thrust fault which flattens out within the sedi- ment Research Council funded this project and pro-
ments. There is no further major deformation vided personal support through a Postdoctoral Fel-
of the elevated ridges until some 60 to 70 km to lowship.
the landward when they are rapidly uplifted Department of Earth Sciences, Cambridge Uni-
until they are raised above sea-level; this is versity, Contribution No. ES 59.

References
ABDEL-GAWAD, M. 1971. Wrench movements in Assoc. Petrol Geol. 29, 319-27.
Baluchistan Arc and Relation to Himalayan- CHAPMAN, R. E. 1974. Clay diapirism and overthrust
Indian ocean tectonics. Bull. geol. Soc. Am. 82, faulting. Bull. geol. Soc. Am. 95, 1597-602.
1235. CHAPPLE, W. M. 1978. Thin-skinned fold-and-thrust
AHMED, S. S. 1969. Tertiary geology of part of south belts. Bull. geol. Soc. Am. 89, 1189-98.
Makran, Baluchistan, West Pakistan. Bull. Am. CLOSS, H., BUNGENSTOCK, H. & HINZ, K. 1969.
Assoc. Petrol. Geol. 53, 1480-99. Ergebnisse seismischer Untersuchungen im
BARKER, P. F. 1966. A reconnaisance survey of the nordlichen Arabischen Meer ein Beitrag zur
Murray Ridge. Philos. Trans. R. Soc. London, Internationalen Indischen Ozean Expedition.
259A, 187-97. Meteor forsch. 2C, 28.
BERBERIAN, M. 1976. Contribution to the seismo- DYKSTRA, J. D. & BIRNIE, R. W. 1979. Segmentation
tectonics of Iran (Part II). Rep. Geol. Surv. Iran, of the Quaternary subduction zone under the
39, 516. Baluchistan region of Pakistan and Iran. In:
BLOT, M. A. 1961. Theory of folding of stratified FARAH, A. & DEJONG, K. A. (eds). Geodyna-
viscoelastic media and its implications in tecto- mics of Pakistan, 319-23. Geol. Surv. Pakistan,
nics and orogenesis. Bull. geol. Soc. Am. 72, Quetta.
1595-620. FARHOUDI, G. • KARIG, D. E. 1977. Makran of Iran
BUFFEER, R. T., SHAUB, F. J., WATKINS, J. S. & and Pakistan as an active arc system. Geology, 5,
WORZEL, J. L. 1979. Anatomy of the Mexican 664-8.
Ridges, southwestern Gulf of Mexico. In: HUDSON, R. G. S., McGUGAN, A. & MORTON, D. M.
WATKINS, J. S., MONTADERT, L. & DICKERSON, 1954. The structure of the Jebel Hagab area,
P. W. (eds). Geological and Geophysical Inves- Trucial Oman. Q. J. geol. Soc. London, 60,
tigations of Continental Margins. Mem. Am. 121-57.
372 R. S. W h i t e

HUNTING SURVEYCORPORATION 1960. Reconnaissance Mar. Geol. 3, 411-27.

Geology of Part of West Pakistan. Maracle Press, STOFFERS, P. & Ross, D. A. 1979. Late Pleistocene
Toronto. and Holocene sedimentation in the Persian
HUTCHISON, I., LOUDEN, K. E., WHITE R. S. & VON Gulf--Gulf of Oman. Sediment. Geol. 23, 181-
HERZEN, R. 1981. Heat flow and age of the Gulf 208.
of Oman. Earth planet. Sci. Lett., in press. STOLE, R. D. t~ BRYAN, G. M. 1979. Physical prop-
JACOB, K. H & QUITrMEYER, R. C. 1979. The Makran erties of sediments containing gas hydrates. J.
region of Pakistan and Iran: trench-arc gap with geophys. Res. 84, 1629-34.
active plate subduction. Geodynamics of Pakis- TAYLOR, P. T. 1968. Interpretation of the north
tan, 305-18. Geol. Surv. Pakistan, Quetta. Arabian Sea aeromagnetic survey. Earth. planet.
KAZMI, A. H. 1979. Preliminary Seismotectonic Map Sci. Lett. 4, 232-6.
of Pakistan 1:2,000,000. Map Set. Geol. Surv. VITA-FINzI, C. 1975. Quaternary deposits in the
Pakistan, Quetta. Iranian Makran. Geog. J. 141,415-20.
KING, G. C. P. 1978. Geological faults: fracture, 1979. Contributions to the Quaternary geology
creep and strain. Philos. Trans. R. Soc. London, of southern Iran. Rep. Geol. Miner. Surv. Iran,
288A, 197-212. 47.
NowRoozl, A. A. 1976. Seismotectonic provinces of WESTBROOK, G. K. 1975. The structure of the crust
lran. Bull. seism. Soc. Amer. 66, 1249-76. and upper mantle in the region of Barbados and
PAGE, W. D., ALT, J. N., CLUEF, L. S. & PI.AFKER, G. the Lesser Antilles. Geophys. J. R. astron. Soc.
1979. Evidence for the recurrence of large- 43, 201-42.
magnitude earthquakes along the Makran coast WtnTE, R. S. 1977a. Recent fold development in the
of Iran and Pakistan. Tectonophysics, 52, 53.3- Gulf of Oman. Earth planet. Sci. Lett. 36, 85-91.
47. 1977b. Seismic bright spots in the Gulf of
PARSONS, B. & SCLATER,J. G. 1977. An analysis of Oman. Earth planet. Sci. Lett. 37, 29-37.
the variation of ocean floor bathymetry and heat 1979a. Gas hydrate layers trapping free gas in
flow with age. J. geophys. Res. 82, 803-27. the Gulf of Oman. Earth planet. Sci. Lett. 42,
PRICE, N. J. 1967. The initiation and development of 114-20.
asymmetrical buckle folds in non- 1979b. Deformation of the Makran continental
metamorphosed competent sandstones. Tecto- margin. In: FARAn, A. & DEJONG, K. A. (eds).
nophysics, 4, 173-201. Geodynamics of Pakistan, 295-3{14. Geol. Surv.
SNEAD, R. 1964. Active mud volcanoes of Baluchis- Pakistan, Quetta.
tan, West Pakistan. Geog. Rev. 54, 545-60. & KLITGORD, K. D. 1976. Sediment deforma-
SONDHI, V. P. 1947. The Makran earthquake, 28th tion and plate tectonics in the Gulf of Oman.
November, 1945. The birth of new islands. Indi- Earth planet. Sci. Lett. 32, 199-209.
an Miner. 1, 146--54. & Ross, D. A. 1979. Tectonics of the western
STEWART, R. A. PILKEY, O. H. & NELSON, B. W. Gulf of Oman. J. geophys. Res. 84, 3479-89.
1965. Sediments of the northern Arabian Sea.

ROBERT S. WHITE, Bullard Laboratories, Department of Earth Sciences, Cam-

bridge University, Madingley Rise, Madingley Road, Cambridge CB3 0EZ,
England.
 The Late Cretaceous-Cenozoic history of western Baluchistan
Pakistan the northern margin of the Makran subduction
complex

R. S. Arthurton, A. Farah & W. Ahmed

SUMMARY: The study area hes some 400 km N of the Makran coast, and extends
westwards for some 450 km from the Chaman Fault, a major sinistral transform fracture
zone which marks the suture between the Eurasian and Indian plates. It lies to the north of
the Makran ranges of Tertiary flysch, and flanks the southern margin of a stable, aseismic
area known as the Dasht-i-Margo block.
The Makran region as a whole has been interpreted as an accretionary prism of Late
Cretaceous to Holocene age, resulting from the northward subduction of oceanic
lithosphere. The geological record of the study area is consistent with such an interpreta-
tion, for the period Late Cretaceous-Palaeocene at least.
Rapid northward subduction during the Late Cretaceous gave rise to profuse andesitic
volcanism on the sub-parallel, ENE-trending arc massifs of the Chagai Hills and Ras Koh
geanticlines. The waning of this volcanism during the Maestrichtian coincided with the
accumulation and deformation of flysch sediments in a trench which extended from
Mirjawa to the south of Ras Koh; also with thick forearc sedimentation in the Dalbandin
Trough, between the geanticlines.
In late Palaeocene times, deformed flysch--interpreted as a subduction complex--was
elevated to form a structural high in the Mirjawa range. In the western Ras Koh, the
flysch, along with ultramafic slices, became emplaced against the arc massif to form a new
landmass. This emplacement may have been a result of a late Palaeocene collision
between the northward-migrating Indian continental plate and the continental lithosphere
of the Dasht-i-Margo block, the Ras Koh flysch belt taking the brunt of the initial impact.
The principal forearc faults have regular, arcuate traces which, in the south-west,
parallel the structural grain of the Mirjawa flysch belt. Movement on the faults influenced
sedimentation in the forearc, particularly during the Eocene and Miocene. There is
evidence of tensional transform movement in Eocene times. The forearc faults were
especially active during the Pliocene. Movement across them during this period was
compressional and associated with major folding, particularly in the Dalbandin Trough.
The trend of this folding may imply dextral transform displacement within the trough.

This paper is concerned with the Late Cre- including the Ras Koh geanticline, and the
taceous and Cenozoic evolution of an area that North Chagai Arch (here referred to as the
is widely considered to include the contempor- Chagai Hills geanticline) (Fig. 1), the latter
ary volcanic arc and forearc components of a being interpreted as an area of stable shelf
consuming plate margin (Farhoudi & Karig underlain by continental basement. B e t w e e n
1977; Sillitoe 1978; Jacob & Quittmeyer 1979). these two structural highs he recognized a ma-
The study area (Fig. 1) extends from the jor depression, filled with Cenozoic marine and
Afghan frontier southwards to the Makran, and continental sediments, the M i r j a w a - D a l b a n d i n
from the Iranian frontier in the west of the Trough (Fig. 1). The Makran region to the
Chaman Fault in the east. The Chaman Fault is south was regarded as lying within the Baluchis-
one of the principal fractures in a zone of tan geosyncline.
left-lateral transcurrent displacement according A u d e n (1974) included the area within his
to A u d e n (1974) active throughout the Terti- Dasht-i-Margo zone, the northern part of which
ary, and according to Sillitoe (1978) initiated in is a poorly exposed, aseismic region t e r m e d the
or before early Miocene times. 'Dasht-i-Margo Block'. This 'Block' has been
Present knowledge of the area is due largely referred to as the 'Afghan Block' (Jacob &
to Jones (1960), following the pioneer work of Quittmeyer 1979); and forms part of the Ceno-
V r e d e n b u r g (1901). Jones established the pre- zoic Iran-Afghanistan microcontinent of Silli-
sent lithostratigraphy, though his mapping was toe (1978). A u d e n t e r m e d the area of the
modified by Bakr & Jackson (1964). H e also present study the 'Chagai Eruptive Z o n e ' in
identified four principal structural components recognition of the recurrent volcanism that has

373
374 R. S. Arthurton, A. Farah & W. A h m e d

,J

Ge'an
Hamun-i-
Koh-i OC h i I L°ra
Sultan'

AR NK
) o

Rec i ",~ "-~',K~ p,

'l J _ " ~ .
~ pies,, F--:::~¢~. rA ......... :[..'"
Illllllllll ~ol~. - F:::::-:~ se~
Hamun : I- Mashkhel ~ ~ -J -J
-,~ 28
~sed u ~ flysch
/ " :.|Plioc. ¢1 ~ I U' bosics
4 . ., sed. ~
~ t:(:)':.:.'i::::'::t M' . . . . . ~ L o t e Cret ;'~-
,-'~sed ' v ~ plut. " /
61~M~oc, 111~:.~. . . ::::-~:::lLate Cret.
intr ~.il vote. I

50 100

Fro. 1. Geological map of study area showing the principal structural components. (1) Recent and
undivided; (2) Pleistocene volcanics; (3) Pleistocene sediments; (4) Pliocene sediments: (5) Miocene
sediments; (6) Miocene intrusives; (7) Palaeogene forearc sediments; (8) Maestrichtian-Palaeogene
flysch; (9) ultrabasic igneous rocks; (10) Late Cretaceous-Palaeogene plutonic igneous rocks; (11)
Late Cretaceous volcanic rocks. AR Alam Reg; Ch Chagai; Da Dalbandin; Mi Mirjawa: NK Nok
Kundi; Nu Nushki; Sh Shor Nalla; Sk Saindak; YM Yak Mach.
Sources: Jones (1960), Bakr & Jackson (1964), and recent geophysical and geological fieldwork by or
for the Geological Survey of Pakistan. Limits of Dalbandin Trough (heavy dots) based on
aeromagnetic data. Light dots show limits of present topographical depressions.

affected particularly the Chagai Hills and north- the volcanic arc associated with this subduction
eastern Ras Koh from Late Cretaceous to (Farhoudi & Karig 1977; Jacob & Quittmeyer
Quaternary times (Jones 1960). He placed the 1979).
Makran ranges within his 'Tertiary Flysch Sillitoe's (1978) suggestion that left-lateral
Trough', but included the Mirjawa and Ras transcurrent faulting affected the southern mar-
Koh ranges, as well as the present topographic gin of the Chagai belt (Ras Koh-Dalbandin
depression of Hamun-i-Mashkhel (Fig. 1), area) during the late Cenozoic, perhaps causing
within the Chagai Eruptive Zone to the north. eastward translation of the Ras Koh, has been
More recently, the entire Makran deformed developed by Lawrence et al. (1981), who have
belt and the contemporary depressions to the linked such movement to left-lateral displace-
north, including the Hamun-i-Mashkhel, were ment on the Chaman Fault.
interpreted by Farhoudi & Karig (1977) as an In the following account, the stratigraphy and
accretionary prism of Late Cretaceous or early structural history of the area as determined by
Tertiary to Holocene age. This accretion results Jones (1960) is revised incorporating data col-
from the northward subduction of oceanic lected during detailed surveys by field parties of
lithosphere (Stoneley 1974). White (1979, 1981) the Geological Survey of Pakistan. The princi-
demonstrates that flysch sediments continue to pal structural features of the area are reassessed
be added to the accretionary prism today, with and recent accounts of the geodynamic evolu-
undeformed abyssal plain sediments becoming tion of this part of Pakistan are discussed in the
abruptly involved in frontal fold deformation light of the new data. Particular attention is
some 150 km south of the Makran coast. The focused on the position and behaviour of the
chain of Quaternary volcanoes stretching from presumed margin of continental lithosphere
south-central Iran through the study area into during the Late Cretaceous and Palaeogene, and
southern Afghanistan has been interpreted as on the structural behaviour of the forearc zone
 Late Cretaceous-Cenozoic history o f western Baluchistan 375

throughout the Cenozoic. Lithostratigraphic which is well-defined both by aeromagnetic

terms and formation ages are those used by data (A. Farah et al., in preparation) and
Jones (1960) unless otherwise stated. geological evidence. It extends from a point
about 100 km E N E of Dalbandin to the north-
ern part of Hamun-i-Mashkhel. The Dalbandin
The structural components Trough may be regarded as a forearc basin
(sensu Seely et al. 1974) in relation to the arc
The North Chagai Arch (Jones 1960) we re- massif of the Chagai Hills geanticline. A further
gard as having geanticlinal status, and this forearc basin ( S a i n d a k - A l a m Reg) (Fig. 2) abuts
structural feature is here termed the 'Chagai against the south-western end of the Chagai Hills
Hills geanticline' (Figs 1 & 2). It has a history of geanticline.

CHAGAI HILLS GEANTICLINE ~E

SAINOAK-ALAM REG BASIN "}Ca

1 Pleist voIc

........... . --:::: ;f:::

+ + + + + + ÷ + + ~..,.~ . . . . . . . . . . . . . . . . . . . . . . . . . ~ . ~ v ~
yz
III
2 Neog cont.

3 E. Mioc. intr
__ ___
--_-_--- 40lig. sh mar/costal

5 EOC. FA seds and

• , , ~ olistoliths
' • ' . 6 EOC sh ImsL
215 5 0 km
I .~:2~V. " 7 Poleoc FA seds

8 Maes Paleoc FA
flysch and u'bs.
~+++4 9Plutons: late Cret, P'g

10Late Cret vole

.~ R A S KOH
CHAGAI HILLS GEANTICLINE O~ FLYSCH BELT

, BE

FIG. 2. Generalized horizontal sections from the Chagai Hills geanticline (a) across the Saindak-
Alam Reg basin and Mirjawa flysch belt, and (b) across the Daibandin Trough and Ras Koh flysch
belt. (1) Pleistocene volcanics; (2) Neogene continental sediments; (3) early Miocene intrusions; (4)
Oligocene shallow marine/coastal plain sediments; (5) Eocene forearc sediments with olistoliths; (6)
Eocene shallow-water limestones; (7) Palaeocene forearc sediments, including limestones; (8)
Maestrichtian/Palaeocene deformed flysch, with tectonically emplaced ultrabasics; (9) Plutons of
Late Cretaceous/Palaeogene age; (10) Late Cretaceous volcanics.

recurrent volcanism and plutonism. In the con- The Ras Koh geanticline (Jones 1960) (Figs 1
text of the arc-trench system proposed by & 2) is well-defined from near the C h a m a n
Farhoudi & Karig (1977), this component may Fault in the east to a point some 40 km E of
be classified as the continental margin 'arc Dalbandin where its surface expression is
massif' (Dickinson & Seely 1979). The geantic- abruptly terminated. Aeromagnetic evidence
line is clearly defined along the length of the (A. Farah et al., in preparation) suggests that a
Chagai Hills to a point some 50 km NE of Nok comparable structure may extend from the
Kundi, where the structure plunges to the west. western end of the Ras Koh range to southern
Its northern limb re-emerges from Quaternary Hamun-i-Mashkhel. The Ras Koh geanticline is
cover to the west of Koh-i-Sultan, but the intruded by plutons, though these are of minor
structure terminates abruptly south-westwards extent compared to those of the Chagai Hills
some 30 km W N W of Nok Kundi. geanticline (Auden 1974). There is evidence of
The Mirjawa-Dalbandin Trough (Jones considerable Cretaceous volcanicity.
1960) is here renamed the 'Dalbandin Trough' Two new structural units are defined. The
(Figs 1 & 2). Cenozoic sediments several south-western part of the Mirjawa range is
kilometres thick fill this structural depression, termed the 'Mirjawa flysch belt', and the west-
376 R. S. A r t h u r t o n , A . Farah & W. A h m e d

ern part of the Ras Koh range, the 'Ras Koh glomerate suggest that much of the eruption
flysch belt'. These belts may be linked under was subaerial. No eruptive centres are known,
upper Cenozoic sediments to the south of though evidence of these may have been des-
Hamun-i-Mashkhel. Both are interpreted as troyed by the intrusion of batholiths. The base-
belonging to the 'subduction complex' (Dickin- ment to these volcanics is nowhere exposed.
son & Seely 1979) of the supposed arc-trench The thickest known sequence is that of the
system. Sinjrani Volcanic Group near Alam Reg in the
western part of the Chagai Hills geanticline
Geological history (Figs 1 & 3a), where at least 10 km of andesitic
The interrelationship of formations and the lavas, tufts and volcanic conglomerates underlie
ages of the principal magmatic and tectonic the Maestrichtian Humai Formation (Arthur-
events are summarized in Table 1. A ton et al. 1979). Similar components make up
palaeogeographical history is suggested in the Sinjrani Volcanic Group at the eastern end
Fig. 3. of the Chagai Hills near Chagai village, where a

TABLE 1. Ages and interrelations of lithostratigraphic units'; dates of igneous and tectonic events within
the study area, and regional geodynarnic events. Asterisk indicates proposed new forrnation name

A(;E LITI IOSTRATI(;RAPItlC UNITS R I-;(;ION A L

m a. WEST CENTRAI./EAST IGNEOUS E V E N I'S IECI'ONIC F.V}.N I'S (;EOI)YNAMIC EVFNTS
PLEIS'rOCENE " "........ ~ : : : >st~:~.;,~, ~ i :7:: . . . . . .... ': ~K',; 7 " g u / i a n v , 7 i c ' a , q i m Re,.eq.e fat h i n g . . . . .
_ am ero ° r , ~ ~4,.19n . . . . . . . . . =:.::=..: : ....... :..: . . . . , ,. ., E Xs~n r .n~J,Van ,,r,,8¢nV U o n e | ,'-'ou)
* [ ) a l ' [ C + : ' n" g ( m F n .*. Y
. . .a.k. .M. . a c :h C o l: g
~rn ...... . . . '. . . . r('vt'rse 1~ u l t m ~ 2 r so cs! i,E -
i ~ d d m l l t l ~ a x l n P a k t m l II
~Jloldbeh
["n~':! .; I ' , , h h n g & .'~u d e n 1 9 7 4

::7 ::t:{:7 ;]7;77r{7{ 7: "W" ~ 7 fh

. . . . . . ~, ~.msml~'., s v we 9~'9
....... = . . . : . ....... ..- R ¢g f ~ ....

~:.;: )~.7,,r.
40 ::J::: : ....... ; | " " "
I I t ? :!2::: :. ;:. ]::: i~:: ;! t{ . . . . . {[c t " t t ........ ti, i l p l i l t . . I .................. ~lls~l
I I . [:: :: ::::: :. 7::7::::7: ' I' I ,,i },hrja~:t , . . . . . . . ii , i, . . . . Re (P we 929
IWa~hal'l Saandak " " v ,camsm l • J4
(Sail~dak I . m :: ,, .,t,, m , , n It~.

~ I _ lI . . . . ~ ) b~ : ,,, ,:, ':- ,~, , , ( .....................................

4_{ ~ ~/- ~ ~ ..............................................
{Pow¢l] 1979)
,,,,,,,, . . . . . . . h w a r d Iligh . . . . ly h a l t e d
I~:: ::I ..... 1 " v " : : : : ' ~ ' .... ' .............. II~ ~ F ~ , n , fXlir .;;.,'7~g~rTd" ;d'i~-z
[ ~ / : : :)] i [ " ')::!'. : I] ~ ~ ],-h.,'.llh,ll,,f R . ~ K,d~ s , i b d u ~ i i o n | ~ % h , ~ i n g , f f l l i d i = l , ~ e a n s e a n o l , r ipreadlng
[':- v<{: :5-:1 / ~[..[~ -:'7:;:5:;: ['] ~ ~ / t ,,lllpit'x] .,~ ~ Mc Kcnzie i Sclatei 1971
60 i}7 Jui'~-za I'm ::: aa E ~,, tallis : - -~ ~-
PALE()CENE / 1 i::t ~ m / t i ~ : l c" ~ . . -z

....... a. [(:
:: u m a i ]""m I.~7~:;{4
:. - ~-- ["::::(: . =~-"~
= = I ~ ~a ~~!~
laf rnr ang r i n ¢ ~ Iron rough S l ~ k i n e t a l 972'
dAESTRI

7:i'7{7:7)i!:.:. r.~c.],~kl, i {!.:~'-:i!:{;717! ~l~j,>. "~ ~= i::;.777 i.. L: ::);:;-: .::: ::, {i.i~{ . 'h>~s'i~'~';<~'~',;'~'ffh
LATE : 4:;:~::: Sinjrani v<)lcanlc !::7 !7~77~:2{;: amlesilic ~ - : :::: ": :T :T :: 7".:" ::':': ;~ "
gRETACEOUS {:;{ :; "i: Volcanic Gp :::.:::.:.{!::{;:{~: v,~ c a n i s m o n :{: :{ |::: :LI'{(::
[pre-Maestnchtiatn) ":::: ~ : :::': Gn '::::'::F :5:"': R:,~ K h : : : L : : N o r l h w ~ d movemen| of I n d i ~ plate
80 :~i :5 5 ~: "~ .:~ :..:~..:< .._ ~.o;,am . : :;7; : : :: (P ~ 979)
~7¢:i:7i~{!i!!:.:t~;:~ ::::::::::::::::::::::: c~,,~ ...... ~ ::::...: :: :. ::: ° °

Late Cretaceous
Andesitic volcanism dominated the area dur- thickness of about 4 km has been estimated. In
ing the Late Cretaceous. Andesite lavas, tufts the north-eastern Ras Koh range, the broadly
and volcanic conglomerates several kilometres equivalent Kuchakki Volcanic Group is esti-
thick accumulated mostly in marine environ- mated to be at least 6 km thick though, accord-
ments as indicated by intercalations of marine ing to Jones (1960), the proportion of lavas is
limestones and conglomerate, the latter includ- less than that in the Sinjrani volcanics. Again,
ing boulders of 'Hippuritic Limestone' no eruptive centres are recorded.
(Vredenburg 1901). However, local red sedi- Jones (1960) suggested that the Sinjrani and
ments and extensive beds of volcanic con- Kuchakki volcanics might be continuous under
 Late Cretaceous-Cenozoic history o f western Baluchistan 377

the Dalbandin Trough but this suggestion is not range, and tuff breccias mark the top of the
supported by the aeromagnetic evidence (A. formation near Alam Reg. The broadly equiva-
Farah et al., in preparation). Their southern lent Rakhshani Formation of the Dalbandin
extension is also speculative. Small inliers of the area is similar to the Juzzak in its lower part,
volcanics were mapped in the northern Mirjawa but the presence in the upper part of the partly
range (Jones 1960); a further possible outcrop reefal Chapper Limestone, the oncolitic Gat-i-
occurs at the extreme western end of the Ras Hamun Limestone, and cross-bedded sand-
Koh range. In the Makran, there are no records stones indicates that shallow-water conditions
of Cretaceous volcanics in situ. were temporarily established. Limestone con-
In the Chagai Hills geanticline, the Late Cre- glomerate lenses and olistoliths occur north of
taceous volcanism was associated with the wide- Yak Mach, whereas coarse conglomerates in-
spread intrusion of granite and granodiorite cluding debris of intrusives comparable to those
batholiths. This plutonism must have been of the Chagai Hills, limestones of 'Humai' type,
accompanied at least locally by substantial up- and quartzites of more distant derivation occur
heaval and erosion. Pebbles of these intrusives within the Palaeocene sediments north of Dal-
have been recorded (Jones 1960) in the basal bandin. The conglomerates within these trough
conglomerate of the overlying Humai Forma- sediments point to the continuing uplift and
tion. erosion of the Chagai Hills geanticline through-
The Humai Formation, which overlies the out the Palaeocene. The quartzites may have
Sinjrani Volcanic Group around much of the been transported from far to the north, around
periphery of the Chagai Hills geanticline, re- the eastern end of the geanticline.
flects an abrupt reduction in the intensity of The plutonism in the Chagai Hills geanti-
volcanism. It includes shallow-water limestones cline, which was associated with the active
(Humai Limestone). Underlying these lime- volcanism of the Late Cretaceous, recurred or
stones in peripheral sections north of Alam perhaps continued well into the Palaeogene.
Reg are several hundred metres of turbi- Sediments of the Juzzak Formation on the
dites, including limestone conglomerates. By north-west flank of the geanticline are them-
contrast, an outlier of Humai Limestone at Sor selves intruded by diorite, followed in places by
Koh, nearer the crest of the structure, rests grandodiorite and granite. Palaeogene pluton-
directly on the Sinjrani volcanics. This lateral ism occurred also in the Ras Koh geanticline.
variation in the sequence is interpreted as in- There, according to Jones (1960), syenite and
dicative of uplift of the geanticline, resulting in diorite intrusions cut the Kuchakki volcanics
widespread shoaling and the erosion of earlier- and the Rakhshani Formation, and were thus
formed limestones from the crest of the struc- dated as post-Palaeocene. Evidence for a post-
ture. Middle Eocene age for the intrusions as claimed
In the Ras Koh range, the limestone-bearing by Bakr (1958) was not confirmed by Jones
upper part of the Kuchakki Volcanic Group (1960), who found no indication that the intru-
was regarded by Jones (1960) as equivalent to sives cut Eocene sediments. The Ras Koh In-
the Humai Formation (Table 1). Although trusions are clearly defined as positive anom-
Jones recorded no indication of a possible alies on the regional aeromagnetic map (A.
shallow-water origin for these limestones, it is Farah et al., in preparation). A zone of similar
probable that both the Ras Koh geanticline and anomalies extending WSW from western Ras
the Dalbandin Trough to the north were estab- Koh towards Hamun-i-Mashkhel may mark an
lished structural features by this time (Fig. 3b). extension of the intrusive belt and perhaps also
the Ras Koh geanticline under thin upper
Palaeocene Cenozoic cover in this tract (Fig. 3b).
Palaeocene sediments were shown by Bakr &
During the Palaeocene, sediments several Jackson (1964) to comprise a discrete structural
kilometres thick and largely free from volcanics unit which forms the western part of the Ras
accumulated in the Dalbandin Trough and Koh range and is here termed the Ras Koh
Saindak-Alam Reg basin (Figs :2 & 3b). The flysch belt (Fig. 1). In the same area, Jones
Juzzak Formation in the west comprises inter- (1960) had previously mapped outcrops of the
bedded mudstones and turbiditic, volcaniclastic Rakhshani Formation and Pishi Group, consi-
sandstones with some thin laminated lime- dered to be similar in part to the Oligocene
stones. It includes lenses of conglomerate with flysch of northern Makran. The Ras Koh flysch
limestone debris including olistoliths (Fig. 5), belt has an arcuate strike and comprises a
and volcanic boulders. Andesite lavas occur cleaved and highly folded and thrust succession
locally within the sequence in the Mirjawa (Ahmed 1961; Bakr 1963, 1965).
 "szoue31o~, ~le3!pu! saelS
q~noaJ~ u!pueqleCl jo sl!tu!i (9) '.se~.te leUO!SOaZpue p~le!lu~J~Jj!pun (~) tslu~tu!p~s lelU~U!ZUO3/U!eld
elseo3 (17) '.s~uols;tU!l a~le~a-~Olleqs (£) '.3ae~aoj jo s3!uu31oa/slu~tu!p~s ; u u e t u (~) tslu~ua!p~s
13S~(ld (1) "s~tu!l zu~3ols!~Id ol sno~3elza3 ~luq tuoaj e~ae ,(pnls jo ~(aols!q le3!qdea~oz~ozele d "E "913
1 I ::::::::::::::::::::::::::::::::::: 3 N 3 0 019131 d
• ......::::+:a:~:.iJ:.:a::~..:+:',i.:]:.
:,,..- ..:.:. ...T:.:...::::.::.-:v.!.!!!::.h-: + ...~ , .o~
Ir;::' 1 I I I q
I I 3'N3OOlld 31V'1
! ~ ~ ' - .... .. . . . . . . . : \
r -~"'L,~'.~.~.....~ ........... ~"~'.-\ % \
I 1 l 1 l
I I I I
'=1N 3 ::)0 IIAI
/ 31V1 - Ol IAI
\
:.~. ,,~ - ..:..::
.;,,'-'..:-';~:.-.......
,, ....,v...:O : :::: ..:.: .::.:
E t I t I "i
I A1 I:IV :1
--7 /
........ . .... ...... .,, ' I.L.i
.~ '-! .i: --.~...
-, ¢',4,.0\
_
::::.:......... ~ "I" :',.,',7.
.--I" ~ ;.....~;~.~I ~~'~'~~~.~.~.
91"~" ~p ~ ~i ~ ~p
gL£
 Late Cretaceous-Cenozoic history of western Baluchistan 379

The Ras Koh flysch belt includes the Bunap including limestone and volcanic debris, which
Intrusions, bedding-parallel lenses up to some must have been transported across contempor-
15 m wide and about 1.5 km long of sheared aneous Juzzak sediments from a source in the
ultrabasic rocks, including pyroxenite and Chagai Hills geanticline. Unlike the Ras Koh
peridotite, and commonly containing chromite Flysch, there is no record of the occurrence of
(Ahmed 1956; Jones 1960). Jones took the view ultrabasic intrusions. Jones (1960) included
that the host strata were newly deposited when these sediments in the Juzzak Formation. We
'invaded' by these ultrabasics. In the context of propose that the new term 'Mirjawa Flysch
the proposed arc-trench model (Farhoudi & Formation' be applied to this deformed flysch
Karig 1977), the ultrabasic slices may be consi- facies. Like the Ras Koh Flysch Formation, the
dered to have been emplaced tectonically into Mirjawa Flysch is believed to owe its deforma-
the flysch deposits of a subduction complex. tion largely to the penecontemporaneous incor-
The host sediments were classified as Rakhsha- poration of its sediments into an accretionary
ni Formation by Jones (1960), but these differ fold belt. The precise sedimentary and structu-
both in lithology and in tectonic style from the ral relationships of the Mirjawa Flysch and
Rakhshani of eastern Ras Koh and the Dalban- the relatively undeformed Palaeocene sedi-
din Trough. In view of the confusion over the ments of the Saindak-Alam Reg basin (Juzzak
classification of the western Ras Koh sedi- Formation) are unclear (Fig. 2).
ments, they are, as a whole, referred to here as The Mirjawa and Ras Koh Flysch Formations
the 'Ras Koh Flysch Formation'. are probably of the same general age and the
A similar deformed flysch facies of Maes- oldest exposed accretionary sediments of the
trichtian-Palaeocene age forms much of the Makran arc-trench system. A major tectonic
south-western part of the Mirjawa range (Fig. phase, perhaps in part associated with further
4). It comprises mudstones with an axial-plane accretion to the south, affected both of these
cleavage and turbiditic, volcaniclastic sand- flysch formations late in the Palaeocene. In the
stones. There are rare lenses of conglomerate Mirjawa range, the Mirjawa Flysch was uplifted

FI6.4. Cleaved and highly folded flysch sediments of the Mirjawa Flysch Formation (Maestrichtian-
Palaeocene) between Mirjawa and Saindak (see Fig. 1 for locations). View to north.
380 R. S. Arthurton, A. Farah & W. A h m e d

Fl(;. 5. Open-folded mudstones and turbidite sandstones of the Juzzak Formation (Palaeocene) near
Shor Nalla. View to south-east from olistolith of probable Maestrichtian limestone.

to form a structural high which defined the Palaeocene to early-Middle Eocene age overlap
south-western limit of the adjoining (Saindak- m61ange with pronounced unconformity
Alam Reg) forearc basin (Table 1). (St6cklin et al. 1972).
By contrast, in the Ras Koh range, deforma- The northwards thinning of the Kharan
tion of the Ras Koh Flysch against the Ras Koh Limestone was interpreted by Jones (1960) as
geanticline produced a complex thrust rela- indicating an Eocene landmass along the Ras
tionship (Figs 1, 2 & 3c). The Ras Koh flysch Koh. By contrast, the Washap limestones
belt with its arcuate fold traces and included appear to have formed as a shoal accumulation
ultrabasic slices, was emplaced against the on the Mirjawa structural high described above
rocks of the geanticline, and the combined (Fig. 3c). This structural high separated the
structures formed a new landmass along the Saindak-Alam Reg forearc basin from a deep
length of the Ras Koh range (see below). flysch basin which extended into eastern Iran,
there termed the 'Eastern Flysch Trough' by
StOcklin (1974). The existence of a further
Eocene
Eocene structural high, within the forearc
basin, is indicated by a narrow zone of shallow-
The upper age limit for the emplacement of water limestones within the largely turbiditic
the Ras Koh flysch belt is fixed by the existence Saindak Formation, flanking the Tozgi Koh
of relatively undeformed early Eocene shallow- Fault (Table 1, Figs 2 & 3). Carbonate sedi-
water limestones (Kharan Limestone) overlying ment, as well as lithified Eocene limestone
the deformed flysch of southern Ras Koh (Fig. debris including boulders and olistoliths, was
3). Comparable limestones, the Washap and introduced by gravity slide and turbidity flow to
Robat formations, were mapped by Jones the forearc basin both from the Tozgi Koh high
(1960) in the Mirjawa range, while in eastern and probably also from the Mirjawa high. It
Iran shallow-water limestones of late seems likely that lithified carbonate debris from
 Late Cretaceous-Cenozoic history of western Baluchistan 381

the Mirjawa structural high also spilled off into Oligocene time, a north-south trending flysch
the flysch trench to the south-west, there be- trough in eastern Iran (the Eastern Flysch
coming deformed in m61ange. Blocks of similar Trough) closed up, thus integrating the Iranian
limestone of Lutetian (Middle Eocene) age are and Afghan microplates. Powell (1979) has
known in the m61ange of eastern and south suggested that this convergence may have been
eastern Iran (Gansser 1960). caused by the counter-clockwise rotation of the
Two other features of the Saindak Formation Indian plate while in contact with the Afghan
are of interest in connection with the possible microcontinent, an event which he dates as late
Eocene activity of the Tozgi Koh Fault, implied Eocene (40 Ma).
by the existence of its flanking structural high.
The first is the abrupt thickness variation within Miocene
the formation adjacent to the fault; at Shor
Nalla (Fig. 1) the formation thickens along In the Makran, the incorporation of contem-
strike from 900 to 3000 m over a distance of poraneous flysch into the accretionary fold belt
5 km. The second is the volcanic component of continued during the Oligocene and Miocene,
the formation: partly tufts and thin andesite resulting in a continuing southerly migration of
flows, but, more particularly, volcanic con- the trench. According to Ahmed (1969) the
glomerates up to several hundred metres thick, Miocene shoreline was probably in the north
and well seen both at Saindak and at Shor Makran. In the present study area, however, a
Nalla. The volcanic conglomerates are locally substantial break in the sedimentary record
mixed with limestone blocks and boulders, and, occurs between the Oligocene Amalaf Forma-
like the components of the limestone debris tion and the Dalbandin Red Formation of
layers, appear to have been transported by supposed mid-late Miocene age (see below).
gravity sliding from structural highs. During this hiatus two important events are
Taken together, the features of the Saindak recorded in the west. One was the regional
Formation may be interpreted as indicating uplift of the forearc and broad folding of its
Eocene instability on the line of the Tozgi Koh sediments. The other was a volcanic episode
Fault (Fig. 3), perhaps arising from contempor- marked now by the intrusive roots (Shot Koh
aneous transform movement (see Reading Intrusions) of andesitic volcanic centres pene-
1980). The incidence of associated volcanics trating Palaeogene forearc sediments, mostly in
could point to a tensional element in such a the Saindak-Shor Nalla area close to the Tozgi
transform system. Koh Fault, but including one in the Dalbandin
Trough at Nohli Koh (Figs 1 & 3e) immediately
Oiigocene N of the Laki Koh Fault. Intrusive rocks at one
of the centres, near Saindak, have been dated
The Eocene sediments described above are at 18-22 Ma (early Miocene) by Sillitoe &
the youngest within the study area to have been Khan (1977). The intrusion of the andesitic
correlated palaeontologically. The approximate Tanki Sill of the Mirjawa range belongs to this
dating of the succeeding Cenozoic sediments phase of volcanism. This sill is notable both for
has been achieved partly by the correlation of its wide distribution, and for its lateral penetra-
tectonic events (Table 1) which have been tion from the open-folded forearc sediments of
dated elsewhere, and partly by a regional com- the Juzzak Formation to the more intensely
parison of lithofacies. Only one radiometric deformed and the cleaved Mirjawa Flysch. This
date is available, for an igneous event at relationship confirms the juxtaposition of the
Saindak in the early Miocene (Sillitoe & Khan Mirjawa Flysch and the Juzzak Formation in
1977). The chronostratigraphy of the younger early Miocene times.
part of the Cenozoic is therefore speculative. Following the main phase of Oligo-Miocene
The Amalaf Formation of the Mirjawa range erosion of the forearc, a sedimentary basin was
was tentatively assigned an Oligocene age re-established to the north of the Mirjawa and
(Jones 1960) on the basis of faunal identifica- Ras Koh ranges. Continental sediments, for
tion by Vredenburg (1901). It marks a wide- which the new name 'Dalbandin Red Forma-
spread shallowing of the forearc basin in the tion' is proposed, accumulated in the Dalban-
Saindak area (Fig. 3d). The red and green, din trough and Saindak-Alam Reg basin (Fig.
cross-bedded sandstones and mudstones were 3f). In the type section, 40 km E N E of Dalban-
interpreted by Jones (1960) as estuarine- din, red conglomerates pass up through cross-
fluviatile, marking the end of the marine condi- bedded, probably fluviatile, red and grey-green
tions. This interpretation is supported by the sandstones to red and green mudstones. The
view of Stocklin (1974) that, in late Eocene to mudstones carry thin dolomitic layers. They are
382 R. S. A r t h u r t o n , A . Farah & W. A h m e d

partly gypsiferous and of probable lacustrine rest with marked angular unconformity on
origin. The basal conglomerates in the vicinity older Cenozoic formations. These conglomer-
of the Laki Koh Fault (Fig. 3f) include much ates have an extensive outcrop immediately
coarse, locally derived material, such as blocks north of Koh-i-Dalil railway station (35 km
of Palaeocene limestone north of Yak Mach west of Alam Reg); they include mostly sub-
(Fig. 1); this is taken as an indication of angular debris, but also scattered well-rounded
contemporaneous activity on the fault. Con- pebbles more typical of the conglomerates of
glomerates higher in the sequence contain well the Dalbandin Red Formation. The Dalil con-
rounded, far-travelled clasts, and include a high glomerates are themselves folded, though con-
proportion of quartzites. The sequence is at siderably less intensely than the Dalbandin Red
least 3000 m thick against the Laki Koh Fault in Formation nearby. Folded conglomerates be-
the Dalbandin area (Figs 2 & 3f), and it is lieved to be of similar age, and for which the
probable that contemporaneous activation of name Yak Mach Conglomerate Formation is
this and similar faults controlled the accumula- proposed, are seen in the Dalbandin Trough, in
tion of the formation. an anticline immediately NE of Yak Mach. The
The extent of this continental 'red bed' sedi- stratigraphic relationship of the Yak Mach con-
mentation is poorly known. The most westerly glomerates to the adjoining Dalbandin Red
occurrence of the Dalbandin Red Formation is Formation is not clear.
in the Amalaf syncline at Shor Nalla (Fig. 1), The sediments of the eastern part of the
where it rests unconformably on the Saindak Dalbandin Trough suffered particularly as a
Formation and on Shor Koh Intrusions. The result of the two Pliocene tectonic phases.
age range of these red beds is speculative. Jones There are two manifestations of the deforma-
(1960) included them in the Kamerod Forma- tion (Figs 2 & 3g). One is folding with axial
tion of Plio-Pleistocene age (see below). They traces which swing from SW-NE west of Yak
were, however, clearly involved in the major Mach to E-W north of Dalbandin. The other is
folding and faulting episode ascribed to the major reverse faulting and associated folding
main Himalayan Orogeny, dated as late along arcuate fractures, convex to the south.
Pliocene-early Pleistocene (Jones 1960; Auden The Great Chappar and Laki Koh faults are the
1974). On lithological grounds, it seems prob- most important fractures for which there is
able that the Dalbandin Red Formation is the clear field evidence (Fig. 3). As suggested by
counterpart of the Mio-Pliocene Upper Red Jones (1960), and later by Lawrence et al.
Formation of Iran (St6cklin 1971). Similar red (1981), a similar reverse fault (Ahmad Wal
beds on the eastern marginal parts of the Lut Fault) may define the present northern bound-
block in east Iran have been assigned a prob- ary of the Ras Koh range (Fig. 2). The aero-
able Mio-Pliocene age (St6cklin et al. 1972). A magnetic and gravity data (A. Farah et al., in
mid-late Miocene age for the Dalbandin Red preparation) provide definite support for a
Formation seems likely, in view of the early major fracture on this line. Along parts of their
Miocene date for the Saindak intrusion, which traces, these faults are low-angle thrusts; for
pre-dates the red beds. example at Laki Koh, 8 km N of Yak Mach,
tightly folded Palaeocene limestone of the
Pliocene Rakhshani Formation rests upon a low-dipping
surface of vertical conglomerates of the Dal-
In the study area, Pliocene deformation bandin Red Formation. The alignment of the
occurred in two main phases. These were sepa- axial traces of the folds in the Dalbandin trough
rated by a period during which there was sub- reflects the general alignment of the trough
stantial erosion, followed, and perhaps locally itself (Fig. 3g), confined between the Chagai
accompanied, by the further accumulation of Hills geanticline and the Ras Koh massif. To
conglomerate. In the Shot Nalla area of the the west of Dalbandin, the axial traces become
Mirjawa range, the earlier phase caused reverse markedly discordant to the trace of the Laki
faulting and related folding about NNW-to- Koh Fault. Both the folding and the faulting are
WNW trending axes, parallel to the strike of indicative of horizontal shortening between the
the Mirjawa flysch belt (Fig. 3g). This earlier geanticlines. The extent to which this compress-
phase involved strata up to and including the ion may have been accompanied by trans-
Dalbandin Red Formation. In the same area, current movement on the faults is unclear. It is
the succeeding conglomerates, which were also arguable that the discordance of the fold axial
included by Jones (1960) in the Kamerod traces with the Laki Koh fault in the Dalbandin
Formation (see below) and for which the name area is best explained in terms of a dextral shear
Dalil Conglomerate Formation is proposed, system. However, NNW-trending minor folds
 Late Cretaceous-Cenozoic history o f western Baluchistan 383

which affect Dalil conglomerates immediately S in the study area, many smaller vents. Their
of the Tozgi Koh fault, 6 km S of Alam Reg distribution broadly coincides with the western
(Fig. 1), may be taken as an indication of part of the Chagai Hills geanticline. No vents
sinistral displacement across the fault. The case are known in the Mirjawa range, though there
for major sinistral movement in this part of the is a cluster just north of the Tozgi Koh Fault,
Makran convergence zone (Lawrence et al. near Alam Reg (Fig. 1).
1981) remains speculative. Further movement on the major faults within
The sediments of the succeeding Kamerod the forearc zone occurred during the Pleis-
Formation (see below), regarded by Jones tocene. Kamerod sediments have been dis-
(1960) as Plio-Pleistocene, overlie the Yak placed across the Tozgi Koh Fault at Alam Reg,
Mach and Dalil conglomerates with angular and across the Laki Koh Fault north of Yak
unconformity. Their deposition ended a period Mach. Near Yak Mach railway station, the
of widespread erosion on the forearc which Kamerod sediments have been gently warped
must have both accompanied and followed the about a pre-existing WSW-trending axis. Dur-
later of the Pliocene tectonic episodes. ing the later part of the Pleistocene, the Kamer-
od sediments were deeply dissected and, over
wide areas, completely eroded. Travertine de-
Pleistocene
posits, post-dating but related to the Koh-i-
The Kamerod sediments are buff silts of Sultan volcanism (Sillitoe 1978), occur in a
lacustrine and perhaps loessic origins, capped number of places on the Chagai Hills geanti-
by fan gravels composed mostly of subangular cline, mostly around the western end of the
volcanic debris (Fig. 3h). They are well- structure, north of Alam Reg.
preserved north of the Laki Koh Fault where
they form terraced surfaces at a number of
levels up to about 50 m above the present
Discussion
alluvial pavement. In the west near Nok Kundi
(Fig. 1), the formation is seen in the terraces The oldest exposed parts of the Makran sub-
south of the Tozgi Koh Fault (Fig. 3g), and is duction complex in Pakistan are of Maestrich-
preserved as isolated patches north of that fault tifin-Palaeocene age, in the Mirjawa and Ras
under conformable flows of andesite belonging Koh Flysch belts. The flysch deposits, which are
to the Koh-i-Sultan Volcanic Group (Table 1, highly deformed, and associated locally with
Fig. 1). ophiolitic slices, are juxtaposed against thick,
There has been confusion over the applica- broadly folded piles of probable Senonian vol-
tion of the term Kamerod Formation to the canics, intruded by plutons, in the Chagai Hills
study area. Jones (1960) included not only and Ras Koh geanticlines (Figs 1 & 2). The
sediments fitting the above Kamerod descrip- junction between the flysch belts and volcanics
tion, but also the strongly folded (Dalbandin) is taken as defining the southern limit of the
red beds of Shor Nalla, and part of the strongly Dasht-i-Margo block in Maestrichtian-
deformed 'Dalbandin Assemblage' of the Dal- Palaeocene times (Fig. 3b). It is possible that
bandin trough. Arthurton et al. (1979) followed older parts of the subduction complex underlie
Jones by classifying what is here termed the the volcanics; if so, the major faults of this part
Dalil Conglomerate Formation as Kamerod, of the forearc, such as the Great Chappar and
but applied the term 'Quaternary Older Allu- Laki Koh faults (Fig. 3), may be inherited from
vium' to the undeformed silts and gravels. fractures dating from earlier accretionary de-
These classifications we now consider to be formation of the subduction complex.
erroneous, not only because the Dalbandin red The extent to which the Late Cretaceous-
beds show little resemblance to those of the Recent magmatic and tectonic events in the
Kamerod type area, but also because they and study area were dictated by external geodyna-
the Dalil Conglomerate clearly pre-date the mic events is open to debate (Table 1). The
later Pliocene tectonic phase. collision of the northward-moving Indian sub-
The Koh-i-Sultan volcanics were erupted continent against the southern margin of
from a number of vents in the north-western Eurasia was dated as Late Cretaceous-early
part of the study area (Fig. 3h). They form part Palaeogene by Powell & Conaghan (1973),
of a Quaternary volcanic arc extending west- though initial contact in Pakistan was placed as
south-westwards to south-central Iran (Gansser late as early Eocene by Stoneley (1974). The
1971). The principal eruptive centres are Kuh-i- Ras Koh flysch belt, which flanked the con-
Tartan and Bazman in Iran, and Koh-i-Sultan tinental lithosphere of Dasht-i-Margo to the
(Fig. 1). Besides these main centres, there are, south-east, may have suffered severely as a
384 R . S. A r t h u r t o n , A . F a r a h & W. A h m e d

result of the collision. The late Palaeocene Counter to the proposal for eastward transla-
thrusting of the Ras Koh flysch against the tion of the Ras Koh during the Pliocene, the
volcanics of the Ras Koh geanticline may have supposed Pliocene structures in the Dalbandin
been caused by this collision (Figs 1, 2 & 3c). Trough indicate dextral, rather than sinistral,
The subsequent, more stable conditions in displacement across the Laki Koh Fault (Fig.
which the early E o c e n e Kharan Limestone 3). However, sinistral displacement across the
accumulated may reflect the late Palaeocene Tozgi Koh Fault is indicated by minor folds of
slowing or stopping of seafloor spreading in the N N W trend near Alam Reg (Fig. 1).
Indian Ocean (McKenzie & Sclater 1971). Whatever the pros and cons of sinistral, as
Sedimentation in the forearc basins of Sain- opposed to dextral, transcurrent m o v e m e n t on
d a k - A l a m Reg and the Dalbandin Trough the forearc faults, it is clear that these fractures
appears to have been influenced from Eocene have had a long history of activity up to recent
times, at least, by m o v e m e n t on major faults times. Their regular, arcuate trends (Figs 1 & 3)
within the forearc. Both the Eocene Saindak suggest that the faults at outcrop are the super-
Formation and the Miocene Dalbandin Red ficial manifestation of major faults within the
Formation accumulated under such influence underlying northern part of the subduction
(Fig. 3), and evidence is cited herein favouring complex, or its transition to the continental
tensional transcurrent m o v e m e n t controlling lithosphere of the Dasht-i-Margo block. The
the accumulation of the Saindak Formation. repeated activation of these faults highlights a
The possibility of major transcurrent move- fundamental structural weakness of even the
ment having occurred on the forearc faults has oldest part of the Makran accretionary prism.
been raised by Sillitoe (1978), who suggested
that the Ras Koh has been transported east-
ACKN(IWLEI)(;MENTS: Much of the new information
wards to its present position by sinistral trans- contained in this paper is the result of fieldwork by
current movement. Lawrence et al. (1981) sup- officers of the Geological Survey of Pakistan, whose
port this idea, and favour a Pliocene age for the invaluable contribution to this paper is gratefully
event, which they relate to the collision of the acknowledged. One of us (R.S.A.) is indebted to R.
Arabian and central Iranian continental slabs G. Davies and K. A. de Jong for an introduction to
(Table 1). Lawrence et al. (1981) have demons- thc study arca. The work was partly supported by the
trated a history of sinistral transcurrent move- U.S, National Science Foundation under Grant INT-
ment on faults further south in the subduction 76-22304, and has arisen in part from the engagement
of one of us (R.S.A.) by the Ministry of Overseas
complex. Sillitoe (1978) has claimed sinistral
Development, U.K. The authors acknowledge the
displacement of some 450 km on the Chaman major contribution in fieldwork by W. J. Barclay, and
Fault since early Miocene times at least; the are grateful to A. J. Reedman for discussion and
Chaman Fault forms the eastern boundary of critical reading of the manuscript. This paper is
both the Dasht-i-Margo block and the Makran published by permission of the Director, Institute of
subduction zone (Fig. 1). Geological Sciences, U.K.

References
AHMED, S. S. 1969. Tertiary geology of part of south Orogenic belts. Data for Orogenic studies. Spec.
Makran, Baluchistan, West Pakistan. Bull. Am. Publ. geol. Soc. London, 4, 235-53.
Assoc. Petrol. Geol. 53, 1480-99. BAKR, M. A. 1958. Geology of part of western Ras
AHMED, W. 1956. Geological report of Kambran- Koh Range, Chagai District, Kalat Division.
Rayo and Nag-Parshad-Charkohan Kullan, west- Pak. J. Sci. 10, 227-36.
ern and eastern Raskoh. Geol. Surv. Pakistan -- 1963. Geology of the western Ras Koh Range,
(unpubl. report). Chagai and Kharan Districts, Quetta and Kalat
-- 1961. Geology of the eastern Ras Koh range, Divisions, West Pakistan. Rec. geol. Surv. Pak.
Chagai and Kharan districts. Geol. Surv. Pakis- 10, 28 pp.
tan (unpubl. report). 1965. Structure of the Western Ras Koh Range,
ARTHURTON, R. S., SARWARALAM, G., ANISUDD1N- Chagai and Kharan Districts, Quetta and Kalat
AHMAD, S. & SAEED IQBAL, 1979. Geological Divisions, West Pakistan. Pak. J. Sci. 17, 1-17.
history of the Alamreg-Mashki Chah area, Cha- - - & JACKSON, R. 0. 1964. Geological map of
gai District, Baluchistan. In: FARAH, A. • DE Pakistan 1:2,000,000. GEOL. SPRY. PAKISTAN.
JONG, K. A, (eds). Geodynamics of Pakistan, DEWEY, J. F., PITMAN, W. C. III, RYAN,W. B. F. &
325-31. GEOL. SURV. PAKISTAN, QUETTA. BONNIN, J. 1973. Plate tectonics and the evolu-
AUDEN, J. B. 1974. Afghanistan-West Pakistan. In: tion of the alpine system. Bull. geol. Soc. Am.
SPENCER, A. M. (ed.). Mesozoic-Cenozoic 84, 3137-80.
 L a t e C r e t a c e o u s - C e n o z o i c history o f western B a l u c h i s t a n 385

DICKINSON, W. R. & SEELY, D. R. 1979. Structure SEELY, D. R., VAIL, P. R. & WALTON, G. G. 1974.
and stratigraphy of forearc regions. Bull. Am. Trench-slope model. In: BURK, C. A. & DRAKE,
Assoc. Petrol. Geol. 63, 2-31. C. L. (eds). The Geology of Continental Mar-
FARHOUDI, G. 8z KARIG, D. E. 1977. Makran of Iran gins, 249-60. Springer-Verlag, New York.
and Pakistan as an active arc system. Geology, 5, SILLITOE, R. H. 1978. Metallogenic evolution of a
664-8. collisional mountain belt in Pakistan: a prelimin-
GANSSER, A. 1960. Ausseralpine Ophiolithprobleme. ary analysis. J. geol. Soc. London, 135, 377-87.
Eclog. geol. Helv. 52, 659-80. -- & KHAN, S. N. 1977. Geology of the Saindak
1971. The Taftan volcano (SE Iran). Eclog. porphyry copper deposit, Pakistan. Trans. Instn
geol. Helv. 64, 319-34. Ming Metall. 86, B27-42.
JACOa, K. H. & QUITTMEYER,R. L. 1979. The Makran STOCKLIN, J. 1971. Stratigraphic lexicon of Iran, Part
region of Pakistan and Iran: trench-arc system 1: Central, North and East Iran. Rep. geol. Surv.
with active plate subduction. In: FARAH, A. & Iran, 18, 338 pp.
DEJONG, K. A. (eds). Geodynamics of Pakistan, 1974. Possible ancient continental margins in
303-17. Geol. Surv. Pakistan, Quetta. Iran. In: BURK, C. A. & DRAKE, C. L. (eds). The
JONES, A. G. (ed.) 1960. Reconnaissance geology of Geology of Continental Margins, 873-87. Sprin-
part of West Pakistan: a Colombo Plan Co- ger-Verlag, New York.
operative Project. Report by Govt of Canada for , EFIEKHAR-NEZHAD, J. & HUSHMAND-ZADEH,
Govt of Pakistan, Toronto. 550 pp. A. 1965. Geology of the Shotori Range (Tabas
LAWRENCE, R. D., KHAN, S. H., DEJONG, K. A., area, East Iran). Rep. geol. Surv. Iran, 3, 69 pp.
FARAH, A & YEATS, R. S. 1981. Thrust and strike , -- & -- 1972. Central Lut reconnaiss-
slip fault interaction along the Chaman trans- ance, East Iran. Rep. geol. Surv. Iran, 22, 62 pp.
form zone, Pakistan. In: MCCLAY, K. R. ~:; STONELEY, R. 1974. Evolution of the continental
PRICE, N. J. (eds). Thrust and Nappe Tectonics, margins bounding a former southern Tethys. In:
363-70. Geol. Soc. London. Spec. Publ. 9. BURK, C. A. & DRAKE, C. L. (eds). The Geology
MCKENZIE, D. & SCLATER,J. G. 1971. The evolution of Continental Margins, 889-903. Springer-
of the Indian Ocean since the late Cretaceous. Verlag, New York.
Geophys. J. R. astron. Soc. 24, 437-528. VREDENBURG, E. W. 1901. A geological sketch of the
POWELL, C. McA. 1979. A speculative tectonic his- Baluchistan desert and part of eastern Persia.
tory of Pakistan and surroundings: some con- Mere. geol. Surv. India, 31, 179-302.
straints from the Indian Ocean. In: FARAH, A. & WHITE, R. S. 1979. Deformation of the Makran
DEJONG, K. A. (eds). Geodynamics of Pakistan, continental margin. In: FARAH, A. & DEJONG, K.
5-24. Geol. Surv. Pakistan, Quetta. A. (eds). Geodynamics of Pakistan, 295-304.
• CONAGHAN, P. J. 1973. Plate tectonics and Geol. Surv. Pakistan, Quetta.
the Himalayas. Earth planet. Sci. Lett. 20, 1-12. - - 1981. Deformation of the Makran accretionary
READING, H. G. 1980. Characteristics and recognition sediment prism in the Gulf of Oman (north-west
of strike-slip fault systems. Spec. Publ. Int. Indian Ocean) (this volume).
Assoc. Sedimentol. 4, 7-26.

RUSSELL S. ARTHURTON, Institute of Geological Sciences, Ring Road, Halton,

Leeds LS15 8TQ, England.
ABUL FARAH, Geological Survey of Pakistan, P.O. Box 15, Quetta, Pakistan.
WAHIDUDDIN AHMED, Geological Survey of Pakistan, P.E.C.H. Society, Karachi,
Pakistan.
 The Makran, Southeastern Iran: the anatomy of a convergent
plate margin active from Cretaceous to Present

G. J. H. McCall & R. G. W. Kidd

SUMMARY: The inland geology of the Iranian Makran, long neglected by geologists
because of its lack of hydrocarbon potential, is now reasonably well known following a
regional mapping programme carried out on behalf of the Geological and Mineral Survey
of Iran. The mountain range can be divided into seven geotectonic provinces forming an
arc round the late Pliocene epeirogenic Jaz Murian Depression at the southern end of the
Lut block. From the edge of the Jaz Murian outwards to the S these provinces are: (1) a
marginal basin in which well-preserved ophiolites formed and deep-water pelagic sedi-
ments were deposited from Jurassic to Palaeocene; (2) a narrow zone of continental crust
of Palaeozoic metamorphics capped by shelf limestones of mainly Cretaceous age but
including Carboniferous, Permian, Jurassic and Palaeocene developments; (3) a zone of
ophiolitic m61ange, the renowned Coloured M61ange; (4) a zone of immensely thick
Eocene-Oligocene flysch; (5) a similar zone of Oligocene-Miocene flysch; (6) a southern
zone of Miocene neritic to molassic sediments and (7) a Miocene-early Pliocene neritic
zone W of the Zendan fault. The structure is dominated by steep inward-dipping reverse
faults forming schuppen. These developed in a series of events which climaxed in the late
middle Miocene with some faults continuing to be active to the present day with
continuing uplift. Intense periclinal folds of dominantly chevron style are developed in the
flysch, synchronous with the faulting. Over substantial areas in the flysch deformation has
been so intense that a m61ange has developed, resembling some Franciscan m61anges
more closely than the Coloured M61ange. It contains exotic blocks which are tectonically
intruded from the Coloured M61ange, which forms the basement to the flysch zones.
The most significant discovery of this programme is the continuation into the Makran of
the Sanandaj-Sirjan zone in the form of the Bajgan-Dur-Kan complexes (province 2
above), thus forming a sliver of continental crust that stretches from the Bitlis Massif in
Turkey to the Makran. This separates the two ophiolitic developments of the Makran: to
the south the Coloured M~lange, the trench margin sequence of a north-dipping
subduction zone formed mainly in Late Cretaceous-Palaeocene time, and to the north a
belt of well-preserved Cretaceous-Palaeocene ophiolites formed in a marginal basin. The
latter ophiolites may be the equivalent of the island arc of this subduction system. The
Makran was not involved with the third group of ophiolites in this region, those of the
Oman, Neyriz and Kermanshah, which were emplaced as a result of collision of the
Arabian continental margin with an intra-oceanic NE-dipping subduction zone in the
Campanian. Any association of these ophiolites with the Coloured Mdlange along the
Zagros is due to the Pliocene collision of Arabia and Iran.
Following uplift of the Inner Makran in the late Palaeocene northward subduction has
continued to the present day with related andesitic volcanism to the north and migration of
the trench to the south following tectonic events in the Oligocene and middle Miocene, so
that the trench is now 300 km from the present andesitic volcanism. The products of
Eocene to Present subduction, including the immense deformed flysch deposits, are thus
superimposed on the condensed Cretaceous-Palaeocene subduction system.

A p a r t from the excellent p i o n e e r w o r k of

Introduction
H a r r i s o n & F a l c o n (1936) and H a r r i s o n et al.
T h e I r a n i a n M a k r a n comprises the 150 k m wide (1935-36) and s o m e u n p u b l i s h e d oil c o m p a n y
area of SE I r a n b e t w e e n the Jaz M u r i a n D e - studies, the g e o l o g y of the M a k r a n was long
pression and t h e coast (Fig. 1). T h e east-west n e g l e c t e d b e c a u s e of its lack of h y d r o c a r b o n
t r e n d i n g m o u n t a i n r a n g e rises from the coast to potential. H o w e v e r , most of the I r a n i a n M a k -
a r o u n d 6000 feet and in a few places to over ran m o u n t a i n r a n g e has n o w b e e n m a p p e d by
8000 feet b e f o r e d r o p p i n g d o w n again into the P a r a g o n - C o n t e c h (an I r a n i a n - A u s t r a l i a n joint
Jaz M u r i a n . W i t h i n I r a n it is 600 k m long f r o m v e n t u r e ) on c o n t r a c t for the G e o l o g i c a l a n d
the coast at M i n a b in the west to the Pakistan M i n e r a l S u r v e y of Iran. T h e m a p p i n g c o v e r e d
b o r d e r in the east. four full and two half 1:250,000 q u a d r a n g l e s (1 °

387
388 G. J. H. McCall & R. G. W. Kidd

1 I '1
j' t
s
(
" ~ \
--
Kuh-e-Birk
- .
-
&rot/[ .J'~lllllll(lll)lll~ . . . . ~£_'_1 ¢" f "% -I ~ t J H , / l ~ mi[~_ ~ %- ~.~

z: I~",O~lllllll['~-------\'/') • ~ IRANSHAHR ~ "~- -- ~ , _ . --¥

~ ~ ~ \) , . L - ~
~ x ; A ~ 'V
' ~-~I'r~F.._~-~---I- ='-r._ -I _~ ble connection N . . ~
~:~5~ ~ - - r . / / / / ~ IIDr~IIIIX-1--/-----_~-,,# - , ) ~ _ ~ -

26"

.... ~ A H B A H A R ~

...... o_0_ 0.. . . . ,oo

2 ~ R~ft hke spreading zone (loner

Makran marginal basin)
3 ~ Carbonate Fore-Arc zone (Bajgan-Our Kan zone) 6 ~ Oligocene-Miocene flysch zone

4 r~ Coloured M~lange 7 ['~ Miocene Neritic molasslc zone

]~] E. . . . . -Olig . . . . . flysch . . . . 8 [~ (Makran Unit) Mio-Pli . . . . ,t,c tool. . . . . . . . .

Fro. 1. Sketch map of the Iranian Makran showing the eight geotectonic provinces and the lines of
the sketch sections (Figs 2 & 3).

by 11o~). In addition, nine maps (~o by ~o) of Sketch sections across the Makran mountain
special interest were mapped on a 1:100,000 range are shown in Fig. 2 (Minab sector--NW)
scale and a 1:500,000 tectonic map accompa- and Fig. 3 (Fannuj sector--centre). The lines of
nied the overall report. In all an area of over these sections are on Fig. 1. The structure is
80,000 km 2 (about the size of Scotland) was dominated by steep inward dipping (toward the
mapped in 2 years using helicopters and more Jaz Murian) faults forming schuppen. Thus the
than 30 geologists. The maps and reports with relationships between the geotectonic provinces
full details of the geology will be published (apart from the Jaz Murian) are largely tecto-
shortly by the Geological and Mineral Survey of nic. From the Inner Makran out towards the
Iran (McCall 1981). coast these provinces are:
This paper is intended to relate the broad
results of this programme to the global tectonics (1) J a z M u r i a n D e p r e s s i o n
of this area. The mapping was a group project
but the conclusions in this paper are those of This is a late Pliocene epeirogenic depression
the two authors alone. G. J. H. McCall was 300 km in length in an east-west direction and
senior consultant to this project and compiled over 100 km across north-south. The only rocks
the reports (McCall 1981) and R. G. W. Kidd are superficial: alluvial fans of gravel, silt
contributed to it in the second field season, and plains, playa lakes, dunes and salt pans.
wrote the section on the plate tectonics and
evolution of the region in the final report. (2) I n n e r M a k r a n s p r e a d i n g z o n e

This is a zone of rifting occupied by largely

Geology of the Iranian Makran
undeformed ophiolites (as defined by the 1972
Penrose Conference). There are three distinct
There are eight geotectonic provinces (includ- ophiolites. One is tholeiitic and resembles the
ing the Jaz Murian) in the Iranian Makran Troodos Complex, Cyprus. It is Early Cre-
(McCall 1981)--these are shown in Fig. 1. taceous to early Palaeocene in age and consists
 An active convergent plate margin 389
SW 16 13 14 13
sheeted dykes which are largely dacitic-rhyolitic
and include micro-trondhjemite. It also con-
tains screens of pillow lavas and of gabbro and
trondhjemite. The third ophiolite, of Early Cre-
[ ii IL
8 5,6,7 4 3 taceous to early Palaeocene age, consists of
fragmented plutonics (ultrabasics, troctolite,
12 1,O 4 9 8 /
7 6
/ 5 4 3 ~ 1, "'E
cumulate gabbro, high-level gabbro and trond-
hjemite) folded into immense flexures, and
diabase dykes above (some sheeted), grading
up into pillow lavas overlain by pelagic sedi-
ments.
it it
3 2 1
The dyke trends in these ophiolites are
roughly parallel to the strike of the ophiolites as
1[~ Pl() 310 Km VERTICAL SCALE EXAGGERATED a whole--following the curvature of this Inner
TOPOGRAPHy GENERALISED
Makran spreading zone round from a north-
FIG. 2. Sketch section across the Makran in south trend in the NW of the area to east-west
the Minab sector along the line shown in near Fannuj. This suggests that these ophiolites
Fig. 1. Geotectonic province numbers are were formed in approximately their present
shown in heavy type beneath the section. positions and have not been tectonically em-
Other numbers indicate geology as follows: placed from elsewhere. Indeed their present
(1) gravel and silt fans of the Jaz Murian
depression; (2) Lower-Upper Cretaceous width of about 50 km in the NW to as little as
calc-alkaline ophiolite: mainly sheeted 15 km near Remeshk (see Fig. 1) may not
dykes of intermediate to acid composition, necessarily be a great deal less than their origin-
screens of pillow lavas, pelagic sediments, al width, as apart from steep bounding reverse
gabbro and trondhjemite; (3) lower- faults they are relatively undeformed. Near
upper Eocene proximal turbidites, acid Fannuj, however, there is a local development
welded tufts, diabase and diorite sills and of metamorphics including blueschists in a
shelly limestone, all overlying ophiolite m61ange within the Spreading Zone. This in-
sequence; (4) Lower Cretaceous-lower cludes Cretaceous Globotruncana limestone
Palaeocene pillow lavas of outer tholeiitic
ophiolite with pelagic limestones; (5) and radiolarites, and metamorphics which may
serpentinite; (6) low-level cumulate gabbro be as old as Palaeozoic.
of outer ophiolite; (7) high-level non-
cumulate gabbro; (8) trondhjemite at top (3) BajgannDur-Kan zone
of 7; (9) sheeted diabase dykes; (10) Dur-
Kan Complex: mainly Cretaceous but in- This is a narrow but continuous zone of
cluding Permian and Jurassic shelf carbon- continental crust. In the NW of the area where
ates structurally overlying 11" (11)Bajgan it is up to 40 km wide (see Figs 1 & 2) it consists
Complex: greenschist and amphibolite of Palaeozoic metamorphics (the Bajgan Com-
facies Palaeozoic metamorphics; (12) tec-
tonic enclaves of ultrabasics with plex). Eastward the Bajgan Complex is overlain
chromite; (13) Coloured M61ange: Lower by a sequence of shelf limestones of the Dur-
Cretaceous-lower Palaeocene block-to- Kan Complex (see Figs 1 & 3), which are
block m61ange; (14) ultrabasic tectonites, predominantly of Early Cretaceous to early
partly layered with chromitite cumulates; Palaeocene age, but have tectonic inclusions of
(15) Eocene-Oligocene flysch; (16) upper Carboniferous, Permian and Jurassic shelf
Oligocene-lower Miocene flysch plus limestones. This continental sliver thins east-
some neritic, gypsiferous Miocene ward and appears to pass into the Kuh-e-Birk
sediments; (17) Burdigalian reefal range through a sigmoid flexure and thence into
limestone" (18) Makran unit: upper
Miocene-lower Pliocene neritic to molas- Pakistan. To the NW of the mapped area it can
sic sediments; (19) coastal mudflats. be traced up through the Sanandaj-Sirjan zone
(see Fig. 4) as far as the Bitlis Massif in Turkey
of cumulate gabbros overlain by high-level gab- (Stocklin 1977). It has been suggested that it
bro, trondhjemite, d/abase sheeted dykes, pil- may be an Alpine-type nappe (Gansser pers.
low lavas with copper shows and pelagic sedi- comm.) but the structure of this region is one of
ments. The other two ophiolites are distinctly steep schuppen not of low-angle thrusts and a
calc-alkaline and may represent separate parts nappe thousands of kilometres long seems im-
of the same ophiolite sequence. One of these, probable. It must be considered a global
of Early to Late Cretaceous age, has only the geotectonic entity and the continuation of the
upper parts exposed and consists mainly of Sanandaj-Sirjan zone into the Makran is impor-
390 G. J. H. McCall & R. G. W. Kidd
SOUTH

7 6 5

8 6 3 4 5 3 NORTH
9 8 6 \ 4 3 5 3 2

II ..It~ I L _ _
5 4 & 3 imbricated 2 1

7 'i' :'I..... VERTICAL SCALE EXAGGERATED

TOPOGRAPHY GENERALISED

Ft6. 3. Sketch section across the Makran in the Fannuj sector along the line shown in Fig. 1.
Geotectonic province numbers are shown in heavy type beneath the section. Other numbers indicate
geology as follows: (1) superficial gravel fans and dunes of Jaz Murian depression; (2) mainly
greenschists and blueschists with Lower-Upper Cretaceous pelagic limestones and radiolarites; (3)
pillow lavas and pelagic sediments; (4) diabase dykes, some sheeted; (5) fragmented and folded
ophiolitic plutonics: layered cumulates at base (ultrabasics, troctolite, gabbro), non-cumulates above
(gabbro, trondhjemite); (6) Dur-Kan complex: mainly Cretaceous, but including Permian and
Jurassic, shelf carbonates; (7) Bajgan Complex of Palaeozoic metamorphics structurally overlain by
6; (8) Coloured M61ange; (9) lower-upper Eocene distal flysch and minor pelagic and shelf
limestones dislocated throughout with exotic blocks; (10) tectonic dislocation and protrusion
m61ange, 80% flysch, 20% exotics; (11) lower-middle Miocene proximal flysch, deep water facies;
(12) upper Oiigocene-iower Miocene mainly distal flysch; (13) lower-middle Miocene neritic
sediments, mainly fine gypsiferous mudstones; (14) exotic raft to harzburgite; (15) middle-upper
Miocene neritic sediments; conglomerates and sandstone in the north, sandstone, siltstone and
mudstone in the south, lateral facies changes to deltaic and estuarine facies with some evaporites.

tant in the consideration of any plate tectonic We suggest that the Coloured M61ange was
model for the region. formed in the trench of a NE- to north-dipping
subduction zone by scraping off fragments of
the downgoing plate, possibly mixing these
(4) Coloured M61ange Zone fragments with small bits of the overriding
This is a tectonic, ophiolitic, block-to-block plate. Whether there is a contribution to the
m61ange (Gansser 1974) consisting of serpenti- m61ange from the overriding plate depends
nite, other ultrabasic and basic ophiolitic rocks, upon establishing the derivation of the exotic
pillow lavas, pelagic limestones, radiolarites components of the m61ange (shallow water
and distal turbidites. It also contains minor limestones and metamorphics). This question
andesite, rhyolite, rhyolitic welded tuff, has not yet been resolved.
trachyte and exotic components (metamorphics The outcrop of Coloured M61ange is about
and Lower Cretaceous reefal limestone). There 30 km wide in the NW of the area but thins
is no true matrix: block boundaries are sheared rapidly to a few kilometres and pinches out to
and often serpentinite acts as a 'lubricant'. The the south of Remeshk (see Fig. 1), again re-
m61ange is not all chaotic: much of it consists of appearing in the area not mapped in this project
stacked slabs that are the right way up and to the south of Iranshahr. However, exotics
consistently dip steeply inward toward the NE. within the flysch of the sediments of provinces
The fossil ages are enigmatic in that the radiola- 5, 6 and 7 indicate that the Coloured M61ange
rites are Jurassic-Coniacian while the pelagic extends 70 km of the south of its area of outcrop
limestone (GIobotruncana) is Cenomanian- beneath these sediments (see below).
Maastrichtian, yet the radiolarites and pelagic The age of onset of m61ange formation can-
limestones sometimes appear to form an inter- not be determined but the abundance of
bedded sequence. The youngest rocks are early Cenomanian-Maastrichtian rocks suggests that
Palaeocene biomicrites. The pillow lavas and much of it did not start forming a m61ange until
sediments are commonly interbedded. Maastrichtian time. The involvement of early
 An active convergent plate margin 391

? , , , , 5ooKm
/ cAsP'A" I

'x IRAN
.",
V

f ~ "
+
• Sabzevar
V

I AFGHANISTAN
1-
% t I V

+
qN~+
'X,
%,\
v Z 'g
+
+
+
+
+ '
+/, [ f " 4-" % +
+
+
+
+
+
+ "+ (i) + + +
q,,'• + ~ N ~ n g/ + Birjand e V ~ /

+",v k+ +" +". _ s, sTA .+

• \ \ VX"+ B L0 C K +\ \+ + + -t-
~"~.o. "%.'2~+\ ".-+/+ , + , , , +

5~./,.,,,Tx~ -:- \ ~\ */jAz \ v

4/ DUR-KAN

UUCTION ZONE
ARABIA
GULF
OF
OMAN OMA N

(iJ)
/
C'}H! if~(~rllH I MHrCJlH

Colouted ill(~ I~l i H~u

VV
/iotlh Sid(ff o~ .~(~(l[ll{3ltl [~?lhy~

("'+'-'~ M~t:ro(ontHlentHf blol:k

,,_.,, .--"

F~6.4. The three distinct ophiolite developments in the Makran and surrounding regions with the
associated microcontinent of Southern Tethys; (i) and (ii) are section lines shown in Fig. 5.

Palaeocene rocks indicates that some m61ange unfaulted sections several thousand metres
formation continued into the Palaeocene. This thick may be traversed in the field. However,
event ended by the late Palaeocene when there for the most part the flysch is highly folded and
was an abrupt change in palaeography mainly faulted. There are box, kink, isoclinal and open
involving uplift of geotectonic provinces 2, 3 folds but by far the predominant type are
and 4. These provinces are patchily obscured by periclines of chevron style. There is a slight
minor younger shelf limestone and flysch but south-vergence so that synclines have long
for the most part marine deposition ceased in south limbs and short north limbs, thus there is
the Inner Makran in the Palaeocene. an overall northward dip of the strata as a result
of the folding. Folds are developed on all scalds
(5) Eocene-Oligocene flysch up to a few kilometres across but frequently
To the south of the Coloured M61ange Zone both limbs and axial regions are cut out by
is an immensely thick flysch sequence. It is north-dipping, steep, reverse faults. There is
mainly distal, calcareous, turbiditic flysch, with dominant faulting out of the anticlines and
thousands of repetitions of classic, complete or preservation of the synclines. This contrasts
partial, Bouma sequences. It has beautifully with the Zagros style of dominant preservation
preserved trace fossils and classic flute and tool of anticlines that occurs west of the Z e n d a n
marks. Continuous sequences may be more fault south of Minab in geotectonic province 8
than 10,000 m thick and certainly occasional (Fig. 1).
392 G. J. H. McCall & R. G. W. Kidd
Commonly the flysch has been so tectonized Aquitanian producing a trough to the north in
that it has become dislocated and often has which flysch deposition persisted in the late
developed into a m61ange. This m61ange is middle Miocene between the emerging south-
distinct from the Coloured M61ange being very ern area and emergent provinces 2 to 5. Shallow
similar to some of the Franciscan m61anges shelf evaporites and reefal limestones were
(Cowan 1978, 1981). It contains exotic blocks formed in the south. The latter have yielded
derived from the underlying Coloured M61ange important and diverse coral collections and
basement and also blocks of Eocene pelagic and accompanying foraminiferal faunas. In the late
shelf limestones. It has been referred to as middle Miocene when the flysch trough finally
wildflysch but it is clearly not. It is very similar filled up, coarse detritus spread over the entire
to the exotic flysch of the Arakan, Burma southern area. These middle to upper Miocene
(Brunnscheiler 1966). The exotic blocks were estuarine, deltaic and shallow shelf sediments
all emplaced tectonically, being squeezed up range from conglomerate to fine gypsiferous
from the basement like pips. Some large blocks mudstones. There are local developments of
are composed of Coloured M61ange itself, not fluviatile fanglomerate (true molasse) at the top
just of blocks that were once constituents of of this sequence. This sequence forms the chain
Coloured M61ange. of immense open synclines (see Fig. 3) which
In the east of the project area the Eocene- are so conspicuous on the Landsat imagery.
Oligocene Flysch Zone widens out to over These were incorrectly interpreted as uplifted
100 km in width. In this region the flysch is less deepwater sequences of the trench slope by
commonly tectonized to a m61ange and contains Farhoudi & Karig (1977). In general the neritic
less exotic components. In the north of the sediments are contorted where they were thinly
eastern part of the project area the flysch bedded and incompetent and are folded into
appears to grade up into a series of shallow- large open folds where they are thickly bedded
water basins. and competent.
The style of folding described above is similar
to that described by White (1977, 1981) off the (8) Mio-Pliocene Neritic Zone
coast of the Makran. In the inland Makran the
folding has been many times more intense and In the project area rocks of this zone (the
has developed on much smaller scales as well as Makran unit) (Huber 1952; Stocklin 1952) are
on a large scale. As a result of the intense only exposed west of the Zendan Fault (see Fig.
deformation it is not possible to demonstrate 1). This sequence is the youngest marine se-
that there was syn-sedimentary deformation of quence of the Makran and includes gypsiferous
basins as White has described off the coast. mudstones, deltaic sandstones and estuarine
conglomerates with minor fluviatile conglomer-
(6) Oligocene-Miocene flysch ates at the top. The term Makran (or Mekran)
has been loosely applied by palaeontologists to
Further uplift occurred in this region in the
virtually any neritic sediments encountered
mid-Oligocene. The trough of flysch deposition
near the Makran coast in Iran and Pakistan
shifted to the south (and to the SW in the
which range in age through the entire Miocene
extreme west of the area). The Oligocene to
and Pliocene. Here the name has no stratig-
earliest Miocene flysch is mainly distal, while
the early to middle Miocene flysch is proximal. raphic significance outside the limited zone
There is no evidence of any co-existing shelf described. Equivalent sediments probably also
occur along the south coast of the Makran
limestone during this time, unlike in the
(Stocklin 1952; Anon 1962) outside the area
Eocene.
mapped in this project.
The deformation of the Oligocene-Miocene
West of the Zendan fault these sediments are
flysch is similar to that of the Eocene-Oligocene
highly folded and even Pliocene beds may be
flysch. There is similar development of m61ange
vertically disposed. The synclines are frequent-
with exotic blocks some of which are of Eocene
ly faulted out so that the anticlines are preferen-
flysch. The pronounced spatial realtionship of
tially preserved. This is the opposite of the
dislocation and m61ange formation to Miocene
dominant process within tectonic provinces 5, 6
faults suggests that this extreme tectonism may
and 7. The different styles of folding may be
be predominantly a Miocene event.
due to the different nature of the basement: to
the west of the Zendan fault the basement is
(7) Miocene Neritic Zone
probably continental whereas to the east of the
In the south of the area rapid shallowing to Zendan fault it is oceanic or comprises oceanic
platform sea conditions commenced in the lithologies broken up into a m61ange.
 An active convergent plate margin 393

Plate tectonics and the evolution of the that show the relationships of the different
tectonic units across two sections (see Fig. 4):
Makran (1) from Arabia NW across the Persian Gulf,
Figure 4 shows the continents, microconti- the Zagros mountains including the Neyriz
nents, ophiolites and ophiolitic m61anges that ophiolite, the Zagros Crush Zone, the Sanan-
make up the essential tectonic units and sutures daj-Sirjan belt, the Baft-Nain belt, the Lut
of the Makran and surrounding region. The block, the Birjand-Zahedan belt into the
plate tectonic development of this region will be Afghanistan-Sistan block, and (2) from Arabia
demonstrated by means of diagrams (Fig. 5) including the Oman ophiolite north across the
Oman Sea, the Makran including the flysch
belt, Coloured M61ange zone, Bajgan-Dur-Kan
ST
...f]JHH!Hlil]I]I]iilHII~I[~]I/IIII]I]I~[E]HIH]I]ILINIIiJI
PT
zone and Inner Makran ophiolite zone into the
Lut block.
The exact positions in the Palaeozoic of the
microcontinental fragments that now form Iran
b. M i d - J u r a s s i c - Early Cretaceous are not likely ever to be established but the
similarities of the Palaeozoic sequences of
~:[],H[]I[t]H]':I~I:] J ~
India, Iran and Arabia (Stocklin 1977) suggest
that along with India, most of lran was once
iL
part of Gondwanaland and separated from
OS z ~? ST IMR
Gondwanaland during the Triassic (Fig. 5a).
By middle Jurassic time a substantial South-
ern Tethys had developed in this region. Sub-
c Cenomanian - CamDanian duction appears to have commenced along the
i
Sanandaj-Sirjan zone, indicated by andesites of

g. Present
NO ZCZ ~,

AZ
---___% AS

i,

ii. HA
d, Maastrtchtian
¢~'~N

.,7

SO
O

CM IMSZ
BNMB BZMB

.
ii.
AO
SO
~U MPNS OM.F EF ~

FIG. 5. Sketch sections along lines (i) and

e. Eocene (ii) in Fig. 4 showing reconstructions of the
NO ~'~ probable relationships of continents,
~' AZ~/llZiIHllrL![[iiliNjrljiEilr~:lli:li,l]Hi;i[ir~ micro-continents, ridges, subduction zones
and back-arc basins from Triassic to the
present:
AO Arabia (Oman), AS Afghanistan--
Sistan Block, AZ Arabia (Zagros), BDKZ
~J'__~lllllll!ilillll;l!!lmMl:lil!llllllllllltllNTl~ Bajgan-Dur-Kan zone, BNMB Baft-Nain
marginal basin, BNR Baft-Nain rift,
BZMB Birjand-Zahedan marginal basin,
CM Coloured M61ange, EF Eocene flysch,
f. Late Miocene HA Hawasina, I Iran, IMR Inner Makran
i. ~ A. A rift, IMSZ Inner Makran Spreading Zone,
LLIT Lut Block, MPNS Miocene-Pliocene
neritic and molassic sediments, NO Neyriz
ophiolite, NSZ Neyriz subduction zone,
OMF Oligo-Miocene flysch, OSZ Oman
.... ~ OMF EF subduction zone, PT Palaeozoic Tethys,
SO Semail ophiolite, SSZ Sanandaj-Sirjan
AO ~[llliIlllllllllllllllllllllllllllllllllllll[IHl~ LUT Zone, ST Southern Tethys, ZCZ Zagros
Crush Zone.
 394 G. J. H. M c C a l l & R. G. W. K i d d
this age (Fig. 5b(i)). There is no direct evidence early as Early Cretaceous or even Jurassic with
of any subduction in the Makran at this time late addition of younger rocks. However, the
although Jurassic pelagics in the Inner Makran great abundance of Campanian-Maastrichtian
(geotectonic province 2) indicate the develop- rocks suggests that most of it was not formed
ment of a rift (Fig 5b(ii)). During Early Cre- until Maastrichtian or even Palaeocene. Thus
taceous time or possibly a little later, an the Coloured M61ange was mainly formed after
intra-oceanic NE-dipping subduction zone de- completion of the emplacement of the Oman
veloped. It was as a result of collision of the ophiolites. Two possible causes for the produc-
Arabian continental margin with this subduc- tion of the Coloured M~lange are: (1) collision
tion zone that the Kermanshah, Neyriz and of the intra-oceanic subduction zone with the
Oman ophiolites were emplaced in the Campa- Arabian continental margin in the Campanian
nian. would cause a major rearrangement of plate
Fig. 5(c) shows the situation in Cenomanian boundaries in this region probably leading to an
to Campanian time. The continental margin of increased rate of subduction in the Makran, and
Arabia was about to collide with the NE- (2) complete subduction of the older oceanic
dipping subduction zone. The Hawasina thrust crust would result in attempted subduction of
slices of the Oman were being stacked up in this young oceanic crust that would be standing
subduction zone--the distal uppermost sheet much higher than the old crust and so would be
first followed by successively lower and more more easily scraped off the downgoing plate.
proximal sheets. Finally, in Campanian time The emplacement of the Oman, Neyriz and
the collision was completed with the Semail Kermanshah ophiolites appears to coincide
ophiolite being emplaced on top of the stacked with changes in the relative motion between the
Hawasina continental margin sequence. The African and Eurasian plates (Dewey et al. 1973)
oceanic crust that was subducted would have from predominantly compressional with a slight
been mainly Jurassic and Triassic, yet the sinistral component in the region of Arabia to
Semail ophiolite is Cenomanian to Coniacian dextral strike slip. However, the existence of
(Glennie et al. 1973) and hence came from the several plates and more than one subduction
ocean to the NE of the subduction zone-- system without preservation of any oceanic
where it was part of a forearc, backarc or crust between Africa/Arabia and Eurasia
even represents the island arc itself. The origin makes it impossible to reconstruct the motion
of the Neyriz and Kermanshah ophiolites (Fig. along a single plate boundary from the motions
5c(i)) is identical to that of the Oman ophiolite. of the major continental plates. The probable
The only difference is that the Zagros (regarded age of Coloured M61ange formation is coinci-
as part of the Arabian continental margin to the dent with the breaking away of India from
SW of the Zagros Crush Zone) collided with Gondwanaland and its rapid acceleration north-
lran along the Zagros Crush Zone during the ward.
Pliocene. There are no typical andesitic rocks in the
Fig. 5(d) shows the situation in the Maas- Iranian Makran and it is suggested that the
trichtian when the ophiolites have been em- Inner Makran spreading zone may be the
placed. There was ocean to the NE of both the equivalent of the arc. As described above the
Oman and Zagros; this ocean may have still orientation of the sheeted dykes along the
been spreading at this time. There was certainly length of the zone and the distinctly calcalkaline
no collision of Arabia with the Makran region nature of the inner (more northward or north-
during the Campanian. eastward) ophiolites support this suggestion. It
In the Makran the Inner Makran spreading is possible that 'arc' rocks may lie beneath the
zone was actively spreading from Early Cre- Jaz-Murian depression (see Fig. 1) or have been
taceous until the early Palaeocene. If this was a tectonically removed, but in the former case the
subduction-related process then it follows that ophiolites would be in front of the arc (rather
there was some subduction of oceanic crust in a than being back-arc) and in the latter it is
NE or northward direction throughout this surprising that no trace should remain when the
period as indicated in Fig. 5(b(ii), c(ii) and ophiolites are not severely deformed. There are
d(ii)). The Baft-Nain, Birjand-Zahedan and small amounts of possible 'arc' rocks in the
Sabzevar belts, were also spreading during the Coloured M61ange but these could equally well
Cretaceous after initiation in the Jurassic. In originate from oceanic island volcanoes or
the Makran the age of formation of the Col- simply be unusual ocean-floor rocks. It is possi-
oured M61ange as opposed to the age of its ble to construct a more complicated plate tec-
constituent rocks cannot be precisely dated. tonic model involving many microplates.
M61ange formation may have commenced as However, in this paper we wish to put forward
 An active convergent plate margin 395

the simplest model compatible with the known The main evidence for this subduction is the
geology although accepting that whole sections great volume of Eocene volcanics 150 km NE
might have completely vanished. of the Crush Zone in a parallel belt more than
In this Makran subduction system the Baj- 1600 km long. There are also minor accumula-
gan-Dur-Kan Zone (geotectonic province 3 in tions of Eocene flysch indicating the presence
Fig. 1) formed a forearc of shelf limestones on of a trough along the Crush Zone at this time.
continental basement. This forearc zone ex- These may be remnants of a large volume of
tends to the NW and appears to have been Eocene flysch that has now vanished by subduc-
emergent during the Cretaceous in the NW of tion or by overthrusting of Central Iran on to
the project area and in the Sanandaj-Sirjan the Zagros. However, it is possible that the
belt. Since there is a continuation of the Mak- Sanandaj-Sirjan zone was mainly emergent and
ran to the NW in the Sanandaj-Sirjan and blocked off the supply of flysch from the vol-
Baft-Nain belts it is inferred that subduction canics to the north.
also took place along the line of the Zagros In the Oligocene there was some adjustment
Crush Zone where there are remnants of Col- to the north-dipping subduction zone in the
oured M61ange. This is supported by the pre- Makran. This resulted in a southward shift of
sence of andesites and flysch of Cretaceous age the trench and uplift of the Eocene flysch,
along the Sanandaj-Sirjan belt. which then provided part of the source of the
The Baft-Nain, Sabzevar and Inner Makran Oligocene-Miocene flysch. This event may be
ophiolite troughs closed up by the late related to a readjustment of plate motions at
Palaeocene (Fig. 5e) so that the Sanandaj- this time which include the cessation of sea-
Sirjan and Bajgan-Dur-Kan microcontinents floor spreading in the Indian Ocean (McKenzie
had coalesced with the Lut block to form the & Sclater 1971). The further readjustment of
Central and East Iran block of Takin (1972). In plate motions when Indian Ocean sea-floor
the Inner Makran there was uplift and com- spreading recommenced in the middle Miocene
pression of the Coloured M61ange, Bajgan- approximately coincides with renewed uplift
Dur-Kan and Inner Makran spreading zones. It and yet further southward migrations of the
is possible that there was some minor subduc- trench in the Makran. The andesitic volcanics
tion of the latter zone indicated by the blue- to the north, the tectonics, and the flysch
schists near Fannuj. deposition, however, are all consistent with the
By the end of the Eocene the Birjand- existence of a north-dipping subduction zone in
Zahedan trough had also closed, joining the the Makran throughout the Tertiary. The dis-
Afghan-Sistan Block to the Central and East crete tectonic events probably relate to changes
Iran microcontinent to form a larger continen- in the rate and/or direction of subduction.
tal block. The resulting tectonic situation was a Fig. 5(f) shows the relationship between Ara-
simple north-dipping subduction zone in the bia and Central Iran in the late Miocene.
Makran with a single continental mass to the Arabia/Zagros is about to collide with Central
north (Fig. 5e(ii)). There was substantial Iran while there is still a substantial width of
Eocene andesitic volcanism to the north of the ocean separating Oman and the Makran. Not
project area, which provided the source for the until the Pliocene collision (Stocklin 1977) did
huge flysch deposits of geotectonic province 5. Arabia/Zagros have any direct relationship with
This flysch was deposited on an already Central Iran. This collision resulted in the
accreted prism of Coloured M61ange up to 70 juxtaposition of two separate subduction sys-
km wide, or possibly partly on oceanic crust, tems: (1) the subduction system represented by
parts of which were later detached from the the Neyriz and Kermanshah ophiolites and
subducting plate and tectonically intruded into related rocks which had been sitting passively
the flysch. It might be possible to distinguish on the Arabian/Zagros continental margin since
between these alternative hypotheses for the their emplacement in the Campanian, and (2)
nature of the basement to the flysch by detailed the subduction system that had been active
study of the exotics within the flysch. However, along the SW edge of the Sanandaj-Sirjan belt
since the oceanic crust subducted during and from the Jurassic through to the Tertiary. Due
subsequent to the Eocene may be similar to that to the deep embayment in the Arabian (plus the
which formed the Coloured M61ange in the Zagros) continental margin represented by the
Maastrichtian or Palaeocene this might prove Gulf of Oman, Arabia has not yet collided with
difficult. the Markan (Fig. 5g(ii)). In the Zagros the
Along the Zagros Crush Zone there also continued covergence resulted in intense fold-
appears to have been substantial NE subduc- ing of the Arabian continental platform sedi-
tion of oceanic crust in the Eocene (Fig. 5e(i)). ments and thickening and possible subduction
396 G. J. H. McCall & R. G. W. Kidd
even of the continental basement. Along the mained approximately stationary. In the pro-
Zagros Crush Z o n e all but a few remnants from cess a substantial accretionary prism of sedi-
the trench zone of subduction system (2) above ment has built up. Accretion of material from
have been obliterated by overthrusting of Cen- the downgoing plate clearly occurred during the
tral Iran from the NE. This effect can be seen in Late Cretaceous and Palaeocene and may have
the NW of the project area near Minab (Fig. 1) occurred since (see discussion above).
where the i m m e n s e flysch belt of the Makran If subduction continues as at present, O m a n
has been thinned against the Z e n d a n fault will collide with the Makran and this accretion-
(which is a continuation of the south side of the ary prism may be overthrust from the north and
Zagros Crush Zone). Going further NW into obscured as has occurred along the Zagros
the Zagros (see Fig. 4) the Coloured M61ange Crush Z o n e (assuming such an accretionary
Z o n e also thins except for a few remnants so prism was once developed there). The result
that the Arabian continental margin is juxta- will be a single suture from Turkey to Pakistan,
posed against the Sanandaj-Sirjan belt (the NW but one marking the disappearance of at least
continuation of the B a j g a n - D u r - K a n Zone). two subduction systems.
Subduction in the Makran is continuing at the
present day with deformation of the sediment ACKNOWLEDGMENTS:This paper was prepared at the
of the Gulf of O m a n off the Makran coast suggestion of the organizers of this symposium to
where the subducting plate initially bends down present an up-to-date model of the tectonics of this
only about 1° (White & Klitgord 1976; White region based on the recent work carried out on behalf
of the Geological and Mineral Survey of Iran. The
1977). Further inland the subducting plate
full descriptions of the geology, only summarized
must bend down more steeply, but the initial low here sufficiently to support the plate tectonic model,
dip is related to the unusually great (250 Am) will be published by the Geological and Mineral
separation between the trench and the 'arc'. Survey of Iran. The contributions of the survey and of
From the Eocene onwards the 'trench' has the geologists who worked on this regional mapping
migrated southwards while the 'arc' has re- programme are acknowledged.

References

ANON, 1962. Unpublished report on the Geology of Nat. Iranian Oil Co., Tehran and the Geol. Soc.
the Southern Makran, AGIP, Milan, filed at the Lond.
Geological and Mineral Survey of Iran, Tehran. - - ALLISON,A., HUNT, J. A., MALING,P. B.
BRUNNSCHEILLER,R. O. 1966. On the geology of the d MCCALL, R. J. C. 1935-36. Geology of the
lndo-Burman ranges. J. geol. Soc. Aust. 13, Landward Makran. Unpubl. maps, filed at the
137-94. Nat. Iranian Oil Co., Tehran and the Geol. Soc.
COWAN, D. S. 1978. Origin of blue-schist bearing Lond.
chaotic rocks in the Franciscan complex, San HUBER, H. 1952. Geology of the Western Coastal
Simeon, California. Bull. geol. Soc. Am. 89, Makran area. Unpubl. report, Iranian Oil Co.,
1415-23. filed at the Iranian National Oil Co., Tehran,
1981. Deformation of partly dewatered and con- No. GR 9lB.
solidated Franciscan sediments near Piedras Blan- MCCALL, G. J. H. 1981. Compiler of reports of
cas Point, California (this volume). Geological and Mineral Survey of Iran:
DEWEY, J. F., PITMANIII, W. C., RYAN, W. B. F. & (1) Report on East Iran Project Area No. 1.
BONNIN, J. 1973. Plate tectonics and the evolu- (2) Explanatory text of the Minab quadrangle
tion of the Alpine system. Bull. geol. Soc. Am. map, 1:250,000.
84, 3137-80. (3) Explanatory text of the Tahenii quadrangle
FARHOUDI, G. & KARIG,D. E. 1977. Makran of Iran map 1:250,000.
as an active arc system. Geology, 5, 664-8. (4) Explanatory text of the Fannuj quadrangle
GANSSER, A. 1974. The ophiolitic melange, a world map 1:250,000.
wide problem, and Tethyan examples. Ecolog. (5) Explanatory text of the Pishin quadrangle
geol. Heir. 67, 479-507. map 1:250,000.
GLENNIE, K. K., BOEUF, M. G. A., HUGHES-CLARKE, (6) Explanatory text of the Nikshahr quadrangle
M. W., MOODY-STUART,M., PILAAR,W. F. H. & (southern half), 1:250,000.
REINHARDT,B. M. 1973. Late Cretaceous nappes (7) Explanatory text of the Saravan quadrangle
in the Oman Mountains and their geologic evolu- (southern half), 1:250,000 (in press).
tion. Bull. Am. Assoc. Petrol. Geol. 57, 5-27. MCKENZIE, D. P. & SCLATER,J. G. 1971. The evolu-
HARRISON, J. V., & FALCON,N. L. 1936. Geology of tion of the Indian Ocean since the Late Cre-
the Coastal Makran. Unpubl. maps, filed at the taceous. Geophys. J. R. astron. Soc. 24, 437-528.
 A n active c o n v e r g e n t plate m a r g i n 397

STOCKLIN, J. 1952. Geology of the Central Coastal WHITE, R. S. 1977. Recent fold development in the
Makran Area. UnpubL report, Iranian Oil Co., Gulf of Oman. Earth planet. Sci. Lett. 36, 85-91.
filed at the National Iranian Oil Co., Tehran, -- & KLrT~ORD, K. 1976. Sediment deformation
No. GR 91C. and plate tectonics in the Gulf of Oman. Earth
1977. Structural correlation of the Alpine planet. Sci. Lett. 32, 199-209.
ranges between Iran and Central Asia. Mere. h. WHITE, R. S. 1981. Deformation of the Makran
Ser. geol. Fr. 8, 333-53. accretionary sediment prism in the Gulf of Oman
TAKIN, M. 1972. Iranian geology and continental drift (this volume).
in the MiddleEast. Nature. London,235,147- 50.

G. J. G. McCALL, 57 Venns Lane, Hereford, U.K.

R. G. W. KIDD, RSMAS-MGG (Marine Geology and Geophysics), University of
Miami, 4600 Rickenbacker Causeway, Miami, Florida 33149, U.S.A.
 The Coastal Belt of the Franciscan: youngest phase of northern
California subduction

Steven B. Bachman
SUMMARY: The Coastal Belt of the Franciscan complex represents the latest Cretaceous
to middle Tertiary part of the accretionary prism of an arc-trench system in northern
California. On the Mendocino coast, both deformed slope and trench deposits and much
less deformed slope basin deposits are recognized. Deformation increases with depth in
the slope deposits, towards the contact with the accretionary prism. In the accretionary
prism, bands of deformed rocks from tens of metres to kilometres in width suggest major
zones of shearing. Stratigraphic tops of beds face eastward, yet palaeontological ages young
westward; imbricate stacking is suggested by these relationships. Stratigraphical tops and
palaeontological ages indicate that the slope basin deposits are in normal stratigraphic
sequence.
Facies in the accretionary prism may represent trench fill, abyssal plain, feeder channel,
and small fan deposits that were accreted to the leading edge of the continent and may also
include slope deposits that were deformed during progressive deformation of the
accretionary prism. The less deformed slope deposits may have been deposited as slope
basin turbidites, small fans, and channels. South- and south-westward-directed palaeocur-
rent indicators and sandstone petrology suggest sediment transport from the Klamath
Mountains and Oregon across the shelf and down the slope and/or along the trench axis.

Geological setting
The Franciscan complex of California and Ore- In the central area of Coastal Belt exposures,
gon is an ancient subduction complex ranging in the Coastal Belt is in contact on the east with
age from Jurassic to Miocene or younger. The either the Franciscan Central Belt m41ange or
Franciscan complex is divided into three belts
on the basis of characteristic lithologies and
structures; the Coastal Belt is the westernmost
of these units (Bailey et al. 1964). The Coastal C~ PE
MENDOCIN(
Belt is exposed in the northern Coast Ranges of
California and extends from near Cape Mendo-
cino in the north to as far south as the San
Francisco Bay area (Fig. 1). East of it is the
Central Belt m41ange.
The Central Belt of the Franciscan is a
m61ange unit that includes blocks of sandstone,
greenstone, chert, blueschist, eclogite, am-
phibolite, and serpentinite incorporated in a
sheared shale and sandstone matrix (Bailey et
al. 1964; Berkland et al. 1972; O ' D a y 1974).
Central Belt rocks range in age from Jurassic to
Cretaceous (Blake & Jones 1974), but near the
contact with the Coastal Belt in this area, only
Lower and U p p e r Cretaceous strata are present
( O ' D a y 1974; K r a m e r 1976).
The Coastal Belt consists of interbedded GR
sandstones, siltstones, shales, and minor con- {~ CO~
glomerates. Volcanic rocks, limestone, and CEI'
chert occur locally. Laumontite is c o m m o n in
veins and on grain overgrowths in the I]TI] EA
sedimentary rocks. The presence of only zeolite
facies m e t a m o r p h i s m indicates that the Coastal FtG. 1. General geology of northern Cali-
Belt experienced shallower burial than some of fornia Coast Ranges after Blake & Jones
the rest of the Franciscan complex. (1978).

401
402 S. B. Bachman
the Leggett peridotite. This eastern contact has styles of deformation are classified on a numer-
been decribed variously as a high angle shear ical scale, with the end members as described
zone (Ogle 1953; O ' D a y 1974), an east-dipping, above. Each outcrop was designated and map-
high angle reverse fault (Kramer 1976, 1977), ped according to its tectonic style, both on
or a more recent landslide depositional contact small- and large-scale maps.
(Kleist 1974a). A broad band of current earth-
quake activity parallel to the contact suggests Style 1
that an earlier reverse and/or strike-slip fault This end member includes the sheared,
zone separating the Coastal and Central boudinaged, and stratigraphically discon-
M61ange belts may be reactivated in sympathy tinuous beds such as those in Fig. 2. The
with movement along the San Andreas fault. majority of these beds consist of dismembered
The San Andreas fault is located 7-17 km sandstone and mudstone, but locally contain
offshore and is the present western boundary of conglomerates, limestones, and volcanic rocks.
the Coastal Belt. In all cases, the mudstone is a matrix that
encases the other material. Individual beds
cannot be traced laterally for more than 100 m
Coastal Belt structure or so, but on an outcrop scale, sandstone/
mudstone ratios are relatively constant lateral-
Analytical techniques
ly, giving a gross sense of stratigraphy. Typical-
Because normal structural mapping techni- ly, the sandstones are deformed into elongate
ques were of limited value in the complexly saucer-shaped or bladed phacoids or boudins.
deformed and stratigraphically disrupted parts The edges of the boudins commonly are
of the Coastal Belt, a different type of structu- sheared, with broken sandstone grains up to
ral mapping scheme was used to resolve some several millimetres in thickness forming the
of the complexities. This technique is described circumference of the boudins. The ends of
in more detail elsewhere (Bachman 1979). The boudins are commonly sheared as well, with
technique is based on analysis of the styles of shale intruding along small fractures.
deformation within the Coastal Belt which Anastomosing shear surfaces disrupt the
range from sheared, boudinaged, and strati- mudstones so that no internal bedding is evi-
graphically disrupted beds to continuous beds dent. The mudstone matrix of these sheared
with little structural disruption. The less de- rocks in some instances contains extensive frac-
formed rocks are commonly involved in local, tures formed parallel to the shear and boudin
open folds and minor faulting. Accordingly, the foliation. Where fracturing is pervasive, the

Fl(;. 2. Structural style 1 deformation with sandstone boudins in sheared mudstone matrix. Stratal
continuity is extensively disrupted.
 The Coastal Belt of the Franciscan 403

fine-grained matrix has a scaly appearance. The the deformation features of style 1, but strata
long axes of the boudins in style 1 areas are have greater lateral continuity (Fig. 3). Pinch
generally aligned with the shear fabric in the and swell and some boudinage are present, but
shales to form a foliation. This foliation is bedding units can be traced laterally. General-
sub-parallel to the shear surfaces in the matrix ly, the mudstone matrix is little or moderately
and the elongation of the boudins is in the plane deformed, with some traces of original bedding
of shearing. The foliation in the boudins and features often evident.
the surrounding matrix generally dips steeply
eastward. Style 3
The elongation of boudins that is typical in This tectonic style includes the least-
style 1 deformation is best measured quantita- deformed rocks, which are continuous beds
tively in conglomerate beds. The aspect ratios with little structural disruption (Fig. 4). Intra-
(long dimension + short dimension) of chert stratal slump folds, local open folding, and
and limestone clasts in undeformed con- minor faulting are the only evidence of de-
glomerate beds and in the adjacent deformed formation in these rocks.
parts of these beds change from 1.5:1 to be-
tween 2 and 3:1. This elongation of well- Structural units
rounded chert clasts strongly suggests a tectonic
cause for many of the boudin shapes (Bachman Mapping and contouring the tectonic styles in
1978a). the Coastal Belt reveal several scales of de-
formation geometry. On the largest scale (Fig.
Style2 5), the Coastal Belt can be divided into two
The rocks of this tectonic style show many of units (Kramer 1976; Bachman 1978a), a west-
~ 4!; ¸. •

FI6. 3. Structural style 2 deformation. Individual beds are often discontinuous, but overall stratal
continuity of units is preserved.

FIG. 4. Structural style 3 rocks. Stratal continuity has not been disrupted.
404 S. B. Bachman
123045 '

COASTAL BELT FRANCISCAN

(Western Unit)
COASTAL BELT FRANCISCAN
(Eostern Unit)
LEGGETT PERI DOTITE

CENTRAL FRANCISCAN
MELANGE

r'''l
0 5 iO KM

....
!,I:
JKc '

39" 30
KTcw

/~ KT¢

t KTcw ~9,

39°15
1 I

Flo. 5. Structural units in the Coastal Belt. Western Unit is primarily style 1 and 2 rocks with minor
style 3, while Eastern Unit is primarily style 3 rocks. Linear bands of short wavy lines are inferred
shear zones represented by style 1 rocks. See text for definition of structural styles.

ern unit of style 1, 2 and minor style 3 rocks, occur (Figs 6 & 7 ) . In a few localities, wide
and a more eastern unit of primarily style 3 bands of sheared rocks also appear to consist of
rocks. Locally, these two units alternate across sets of these narrow deformed bands as well.
strike (Fig. 5). Smaller scale maps (Fig. 6) show These narrow bands are characterized by a
more detailed structure. transition zone with a progressive increase in
the degree of deformation from more-coherent
Western M~lange Unit beds into the boudinaged zone over a width of
A western unit exposed along the coast and 10 to 100 m (Figs 6 & 7). The narrow bands of
adjacent inland areas (Fig. 5) is recognized by highly sheared rocks probably represent indi-
the abundance of tectonic style 1 and 2 rocks. In vidual fault zones of the Coastal Belt.
current usage, the tectonic style of this unit
would categorize it as a m61ange. The remark- The structural trend in the Mdlange Unit is
able aspect of the Western Unit is the in- N W - S E . Beds dip predominantly to the east at
homogeneity across strike in structural style a high angle or are slightly overturned. Strati-
over a distance of tens to hundreds of metres graphic tops of beds face eastward. The overall
(Figs 5 & 6). Bands of sheared rocks several age progression in the M61ange Unit youngs
kilometres wide envelop less-sheared rocks westward from Cretaceous to Palaeogene,
(Fig. 5). Within the less-sheared rocks, narrow- however, suggesting imbricate stacking of
er bands of deformed rocks tens of metres wide sedimentar), units.
 The Coastal Belt o f the Franciscan 405

°'- <:.x,,.,,x~ ~" "~ A'

~.I" -

_~i _ I

0 500 m
I ~ ~ , , I

\\ \
\
\
\
\
°~ \
\\ \ \
\
I
\".2 \ 3 \
\
\ vot\X 4 Decreosing ~........,,
\
~ Deformation 0 ~'"J::..:".;,
:'!

FIG. 6. Structural styles at Mendocino Headlands, showing deformation changes .across strike at
more detailed scale than Fig. 5. See text for definition of structural styles. Section A-A' is Fig. 7.

A
N61E~
A'
synclines and eastward-dipping homoclines
(Fig. 8). Unlike the M61ange Unit, the eastern
beds are in a normal stratigraphic sequence;
they young in the direction of stratigraphic
-2OO
facing. The degree of deformation increases
-3OO
with stratigraphic depth in the Eastern Unit, as
well as towards the contact with the M61ange
Unit. The contact b e t w e e n the units ranges
FIG. 7. Structure section across Fig. 6.
Elongate pods represent structural style 1 from abrupt to gradational; it varies from gra-
rocks, short wavy lines represent structural dational over a distance of hundreds of metres,
style 2 rocks, and straight or folded con- N6BE ~ COASTAL BELT CENTRAL
tinuous lines represent structural style 3 IB {Western) (Eastern) (Weslern) {Eastern MELANGE
B'

rocks. Stippled pattern is massive sand-

stone.

Eastern Unit o
, . . . .
Wm
A n eastern unit in the Coastal Belt (Fig. 5),
defined by large areas of style 3 deformation, FIG. 8. Structure section B-B' referenced
has been d e f o r m e d locally into open folds, both in Fig. 5. Patterns are the same as Fig. 7.
406 S. B. Bachman

with gradually increasing deformation of the High-angle NW and NE trending faults

Eastern Unit toward the contact with the (trends 3 and 4) are the most abundant faults in
M61ange Unit, to faulted (Kramer 1976) with a the Coastal Belt. They are the youngest faults
complex combination of Quaternary faulting recognized. Slickensides indicate that strike-slip
(brecciated zone) and smaller-scale alternation movement predominates, with offset beds and
of tectono-stratigraphic units (Bachman 1979). conglomerate clasts suggesting right-lateral
movement. Fault gouge, up to tens of cen-
timetres thick, is common along these faults.
Faults
The origin of the shear foliation in the
In addition to the shear zones recognized in M61ange Unit apparently is unrelated to the
the M61ange Unit, discrete fault planes are also San Andreas fault system. The foliation,
evident in both the M61ange and Eastern Units. although subparallel to the NW trending San
The orientation of several hundred of these Andreas fault, is clearly offset by the youngest
faults was measured (Fig. 9) in the M61ange faults (trends 3 & 4) that may be sympathetic to
Unit. the San Andreas fault. The shear foliation also
The low-angle north- and NE dipping faults appears to pre-date the north- and NE dipping
(trend 1, Fig. 9) are offset by faults of other faults in most areas.
trends and thus appear to be the oldest faults.
Where it is apparent, the sense of movement on
faults with these trends is underthrusting of Folds
western blocks under eastern blocks. The rela-
tive ages of the north- and N E dipping faults are Folding on several scales can be recognized in
not known; the two trends were never seen with the Coastal Belt. On the largest scale,
cross-cutting relationships. kilometre- and larger-wavelength NW trending
folds expose melange and volcanic rocks in the
cores of anticlines with less-deformed Eastern
N Unit rocks on the flanks of the folds (Fig. l(I).
Smaller-scale folding is more difficult to recog-
nize except in well-exposed areas; outcrop-scale

N62E ~ COASTAL BELT LEGGETT

C (Weslern) (Eostern) (Western) (Eastern) PERIDOTITE

t ~" ~-T~-( / ~--.~ ~- t

0 5

km

F1o. 10. Structure section C-C' referenced

in Fig. 5. Patterns are the same as Fig. 7.

folds in well-exposed melange rocks fall into

two groups. Folds with SE plunging axes often
Fio. 9. Lower hemisphere equal area pro-
jection of poles to fault planes, Melange verge to the SW. The plunge of these folds
Unit. Densities of orientations are con- varies from less than 5 ° to 30 ° . The second
toured as 1, 3, 5, and 7% of total popula- group of folds plunge 10°-50 ° to the NE. Two of
tion in 1% of hemisphere area. these folds verge to the NW; the others are
upright. Many of these smaller-scale folds are
isoclinal.
The moderately NW dipping faults (trend 2, Folds in the melange are commonly sepa-
Fig. 9) offset the north- and N E dipping faults rated by faults and shear zones. The limbs of
(trend 1), but are in turn offset by the other isoclinal folds are commonly sheared, leaving
fault trends. The sense of displacement of these overturned limbs boudinaged (Bachman 1978a)
faults is not evident. or fold hinges floating in the sheared matrix.
 The Coastal Belt of the Franciscan 407

The folding and shearing appear to be contem- precipitation after the formation of the volcanic
poraneous. Early-formed fractures are folded rocks cannot be ruled out.
during recurring folding and movement along Volcanic rocks are present in the M61ange
the shear zones. Locally, sand has flowed into Unit in blocks as large as 100-200 m across.
the mudstone matrix in the isoclinal fold hinges. They are arranged in somewhat linear bands
The latest period of folding, presumably parallel to the tectonic fabric. The volcanic
Quaternary in age, slightly tilted the uplifted rocks include pillow lavas, pillow breccias, and
marine terrace deposits of probable Plio- hyoloclastics with isolated pillows. One such
Pleistocene age exposed along the coast (Kram- sequence is 10 m thick. As already described,
er 1976). The youngest, NW-trending faults limestone blocks are prevalent in the rocks.
offset these terraces and truncate all the earlier O ' D a y (1974) reported that chemical analyses
folding. of some of the rocks are equivalent to the
compositions of spilite, albite diabase, and lab-
Soft sediment deformation radorite diabase. Some of the rocks are por-
phyritic, with albite, olivine, and augite usually
The wispy, pull-apart, and somewhat chaotic replaced by carbonate and magnetite. The
appearance of many Coastal Belt m61ange groundmass is chloritized with minor amounts
rocks suggests deformation of somewhat uncon- of pumpellyite and pyrite. Quench crystals are
solidated sediments (Kleist 1974b). The elonga- common. Zeolite and silica veins have been
tion of chert clasts, thrust stacking origin of partially replaced and are cut by calcite veins.
some boudins (Bachman 1979; Bachman & The contact between the igneous rocks and
Allmendinger 1980), and lateral extent of de- the Coastal Belt sediments is usually obscured,
formed rocks indicate that the boudinage and but in several places shearing along a fault zone
pull-apart features are probably tectonic rather is evident. Within 6-8 m of the contact, both
than sedimentary in origin. However, intra- sediments and volcanic rocks are brecciated,
stratal folding, chaotic beds, and the swirled with a reduction in sand grain size along dis-
mixing of sediments were probably caused by crete shear zones. The sheared volcanic mate-
slumping and gravitational flowage on the mar- rial along the contact contains slickensided pil-
gin slope (cf. Cowan 1981). lows and sandstone and shale boudins. The
In the Eastern Unit, the only soft sediment igneous rocks may be displaced exotic blocks
deformation features apparent are slump folds, tectonically emplaced into the Coastal Belt
load structures, and flame structures. sediments.
The thickness of the Coastal Belt has been
variously estimated as up to 7500 m (O'Day
Coastal Belt stratigraphy 1974), up to 10 000 m (Kleist 1974a), and
6000 m (Kramer 1976). In convergent margin
Sandstone and shale comprise the vast majority depositional settings such as subduction com-
of Coastal Belt stratigraphy. Minor amounts of plexes, deformation and sedimentation are
conglomerates, volcanic rocks, limestone and often synchronous; consequently, the thickness
chert are present also. Sandstone to shale ratios of tectono-stratigraphic units cannot be mea-
exceed 1:1 in most areas; in local areas, they sured directly. The structural thickness of stack-
approach 8:1. ed imbricate sheets can far exceed the strati-
Limestones occur as clasts in conglomerates, graphic thickness. The structural thickness of
as blocks in volcanic rocks, as thin, discon- the M61ange Unit may exceed 20 km (Fig. 8).
tinuous beds or boudins in shale units, or as Kramer's 6000 m calculation is the apparent
lensoid- to ball-shaped concretions. Grey and thickness of the homoclinal Eastern Unit west
brick-red limestone blocks up to 1 m across are of Willits. With structural repetition primarily
found within the volcanic rocks described be- by open folding, this 6000 m represents a max-
low. The limestones have been recrystallized imum present stratigraphic thickness of the
and are now primarily sparite. The contacts Eastern Unit. The thickness of sediment lost to
between the limestones and igneous rocks are erosion is unknown. Thermal alteration of
somewhat irregular and have the appearance of organic matter in the Eastern Unit (Bachman
intrusive contacts, with limestone blocks float- 1978a, 1979) suggests that several kilometres of
ing in the igneous rock, although no chilled overburden of unknown age has been removed
margins are evident in the igneous rocks. The by erosion.
brick-red limestones may owe their colour to The age of the Coastal Belt has been deter-
mineralization during igneous activity. Second- mined using a combination of palynomorph
ary formation of carbonate from pore-water assemblages (Evitt & Pierce 1975; Kleist 1975;
408 S. B. B a c h m a n

Bachman 1978a, 1979) and foraminifera (Bach- The beds of facies A range from 2 m to 30 m
man 1978a, 1979; McLaughlin et al. 1979). or more in thickness. They are mostly medium
These fossil data suggest that the M61ange Unit or coarse sandstone, but minor conglomerates
rocks range in age from Late Cretaceous on and pebbly sandstones also are present. The
the east side to Oligocene or Miocene on the beds commonly are amalgamated; locally no
west side. Stratigraphic tops of beds face east- internal features or textural changes are evident
ward, suggesting imbricate stacking of in the beds. Facies A conglomerates range from
sedimentary units. Even younger Coastal Belt clast supported to matrix supported; both dis-
rocks may be found west of the Mendocino organized and organized beds occur.
coast area, because subduction continued into Channelling of facies A rocks into the under-
the Plio-Pleistocene (Atwater & Molnar 1973). lying sediments is not uncommon. Rip-up clasts
The Eastern Unit ranges in age from Cre- of the underlying sediments are often found
taceous to Eocene. The youngest fossil assem- near the base of the A beds.
blages occur in the structurally highest posi- Facies B sandstone beds are difficult to dis-
tions, suggesting that normal stratigraphic rela- tinguish from those of facies A in many cases;
tionships exist in the Eastern Unit. In all cases dish structures, diffuse laminations, fluid-
where the resolution of palaeontological ages escape pipes and more frequent shale partings
permits interpretation, the Eastern Unit is are indicative of facies B rocks when present.
younger than the underlying or immediately Facies B beds are up to 6 m thick, but the
adjacent m61ange. thicker beds are amalgamated. Commonly,
facies A and B rocks are found in the same
Facies analysis locality. Channellized facies A rocks are often
associated with facies B, C, and D rocks (Fig.
I have used the facies classification scheme 11). These facies may occur in megasequences
for turbidites and related resedimented deposits (Ricci-Lucchi 1975), with beds thinning and
(e.g. Walker & Mutti 1973; Bouma & Nilsen fining upwards.
1978) to characterize Coastal Belt sediments. Facies E overbank and levee deposits also
However, in most cases it is not appropriate to occur with some of the A, B, C, D associations
interpret the associations, or groupings, of (Fig. 11), or alone with facies A rocks. Facies F
these facies using a deep-sea fan facies associa- slumps and pebbly mudstones also are associ-
tion model: submarine fan facies terminology is ated with thick facies A rocks (Fig. 11). The
only appropriate if the geometry of a fan com- pebbly mudstones range in thickness from
plex can be reasonably documented. Instead a 1.5 m to 8 m. A section that contains the
convergent margin sedimentation and facies largest volume of fine-grained rocks in the
model (Bachman 1979; Underwood et al. 1980; Western Unit has the following facies sequence:
Underwood & Bachman 1981) is used for inter- > 2 0 m A , B--* 1 0 0 m G , D ~ 1 0 m A ,
pretation of facies associations. B - - ~ 8 0 m G , minor D ~ > 1 0 0 m D , C , B
(coarsening- and thickening-upward sequence).

M~lange Unit facies Typical facies B beds in the M61ange Unit are
Bedding disruption restricts measurement of associated with facies C beds. These facies may
detailed stratigraphic sections to less-deformed be separated by facies G mudstones, or inter-
areas, although sequences of facies are very bedded with facies D rocks, usually in a thin-
often on a scale smaller than the size of the ning-upward sequence. Facies B, C, F and G
coherent blocks (tens to hundreds of metres rocks are also associated with channels,
thick) within the long linear shear zones. Thus, grooves, and mudstone rip-up clasts up to 4 cm
gross stratigraphic patterns, traceable laterally, by 30 cm in size at the base of B beds. The
remain despite the shearing. The sub-parallel facies F slumps have sets of discrete glide planes
orientation of shear surfaces with bedding may at the base of each slump that cut across lower
aid preservation of these patterns. chaotic beds. Flame structures and ball and
The full range of turbidite and associated pillow loading are common in the sandstone
facies is present in the Western M61ange Unit. units.
The thick, massive sands of facies A and B are Facies B rocks may also have the association
particularly abundant in this unit, as well as the B, C, D, E and minor F. The facies E levee
Eastern Unit. Sandstone/mudstone ratios in the deposits have abundant starved ripples and very
M61ange Unit range from somewhat less than low-angle reverse fault blocks that are inter-
one to as high as 16. The overall ratio is preted as levee slumps (T. Nilsen, pers.
approximately 2 or 3. comm.). Small intrastratal folds also indicate
 Upper Eocene-Lower Oligocene.?
_l Age
67 meters
I0 m ~ I I I 1 l
wt Thickness
m m ~ Fan facies
r~

Chonnelizedupperfan-slope or trench Depos|tionol

environment

(%

FIG. l 1. R e p r e s e n t a t i v e s t r a t i g r a p h i c s e c t i o n in a p o r t i o n o f the M 6 l a n g e Unit.

410 S. B. Bachman

slumping of these levee deposits. Levee de- surface. This antecedent has not been reported
posits and facies B rocks are also found in elsewhere in turbidite facies and the process of
association with C and D beds in the following formation is not clear.
sequence: A 5 m thick facies F pebbly mudstone that
fills a channel cut into lower beds contains a
C, D - + B, E--+ C. basal layer of mudstone rip-up clasts overlain
by blocks of deformed sandstone in the matrix.
The sandstone clasts apparently were eroded
Eastern Unit facies from only semi-lithified beds. A common associ-
Facies relationships in the Eastern Unit are ation is channellized facies A and/or B with
clearer because the rocks are little deformed. facies E that thin- and fine-upward to C and/or D
along the South Fork of the Eel River, facies rocks (Fig. 13) and may correspond to
lateral exposures can be traced over several progressive channel fill and abandonment. The
kilometres as the river meanders. The full range sequence may be capped by facies D and G
of turbidite and associated facies also is present (Fig. 13) and the fining-upward sequence may
in the Eastern Unit. Thick massive sands of go directly from A, E, D, G. A slight modifica-
facies A and B are abundant. Sandstone/mud- tion of this trend involves the sequence chan-
stone ratios of sequences in the Eastern Unit nellized A, G, E, B, C with a thinning-u~_ward
range from 2.5 to 13. Amalgamated sandstone sequence. The facies C and D beds are laterally
beds, with thin shale partings, approach 85 m in continuous over tens of metres; exposures pre-
thickness (Fig. 12). clude tracing them further. Facies A beds may
Facies A conglomerates are similar to those also be encased in hemipelagic deposits, giving
in the Melange Unit. In one area, a 25 m thick a G-A-G sequence.
facies A sandstone onlaps against an erosional Several thickening- and coarsening-upward
surface cut several metres into overbank de- sequences are evident in the Eastern Unit.
posits. Elsewhere, a channel cuts into folded, Laterally continuous facies C and D sediments
disrupted sandstone beds that were apparently are overlain by G and E beds, then covered by
part of a slump. thick channelled facies A and B beds.
Dune cross-bedding up to 55 cm thick is Similar to this sequence is the coarsening-
common within levee and overbank deposits upward sequence facies B, C, D, G overlain by
(facies E). Locally, a sequence of graded sand, channellized facies A and B rocks. The Eastern
parallel laminated sand, convolute laminations, Unit consists of many mixed facies as well: A,
and silt having the appearance of a Bouma T~, d B, C with minor D, E; A, B, F, C, D; A, F, G,
sequence is deposited on an inclined dune bed E with some B; and A, B, C, E, G.

Fro. 12. Thick massive sandstones of Eastern Unit. Note automobile in lower left for scale.
 U, C r e t a c e o u s to E o c e n e ? Age

56 meters Thickness
E I
3
~. o
Fan facies
o

Depositional
Channelized mid-fan
environment
. - ....

Thinning

FIG. 13. Stratigraphic section in a portion of the Eastern Unit.

4~
 412 S. B. B a c h m a n

the Eastern Unit appears to have a composi-

Sedimentary petrology
tional sub-field with more feldspar and less total
Using the petrographic techniques of Dickin- quartz than the bulk of the rocks; these rocks
son (1970) and Graham et al. (1976), 99 sand- plot within the plutonic arc provenance field of
stones were analysed (Bachman 1979). In addi- Dickinson & Suczek.
tion, a detailed study of quartz types aided in The remaining rocks appear to have a mixed
provenance interpretation. Triangular diagrams provenance. Volcanic detritus is present in vir-
of modal percentages of the various mineralo- tually all the rocks in varying amounts. Most
gical components were compared to Dickinson polycrystalline quartz grains are finely crystal-
& Suczek's (1979) tectonic provenances for line with metamorphic textures. Chert is a
sandstones; this comparison indicates a mixed ubiquitous component. Quartz-mica schist frag-
tectonic provenance for Coastal Belt sand- ments and detrital epidote are not uncommon;
stones. blueschist fragments are present, but rare.
Data from Tertiary sandstones are plotted in The variety of conglomerate clasts includes
two separate groupings (Fig. 14). Most sam- widely varying proportions of red, green, and
ples, from both M61ange and Eastern Units, are black chert, sandstones of varying composi-
quartz- and feldspar-rich; some m61ange sam- tions, mudstone and siltstone, volcanic rocks of
ples are lithic-rich and plot nearer the L pole. varying compositions, plutonic rocks, lime-
Cretaceous sandstones, not shown, display stone, and vein quartz. Sandstone clasts vary in
similar groupings. composition within any one conglomerate.
Many of the sandstone clasts are similar in
Q composition to the sandstone matrix. In other
@ Eoc.-Olicj.M~lanqe cases, compositions are dissimilar; a M61ange
Unit conglomerate matrix enriched in volcanic
lithic fragments commonly contains sandstone
2"_~' Eoc.-Oliq. Slope Basin clasts with low volcanic content. The sandstone
clasts have elongate or irregular shapes, sug-
L.Tertiary Slope Basin gesting erosion of only partly lithified sedi-
~ ~ VolcanicLithic-Rich ments.
The volcanic clasts are for the most part
porphyritic and range in composition from aci-
dic to basaltic. The majority of the clasts are
F L probably andesitic. More mafic clasts common-
Fro. 14. Triangular diagram of modal per- ly have remnants of rinds suggesting a pillow
centage of quartz, feldspar, and lithic frag- origin. The volcanic clasts are included in sand-
ments in Coastal Belt sandstones of Terti- stone matrices of varying compositions; there is
ary age. no correlation between volcanic lithic-rich
matrices and the relative abundance of volcanic
clasts of any composition.
The anomalous Mdlange Unit rocks are
abundant in volcanic lithic fragments and plot Palaeocurrent analysis
in the volcanic-arc to transitional-arc fields of
each of Dickinson & Suczek's (1979) diagrams. The amount of deformation in the M61ange
These volcanic-rich sandstones are all located in Unit precludes a straightforward analysis of
the M61ange Unit in somewhat linear outcrop palaeocurrent data. Only data from the less-
patterns. These sandstones also contain small deformed rocks (style 2 and 3) were used for
amounts of polycrystalline quartz types such as analysis. Measured palaeocurrent indicators in-
polygonized, partly recrystallized (bimodal), clude flute casts, groove and tool casts, and
and schistose texture quartz that are most likely ripple and dune cross-laminations. The direc-
from metamorphic terranes; volcanic-rich con- tions of overturning of intrastratal folds were
glomerate matrices contain clasts of varying measured as palaeoslope indicators. The direc-
lithologies. Thus, the volcanic lithic-rich sand- tion of overturning of flame structures may
stones appear to be derived from a distinct indicate current directions or down-slope direc-
volcanic source, and other detrital material tions. Fig. 15 shows the calculated palaeocur-
from non-volcanic sources was mixed locally. rent directions. Grooves and symmetric ripples
This volcanic source contributed detritus spor- are plotted with bimodal directions.
adically from Late Cretaceous into the Oli- Sole marks are poorly preserved in the
gocene or later. The Eocene-Oligocene part of M61ange Unit because of shearing along the
 The Coastal Belt o f the Franc&can 413

Interpretation of structural styles and trends

~(28) ~(241 The M61ange Unit is probably a Late Cre-

Ripple Cross Lominalions , M~lonqe taceous to late Eocene or younger accretionary
prism, with the youngest sediments towards the
seaward side of the prism (Bachman 1978a).
The Eastern Unit has many of the characteristics
of trench slope basins (e.g. Moore & Karig
~(28) W~IP ell)
1976). It is much less deformed than the
FCbpplCross
e Lorn,nohons, Slope Bosms Slump Folds , M~loncje
M61ange Unit, slightly younger than the adja-
( Downslope Di,echon )
cent m61ange, and its contacts with the m61ange
IN are partly faulted, partly gradational in degree
of deformation. It is unlikely that many seg-
Z(IS) PALEOCURRENT ments of the Eastern Unit were juxtaposed with
FluleCosts,Slope Bosins
AND PALEOSLOPEPATTERNS the melange after cessation of subduction be-
cause of the complex structural patterns shown
FIG. 15. Palaeocurrent and palaeoslope in Figs 5 & 10, the complementary age rela-
measurements from the Coastal Belt. The
number of measurements is shown in par- tionships of the Eastern Unit overlying the
entheses. m61ange, and the partly gradational contacts
between units. However, such a juxtaposition
of portions of the outcrops by either later thrust
bases of the beds. Only two or three groove or strike-slip faulting cannot be entirely ruled out.
marks could be measured; the small number are The linear patterns of boudinaged and
statistically insignificant. The much more abun- sheared rocks in the M61ange Unit probably
represent shear zones in the accretionary prism
dant ripples suggest transport to the south. Less
than one-fourth of the ripple directions were (Figs 5 & 6). The pattern of age determinations
measured on inferred overbank deposits; cur- in the unit suggests imbricate stacking; these
rent directions of flows that overtop channels sheared rocks may be wide zones of imbricate
would be expected to flow in divergent direc- thrusts along which sediments were detached
tions from the channel (Walker 1977), but from the oceanic plate and accreted to the
generally downslope. The general sense of cur- leading edge of the overriding continental
rent flow in the M61ange Unit apparently was to plate. The shear foliation now dips steeply to
the south and SW. the east or is slightly overturned; if landward
Flute casts, groove casts, and ripple lamina- rotation during progressive uplift and deforma-
tions all suggest current direction was toward tion of the accretionary prism is taken into
the SW in the Eastern Unit (Fig. 15). The account, the imbricate thrust zones dipped
gently towards the east before tilting. Palaeon-
minor NW directed ripple indicators may repre-
sent local turbidity flow towards the NW or may tological age resolution does not permit identi-
fication of individual thrust sheets in the
be divergent overbank currents.
m61ange.
Isoclinal folds in the M61ange Unit probably
developed contemporaneously with the shear-
ing. The seaward-verging folds are similar in
Palaeogeography and tectonics of the orientation to those recognized developing near
Coastal Belt the toe of modern accretionary prisms, im-
mediately after sediments are detached from
Reconstructed Late Mesozoic and early Tertiary the oceanic plate (Seely et al. 1974; Kulm &
plate motions indicate oblique subduction of Fowler 1974; Moore, J. C. & Karig 1976).
the Farallon plate beneath the North American Slump folds in the M61ange Unit, presumably
plate (Coney 1981), with rates of convergence formed on the oversteepened, tectonically ac-
varying from 14 cm yr -t in the period 80 to tive trench slope, are also overturned seaward
45 Ma to 8 cm yr -1 after 45 Ma. The increased (SW).
convergence rate may have resulted in flatten- The volcanic and intrusive rocks in the
ing of the dip of the Benioff zone, eastward Coastal Belt are all located in the most sheared
shift (Coney & Reynolds 1977) or temporary parts of the m61ange, forming linear bands
disappearance of the volcanic arc, and eastward parallel to the shear foliation. Volcanic and
shift or disappearance of the locus of blueschist intrusive rocks can be introduced into the accre-
metamorphism (Bachman 1978b). tionary prism as offscraped fragments of sea-
414 S. B. B a c h m a n

mounts, aseismic ridges, oceanic fault blocks, Bachman 1981), representing either filling of
or as intrusions into the trench or prism along the trench or filling of elongate slope basins.
leaky transforms or triple junctions (see discus- The deformation and imbricate stacking sequ-
sion in Bachman 1978a). The association of ence of these M61ange Unit rocks suggests an
carbonates with volcanic blocks suggests that origin in the trench or on the lower slope where
some of the igneous rocks are offscraped sea- deformation is greatest; lower slope basins are
mounts or oceanic crust formed above the generally not filled with thick, coarse clastics
CCD. (see Underwood & Bachman 1981), so the
The pattern of Eastern Unit exposures along channellized m61ange sandstones are most like-
section B-B' (Fig. 5) indicates basin erosional ly trench deposits. No definitive criteria are
remnants exposed between outcrops of present for interpreting many of the other
m61ange rocks. These Eastern basin deposits M61ange Unit sediments. The absence of any
(Figs 8 & 10) may be either erosional remnants thick pelagic or hemipelagic deposits suggests
of a large slope basin that once covered much of that any fine-grained slope or trench deposits
the Western M61ange Unit accretionary prism, were overwhelmed by coarse clastics and that
or they may have been deposited as several abyssal plain deposits may have been selectively
discrete basins, each ponded behind a ridge of subducted (Moore 1975) and/or underplated
the uplifted accretionary prism. On modern beneath the margin at a level deeper than the
trench slopes, several ponded basins may even- present erosional surface.
tually fill and overtop an uplifted accretionary Eastern Unit sedimentary facies associations
ridge to form a larger basin (G. F. Moore & resemble well-developed slope basin and chan-
Karig 1976). nel facies (Underwood & Bachman 1981). The
Other tectonic settings for part of the Eastern thick channel sands may have been deposited in
Unit could be possible if the postulated oblique feeder channels that led into the trench and
subduction along the coast was decoupled into a distribution networks within the slope basins;
subduction zone/transcurrent fault pair evident the outcrop pattern of several conglomerate
in some western Pacific subduction complexes bodies suggest transport at an angle to the
(Fitch 1972; Walcott 1978; Lewis 1981). Part of margin. The SW directed palaeocurrent indica-
the Eastern Unit would then have been depo- tors in the Eastern Unit probably represent
sited in a basin similar to the present continental flow in basins parallel to the margin, with local
borderland basins of southern California (e.g. feeder channels at higher angles to the margin.
Crowell 1976). However, no evidence of older The Eastern Unit also has well-developed
large-scale strike-slip faults within the Eastern coarsening- a.nd thickening-upward sequences
Unit has been reported. Borderland basins and that can be interpreted as prograding deposi-
marginal ocean basins must be enclosed on the tional lobes of small fans developed in slope
seaward side by an uplifted basement terrane basins. The small amount of typical slope and
older than the accumulated sediments. No canyon facies (i.e. G -+ A ---+ G), particularly
rocks older than Upper Cretaceous Coastal hemipelagic muds, suggests that large amounts
Belt rocks have been found on the seaward side of coarse clastic material were available from
of the Eastern Unit. the source area and that sedimentation rates
were fairly high. Continuing fieldwork in the
northern section of the Coastal Belt expo-
Interpretation of Coastal Belt lithofacies sures confirms the presence of more typical
thick hemipelagic slope muds to the north of
With the aid of a covergent margin sedi- the sand-rich slope basins. Poor palaeontologic-
mentation and facies model (Bachman 1979; al age resolution and structural complexities
Underwood et al. 1980; Underwood & Bach- preclude sedimentation rate calculations. High
man 1981), the depositional setting of many of energy flow in the basins could have prevented
the Coastal Belt sediments can be postulated. low energy deposition, and the fine material
Many of the M61ange Unit sedimentary facies was transported to deeper water (Bachman
associations represent channellized flow, with 1978a). High sand/mud ratios are also present
local channel cutting and thinning- and fining- in another inferred Franciscan trench slope
upward sequences. basin in central California (Howell et al. 1977).
The areal distribution and current direction
indicators of these sandstones are parallel to the Interpretation of provenance and sediment
margin. Both trench and slope environments
distribution
can have channellized sediment dispersal pat-
terns parallel to the margin (Underwood & It has been shown that transport of sediments
 The Coastal Belt o f the Franciscan 415

in the Coastal Belt was generally toward the entire arc would later be subducted); (4) a
the south and SW. Several rock types, in- western arc is not compatible with the evidence
cluding metamorphic, plutonic and volcanic for a west-facing Coastal Belt subduction
rocks, and recycled oceanic sediments, contri- complex; and (5) volcanic detritus is found in
buted detritus to the Coastal Belt. Inferred rocks of Late Cretaceous, early Tertiary, and
trench deposits are locally very volcanic lithic- Eocene to Oligocene age, suggesting that the
rich. Source areas previously suggested for source was stable over a long period of time.
Coastal Belt sediment include the Sierra Neva- However, arc volcanism was widespread in
da, Klamath Mountains, and uplifted older central and southern Oregon during Coastal
Franciscan terranes (O'Day 1974; Kleist 1974; Belt deposition (Snyder et al. 1976) and the
Kramer 1976). area was topographically high (Axelrod 1968).
The Klamath mountains to the NE of In the lower Tertiary forearc basin of western
the Coastal Belt contain source rocks that Oregon, volcanic detritus was transported long
would account for virtually all Coastal Belt distances northward from this southern Oregon
detritus. The abundance of metamorphic rock arc (Chan & Dott 1980). It is probable that
types over plutonic types in the Klamaths would these volcanic arc rocks also shed detritus to-
also account for the predominance of metamor- wards the trench; during volcanic pulses, cur-
phic quartz types in the Coastal Belt. The rents may have brought this detritus similar
direction of palaeocurrent indicators also distances south along the trench to the site of
suggests transport from the NE, the present Coastal Belt deposition. In the central part of
position of the Klamath Mountains relative to Coastal Belt exposures, the volcanic-rich sand-
the Coastal Belt assuming no large-scale lateral stones are restricted to possible trench environ-
transport of Coastal Belt rocks. Lower Tertiary ments in the Western M61ange Unit.
current directions in streams draining the Kla-
math Mountains were towards the south (Dick-
inson et al. 1979). Thus, the Klamath Moun- Discussion
tains must be considered as a possible major
source for Coastal Belt sediments. However, The Coastal Belt subduction complex shares
large-scale lateral translations parallel to the many structural and stratigraphic relationships
margin may have separated the Coastal Belt with other convergent margins. Imbricate
from source rocks similar in composition to the stacks younging oceanward, m61ange terrane,
Klamath Mountains, but at large distances from and slightly younger less-deformed rocks repre-
the present site of the Coastal Belt. senting trench slope basins are all present. The
Palaeomagnetic studies by K. Verosub and structural mapping technique used here has
myself indicate resetting of remnant magnetism further defined the geometry of deformation
after folding, possibly at slightly elevated within this subduction complex.
temperatures (100°C or more) for a long period The recognition of trench and trench slope
of time (Bachman 1979). Thus, there are at deposits in ancient subduction complexes is a
present no data suggesting large lateral transla- difficult problem. Using structural and
tions of Coastal Belt rocks. sedimentary facies models it is possible to de-
The source for the volcanic lithic-rich sand- termine the depositional setting of some of the
stones has been proposed to be a western less-deformed subduction complex rocks.
andesitic arc that has subsequently been sub- When the rocks are highly deformed and
ducted in part (Beutner 1977), based on one depositional sequences become fragmented,
DSDP site drilled off northern California that these criteria are less useful. The sandstone
penetrated volcanic rocks similar in composi- petrology, however, may reflect differing
tion to andesite. This hypothesis is unlikely source areas and distribution patterns of sedi-
based on several arguments: (1) no other evi- ments deposited on the abyssal plain-trench and
dence, sedimentary or tectonic, exists to sug- trench slope. In the Coastal Belt, the volcanic
gest a western arc; (2) the DSDP site is west of lithic-rich sandstones of the Western M61ange
the San Andreas fault, so the rocks formed as Unit may have received volcanic detritus by
much as 1000 km south of their present position long distance transport along the trench. Thus,
relative to the Coastal Belt; (3) if the arc was with the proper sedimentological setting and
1000 km long so that the northern end could lack of extensive stratigraphic dismemberment,
contribute detritus to the Coastal Belt, frag- the complex structural and stratigraphic rela-
ments of the arc would now be exposed all tionships in subduction complexes may be part-
along the coast of California (it is unlikely that ly reconstructed.
416 S. B. Bachman

References
ATWATER, T. & MOLNAR, P. 1973. Relative motion of plate tectonics. Spec. Pap. geol. Soc. Am. (in
the Pacific and North American plates deduced press).
from sea-floor spreading in the Atlantic, Indian - & REYNOLDS, S. J. 1977. Cordilleran Benioff
-

and South Pacific oceans. In: KOVACH, R. L. & zones. Nature. London, 270, 403-6.
NUR, A. (eds). Proc. Conf. Tectonic Problems COWAN, D. S. 1981. Deformation of partly dewatered
San Andreas Fault System. Univ. Stanford Publ. and consolidated Franciscan sediments near
geol. Sci. 13, 136-48. Piedras Blancas Point, California (this volume).
AXELROD, D. I. 1968. Tertiary floras and topographic CROWELL, J. C. 1976. Implications of crustal stretch-
history of the Snake River Basin, Idaho. Bull. ing and shortening of coastal Ventura basin,
geol. Soc. Am. 79, 713-34. California. In: Aspects of the geologic history of
BACHMAN, S. B. 1978a. A Cretaceous and early the California continental borderland. Pacif. Sec.
Tertiary subduction complex, Mcndocino coast, Misc. Publ. Am. Assc. Petrol. Geol. 24, 365-82.
northern California. In: HOWELL, D. G. & DICKINSON, W. R. 1970. Interpreting detrital modes
McDOUGALL, K. A. (eds). Mesozoic of graywacke and arkose. J. sediment. Petrol. 40,
Paleogeography of the Western United States. 695-707.
Pacif. Coast Paleogeogr. Syrup. Pacif. Sec. Soc. --~Q SUCZEK, C. A. 1979. Plate tectonics and
econ. Paleontol, Mineral. 2, 419-30. sandstone compositions. Bull. Am. Assoc.
1978b. Latest Cretaceous and early Tertiary Petrol. Geol. 63, 2164-82.
subduction in northern California--fundamental -- , INGERSOH., R. V. & GRAIIAM, S. A. 1979.
changes from Late Mesozoic. Abstr. Prog. geol. Paleogene sediment dispcrsal and paleotectonics
Soc. Am. 10, 361-2. in northern California. Bull. geol. Soc. Am. 90,
1979. Sedimentation and margin tectonics of the 1458--528.
Coastal Belt Franciscan, Mendocino coast, north- Evrrr, W. R. & PU-R('I-, S. T. 1975. Early Tertiary
ern California. Thesis, PhD, Univ. California, ages from the coastal belt of the Franciscan
Davis, 166 pp. (unpubl.) Complex, northern ('alifornia. Geology, 3, 433-
& AI.IMENDINGER, R. 1980. Subduction com- 6.
plex deformation related to thrust ramps--a FITCH, T. J. 1972. Plate convergence, transcurrent
possiblc model. Abstr. Prog. geol. Soc. Am. 12, faults, and internal delk)rmation adjacent to
380-1. Southeast Asia and the western Pacific. J. geo-
BAILEY, E. H., IRWIN, W. P. & JONES, D. L. t964. phys. Res. 77, 4432-60.
Franciscan and related rocks, and their signifi- (}RAIIAM, S. A., IN(;I-RSDI.I., R. V. & DICKINSON, W.
cance in geology of Western California. Bull. R. 1976. Common provenance fl)r lithic grains in
Calif. Div. Mines Geol. 183, 177 pp. Carboniferous sandstoncs from Ouachita Moun-
BFRKLAND, J. O., RAYMOND, L. A., KRAMI-R, J. C., tains and Black Warrior Basin. J. sediment.
MOORES, E. M. & O'DAY, M. 1972. What is Petrol. 46, 62(~32.
Franciscan? Bull. Am. Assoc. Petrol. Geol. 56, HowvJJ, D. G., VI'I)DI'~R, J. C., M('LF.AN, H., JOYCE,
2295-302. J. M., CLARK, S. t[. JR & SMrFtt, G. 1977.
BI-:UrNER, E. C. 1977. Evidence and implications of a Review of Cretaceous geology, Salinian and
Late Cretaceous-Paleogene island arc and mar- Nacimiento blocks, Coast Ranges of central
ginal basin along the California coast. Abstr. California. In: HDWELi~, D. G. (ed). Pacif.
Prog. geol. Soc. Am. 9, 389. Coast Paleogeogr. Field Trip Guide. Pacif. Sec.
BLAKE, M. C. JR & JONES, D. L. 1974. Origin of Soc. econ. Paleontol. Mineral. 2, 1-46.
Franciscan melanges in northern California. In: KLV:lST, J. R. 1974a. Geology of the Coastal Belt,
DOTr, R. H. JR & SHAVER, R. H. (eds). Modern Franciscan Complex, near Ft. Bragg, California.
and Ancient Geosynclinal Sedimentation. Spec. Thesis, PhD, Univ. Texas, Austin, 133 pp.
Publ. Soc. econ. Paleontol. Mineral. Tulsa, 19, 1974b. Deformation by soft-sediment extension
345-57. in the Coastal Belt, Franciscan Complex. Geo-
& -- 1978. Allochthonous terranes in north- logy, 2, 501-4.
ern California?--a reinterpretation. In: - 1975. Variation and distribution of sandstone
-

HOWl.;LL, D. G. & McDOUGALL, K. A. (eds). composition in the Coastal Belt, Franciscan

Mesozoic Paleogeography of the Western United Complex, near Ft. Bragg, California. Abstr.
States. Pacif. Sec., Soc. econ. Paleontol. Miner- Prog. geol. Soc. Ant. 7, 335-6.
al. Pacif. Coast Paleogeogr. Syrup. 2, KRAMER, J. C. 1976. Geofl~gy and tectonic implica-
397-400. tions of the Coastal Belt Franciscan, Ft. Bragg-
BOUMA, A. H. & NILSEN, T. H. 1978. Turbidite facies Willits area, northern Coast Ranges, California.
and deep-sea fans--with examples from Kodiak Thesis, PhD, Univ. California, Davis, 128 pp.
Island, Alaska. Proc. Offshore Technol. Conf. - 1977. Cenozoic tectonics (plate motions) of the
-

1,559-70. Coastal Belt Franciscan, northern California

CHAN, M. A. & DON, R. H., JR 1980. Deep-sea fan Coast Ranges. Abstr. Prog. geol. Soc. Am. 9,
deposition in an Eocene forearc basin, western 448-9.
Oregon. Abstr. Prog. geol. Soc. Am. 12, 401. KVLM, L. D. & FOWLER, G. A. 1974. Oregon con-
CONEY, P. J. 1981. Mesozoic-Cenozoic Cordilleran tinental margin structure and stratigraphy: a test
 The Coastal Belt o f the F r a n c i s c a n 417

of the imbricate thrust model. In: BURK, C. A. turbidite formations of northern Apennines
& DRAKE, C. L. (eds). Geology of Continental (Italy). J. sediment. Petrol. 45, 3-43.
Margins, 261-83. Springer-Verlag, New York. SEELY, D. R., VAIL, P. R. & WALTON, G. G. 1974.
LEWIS, K. B. 1981. Quaternary sedimentation of the Trench slope model. In: BVRK, C. A. & DRAKE,
Hikurangi oblique-subduction and transform C. L. (eds). Geology of Continental Margins,
Margin. In: BALLANCE, P. F. & READING, H. G. 249-60. Springer-Verlag, New York.
(eds). Sedimentation in Oblique-slip Mobile SNYDER, W. S., DICKINSON, W. R. & SILBERMAN, M.
Zones. Spec. Publ. int. Assoc. Sediment. 4, L. 1976. Tectonic implications of space-time
171-89. Blackwell Scientific Publications, patterns of Cenozoic magmatism in the western
Oxford. United States. Earth planet. Sci. Lett. 32, 91-
MCLAUGHLIN, R. J., KLING, S. A., POORE, R. Z., 106.
McDoUGALL, K., BEUTNER, E. C. & OHLIN, H. UNDERWOOD, M. B. & BACHMAN, S. B. 1981.
N. 1979. Post-Middle Miocene microplate accre- Sedimentary facies associations within subduc-
tion of Franciscan Coastal Belt rocks to northern tion complexes (this volume).
California. Abstr.Prog.geol. Soc. Am. 11,476-7. UNDERWOOD, M. B., BACHMAN, S. B. te'zSCHWELLER,
MOORE, G. F. & KARIG, D. E. 1976. Development of W. J. 1980. Sedimentary processes and facies
sedimentary basins on t h e lower trench slope. associations within trench and trench-slope set-
Geology, 4, 693-7. tings. In: FIELD, M. E., BOUMA, A. H. &
MOORE, J. C. 1975. Selective subduction. Geology, 3, COLBURN, I. (eds). Quaternary Depositional En-
530-2. vironments on the Pacific Continental Margin,
-- & KARIG, D. E. 1976. Sedimentology, structural Pacif. Sec. Soc.econ. Paleontol. Mineral. 211-29.
geology, and tectonics of the Skikoku subduction WALCOrr, R. I. 1978. Geodetic strains and large
zone, southwestern Japan. Bull. geol. Soc. Am. earthquakes in the axial tectonic belt of North
87, 1259-68. Island, New Zealand. J. geophys. Res. 83, 4419-
O'DAY, M. S. 1974. The structure and petrology of the 29.
Mesozoic and Cenozoic rocks of the Franciscan WALKER, R. G. 1977. Upper Cretaceous re-
Complex, Leggett-Piercy area, Northern Califor- sedimented conglomerates at Wheeler Gorge,
nia Coast Ranges. Thesis, PhD, Univ. California, California: description and field guide--reply. J.
Davis, 152 pp. sediment. Petrol. 47, 928-30.
OGLE, B. A. 1953. Geology of Eel River Valley area, -- & MuTn, E. 1973. Turbidite facies and facies
Humboldt County, California. Bull. Calif. Div. associations. In: Pacif. Sec. Short Course, Turbi-
Mines Geol. 164, 128 pp. dites and Deep-water Sedimentation, 119-57.
RIccI-Luccm, R. 1975. Depositional cycles in two Soc. econ. Paleontol. Mineral.

STEVEN B. BACHMAN, Department of Geological Sciences, Cornell University,

Ithaca, New York 14853, U.S.A.
 The Franciscan Complex of northernmost California:
sedimentation and tectonics

K. R. Aalto
SUMMARY: The Franciscan Complex consists, from east to west, of a belt of foliated
(textural zone 2 and 3) rocks deformed beneath the Coast Range thrust (CRT), a belt of
broken formation (BF) and a belt of sheared olistostromes (SO) consisting of conglomer-
ate, greywacke, limestone, chert, greenstone, plutonic and glaucophane schist blocks in
chiefly argillaceous matrix. South of the Klamath River, zone 2 and 3 rocks are of an outer
fan or basin-plain facies and have overthrust midfan unfoliated (textural zone 1) BF in
which suprafan lobes are preserved as megaboudins. To the north, foliated rocks are in
gradational contact with similar 13F and two episodes of thrusting were confined to the
CRT. Blocks of chert, gabbro and metavolcanic rock were tectonically transported along
thrust faults. Greywacke and conglomerate of the SO belt are inner fan deposits. The
BF-SO contact is gradational, delineated by the westward appearance of greenstone, chert
and plutonic blocks increasing progressively in size and/or abundance. Olistostrome units
are intercalated with isoclinally folded, recycled sediments. BF and SO greywackes are
commonly feldspathic litharenites with mean parameters (Dickinson 1970): Q = 54.6,
C/Q = 0.28, F = 21.6, P/F = 0.97, L = 23.8 and V/L = 0.82. Sediments were deposited
in a trench slope basin and intercalated with olistostromes derived from elevating portions
of the accretionary basement. Progressing upwards in the local sequence north of the
Klamath River, the gradational SO-BF transition, decrease in overtuned beds and a
retrograding fan sequence reflect basin filling, diminished faulting and lessened relief
concomitant to basin elevation.

Geological setting tion. The Redwood Mountain and Patricks

Point outliers (Fig. 1, Irwin 1966) represent
The Franciscan Complex is exposed throughout klippen of zone 2 and 3 rocks that are thrust
the Coast Ranges geological province (Bailey et over a western belt of Franciscan m61ange with
al. 1964). It is juxtaposed against ultramafic zone 1 greywacke (Aalto 1976, 1978; Talley
rocks of the western Klamath Mountains belt 1976). Talley (1976) subdivides zone 3 rocks of
(Irwin 1977) at the Coast Range thrust, along the Redwood Mountain outlier into graphite
which Klamath rocks may have overriden and lawsonite schist members. These members
the Franciscan Complex by tens of kilometres are petrologically different from zone 3 rocks
(Fig. 1, Blake & Jones 1977). Immediately beneath the Coast Range thrust immediately to
beneath the thrust, Franciscan rocks have suf- the east (Monsen & Aalto 1980). North of the
fered cataclasis. For field mapping, the degree Klamath River (Fig. 1, location D dashed line),
of textural reconstitution of greywacke sand- zone 2 rocks are in gradational contact with
stone is assessed following the scheme of Blake zone 1 broken formation with the degree of
et al. (1967): textural zone 1 shows no foliation, textural reconstitution increasing progressively
zone 2 has platey cleavage with bedding fea- upwards towards the Coast Range thrust. The
tures preserved, and zone 3 has well-developed nature of the transition between the northern
foliation largely obliterating bedding features, and southern regions has not been determined,
and quartz segregation laminae exceeding but it is likely that the zone 3-zone 1 thrust is
i mm in thickness. Units with zone 3 greywack- somewhere overlapped by the Coast Range
es are formally termed 'South Fork Mountain thrust as one moves north.
Schist' (see Lanphere et al. 1978, for a review of The zone 1 Franciscan Complex may be
stratigraphic nomenclature). subdivided into an eastern belt of solely broken
In the region portrayed in Fig. 1, textural formation and a western belt of m61ange (termi-
grade beneath the Coast Range thrust de- nology of Hsii 1974) containing sedimentary,
creases gradationally northward. To the south, metamorphic and igneous tectonic blocks dis-
zone 3 metavolcanic and metasedimentary persed in an argillaceous matrix. The boun-
rocks immediately below the Coast Range dary between these two units is approximately
thrust have overthrust zone 1 broken forma- located in Figs 1 (lines labelled B), 2 & 3.

419
420 K. R. Aalto
Fro. 1. Location map showing distribution
of textural zones and regional geology.
CRT is the Coast Range thrust fault. Zone
1 Franciscan basement: no pattern; zone
2: dots; zone 3: wavy dashes; undifferenti-
ated zone 2/3: mixed pattern. Letters A-F
refer to features discussed in text. Beta (/3)
KLAMATH ~ denotes approximate locations of exotic
STUDY MAP • Western glaucophane schist blocks within zone 1
•AREA ~, AR EA terrane Sedimentological data of Table 1
Klamath are derived from all complete beds exposed
belt . in place along Highway 299 to Area A, the
coast from the Trinidad map area (Aalto
1976) to Crescent City and in the outlined
Klamath map area (Fig. 2). Regional geol-
ogy from Aalto (1976), Monsen (1979,
location A), Shimamoto (1976), Tailey
(1976), Young (1978), the Weed and Red-
ding 1:250,000 sheets of the Geologic Map
Redwood of California (1964) and near the Klamath
Mtn. , River mouth, BSc thesis work of Ensrud
outlier ) (1978), Jepsen (1979), Lind (1979) and
Lozinsky (1979).
C ".;~.
Patricks Pt. parallel to the foliation. Axial planes of F2 folds
outlier.. are orientated at an angle to Si foliation and are
associated with poorly developed S 2 strain-slip
cleavage.
The zone 3-zone 1 boundary thrust in the
/r3 south dips gently to the east (Fig. 1, locations
F
A - C ) . A marked contrast exists between iso-
clinically folded zone 3 schists with abundant
quartz veins and pods, and the subjacent zone i
argillite and greywacke, which is mylonitized
and overturned but unfoliated (Monsen 1979).
Tectonic blocks of greenstone and ribbon chert
up to tens of metres in maximum dimension are
q
dispersed within the gouge zone of the zone
3-zone 1 thrust fault. They have no obvious
source in the upper or lower thrust plates.
~OKM Monsen & Aalto (1980) believe that they were
tectonically transported along the zone 3-zone
1 thrust fault from a m61ange terrane now
underlying the western Klamath Mountains
Structure
belt. Similar tectonic blocks increase in abund-
All units of the Franciscan Complex strike NW ance as one nears the Coast Range thrust north
and dip predominantly to the NE, whether in of the Klamath River (Figs 2 & 3). They are
fault or depositional contact. From east to west most concentrated in zone 2 metasediments
(ignoring the zone 2/3 klippe), units become (Fig. 4A). However, some are found in the
structurally lower. underlying zone 1 broken formation as much as
1000 m west of the zone 1-zone 2 boundary
(Fig. 2). Blocks of chert and greenstone can
Zone 2/3 rocks below Coast Range thrust
exceed 30 m in maximum dimension. The
Franciscan zone 2/3 metasedimentary and cherts are brecciated and partially recrystal-
metavolcanic rocks beneath the Coast Range lized, but contain radiolaria. Greenstone blocks
thrust exhibit at least two periods of deforma- are brecciated and consist of silicified felsic
tion. Primary $1 foliation of metasedimentary crystal tuff, porphyritic felsite and basalt.
rocks in all regions beneath the thrust is parallel Chloritization and quartz veining are common
to bedding. Associated with this foliation are F 1 and blocks have foliated alteration rinds paral-
folds whose axial planes are approximately lel to their long dimensions. Highly deformed,
 The Franciscan Complex of northernmost California 421

__ 41045 '

SO'B!
?

} "# "I N

30 ~ ~

-- ~T-~ ,,~ ~ ~- ,,~ & V°

gN~

.........

c ~V.>. S.=,~=7~ ,, c

K ~ ~:' ~ ~ ~ "'"

,:
. . . . . . . ,.C "W i;
. VZ' ~ ;!:!): ~ x'
mile
l

o I km g' ,~ p,, .<~ ,:,

i |

! .<o "~ \ ,A (.;."..,::~ ~ ,

4
1 o 3 0 ~t ~&o \ .d .~ ,,~, ,.,

FI6.2. Klamath map area. From east to west: Umf = ultramafic rock of the western Klamath belt
which is bounded by the Coast Range thrust (solid line), z2 = textural zone 2, Franciscan
metasedimentary rock (western contact dotted) containing boudins of greenstone (g) and the chert
(c), BF = textural zone 1 Franciscan broken formation belt (western contact dashed), and
SO = textural zone 1 Franciscan sheared olistostrome belt. Quaternary fluvial sediment and marine
terrace cover is depicted by stipple near the Klamath River and Crescent City (CC).

saussuritized, albitized cumulate gabbro blocks the south, a second period of thrust faulting
exist at location D (Fig. 1) in zone 1 sedimen- transported zone 3 rocks westward over zone 1
tary rocks. broken formation. To the north, this second
It is likely that at least two periods of thrust period of motion was largely confined to the
faulting account for many of these geological Coast Range thrust; however, subsidiary re-
relationships. In both the north and south, a verse faults within the lower plate may have
first period of motion along the Coast Range provided pathways for transport of the chert,
thrust cataclastically d e f o r m e d Franciscan rocks greenstone and gabbro blocks. The i m m e n s e
and produced the S 1 foliation and F1 folds. To blocks distant from the thrust were probably let
422 K. R. Aalto

,.o
SO
,.::-.--:...-_.
BF
, .
.. ...
i 0

0 1kin
i J

A A"

SO BF ~\~.G~ ~ 1

~ " " "- "- "'~1~ - " c h e r t " - I

~0,,' • , . zone

Fl~;. 3. Topographic and geological cross-sections, Klamath map area. Lines of section shown on Fig.
2. CRT = Coast Range thrust, 2 = textural zone 2 Franciscan metasedimentary rock. Other letter
symbols as in Fig. 2. Distribution of greenstone, chert and plutonic boudins shown in black. The
position of a linear chert zone within the broken formation belt is indicated on A-A'. Sense of
motion and orientation of fault adjacent to this zone is conjectural due to poor exposure.

d o w n d u r i n g H o l o c e n e land d e g r a d a t i o n . In zone of m y l o n i t i z e d argillite and serpentinite up

both regions, the s e c o n d m o t i o n may have to 100 m thick, c o n t a i n i n g dispersed tectonic
p r o d u c e d the S 2 cleavage and F, folds. blocks of b r e c c i a t e d chert, g r e e n s t o n e , con-
g l o m e r a t e and g r e y w a c k e derived from the
lower plate of zone 1 m61ange (Aalto 1978).
Zone 2/3 outliers
Talley (1976) describes c o m p l e x thrust faulting
T h e thrust fault b e n e a t h the R e d w o o d M o u n - and e m p l a c e m e n t of exotic g l a u c o p h a n e schist,
tain and Patricks Point outliers is m a r k e d by a a m p h i b o l i t e and g r e e n s t o n e blocks associated

Fl~. 4. (A) Textural zone 2 Franciscan Complex, Klamath map area. Foliated argillite and
greywacke containing a large greenstone boudin (under hammer). Exposure within 1(10 m of the
Coast Range thrust in central part of area.
(B) Franciscan Complex of the sheared olistostrome belt, Klamath map area. Block of highly
indurated, sheared tuff (labelled T) and dark grey argillite (A) set in the weakly indurated
argillaceous matrix (Mx) of a sheared olistostrome unit. Note the veining and near vertical foliation
in the foreground block. Brecciated appearance of tuff due to offset of complexly intercalated tuff
and argillite. Indurated blocks are transported to the beach in debris flows generated by the failure of
olistostrome matrix. Greenstone block (V) at right. Notebook in lower right for scale. Exposure
along the coast approximately 1 km north of the southern boundary of the map area.
(C) Franciscan Complex of the Trinidad area. Sheared olistostrome containing boudinaged blocks
of greenstone (intercalated tuff and argillite, labelled V), greywacke (S) and limestone (L) dispersed
in foliated argillite. The arrow above the elongated tail of the limestone block depicts a block
fragment severed'by matrix intrusion during deformation. The olistostrome unit is intercalated with
turbidites which dip into the cliff. Dashed lines show bedding in the turbidites, offset at lower right.
The entire sequence is overturned. Notebook for scale. Exposure located approximately 1.5 km
north of Elk Head (Fig. 1, location El.
(D) Franciscan Complex of the broken formation belt, Klamath map area. Boudinaged classic
proximal turbidites (facies C). In upper right some boudins are bounded by compressional shears
and have assumed phacoidal form. 30 cm rule for scale. Exposure midway along coastal exposure of
the BF belt (1 km south of Damnation Creek).
(El Franciscan Complex of the sheared olistostrome belt, Klamath map area. Overturned
interstratificd turbidities and diamictite containing angular clasts of sandstone, greenstone (at pen
point) and chcrt (left of pen point). The contact (above pen) between the overturned Tab e bed and
diamictite is depositional. Unsheared diamictites with unoriented clases exist in adjacent areas.
Turbidites are overbank deposits (facies E) and intercalated diamictites (facies A) debris flow
deposits. Exposure along coastal cliffs approximately 2 km NNW of the Klamath River mouth.
 The Franciscan Complex of northernmost California 423
with serpentinite within the Redwood Moun- Along the eastern margin of the Redwood
tain outlier. The internal thrusting described by Mountain outlier, zone 1 Franciscan rocks have
Talley has no counterpart in area A (Fig. 1) to been emplaced over zone 2 greywacke and
the east, supporting the contention that a direct argillite by reverse faulting (Lind 1979; Talley
correlation of zone 3 rocks of the two regions 1976). Such faulting presumably postdates em-
cannot be made (Monsen & Aalto 1980). Zone placement of the klippe and is likely to be
2 metasedimentary rocks of the Patricks Point related to Cenozoic compression (Silver 1971;
outlier have been overturned by drag folding Simila et al. 1975). Monsen & Aalto (1980) and
during thrusting (Aalto 1978). Talley (1976) suggest that the faults on the

~ ~ ~ ~ii~, ~ ~i~.~÷~-~i~ ~ ~ •~ ~i~ - - ~

~ ~!!~i••!~
¸ ~•~

!i~!ii• ~ ~

~i~~IIII~
~i~ ~
424 K. R. Aalto
eastern margins of both outliers may have a & 2) consists of highly deformed, boudinaged
significant strike-slip component, thus explain- individual greywacke beds and successions of
ing the lack of continuity between zone 3 rocks beds in sheared argillite (Fig. 4D). Beds are
west of the Coast Range thrust. Such motion is commonly upright, strike N30-40°W and dip
in accord with Cenozoic plate motions (Dott 25-35 ° either NE or SW (Fig. 5). This pattern is
1979; Herd 1978). indicative of a chevron style of folding, but fold
hinges are rarely exposed. This folding appears
Zone I rocks
to be unrelated to the events induced by motion
The zone 1 broken formation terrane (Figs 1 on the Coast Range thrust.

A B

FIG. 5. Poles to bedding within: (A) the sheared olistostrome belt (n = 102), and (B) the broken
formation belt (n = 467) of the Klamath map area (Fig. 2). Upright beds are depicted by dots and
overturned beds by x's. Contours show 3 (encircled white) and 6 (encircled black) points per unit
area.
 The Franciscan Complex of northernmost California 425

In the south-central part of the zone 1 broken have a preferred orientation parallel to the
formation belt in the Klamath map area (Figs 2 dominant NW strike and NE dip of the Francis-
& 3), several elongate bodies of red ribbon can Complex. Recumbent or isoclinal folds
chert are arranged in a line trending NNW. exist in olistostrome matrix material in the
Where observable, western (stratigraphically vicinity of larger blocks, which commonly have
lower) contacts between the chert and sur- rinds of mylonitized argillite. Such observations
rounding broken formation are either thrusts imply tectonic deformation. As Hsti (1974)
or high-angle faults. Upper contacts in the suggests, however, such deformation may be
two northern localities dip gently NE and could imposed on olistostromes and one must look to
be thrust faulted. However, the presence of other criteria to discover the initial mode of
intercalated chert and argillite at the upper mixing of blocks. The gradational contact be-
contact in the northernmost locality suggests tween the broken formation and sheared olis-
conformity. tostrome belts and the depositional contacts of
The contact between the eastern broken thin mixed units described above support the
formation and western m61ange belts within interpretation of initial sedimentary mixing.
zone 1 Franciscan is apparently gradational. In Immediately to the west of the Patricks Point
the Klamath map area (Figs 2 & 3), the contact and Redwood Mountain outliers, glaucophane
is marked by appearance of blocks of green- schist and associated eclogite and amphibolite
stone, chert and conglomerate which increase occur in isolated tectonized blocks within the
in size and abundance to the SW. The blocks melange (sheared olistostrome?) units (Fig. 1).
are dispersed in a highly fragmented argil- The petrology and structural setting of these
laceous matrix with penetrative shear-fracture blocks have been studied in some detail by
fabric. The degree of fragmentation has les- Shimamoto (1976), who feels that they repre-
sened the rock strength and m61ange commonly sent the northern continuation of a zone of such
fails by debris flow (Fig. 4B, Aalto 1977). The blocks extending as far south as Santa Barbara
topography is consequently more subdued than (Coleman & Lanphere 1971). They have under-
that of the broken formation belt. gone higher grade metamorphism than the
M61ange units vary in lateral extent and rocks with which they are associated, are com-
thickness. Thinner units are clearly subaqueous monly spheroidal, have chlorite-actinolite-talc
debris flow (Middleton & Hampton 1973) de- rinds and in some localities are associated with
posits for they are interstratified with turbidites serpenitized peridotite bodies.
(Fig. 4C, E). These deposits provide evidence
of an olistostrome mode of emplacement of
m61ange units hundreds of metres in thickness,
which are intercalated with sections of broken Petrology
formation composed of turbidites and associ- Zone 2/3 rocks
ated redeposited sedimentary rocks. In most
regions, however, post-emplacement structural Beneath the Coast Range thrust, zone 3
activity has in most cases obliterated clear rocks are predominantly metapelitic, with sub-
stratigraphic evidence of olistostromes. The de- ordinate amounts of metagreywacke,
gree of deformation within the m61ange (here- metachert, metatuff and metabasalt. The meta-
in termed sheared olistostrome) belt is clearly morphic mineral assemblage is of the
greater than in the eastern belt of broken glaucophanitic-greenschist facies in most of the
formation. Isoclinal folding and overturning of region north of area A (Fig. 1, Monsen & Aalto
beds in the sheared olistostrome belt is common 1980). Young (1978) reports small amounts of
(Figs 4C, E & 5). Axial planes of isoclinal folds lawsonite within zone 3 rocks to the south of
commonly strike NW and dip 40-50 ° NE, and area A (Fig. 1), and it is likely that lawsonite
axes plunge SE at shallow angles. content diminishes northward, implying a de-
Blocks in olistostrome units may be angular, crease in metamorphic grade beneath the
rounded or tectonically elongated. There is no thrust. Such a contention is supported by the
doubt that blocks have been sheared subse- absence of zone 3 rocks in the Klamath map
quent to sedimentary mixing for they are com- area (Fig. 2).
monly bounded by compressional shears and Medium to dark grey, foliated greywacke and
traversed by extensional shear fractures. Argil- argillite of textural zone 2 crop out north of the
laceous matrix commonly intrudes into cracks Klamath River immediately beneath the Coast
and fractures of disintegrating blocks (Fig. 4C). Range thrust, in a thrust plate along the eastern
Irregular blocks show random orientation of margin of the Redwood Mountain outlier (Tal-
slickensided shear surfaces, and elongate blocks ley 1976; Young 1978), and in the Patricks
426 K. R. Aalto

Point outlier. Point-count analysis shows no X-ray analysis of 15 samples indicates that
significant compositional difference between the matrix composition of sandstones and the
greywacke of the Patricks Point outlier and composition of argillite within the sheared olis-
subjacent zone 1 Franciscan (Aalto 1976). tostrome and broken formation belts is identi-
North of the Klamath River (Fig. 2), slightly cal, consisting chiefly of varying proportions of
foliated zone 2 rock is in gradational contact quartz and sodic plagioclase silt, illite/sericite
with zone 1 broken formation and of the same and chlorite. The average specific gravity of
composition. Grain components are the same zone 1 greywackes is 2.66 (range: 2.52-2.72;
as those of zone 1 greywacke discussed below. standard deviation: 0.46), based upon analysis
However, pumpellyite is more common in pla- of 50 samples from the Klamath map area.
gioclase and chlorite more pervasive in the The detrital framework mode proportions of
matrix. Dickinson (1970) may be used to assess prove-
nance, although values to some degree reflect
Zone I greywacke and argillite reduction of unstable components by diagenetic
matrix generation. For the average zone 1
Point counts of 300 grains per thin section greywacke, Q is moderate (54.6), F is low
show no statistically significant compositional (21.6), L is low to moderate (23.8), C/Q is low
difference between zone 1 greywacke of the (0.28), P/F is high (0.97) and V/L is moderate
broken formation (n = 22) and sheared olisto- (0.82). Additional framework modes (Dickin-
strome (n = 16) belts in the Klamath map area, son & Suzcek 1979) include Qm (39.5), Qp
thus they will be discussed together. Aalto (15.1), P (21.0), K (0.5), Lv (19.5), Ls (4.3)
(1976) and Young (1978) provide descriptions and Lt (38.9). Ternary plots of framework
of zone 1 greywacke to the south. Although modes proposed by Dickinson & Suzcek (1979)
most greywackes are lithic arkoses or felds- suggest a mixed subduction complex/dissected
pathic litharenites (sensu Folk 1968), the range arc source.
of composition is wide. No consistent regional
trends of composition exist.
Zone 1 sandstones are commonly texturally Zone I conglomerate
immature to submature, with poor sorting,
angular to subangular grains and an average of Conglomerate among zone 1 Franciscan
34% matrix and 6% calcite cement. Monocrys- rocks is most common in the sheared olisto-
talline quartz commonly has undulose extinc- strome belt. Counts of 100 points made on
tion, although non-undulose grains of inclusion- granules and small pebbles show no consistent
free, idiomorphic volcanic quartz exist in nearly regional variation in composition throughout
every sample. All varieties of potycrystalline the area portrayed in Fig. 1. Indeed, greater
quartz described by Young (1976) exist: grains variation exists between two samples taken at
which are polygonized, with elongated original exposures less than 100 m apart than through-
crystals, new crystals and polygonal crystals out the entire region. A study of 10 samples
(although those with new crystals are most com- taken south of the Klamath River (Fuller 1977)
mon). Chert may contain ghosts of radiolaria and six from the Klamath map area indicates
and is commonly recrystallized or replaced by that the average conglomerate contains 38%
calcite. Feldspars include perthite, untwinned red and green chert (range: 25-51), 21% black
albite and trace amounts of K-feldspar which chert (range: 8-37), 19% vein quartz (range:
are commonly extensively replaced by sericite 4-34), 9% dacite prophyry and quartz kera-
or calcite. Twinned plagioclase (An17_58) is less tophyre (range: 1-41), and less than 4% each
altered but commonly contains pumpellyite. of basalt, andesite, tonalite, diorite, quartz-
Volcanic rock fragments include felsitic, micro- arenite, greywacke and limestone. Following
litic, lathwork and vitric grain types (Dickin- Bailey et al. (1964) and Cowan & Page (1975),
son 1970) which are commonly partially to clasts can be considered either intra- or extra-
wholly replaced by chlorite or calcite. formational, depending upon whether or not
Metasedimentary grain components include they could have been derived by recycling
quartz-mica schist and phyllite. Sedimentary Franciscan rock. The most common intraforma-
grains include siltstone and claystone. Mic- tional lithologies include red and green chert,
rophaneritic grains contain quartz and plagio- basalt, andesite and greywacke. Likely extra-
clase. Minor grain components include detrital formational lithologies include quartzarenite,
zircon, clinopyroxene, muscovite, hornblende, black chert, limestone and diorite. Vein quartz,
serpentinite and accessory haematite, felsitic volcanic rock and tonalite could fall in
leucoxene and goethite. either category, since they occur locally within
 The Franciscan Complex o f northernmost California 427

olistostrome blocks. A tectonic landmass, such Palaeomagnetic study of nine chert samples
as an uplifted, eroding subduction zone com- from a single block with red radiolarian chert
plex, could have supplied any of the extrafor- comformably overlying pillow greenstone at the
mational (as well as the intraformational) clast mouth of the Klamath River suggests a low
components. latitude derivation (Jeff Johnston, Woodward-
Clyde Consultants, pers. comm. 1980). The
Composition of m61ange blocks chert magnetization directions are tightly
grouped with an average inclination of - 5 2 ° ,
For purposes of field mapping, block types corresponding to latitude 33 ° . Since the beds
are recognized throughout the sheared olisto- were likely deposited during the predominantly
strome belt in average amounts as follows: clas- normal polarity period of the Late Mesozoic it is
tic sedimentary (conglomerate and sandstone, probable they formed south of the equator.
= 50%), bedded chert (2 = 12%), greenstone Chert was sampled at intervals of approxi-
( 2 = 3 8 % ) , partially recrystallized limestone mately 1, 10 and 16 m above the greenstone,
(trace), plutonic (trace) and exotic, glauco- representing a depositional interval of 3-18 Ma
phane-bearing metamorphic (trace). There are at sedimentation rates of 1-5 m m l 0 -3 yr, and it
no significant lateral variations of block com- is unlikely all samples originated in a reversed
position, as determined by field study and by epoch. Furthermore, most published tectonic
tallying the distribution of blocks visible on models call for a northward oceanic plate mo-
1 : 12 000 air photos, other than the restriction of tion in the Pacific basin during the Late Meso-
exotic blocks to the south (Fig. 1). zoic and a limestone block in the central Fran-
Greenstone blocks commonly contain the ciscan belt some 150 km SSE of this block was
metamorphic mineral assemblage quartz _.+ shown to be derived from the southern hemi-
albite + chlorite + epidote + pumpellyite + sphere (Alvarez et al. 1980).
carbonate. Felsic tuffs range from those com-
posed nearly entirely of crystals, commonly
fragmented quartz and plagioclase, to those Sedimentology
composed entirely of vitric, microlitic or lath-
work lithic fragments in a chloritic matrix. Depositional setting of the different units with-
Redeposited (epiclastic?) tuff contains clasts of in the Franciscan Complex is determined by
crystal and crystal-lithic tuff in a chloritized recording bedding features and measuring short
argillaceous matrix. In some highly indurated, stratigraphic sections in areas in which beds
foliated blocks, wisps and lenses of tuff have have survived deformation. Table 1 provides a
intruded into argillite and later been offset, comparison of bed features of the zone 1
suggesting ductile deformation followed by de- sheared olistostrome belt, zone 1 broken
velopment of foliation (Fig. 4B). Such deforma- formation belt and zone 2/3 metasedimentary
tion preceded sedimentary mixing in an olis- rocks throughout the region portrayed in Fig.
tostrome. Epiclastic breccia contains angular 1. Features of coarser-grained beds survive
clasts in varying proportions of greywacke, deformation better, thus coarser facies data are
argillite, greenstone and chert in a tuffaceous summarized in separate columns of Table 1.
matrix. Basalt with flow banding and pillow Facies terminology of Walker & Mutti (1973) is
structure is commonly composed of microcrys- employed in classifying classical turbidites. Ter-
talline plagioclase laths (An60~5) with accessory minology of Aalto (1976) better describes bed-
clinopyroxene. Quartz keratophyre contains ding features among conglomeratic sediment
quartz and plagioclase (An3_5) phenocrysts in a gravity flow deposits and thus is used in this
groundmass of plagioclase, epidote, chlorite analysis.
and chalcedony.
Plutonic blocks include plagiogranite con- Sheared olistostrome belt
taining quartz, plagioclase (An13_16) and
hornblende; diabase; serpentinized peridotite Sandstones and conglomerates within the
and highly altered cumulate gabbro containing sheared olistostrome belt have coarser beds,
chiefly albitized plagioclase and clinopyroxene. characterized by more proximal features, than
Blocks were derived from a variety of source those of the broken formation belt. Bedded
terranes, supporting the contention (Alvarez et diamictites (Fig. 4E) and abundant conglom-
al. 1980; Blake & Jones 1978) that a diverse eratic sediment gravity flow deposits containing
assemblage of oceanic crustal plate and arc(?) Aalto (1976) divisions II, III and/or IV (facies
remnants accreted to the North American plate A of Walker & Mutti 1973) suggest a slope
margin during Franciscan accumulation. channel or inner fan depositional setting. The
428 K. R. A a l t o

TABLE 1. Sedimentological comparison of bed types of the zone 1 sheared olistostrome belt, zone 1
broken formation belt and zone 2-3 Franciscan Complex rocks, calculated with and without distal
facies data
Facies (after Aalto 1976"; Sheared olistostrome Broken formation Zone 2 and 3
Walker &Mutti 1973) belt (N = 1047) belt (N = 1354) rocks (N = 470)
(%) (%) (%)
Facies A: bedded diamictite Trace Trace
Facies A: sandy conglomeratic
beds with divisions II, III
and/or IV 17 34 1 1 Trace 1
Facies B: thick, massive to
laminated sandstone beds with
divisions V and/or VI 10 19 16 24 5 22
Facies C: sandstone beds with
Tae or Tabe sequences 24 47 49 75 16 77
Facies D, E and G: thinly bedded
laminated sandstones and argillites 49 34 79
* Division I: massive sandstone or inversely graded pebbly sandstone, II: pebbly sandstone or conglomerate
with or without imbrication, III: normally graded pebbly sandstone, IV: diffusely laminated and/or cross-
stratified pebbly sandstone, V: massive sandstone, VI: laminated sandstone with dish and pillar structures.

bedded diamictites are intercalated with thinly deposits. Each megaboundin may represent
bedded overbank turbidities of facies E (Wal- part of a suprafan lobe which prograded over
ker & M u t t i 1973) and possibly represent lateral outer fan deposits. Migration of middle fan
spills from debris flows which topped a sub- distributary channels due to channel and lobe
marine channel margin. The thicker strati- aggradation led to cessation of deposition of
graphic sections which have survived deforma- coarser-grained sediment and resumption of
tion (Fig. 6A, Aalto 1976) have high sandstone/ hemipelagic and pelagic sedimentation. Thus
shale ratios and lack well-developed vertical each lobe was encased in argillaceous sediment
trends in bed thickness or grain-size distribu- and deformed as a separate entity.
tion. Channelling is common beneath thicker,
coarser-grained beds. Regional synthesis
In the Klamath map area (Fig. 2), the eas-
Broken formation belt
ternmost exposures which gradationally under-
The eastern belt of zone 1 broken formation lie zone 2 rocks are fine-grained and appear to
contains far fewer conglomeratic beds and more be outer fan and basin-plain deposits. As one
classic proximal turbidities and beds with Aalto progresses from west to east across the Francis-
(1976) divisions V and/or VI (Fig. 4D, facies B can Complex in this region, sediment facies
and C of Walker & Mutti 1973). Coarsening- become progressively more distal, suggesting a
and thickening-upward sequences (Fig. 6B) retrograding fan sequence if sediments were
are preserved within megaboudins surrounded deposited within a single basin. Because of the
by sheared argillite. The megaboudins are len- gradational nature of the zone 1-zone 2 contact
ticular in shape, tens of metres thick and hun- in the north no sedimentological distinction can
dreds of metres long. They consist of many be made between zone 1 and zone 2
separate boudinaged beds and short sequences sedimentary rocks.
of beds bounded by compressional shears, in- To the south, Aalto (1976) and Monsen &
dicating both extension and compression. Bed Aalto (1980) have demonstrated that there is a
types (Table 1) and vertical trends suggest a marked difference in depositional setting of
middle to outer fan depositional setting. The zone 1 versus zone 2/3 rocks. The comparative
megaboudins are commonly surrounded by absence of coarse facies bed types (Table 1),
argillite and thinly bedded sandstone with base- low sandstone/shale ratio and existence of
cut-out Bouma sequences which, though com- thickening- and coarsening-upward sections
monly highly sheared, exhibit characteristics of suggest an outer fan/basin plain depositional
outer fan turbidities, hemipelagic and pelagic setting for zone 2/3 Franciscan. The zone 1
 The Franciscan Complex of northernmost California 429

--),

)
E '~'1 //
j"
<..
>
.-~
->

I0 i
/"
/
/

I
.f
o
g .
,.

[-....

J im

g ~II i
rrr

l 3m

(A) (B)

FIG. 6. (A) Stratigraphic section at Crescent City (Battery Point Lighthouse), Klamath map area.
Bedding features are depicted by Roman numerals following the Aalto (1976) model (see Table 1).
The graph portrays thicknesses of successive beds, from base to top.
(B) Graphic portrayal of stratigraphic section in the broken formation belt, Klamath map area. The
graph portrays thicknesses of successive beds, from base to top of the section. The slanted arrows
depict four coarsening- and thickening-upward sequences (a-d). Section located near the midpoint
of the coastal exposure of the BF belt (3/4 km north of Damnation Creek).

greywackes over which these rocks are thrust Maxwell 1974; Suppe 1979, Blake et al. 1981).
faulted are much coarser inner to middle fan It is not the purpose of this paper to review
deposits (Aalto 1976). This sedimentological them in detail. I shall present the one of many
contrast supports the contention that the out- possible models (Fig. 7) for this region which I
liers are klippen, and suggests that zone 3 rocks feel best explains the geology described above
of location A (Fig. 1) have been transported by (Fig. 7), and which is largely based upon mod-
faulting some distance from a distal Franciscan els of accretionary prism development prop-
terrane now underlying the Klamath block. osed by Karig & Sharman (1975) and Moore &
Karig (1976).
The sheared olistostrome belt is composed of
Sedimentary tectonics and geological material derived from two sources. Clastic
history sedimentary rocks of the belt, including much
of the olistostrome matrix, were most likely
The Franciscan Complex is considered to have derived from an elevated subduction zone com-
originated by accretion at a convergent plate plex which could have been exposed at either
margin. Differing tectonic models for the Fran- location 2 or 3 in Fig. 7. This would account for
ciscan have been proposed by many workers the quartzofeldspathic composition of sheared
(e.g. Blake & Jones 1974, 1978; Ingersoll 1978; olistostrome belt sediment and the abundance
430 K. R. Aalto

/
i -,q

AcTvE J
\ 2
k ~ 1 m _

FI6.7. Sedimentary-tectonic model for the Franciscan complex, Klamath map area. The active mode
is characteristic of newly developing trench slope basins in which sediments derived from upslope
portions of the accretionary prism (sources 1 or 2) and/or the arc complex (source 3) are intercalated
with prism-derived gravity slides (source 1) on small fans affected by continuous compressional
faulting. The passive mode is characteristic of older, elevated basins which are filled (sources 2
and/or 3), have diminished relief and are no longer subjected to intense faulting. The asterisk shows
the position of an upfaulted sliver of accreted material in the active basin relative to its future
position in the passive mode. The dashed line represents the present-day cross-section of the
Klamath map area Franciscan.

of non-volcanic lithic components. Intercala- area, in which there is clear evidence favouring
tion of olistostromes containing previously the olistostrome mode of mixing of blocks.
accreted clastic sedimentary, chert, greenstone Should the model be correct, a restored cross-
and plutonic blocks occurred. Olistostromes section of the filled Klamath map area trench
were probably derived from structurally ele- slope basin could be drawn along the dotted
vated portions of the accretionary prism, and line shown in Fig. 7 (passive mode). The oldest
were transported downslope to a developing basin sediments are complexly faulted against
trench slope basin (Fig. 7, active mode). Steep slivers of accretionary prism tectonic m6tange
slopes and continuous fault activity account for which constitutes the trench slope basin base-
the isoclinal folding and coarseness of these ment. Thus, to the SW of the Klamath map
basin deposits. area one might encounter belts of the actual
As the basin filled and became elevated by accretionary basement. Aalto (1978) consi-
continued accretion, fault activity diminished, dered this to be the case in the Trinidad area, in
local relief diminished and olistostromes were which an isoclinally folded broken formation
less frequently emplaced (Fig. 7, passive terrane is situated between m61ange terranes
mode). The broken formation belt has, by each containing exotic glaucophane schist
definition, no olistostrome units and contains blocks as well as the usual greenstone and
chiefly middle to outer fan deposits. This more chert. While most of these m61ange are
distal facies reflects sediment bypassing of the apparently sheared olistostromes, in places
basin. Diminishing fault activity within the clearly intercalated with turbidites (Fig. 4C),
prism accounts for the lack of overturning of Elk Head (Fig. 1, location E) immediately
beds. Thus, the retrograding fan sequence pre- north of Trinidad is most likely primary tectoni-
sent across the Franciscan Complex in the cally mixed prism material. The headland con-
Klamath map area resulted from the elevation, sists chiefly of overturned pillow greenstone,
filling and eventual sediment bypassing of a highly indurated argillite and coarse turbidites
trench slope basin (see Underwood & Karig (Aalto 1976). All contacts between lithologies
1979, for a modern example of sediment bypas- are faults. Trinidad Head (Fig. 1, location F;
sing on the uppertrench slope). S h i r e m a n , pers. comm. 1981) is chiefly an
This model best applies to the Klamath map intrusive complex, zoned from east to west, of
 The Franciscan Complex of northernmost California 431
diabase, g a b b r o and tonalite with plagiogranite Monsen and Mike Shimamoto for discussions which
a n d q u a r t z k e r a t o p h y r e dykes. B o t h h e a d l a n d s contributed to the development of ideas presented in
m a y be part of an a c c r e t i o n a r y b a s e m e n t slab this paper. Also, I thank Sue Cashman, Susan Mon-
w h i c h separates d e f o r m e d slope basin rocks. sen and John Longshore for reviewing and Jacque
Overton for typing the manuscript. Thanks to Lori
This i n t e r p r e t a t i o n is s u p p o r t e d by the p r e s e n c e
Dengler for advice on X-ray work. The success of this
of m61ange with s e r p e n t i n i t e matrix a d j a c e n t to project was furthered by the cooperation of Simpson
T r i n i d a d H e a d . Such a mix is m o r e likely Timber Company and Arcata Redwood Company in
tectonic than s e d i m e n t a r y , and thus likely to granting access to their lands. Portions of this study
constitute a c c r e t i o n a r y b a s e m e n t . were funded by Humboldt State University Founda-
tion grants #115-75 and 115-77. This study is part of
I thank my colleagues and stu-
A C K N O W L E D G M E N T S : a Franciscan Complex research project supported by
dents at Humboldt State University, and especially the Division of Earth Sciences of the National Science
Clark Blake, Sue Cashman, John Longshore, Susan Foundation. NSF Grant # E A R 7809957.

References
AALTO, K. R. 1976. Sedimentology of a m61ange: of graywacke and arkose. J. sediment. Petrol. 40,
Franciscan of Trinidad, California. J. sediment. 695-707.
Petrol. 46, 913-29. & SUZCEK, C. A. 1979. Plate tectonics and
- 1977. Franciscan m61ange-Quaternary uncon-
- sandstone compositions. Bull. Am. Assoc. Petrol.
formities and terrace stability, Trinidad, Califor- Geol. 63, 2164-82.
nia. Abstr. Prog. geol. Soc. Am. 9, 377. DOTT, R. H., JR. 1979. Intracontinental plate bound-
1978. Sedimentology of a m61ange: Franciscan ary east of Cape Mendocino, California: com-
of Trinidad, California: a reply. J. sediment. ment. Geology, 7, 322-3.
Petrol. 48, 677-9. ENSRUD, D. L. 1978. The geology of area B, sections
ALVAREZ, W., KENT, D. V., SILVA, I. P., 8, 9, 16, 17, 20, 21, T13, N, R 2E, Ship Mountain
SCHWEICKERT, R. A. & LARSON, R, A. 1980. quadrangle, Del Norte County, California.
Franciscan Complex limestone deposited at 17° Thesis, BSc, Humboldt State Univ. (unpubl.).
south paleolatitude. Bull. geol. Soc. Am. 91, FARHOUD1, G. • KARIG, D. E. 1977. Makran of Iran
476-84. and Pakistan as an active arc system. Geology, 5,
BAILEY, E. H., IRWIN, W. P. & JONES, D. L. 1964. 664-8.
Franciscan and related rocks, and their signifi- FOLK, R. L. 1968. Petrology of Sedimentary Rocks.
cance in the geology of western California. Bull. Hemphill's, Austin, Texas. 170 pp.
Calif. Div. Mines Geol. 183, 177 pp. FULLER, S. 1977. The origin of the major constituents
BISHOP, D. G. 1977. South Fork Mountain Schist at of the Franciscan melange conglomerates in
Black Butte and Cottonwood Creek, northern Humboldt County, California. Thesis, BSc,
California. Geology, 5, 595-9. Humboldt State Univ. (unpubl.).
BLAKE, M. C., JR., IRWIN, W. P. & COLEMAN, R. G. HERD, D. G. 1978. Intracontinental plate boundary
1967. Upside down metamorphic zonation, east of Cape Mendocino, California. Geology, 6,
blueschist facies, along a regional thrust in Cali- 721-5.
fornia. Prof. Pap. U.S. geol. Surv. 515, 1-9. Hs0, K. J. 1974. Melanges and their distinction from
- & JONES, D. L. 1974. Origin of Franciscan
- olistostromes. Spec. Publ. Soc. econ. Paleontol.
melanges in northern California. Spec. Publ. Mineral. Tulsa, 19, 321-33.
Soc. econ. Paleontol. Mineral. Tulsa, 19, 345-57. INGERSOLL, R. V. 1978. Paleogeography and
- & --
- 1977. Plate tectonic history of the Yolla paleotectonics of the late Mesozoic forearc basin
Bolly junction, northern California: Guidebook, of northern and central California. Pacif. Sec.,
73rd Ann. meet, Cordilleran Sec. geol. Soc. Am. Soc. econ. Paleontol. Mineral. Tulsa, Pacif.
14 pp. Coast Paleogeogr. Symp. 2, 471-82.
- & -- - 1978. Allochthonous terranes in north- IRWIN, W. P. 1966. Geology of the Klamath Moun-
ern California?--a reinterpretation. Pacif. Sec., tains province. Bull, Calif. Div. Mines Geol. 190,
Soc. econ. Paleontol. Mineral. Pacif. Coast 19-38.
Paleogeogr. Syrup. 2, 397-400. 1977. Review of the Paleozoic rocks of the
COLEMAN,R. G. & LANPHERE,M. A. 1971. Distribu- Klamath Mountains. Pacif. Sec., Soc. econ.
tion and age of high-grade blueschist, associated Paleontol. Mineral. Pacif. Coast Paleogeogr.
eclogites, and amphibolites from Oregon and Symp. 1,441-54.
California. Bull. geol. Soc. Am. 82, 2397-412. JEPSEN, K. O. 1979. Geology of the Klamath Glen
COWAN, D. S. & PAGE, B. M. 1975. Recycled Francis- area, southern Del Norte County, California.
can material in Franciscan melange west of Paso Thesis, BSc, Humboldt State Univ. (unpubl.).
Robles, California. Bull. geol. Soc. Am. 86, KARIG, D. E. & SHARMAN, G. F. 1975. Subduction
1089-95. and accretion in trenches. Bull. geol. Soc. Am.
DICKINSON, W. R. 1970. Interpreting detrital modes 86, 377-89.
432 K. R. A a l t o

LANPHERE, M. A., BLAKE, M. C., JR & IRWIN, W. P. bearing tectonic blocks from the Weott, Iaqua
1978. Early Cretaceous metamorphic age of the Buttes, Blue Lake and Trinidad quadrangles,
South Fork Mountain Schist in the northern California. Thesis, BSc, Humboldt State Univ.
Coast Ranges of California. Am. J. Sci. 278, (unpubl.).
798-815. SILVER, E. A. 1971. Transitional tectonics and late
LEATHERS, S. 1978. Petrology, metamorphic grade Cenozoic structure of the continental margin off
and structure of the Redwood Mountain outlier of northernmost California. Bull. geol. Soc. Am.
the South Fork Mountain Schist on the coast of 82, 1-22.
northern California. Thesis, BSc, Humboldt SIMILA, G. W., PEPP~N, W. A. & McEvILLV, T. V.
State Univ. (unpubl.). 1975. Seismotectonics of the Cape Mendocino,
LIND, E. K. 1979. Geology of central belt Franciscan California, area. Bull. geol. Soc. Am. 86, 1399-
north of Gold Bluffs Beach, California. Thesis, 406.
BSc, Humboldt State Univ. (unpubl.). StJPPE, J. 1979. Structural interpretation of the south-
LOZINSKY, R. P. 1979. The geology of the Franciscan ern part of the northern Coast Ranges and
Coast Ranges in the vincinity of Klamath, Califor- Sacramento Valley, California: summary. Bull.
nia. Thesis, BSc, Humboldt State Univ. (un- geol. Soc. Am. 90, 327-30.
publ.). TALLEr, K. L. 1976. Descriptive geology of the Red-
MAXWml., J. A. 1974. Anatomy of an orogen. Bull. wood Mountain outlier of the South Fork Moun-
geol. Soc. Am. 85, 1195-204. tain Schist, northern Coast Ranges, California.
MII)DI.ETON, G. V. & HAMP~rON, M. A. 1973. Sedi- Thesis, MSc, Southern Methodist Univ. (un-
ment gravity flows: mechanics of flow and de- publ.).
position. In: MIDDLHON, G. V. & BOUMA, A. H. UNDERWOOD,M. B. & KARl(;, D. E. 1979. Submarine
(co-chairmcn). Turbidities and Deep Sea Sedi- canyons as a key to sedimentation in subduction
mentation. Pacif. Coast Sec., Soc. econ. Paleon- complexes. Abstr. Prog. geol. Soc. Am. 11,531.
tol. Mineral. Tulsa, 1-38. YOUNG, J. C. 1978. Geology of the Willow Creek
MONSEN, S. A. 1979. Petrology and structure of the quadrangle, Humboldt and Trinity Counties
South Fork Mountain Schist and subjacent Fran- California. Map Sheet Calif. Div. Mines Geol.
ciscan ('omplex, Pine Ridge, California. Thesis, 31.
BSc, Humboldt State Univ. (unpubl.). YOUN(;, S. W. 1976. Petrographic textures of detrital
-- & AAl:rO, K. R. 1980. Petrology, structure and polycrystalline quartz as an aid to interpreting
regional tectonics: South Fork Mountain Schist crystalline source rocks. J. sediment. Petrol. 46,
of Pine Ridge Summit, northern California. Bull. 595-61)3.
geol. Soc. Am. 91,369-73. WAt,K~:R, R. G. & Mufti, E. 1973. Turbidite facies
Moore.:, G. F. & KARl(;, D. E. 1976. Development of and facies associations. In: MIDDLE'tON, G. V. &
sedimentary basins on the lower trench slope. BOUMA, A. H. (co-chairmen). Turbidities and
Geology, 4, 693-7. Deep Sea Sedimentation. Pacif. Coast See., Soc.
SlmaArvtOTO, M. K. 1976. Petrology of glaucophane- econ. Paleontol. Mineral. Tulsa, 119-57.

K. R. AAl.ro, Department of Geology, Humboldt State University, Arcata,

Calih~rnia 95521, U.S.A.
 Sedimentation, metamorphism and tectonic accretion of the
Franciscan assemblage of northern California

M. C. Blake, Jr, A. S. Jayko & D. G. Howell

SUMMARY: The Yolla Bolly terrane of the Franciscan assemblage in northern California
is seemingly a typical subduction complex, having undergone penetrative deformation and
metamorphism to the high pressure-low temperature blueschist facies. Detailed mapping
combined with sedimentological analysis has enabled us to: (1) reconstruct a probable
palaeosedimentary environment, (2) analyse the interaction during and after subduction
between deformation and metamorphism, and (3) speculate on subsequent deformational
history including tectonic accretion of the terrane to North America.
Rocks of the Yolla Bolly terrane consist of three thrust-fault-bounded lithological units: a
lower unit of disrupted mudstone and thin-bedded sandstone (broken formation) contain-
ing scarce volcanic and radiolarian chert horizons, a middle unit predominantly of
thick-bedded to massive sandstone (metagreywacke) that includes several horizons of
radiolarian chert, and an upper unit of mudstone and thin-bedded sandstone (broken
formation) with numerous intrusive and extrusive volcanic rocks plus rare radiolarian
chert. Radiolarians from all three units are of the same age (Tithonian to Valanginian) and
together with the sedimentological data, suggest that the rocks represent a continent-
derived submarine fan, deposited in a complex transform graben possibly similar to the
present-day Gulf of California or basins of the California Continental Borderland, rather
than a trench setting.
Metagreywacke containing lawsonite and aragonite yields radiometric ages of approx-
imately 110 Myr and indicate that these rocks were subducted to depths of 20-30 km
about 30 or 40 Myr after they were deposited. Shortly after subduction, the rocks were
probably involved in a collision that imbricated and tectonically returned the subduction
complex to the surface.

The Franciscan assemblage in northern Cali- South Fork Mountain Schist, and the Yolla
fornia encompasses three major NW-trend- Bolly terrane. The South Fork Mountain Schist
ing belts. The coastal belt consists almost was originally defined on the basis of metamor-
entirely of sheared sandstone, mudstone, and phic grade (Blake et al. 1967) but subsequent
conglomerate (broken formation) generally in- work (Bishop 1977; Worrall 1979; Blake et al.
terpreted as the youngest (late Cretaceous to unpublished data), indicates that the primary
Miocene) part of the Franciscan accretionary lithology is significantly different from adjacent
prism (Bachman 1978; Beutner et al. 1980). fossiliferous Franciscan rocks.
The central belt consists largely of tectonic This paper briefly describes the lithology, age,
mdlange and broken formation and includes metamorphism, and structure of the rocks of
most of the high-grade blueschist, eclogite, and the Yolla Bolly terrane and relates these charac-
exotic limestone knockers in northern Califor- teristics to plate tectonic models.
nia. Palaeomagnetic data suggest that at least
some of the terranes within the central belt
formed in southern latitudes and were subse-
Lithology and sedimentology
quently accreted to North America by large-
scale transcurrent plate motions (Alvarez et al. The Yolla Bolly terrane (Fig. 1) consists of three
1980). lithological units separated by thrust faults. The
The eastern belt consists largely of clastic lower unit consists largely of disrupted mud-
sedimentary rocks locally containing interbed- stone and thin-bedded quartzofeldspathic sand-
ded volcanic rocks and radiolarian chert. All of stone (broken formation) but also contains
these rocks have a faint to pronounced meta- coherent lenses of medium- to thick-bedded
morphic fabric and contain high-pressure sandstone and conglomerate plus minor
minerals such as lawsonite, aragonite, and amounts of volcanic rocks and radiolarian
glaucophane. The eastern belt contains at least chert. The structural thickness is on the order of
two distinctive tectonostratigraphic units, the 1000 m.

433
434 M. C. Blake, Jr, A. S. Jayko & D. G. Howell
....- .- .........
CENTRAL BELT EASTERN BELT KLAMATH PLATE
. . . . . . . .
. . . . . . . . Tectonic melange ~South Fork Mountain Sch|ut F?-3
. . . . . . . . . Yolle Belly terrene
. . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . . . ~UDper mudotono unit
GREAT VALLEY PLATE
~. ". •. ". ". ". ". ". ". ". ". ". • • . ,
[ JGrnywocke end chert unit
~-"~Lower mudstone unit

THRUST FAULT
10KM A A •
.'.'.'.','.'.'.'.'.'.'.'.'.'i'i'i'i'.'.
40e15 '

m _ _

• ~,,~ .. ~,, - . . . . • . •

40"06
6' 123;00' 122"45'

Fn(;. 1. Generalized geological map of the Yolla Bolly area, northern California. Geology from
Worrall (1979) and unpublished mapping by M. C. Blake et al.

The middle unit is made up predominantly of The principal type of clast in the coarser
thick-bedded to massive quartzofeldspathic grained beds are intrabasinal mud rip-up frag-
sandstone that includes several thick horizons ments, though extrabasinal felsic volcanic and
of thin-bedded radiolarian chert. Igneous rocks granitic clasts are also present• Flame struc-
are rare and consist of small diabase-gabbro tures, dewatering pillars, and slurry bedding are
sills, largely intrusive into the chert-rich parts of seen within most sandstone beds, but, surpris-
the unit. The thickness of the unit, as deduced ingly, dish structures have not been observed•
from the structure sections is 1000-2000 m. Load structures, flutes and grooves are display-
The upper unit is composed predominantly of ed on the basal surfaces of some beds.
mudstone and thin-bedded quartzofeldspathic The classification of lithofacies for redepo-
sandstone, but includes abundant intrusive and sited mass flow sediments developed by Mutti
extrusive bodies of basalt and quartz kera- & Ricci Lucchi (1972), and Ricci-Lucchi (1975)
tophyre. Lenses of radiolarian chert are less is a convenient means of describing the
common than in the other units. The structural sedimentary rocks of the Yolla Bolly terrane.
thickness is about 1000 m. Examples of all seven lithofacies ( A - G , listed
Smooth, water-worn exposures in all three in Underwood & Bachman 1981) can be found
units display an abundance of primary in this terrane, though their abundance is highly
sedimentary structures. The observed struc- variable.
tures are common characteristics of flysch, and The lowest stratigraphic unit is characterized
in particular, of turbidites and associated mass- by ripple drifted siltstone (facies E) in associa-
flow deposits. Many beds can be related to the tion with thin-bedded turbidites (facies D) and
Bouma sequence, and small-scale cross- minor amounts of massive mudstone (facies G).
bedding (ripple-drifted siltstone) is ubiquitous. In the lower part of this unit is a 100-200 m
Sandstones commonly show normal grading thick interval of sandstone (facies B and C) with
and some pebbly beds are inversely g r a d e d interbeds of conglomerate (facies A), thinner
 Sedimentation of the Franciscan assemblage, California 435

F~. 2. Examples of coherent strata from the Yolla Bolly terrane, textural zone 1. (A) Lithofacies B,
composite (amalgamated) beds of thick-bedded sandstone. (B) Lithofacies D and E thin bedded
turbidites. (C) Lithofacies E, ripple-drifted sandstone (Bouma T~e and Tce turbidites).

bedded turbidites (facies D and E) and pebbly intervals of thin-bedded turbidites (facies D or
mudstone and slump folded strata (facies F). E). No vertical asymmetric cycles of bedding
Within this sandstone interval there are at least thicknesses are evident. Because of the abun-
four thinning- and fining-upward megase- dance of sandstone in this unit and the absence
quences, each 30-50 m thick. Photographs of of more ductile interbeds this is the most cohe-
these facies are shown in Fig. 2. rent of the three lithological sequences. Because
The above association of lithofacies specifies of this coherency we have been able to map
an environment of principally tranquil, dilute beds of ribbon chert for as much as 20 km along
turbidite deposition (overbank?) with episodes strike (Fig. 3A). Most chert units are 20-50 m
of much higher energy turbidite deposition thick. The lower contacts of the chert beds are
(feeder channels). The conglomerate in basal commonly gradational; sandstone passes up-
parts of the inferred channelled sequences is ward into siliceous shale (1-5 m thick) which
coarser grained than any other conglomerate in grades into pure chert (Fig. 3B). Coarse-
the Yolla Bolly terrane. The preponderance of grained sandstone usually overlies the chert,
mudstone cut by channels filled with coarse commonly separated by a thin (50 cm) basal
material suggests either an inner fan or slope breccia of chert and mudstone (Fig. 3C, D).
with feeder channels as a depositional setting. Within the context of a submarine fan model,
The lower unit is probably the most proximal of the above association of lithofacies indicates a
the three stratigraphic sequences to be discus- middle or supra-fan setting, though the occur-
sed. rence of interbedded chert is enigmatic. Chert
The middle unit consists principally of mas- occurs in all three of the stratigraphic units but
sive bundles of composite (amalgamated) stands out in the middle unit because of its
thick-bedded turbidites, fluxoturbidites, or structural competency. The chert beds must
grain-flow deposits (facies B and C), with thin represent periods of diminished to non-existent
436 M. C. Blake, Jr, A. S. Jayko & D. G. Howell
clastic deposition in areas between places of environment of this unit; a middle to outer fan
active deposition in a supra-fan setting, analo- fringe setting is possible as is a slope setting. In
gous to mud blankets between depositional either case, the intrusive rocks are puzzling.
lobes, or chert deposition may have occurred To summarize, the sedimentological data
more uniformly over the entire fan during offer some constraints for the depositional en-
intervals of high sea-level that trap detrital vironment of strata in the Yolla Bolly terrane.
material in inner shelf, estuarine, and bay set- All the clastic strata derive from mass-flow
tings. processes. Deposition probably occurred in a
The highest lithological unit is principally rip- deep-marine setting. Lithofacies associations
ple-drifted siltstone (facies E) and thin-bedded suggest submarine fan styles of deposition
turbidites (facies D and E). Near the top of this though there are not enough data to specify fan
unit are several thinning-upward cycles of geometries or basin configurations. A possible
greywacke (facies B and D giving way to D and depositional environment for the strata of the
E) which provide evidence of some channellized Yolla Bolly terrane would be restricted sub-
turbidite deposition. The channel fill is finer marine fans within a region subjected to crustal
grained than the analogous detritus in the extension. The lithofacies associations can be
lowest stratigraphic unit. However, the princi- explained as basin-trough (middle unit) and
pal distinguishing feature of this unit is the basin margin facies (lower and upper units).
abundance of greenstone bodies. Many of the The bimodal submarine volcanic activity is also
greenstone blocks are essentially knockers in a consistent with such a regime (Crowell 1976;
sheared matrix of facies E siltstone; numerous Lonsdale & Lawver 1980). The occurrence of
intrusive rocks, however, of basalt and quartz the thick chert interbeds are problematical, but
keratophyre suggest a history of a volcanism would seem to fit a restricted basin undergoing
possibly related to the depositional environ- extension better than a more dynamic trench
ment of this upper unit. The lithofacies associa- setting.
tions do not rigidly constrain the depositional

Fl(;. 3. Depositional relationships between radiolarian chert and turbidites. (A) Through-going chert
horizons on south Yolla Bolly Mountain. (B) Interbedded chert and mudstone at the base of a 50 m
chert bed along Balm of Gileod Creek. (C) Same chert bed showing overlying coarse-grained
sandstone. (D) Same bed showing chert breccia at base of overlying sandstone.
 Sedimentation o f the Franciscan assemblage, California 437

Age with well-developed quartz segregation (textu-

ral zone 3) (Blake et al. 1967). Brittle breakage
Radiolaria from all three subunits are nearly of sandstone beds in the broken formations
identical, and indicate a latest Jurassic or ear-
along the limbs of folds parallel to the tectonic
liest Cretaceous age (David L. Jones, pers. foliation (S1) is manifested by sheared phacoids
comm. 1981). Megafossils were found only in or boudins. Small-scale tectonic structures are
the lowest structural unit and include Buchia folded around the F 2 hinges, but seemingly not
okensis (upper Tithonian) and Buchia pacifica around F 1 hinges. From these observations we
(Valanginian) (Blake 1965). Rb-Sr and 4°Ar/
infer that the deformational fabric seen in the
39Ar measurements on metagreywacke from the
broken formation (Fig. 4), and the textural
Yolla Bolly terrane give metamorphic ages of zonation in the meta-greywacke, formed simul-
about 110 Myr (Lanphere et al. 1978). It there- taneously during the earliest period of deforma-
fore appears that sedimentation occurred about tion.
30-40 Myr prior to blueschist metamorphism. A second set of folds (F2) trend NNE and has
a moderately well-developed axial plane schis-
tosity ($2). These folds are strongly asymmetric
Structure showing vergence to the ESE.
At least three periods of folding have been The third set (F3) is characterized by open,
recognized. The earliest is characterized by NW-trending, and locally overturned folds with
WNW-trending isoclinal folds (F 0 and associ- vergence from NE to SW. These folds are not
ated axial-plane cleavage ($1). Within the cohe- accompanied by axial plane cleavage or schis-
rent sandstone unit, this cleavage defines a tosity.
textural zonation that ranges from non-foliated The thrust faults shown on the geological
greywacke (textural zone 1) through foliated map (Fig. 1) are located on the basis of the
semischist (textural zone 2) to foliated schist juxtaposition of textural grade and lithological

~.~. ,=~ ,~,~, ~ , ~ , ~

FIG. 4. Examples in the stages of development of broken formation or m61ange. (A) Slightly
stretched thin-bedded turbidites. (B) Thin bedded turbidites stretched and broken by normal faults.
(C) Boudins and phacoids of sandstones in a mud (Te part of a turbidite) matrix, bedding still
discernible. (D) Totally disrupted bedding, blocks of sandstone in a homogenized mud matrix.
 438 M. C. B l a k e , Jr, A . S. J a y k o & D. G. H o w e l l

facies. They appear to have been folded by F 2 suggest the following plate tectonic history.
and clearly pre-date F 3. During latest Jurassic and earliest Cretaceous
Metamorphism time the clastic sediments were deposited in a
restricted basin that was also the site of deposi-
In addition to the previously described tex- tion of thick interbeds of radiolarian chert.
tural zonation, all of the rocks contain high- Sandstone petrography and the abundance of
pressure minerals typical of the blueschist intrusive and extrusive volcanic rocks suggest
facies. Within the metagreywacke, crystals of that the basin was on a continental margin
lawsonite, phengite, and chlorite have grown undergoing rifting, perhaps an environment
parallel to S~, particularly in the higher textural similar to the present-day Gulf of California or
grades. The greenstones contain, in addition, the California Continental Borderland.
glaucophane and pumpellyite. Limited oxygen Some 30 or 40 Myr later, all of these rocks
isotope data (Taylor & Coleman 1968) suggest were involved in a subduction event that pro-
that these rocks formed at temperatures of duced metagreywacke and broken formation,
about 100°-300°C. The widespread presence of with the attendant development of high-
aragonite indicates that pressures were on the pressure metamorphic minerals such as lawso-
order of 4--6 kb, corresponding to a depth of nite, aragonite, and glaucophane. Obduction of
15-20 km. the subducted complex back to the surface
apparently occurred soon after preserving the
Conclusions high-pressure-low-temperature metamorphic
mineral assemblage.
Keeping in mind the many uncertainties, we

References
AI.VAREZ, W., KENr, D. V., PREMOL! SILVA, ing and shortening of Coastal Ventura basin,
ISABELI.A,SCIlWEICKERT,R. A. & LARSON,R. A. California. In: Aspects of the Geologic History of
1980. Franciscan Complex limestone deposited the California Continental Borderland. Pacif.
at 12° south paleolatitude. Bull. geol. Soc. Am. Sec. Am. A~soc. Petrol. Geol., Misc. Publ. 24,
91,476-84. 365-82.
BACHMAN,S. B. 1978. A Cretaceous and early Terti- LANPHERE, M. A., BI.AKE, M. C., JR & IRWIN, W. P.
ary subduction complex, Mendocino coast, 1978. Early Cretaceous metamorphic age of the
northern California. In: HOWELl., D. G. & South Fork Mountain Schist in the northern
McDOUGALL, K. A. (eds). Mesozoic Coast Ranges of California. Am. J. Sci. 278,
Paleogeography of the Western United States. 798-815.
Pacif. Sec., Soc. econ. Paleontol. Mineral. LONSDALE, P. & LAWYER, L. A. 1980. Immature
Tulsa, Symp. Pacif. Coast Paleogeogr. 2, 419- plate boundary zones studies with a submersible
30. in the Gulf of California. Bull. geol. Soc. Am.
BEUTNER, E. C., MCLAUGHLIN,R. J., OHLIN, H. N. 91,555-69.
& SORG, D. H. 1980. Geologic map of the King MurrI, E. & Rico Lvccnl, F. 1972. Le torbiditi
Range and Chemise Mountain Instant Study dell'-Appennin settentrionale: introduzione
Area, northern California. U.S. geol. Surv. Map ail'analisi di facies. Mere. Soc. geol. ltal. 11,
MF-1196A. 161-99; English translation by T. H. Nilsen,
BIsHop, D. G. 1977. South Fork Mountain Schist at 1978, Intern. geol. Rev. 20, 125-66; AGI Reprint
Black Butte and Cottonwood Creek, northern Ser. 3.
California. Geology, 5, 595-9. PESSAGNO, E. A., JR 1977. Upper Jurassic radiolaria
BLAKE, M. C., JR 1965. Structure and petrology of low and radiolarian biostratigraphy of the California
grade metamorphic rocks, blueschist facies, Coast Ranges. Micropaleontology, 23, 56-113.
Yolla Bolly area, northern California. Thesis, RICcl-Luccm, F. 1975. Depositional cycles in two
Ph.D., Univ. Stanford. 91 pp. turbidite formations of northern Apennines. J.
, IRWIN, W. P. & COLEMAN,R. G. 1967. Upside- sediment. Petrol. 45, 3-43.
down metamorphic zonation, blueschist facies, TAYLOR, H. P. & COLEMAN, R. G. 1968. t80/160
along a regional thrust in California and Oregon. ratios of coexisting minerals in glaucophane-
Prof. Pap. U.S. geol. Surv. 575-C, C1-9. bearing metamorphic rocks. Bull. geol. Soc. Am.
-- , JAYKO, A. S., NEUMANN, R., WILSON, B. & 79, 1727-55.
WORRALL, D. M. no date. Geologic map of the WORRALL, D. M. 1979. Geology of the south Yolla
Yolla Bolly Wilderness Area, northern Cali- Bolly area, northern California, and its tectonic
fornia, scale 1:48,000 (unpubl.). implications. Thesis, Ph.D., Univ. Texas,
CROWELL, J. C. 1976. Implications of crustal stretch- Austin. 250 pp.
M. C. BLAKE, A. S. JAYKO & D. G. HOWELL, U.S. Geological Survey, Menlo
Park, California 94025, U.S.A.
 Deformation of partly dewatered and consolidated Franciscan
sediments near Piedras Blancas Point, California

Darrel S. Cowan

SUMMARY: Bedded sequences of turbidite sandstone and mudstone and associated

chaotic, polymict pebbly mudstone record a post-depositional, pre-metamorphic deforma-
tion. Sandstone layers were extended and locally disrupted and fragmented and most
clasts in pebbly mudstone were deformed into oblate ellipsoids. On the scale of outcrops
and hand-specimens, extension in sandstone was accomplished by: (1) mesoscopically
ductile pinch-and-swell behaviour and boudinage; (2) extreme necking, which locally
resulted in extension fractures or brecciation; and (3) sets of parallel shear fractures
inclined to bedding. On smaller scales, ductile changes in shape were accommodated by
both intergranular flow of sand-rich sediment and slip along zones typically 0.1 mm or less
in thickness. Microfracturing, granulation, and cataclasis of clastic grains are less
commonly observed, even adjacent to microfaults. Mudstone was also mesoscopically
ductile but generally it flowed by displacements on penetrative, anastomosing surfaces of
slip and was transformed locally into 'scaly clay'. These Franciscan sediments probably
contained appreciable pore fluids when they were deformed, since the deformation of
unconsolidated sands by homogeneous, intergranular flow rather than widespread catacla-
sis is fostered by high interstitial fluid pressures. More localized discontinuities, such as
extension fractures, shear fractures, and microfaults, formed where sand had more
completely dewatered and consolidated. Thus, material response may partly reflect water
content and consolidation history. The axially symmetric extension recorded by small-
scale structures suggests that sediments were deformed in a shallow-level, laterally
unconfined environment on the trench slope, perhaps in response to vertical shortening
and horizontal elongation that accompanied gravitational collapse and spreading.

An important question that commonly arises in polymict, blueschist-bearing olistostromes SE

subduction complexes is whether sediments of San Simeon. The results of this research bear
were deformed while incompletely lithified or on the general problem of whether the de-
after complete consolidation. At active subduc- formation accompanied gravity-driven down-
tion zones, 'soft-sediment' deformation theor- slope mass transport affecting the shallow levels
etically could occur at the base of the trench of a trench slope, or instead was more closely
slope as sediments are scraped off and accreted, confined within the deeper levels of an accre-
or deep beneath the inner wall of the trench if tionary prism and therefore 'tectonic' in charac-
sediment is hauled down and subducted, or on ter.
the inner wall by gravitationally driven sliding The reader should consult Hsii (1968, 1969),
or spreading. In the late Mesozoic-early Ceno- Hall (1976), Cowan (1978), and Smith et al.
zoic Franciscan subduction complex of the Cali- (1979) for general background on the geological
fornia Coast Ranges, only a few studies, such as setting.
those by Kleist (1974) and Bachman (1978), Three major rock types are exposed along
have described soft-sediment structures, prob- the shore between Piedras Blancas and San
ably because much of our attention has focused Simeon: well-bedded turbidites comprising
instead on more highly metamorphosed and 90% sandstone layers displaying a variety of
multiply deformed terranes where early struc- well-preserved B o u m a sedimentary structures;
tures are rarely preserved. The principal objec- lithologically similar but more mudstone-rich
tives of this paper are to describe small-scale rocks in which bedding has been variably dis-
structures in Franciscan rocks that are superbly rupted by the early deformation described in
exposed along the shoreline from 2 km north of detail below; and internally chaotic, polymict
Piedras Blancas Point to San Simeon Point, pebbly mudstone which locally records the
California (Fig. 1) and to show that they can be same deformation. The rocks bear a close pet-
used in conjunction with microfabric to infer rological, structural and metamorphic kinship to
the properties and behaviour of sand and mud the Franciscan exposed SE of San Simeon
during deformation. An earlier phase of this (Cowan 1978; Hall 1976), even though they are
study (Cowan 1978) dealt with the origin of separated by the San Simeon fault zone, along

439
440 D. S. Cowan
35 ° ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
45'

"::i iliiiiii:iiiiiii?i?ilii!il ?ii? ii:)::?iii?iil)i)ili?iii?i?iii)i!i)i)i!i!ii

Piedros Blaocos -
mo i n t
i!~SAN S!MEON~

F RA N CISCA N N : :i: ::i::!::i

~ JURASSIC OPHIOLITE ~ 1
OVERLYING STRATA
I
121 ° 15'

FI(;, l. Generalized geological map of Piedras Blancas-San Simeon area, modified from I lall (I 976).
'Jurassic ophiolite and overlying strata" are part of the Great Valley sequence, which formed in a
lk)rearc basin associated with the Franciscan subduction complex.

which Hall (1975) has postulated at least 80 km posed between San Simeon and Piedras Blancas
of late Cenozoic right-lateral strike-slip. display a pervasive deformation which was
superimposed on So and related syn-
sedimentary features (Fig. 2). Extensional
Small-scale structures
P r i m a r y structures

s o m e primary sedimentary structures, pre-

dating deformation, formed at or immediately
below the sediment-water interface in turbi-
dites and are classified here as syn-sedimentary
or penecontemporaneous. Fine bedding-
parallel laminations, designated S0, range from
<1 mm to several mm in thickness and are
defined by changes in grain size and composi-
tion. Dark laminae are clay-rich; lighter lami-
nae are sandy. So is prominent in sandstone
layers less than 20 cm thick. In mudstone, it is
parallel to a weak fissility, which is also syn-
sedimentary as defined above. At the base of
some sandstone layers, classical examples of
disruption accompanying turbidite deposition
were observed, including ripped-up mudstone
clasts that had already developed bedding- Fl6. 2. Diagrammatic summary of mesos-
parallel fissility due to compaction at the de- copic structures recording layer-parallel ex-
positional interface. Some sand irregularly in- tension, correctly oriented with respect to
truded downward or was swirled into less con- one another but displayed as they would
solidated mud. The dispersal and mixing within appear if they were together in one large
outcrop. Extended layers are grossly para-
associated polymict pebbly mudstones can also
llel to undeformed beds. Structures are
be considered as a primary event. described in the text and actual examples
are shown in Figs 3-8. Deformed clasts
Mesoscopic structures near bottom are in polymict pebbly mud-
stone associated with interbedded sand-
Most of the Franciscan sedimentary rocks ex- stone (shaded) and mudstone.
 Deformation of Franciscan sediments, California 441

FIG. 3. An extended sandstone layer underlain and overlain by more disrupted, thinly bedded,
mudstone-rich section. In addition to the necked area near the right edge of the photograph, note the
irregularly bulbous protrusions along the upper and lower surfaces of the layer. Layering is parallel
to hammer handle, which is 26 cm long.

FIG. 4. Extensional shear fractures developed in a thick sandstone layer. Hammer handle is parallel
to bedding laminations within lozenges, which have rotated with respect to unaffected overlying
sandstone layers. See text for kinematic interpretations.
442 D. S. Cowan

Ft~i. 5. Adjacent sandstone layers exhibiting different mesoscopic deformational behaviour. The
hammcr rests on part of a layer that flowed ductilcly and separated into boudins. The upper surface
of the underlying, thicker layer is offset by an extensional shcar fracture, recording brittle bchaviour,
that does not extend into the boudin. Intcrbedded mudstone also flowed ductilely to accommodate
overall flattening.

F,~. 6. Sawn surface of hand specimen, cut normal to layering. White scale bar is 2 cm long. Thicker
extended layers are separated by mudstone-rich intervals in which thin sandy laminae are more
thoroughly disrupted and disaggregated and locally folded. This sample is representative of the
deformation in the thinly layered sections associated with the bed illustrated in Fig. 3.
 Deformation of Franciscan sediments, California 443

FIG. 7. Close-up of part of specimen in Fig. 6. Scale bar = 1 cm. Deformation mechanisms vary
within a single layer, and from layer to layer. Extension in the prominent bed in centre was
accommodated principally by slip along semi-penetrative shear fractures in the strongly necked
region, and by intergranular flow at the right edge of the photo. Delicate swirling and incipient
disaggregation of thin sandy laminae above involved homogeneous intergranular flow, rather than
fracture and cataclasis of individual grains, on a microscopic scale.

Fro. 8. Another close-up of part of specimen in Fig. 6. Scale bar = 1 cm. Small-scale analogues of
extensional shear fracture shown in Fig. 4. Microfractures are antithetic normal faults (see text for
discussion). This mesoscopic behaviour is 'transitional' between ductile flow and brittle fracture. See
Fig. 11 for a photomicrograph of a typical microfault.
444 D. S. Cowan

structures including pinch-and-swell, necking, swelling attests to mobility of sand within layers
boudinage, and shear fractures are obvious in and Smith (1977) maintains that 'growing'
outcrops (Figs 3, 4 & 5) and hand specimens boudins are evidence for non-Newtonian flow.
(Figs 6, 7 & 8). They are most pronounced in Some sandy protrusions are hook-shaped or
pseudo-stratiform sequences of mudstone and lobate (Fig. 3). Others are partly separated
sandstone that probably were once bedded from their parent layers by narrow injected
turbidites. In some continuous, 500 m long tongues of mudstone. In detail, margins of
exposures, rocks are variably deformed deformed sand layers range from smooth to
throughout and structures are not confined to irregularly scalloped or hummocky. Pinch-and-
localized, stratigraphically truncated intervals swell and associated structures in unmeta-
as slump folds typically are (e.g. Grant-Mackie morphosed or very low-grade sedimentary
& Lowrie 1964; Gregory 1969; Woodcock rocks are described or illustrated by Horne
1976). (1969), Gregory (1969), Corbett (1973), and
A general observation on both outcrop and Kleist (1974), who all interpreted them as evi-
hand-specimen scales is that semi-continuous, dence for deformation of incompletely consoli-
thicker sandstone layers are separated by mud- dated sediments during submarine slumping or
stone-rich intervals where thinner sandy layers sliding. In hand specimens, chaotic zones be-
and laminae are more thoroughly disrupted tween extended semi-continuous layers contain
(Figs 3 & 6). Possible explanations for this variably shaped wispy, elongate, or contorted
disharmonic behaviour are presented below. fragments of sandstone (Fig. 6). Most have
Although nearly all of the mesoscopic struc- highly irregular and locally diffuse boundaries.
tures described here are evidence of layer insta- Many appear to have been injected by mud-
bility during extension, they can be divided into stone matrix or to have sent tiny dykelets of
two gradational types on the basis of deforma- sand into matrix.
tional modes observable at this scale. During Mesoscopic ductile swelling and necking, and
mesoscopically ductile deformation, sand layers irregular distortion of wispy fragments, were
changed shape by mechanisms producing accommodated internally by two principal
apparently homogeneous flow or by small dis- mechanisms. Delicate So bedding laminations
placements along internal slip surfaces (Figs 7 are a useful marker in this regard (Figs 6, 7 &
& 8). Mesoscopically brittle behaviour involved 8). They are either truncated by slip-surfaces
discontinuities extending completely across (microfaults) less than 1 mm thick or irregularly
layers; extensional shear fractures are a familiar swirled. Some distinct laminations become
example (Figs 4 & 5). HsiJ (1968, 1974) first progressively vague or distorted and pass into
recognized and illustrated some of these meso- thoroughly churned, homogenous sandstone or
structures in the study area. Cowan (1978) siltstone. In most sandy layers, variably
suggested that most of the brittle fractures in oriented microfaults and swirled domains are so
olistostromes SE of San Simeon were superim- intimately associated that they must have
posed on earlier ductile structures (D1) during a formed contemporaneously. Thick (>20 cm)
subsequent deformation (D2). A more exten- layers of sandstone are generally devoid of So
sive study of the bedded sequences at Piedras laminations, but similar mechanisms probably
Blancas indicates that both types of behaviour operated in them as well. Many outcrops, hand
were largely contemporaneous. specimens and thin sections of massive, non-
Ductile pinch-and-swell structures, ranging laminated fine- to coarse-grained sandstones
from geometrically ideal to highly irregular, are are crossed by an irregular network of curvipla-
abundant in semi-continuous layers. Necking is nar dark 'veins', generally <2 mm thick, which
locally extreme, but true boudinage, involving I call 'web structure'. They contain more clay
complete separation of a continuous layer into and silt than surrounding sandstone and prob-
isolated fragments or boudins, is rare. Illustra- ably served as both dewatering conduits and
tions of pinch-and-swell structure and boudin- slip surfaces, since some strands are offset a few
age in metamorphic rocks (e.g. Ramsay 1967; millimetres by others.
Weiss 1972) show that the thickness of most Mesoscopically brittle structures include ex-
affected layers does not increase during de- tension ('tension') fractures, oriented approxi-
formation, but some sand layers at Piedras mately normal to layering, and extensional
Blancas apparently swelled between thinned shear fractures. The former occur in thin necks
necks. Bulbous protrusions resembling sandfil- between boudins and contain mudstone that
led aneurysms occur along both upper and flowed from adjacent layers as fractures
lower surfaces and thus cannot be explained as opened. I arbitrarily classify the internal micro-
load casts or scour-and-fill structures. True faults described above as extensional shear
 Deformation of Franciscan sediments, California 445

fractures where they affect both boundaries of a most or all fractures are inclined in the same
sand layer. They are common structures here as sense, but faults within thinner layers common-
well as in some other terranes that record ly have opposing dips (see Fig. 7 in Cowan
soft-sediment deformation (e.g. Gregory 1969; 1978). In some continuous sequences up to
H o m e 1969; see Weiss 1972, plate 141A for an 50 m thick, I noted several layers that are
example in a gneiss). Rast (1956) described each characterized by either sense of rotation.
analogous fractures, which he termed 'joints', Gregory (1969) described the same situation
in metamorphic rocks and considered the in- in lower Miocene strata in New Zealand.
tervening segments of layers as 'lozenge-shaped Another less common structure records a
boudins'. Nearly all such fractures in the Pied- deformational mode distinct from the brittle
ras Blancas-San Simeon area are inclined >45 ° and ductile end-members described above.
from the top surface of sandy layers and hence Necks in some sandstone layers are chaotic
are normal faults if viewed in sections perpen- breccias consisting of angular to subrounded
dicular to layering. I interpret the faults as fragments surrounded by a structureless matrix
Coulomb shear fractures, and displacements of sandstone. Extension was accommodated
along them necessarily resulted in layer-parallel partly by brittle fracture, but fragments rotated
extension. Maximum (ol) and minimum (0"3) randomly within a homogeneously flowing sand
principal stresses were oriented respectively which was derived from the same part of the
normal and parallel to layering at the instant deforming layer.
shear fractures formed. Distinctively angular There are two remarkable aspects of the de-
lozenges (Fig. 4) are typically equidimensional formation in the Piedras Blancas-San Simeon
rather than elongate indicating that the spacing area. In my earlier study (Cowan 1978) I
of fractures is directly proportional to layer showed that clasts in olistostromes SE of San
thickness. I agree with Whitten (1966) that Simeon were variably flattened into oblate
lozenges formed in this way should be distin- ellipsoids and correspondingly extended in all
guished from true boudins resulting from meso- directions in the XY-plane of the total strain
scopically ductile necking or extension fractur- ellipsoid. The strong preferred orientation of
ing. Hsii (1968, 1974) suggested that some deformed clasts defines a foliation, Sl. Most of
wedge-shaped lozenges intimately associated these clasts were mesoscopically ductile, but
with pinch-and-swell and extensional fractures external changes in shape were accommodated
in thickly bedded strata are bounded by 'com- by brecciation, granulation, and cataclastic flow
pressional fractures' (reverse faults), but I on a smaller scale. Bedded sequences and peb-
found no evidence to support his conclusion bly mudstones at Piedras Blancas experienced
when I examined the same exposures. exactly the same type of total strain. Axially
Rast (1956) pointed out that in his examples, symmetric layer-parallel extension, accommo-
lozenges rotated during slip along fractures. dated by pinch-and-swell structure and shear
This behaviour is a consequence of both frac- fractures, is apparent in all cross-sections
ture geometry and the fact that faults are viewed normal to layering (Fig. 2). There is no
confined to sandstone layers and very rarely detectable lineation, defined by either the long
extend across adjacent mudstone. Fragmented axes of boudins or regions of extreme necking,
layers are typically overlain and underlain by as would be expected if a unique direction of
unfaulted or ductilely extended sandy layers maximum elongation (X > Y) had persisted
which preserve the local orientation of bedding throughout the deformation. Surf-eroded expo-
prior to deformation (Fig. 4). The normal faults sures approximately parallel to deformed layer-
are antithetic. SOin the fractured layers under- ing show irregularly rounded, slightly elevated
goes a rotation equal to the angle between hummocks of sandstone and depressions of
bedding laminae (So) in individual lozenges and mudstone.
So in adjacent beds, while the faults themselves Equally striking, in view of the abundance of
externally rotate in the same sense. (Tyler extensional structures, is the near absence of
(1975) analysed an analogous kinematic situa- folds and other evidence for layer-parallel
tion in fractured brittle clasts in ductile mat- shortening. In intermittent seacliff exposures
rices.) As a result, the long axes of lozenges from Piedras Blancas to a point 20 km SE, I
become parallel to So in adjacent layers, but found one 3 m thick interval of disharmonic,
bedding laminae within lozenges are noticeably mesoscopic folds in the midst of a regularly
oblique. In spite of the deformation, meso- bedded sequence of sandstone and mudstone.
scopic layering defined by layer boundaries is Their geometry and setting recall folds inter-
preserved rather than chaotically obliterated. preted as submarine slump-related structures
In any particular bed thicker than about 10 cm, elsewhere (e.g. Grant-Mackie & Lowrie 1964;
446 D. S. C o w a n

Gregory 1969; Rupke 1976; Woodcock 1976). thickness. Other sandstones, particularly those
A compaction fissility in sandstone and siltstone displaying mesoscopic web structure, comprise
developed parallel to bedding laminae prior to not only a network of planar zones but also
folding, but a later, weaker fissility formed in irregularly shaped domains of granulated mater-
the mudstone cores of some folds. A single, ial that invaded pristine sand (Fig. 9). How-
20 cm thick, tightly folded sandstone layer ever, in spite of the drastic changes in thickness
occurs in the midst of deformed strata at Pied- of extended, semi-continuous layers and the
ras Blancas Point. In both localities, the later irregular distortion of delicate, wispy laminae
fissility is nearly parallel to the axial surfaces of only a few millimetres thick, most samples
the folds, which are themselves approximately examined petrographically show little evidence
parallel to layering in over- and underlying for cataclasis or granulation of individual sand
strata. On a hand-specimen scale, shreds of grains. The modest alteration of original clay
sandy layers <1 cm thick in highly disrupted matrix to chlorite and sericite does not mask the
mudstone-rich intervals locally define isolated well-preserved, easily recognizable detrital tex-
fold closures (Fig. 7). ture of the sandstones. Sand grains must have
merely rotated and jostled their neighbours,
Microscopic structures even in some regions of the most intense distor-
tional strain. For example, in the sharply taper-
Mesoscopically ductile flow of sand was ing edges of several flattened clasts of sand-
accommodated on a microscopic scale by either stone in pebbly mudstone, I could detect no
relative intergranular movements or cataclasis. greater incidence of grain breakage or micro-
Cataclasis in sandstones produces microscopi- cracking than in undeformed layers or clasts,
cally distinctive, poorly sorted gouges in which where they are insignificant.
overall grain size has been reduced and angular Extension and flow are also recorded by
remnants of sand grains are dispersed in a dark, microscopic analogues of the pinch-and-swell
irresolvable matrix (e.g. Dunn et al, 1973: structure, injection features, and irregular
Aydin 1978). In some sandstones, glanulation swirling noted in hand-specimens. Detrital
is confined to isolated, planar zones of slip less sand-size clasts of quartz and feldspar in wisps
than 2 mm, but commonly averaging 0.5 mm in that are only one or two grains thick are

Fro. 9. Photo of a thin section from an elongated sandstone clast in pebbly mudstone; field is
20 x 30 mm. Dark zones consist of cataclastic 'gouge' in which sand grains have been fractured and
comminuted. Non-cataclastic flow was probably an important mechanism in the remainder of the
specimen. Veins are filled with calcite.
 Deformation of Franciscan sediments, California 447

FIG. 10. Photomicrograph of disrupted sandy laminae in siltstone. Swirling and local disaggregation
and mixing were accomplished without microfracturing and granulation. Field is 1.4 × 2 mm.

undeformed, as are isolated grains floating in rounded rather than sharp. Some faults merge
mudstone that has flowed extensively (Fig. 12). with bedding-parallel zones of slip.
D. S. Wood (pers. comms. 1978, 1979) found
identical microscopic relations in the Gwna
Mudstones
olistostrome on Anglesey, Wales, and empha-
sized the complete lack of intragranular In undeformed bedded sections, dark brown
deformation in clasts of what are now ortho- siltstones and claystones are finely laminated
quartzites. I should stress here that none of the and fissile. The bedding-parallel orientation of
microscopic fabrics and textures associated with detrital clays and micas is probably a primary
mesoscopic ductile flow and transposition- settling or compaction fabric. Mudstone in de-
related fragmentation in low-grade metamor- formed sequences (e.g. Figs 3 & 6) is weakly
phic rocks, such as domainal slaty cleavage fissile and breaks less readily along irregular
(Hobbs et al. 1976), metamorphic differentia- surfaces that grossly parallel mesoscopic layer-
tion or in situ intragranular plastic strain in ing but closely conform in detail to the surfaces
quartz, are evident at Piedras Blancas and San of deformed sand layers. Apparently, this fab-
Simeon. ric is most strongly developed adjacent to sand
Microfaults are narrow slip surfaces that are layers >5 cm thick, where it is possible to break
typically less than 0.1 mm in thickness along mudstone into progressively thinner lenticular,
which displacements appear to average a few but non-polished, chips. In thin sections cut
millimetres. In thin section, they are defined by normal to layering, these mudstones display
either films of dark brown, irresolvable matrix vaguely lenticular domains that are defined by
as thin as 0.01 mm or thicker zones, averaging slightly different preferred orientations of fine-
0.1 mm, in which platy grains of clays and grained platy materials. Some domain bound-
phyllosilicates are oriented parallel to zone aries are gradational but most are narrow pla-
boundaries (Fig. 11). Unlike cataclastic zones, nar shear (slip) zones or micro-kinkbands.
sand grains within and adjacent to these micro- Other groups of individual micro shear-zones
faults are not fractured or internally strained. <0.05 mm thick form lacy, anastomosing net-
Where faults emerge from sandy layers into silt- works. Similar microstructures in experimental-
or clay-rich layers, they commonly become less ly and naturally deformed clays are described
distinct as they flatten into parallelism with by Morgenstern & Tchalenko (1967),
overall layering; corners of sandy layers are Tchalenko (1968, 1970) and Maltman (1977).
448 D. S. C o w a n

FIG. 11. Photomicrograph of a microfault similar to those illustrated in Fig. 8. Field is 1.4 × 2 mm.
Separation of dark, mud-rich laminae is about 0.5 mm in this section. Fault zone is -0.05 mm wide
and individual anastomosing strands contain a film of dark brown clay. Note thc absence of
cataclastic effects adjacent to fault.

Birefringent phyllosilicates within each zone development of a true slaty cleavage defined by
are strongly oriented parallel or <30 ° to the microscopically penetrative planar preferred
zone boundaries--an apparently diagnostic orientation of inequant detrital or metamorphic
criterion for slip noted by Morgenstern & minerals. In fact, petrographic observations
Tchalenko (1967) and Tchalenko (1968). suggest recrystallization of detrital clay was
Black scaly mudstone or argille scagliose is negligible. Of four mudstone samples analysed
matrix in some deformed pebbly mudstones by X-ray diffraction (S. Johnson, analyst), one
(Cowan 1978). Hand specimens cleave readily of scaly argillite from San Simeon contains an
parallel to a pronounced mesoscopic foliation expandable mixed-layer (montmorillonite-
and samples can be broken into lenticular chips illite) clay and three from Piedras Blancas
that invariably have shiny, polished and locally contain a non-expandable 10 ,~ clay (presum-
striated surfaces in contrast to the weakly fissile ably illite); all contain chlorite and kaolinite as
brown mudstone described above. In this sec- well.
tion, a primary settling or compaction fabric
defined by silt and sand laminae and oriented
phyllosilicates is locally reoriented by planar State of sediments during deformation
kinkbands, or, more commonly, irregular to
swirled patterns of flow; a superimposed set of Several authors, notably Woodcock (1976),
anastomosing slip surfaces, intersecting in acute Helwig (1970), Horne (1969), Kleist (1974),
angles <60 ° is probably responsible for the and Gregory (1969), have recently addressed
distinctive foliation. the long-standing but vexing problems of how
In summary, mesoscopic flow in deformed to differentiate between structures formed dur-
mudstone was accommodated by the disruption ing gravity-driven slumping of semi-
of a primary sedimentary fabric along a variety consolidated ('soft') sediments and those result-
of semi-penetrative, microscopic structural dis- ing from tectonic deformation of consolidated
continuities that are reflected in a crude, layer- (hard) rocks. Much of the debate concerns the
parallel mesoscopic foliation. Although very geometry of mesoscopic folds and their relation
fine-grained, presumably detrital phyllosilicates to adjacent strata. In some literature (e.g.
are strongly reoriented in narrow shear zones, Williams et al. 1969; Potter & Pettijohn 1977)
there is no evidence for the post-disruption there is a tendency to interpret most soft-
 D e f o r m a t i o n o f Franciscan sediments, California 449

sediment structures as 'penecontemporaneous' flow is counteracted by the 'cushioning' effect

with deposition (hence load or slump phe- of high fluid pressures favouring purely inter-
nomena) and to consider those structures post- granular adjustments. Other experiments on
dating lithification as 'tectonic'. Two concepts dry lithified sandstone (Handin et al. 1963) and
are involved here: the degree of consolidation unconsolidated sands (Borg et al. 1960; Fried-
or physical state at deformation, and the overall man et al. 1980) show that localized or perva-
environment of deformation. These should be sive cataclastic flow is the predominant de-
considered independently, if for no other formation mechanism over a wide range of
reason than incompletely lithified sediments confining pressures (50-200 MPa). From these
could potentially be affected by what all would results, it is clear that much of the sand in the
agree is tectonic defomation as they are carried study area was deformed while it was not only
beneath the toe of the trench slope. My analysis water-rich but also under low effective confin-
of physical state is based mainly on microfabrics ing pressures, since cataclastic mechanisms
rather than mesoscopic structures, which I feel were impeded in regions of extreme mesoscopic
are more useful as a record of local deforma- flow and adjacent to microfaults.
tional environment. Other evidence suggests that some sands
were not only fluid-rich but unconsolidated as
well. Individual sand layers or remnants of
Sandstone
layers commonly display several types of be-
At outcrop and hand-specimen scales, both haviour, even within a single hand specimen.
sand layers in originally bedded sequences and For example, intergranular cohesive flow may
sandstone c!asts in pebbly mudstones show two give way to small offsets along closely spaced
contrasting types of behaviour. In the parlance microfaults, especially in necked regions (Fig.
of rock deformation (Griggs & Handin 1960), 7). In outcrops, layers that have deformed
most have deformed by mesoscopically ductilely are generally interspersed with and
homogeneous, cohesive ductile flow, and some adjacent to layers recording brittle failure along
have failed by brittle extension and shear frac- through-going fractures (Fig. 5). Most of these
tures. Three deformation mechanisms structures were probably contemporaneous,
accommodated cohesive flow at smaller scales. since they are intimately associated throughout
Displacements along closely spaced microfaults substantial volumes of sediment. What
are actually observable without a microscope accounted for these abrupt, local changes in
(Fig. 8) and might be considered 'transitional' behaviour? Overall confining pressure and
behaviour. Cataclastic flow, involving micro- strain rate were probably homogeneous and
fracturing, granulation, and general reduction constant, at least on a mesoscopic scale. Local
in grain size, was also locally important. differences in pore pressure could have been
However, much of the mesoscopic distortion responsible. I suggest that the most likely ex-
was accomplished by intergranular movements planation is that sands were already variably
without concomitant cataclasis. dewatered and consolidated at the time of
These observations can be interpreted in light deformation. Unconsolidated, water-rich
of experimental deformation of sandstone and sands, little modified since deposition, de-
unconsolidated sand. In triaxial compression formed by non-cataclastic, intergranular flow;
tests on water-saturated, porous, consolidated semi-penetrative microfaults, general catacla-
sandstones, Handin et al. (1963) found that sis, and finally brittle shear fractures developed
deformational behaviour was strongly depen- in progressively more consolidated materials. I
dent on effective confining pressure (P~), equal consider all of these structures to have formed
to (confining p r e s s u r e - fluid pressure). At in 'soft' sediment. Behavioural differences re-
high Pe (low fluid pressures), samples deformed flect local variations in water content (and
ductilely by wholesale cataclastic flow on the perhaps Pe), since early consolidation or lithi-
microscale. At progressively lower Pe, grain fication basically involves compaction, reduc-
breakage became localized and finally res- tion in porosity, and dewatering. If, instead,
tricted to narrow, through-going, sharply de- differences in Pe were solely responsible, all of
fined shear zones as the samples experienced the layers would have to have been consoli-
transitional, then brittle behaviour. When con- dated equally and fluid pressure then selectively
fining pressure equalled fluid pressure (P~ = 0) raised in small domains to promote intergranu-
samples were shortened up to 25% without any lar flow. I postulate that early, progressive
evidence for cataclastic flow or granulation, changes in water content in consolidating sands
even along shear zones. Handin et al. (1963) can have as much influence on behaviour and
deduce that frictional resistance to sliding and mechanism as do changes in pore pressure once
45(1 D. S. C o w a n

sands have passed through some critical, but as with sand layers <2 cm thick that were loose
yet experimentally undetermined, threshold of enough literally to be poured out of the core
lithification. barrel. But it seems paradoxical that stiff mud
would behave more ductilely than unconsoli-
dated sand, as the mesoscopic structures at
Mudstone
Piedras Blancas and San Simeon imply. When
It is much more difficult to decipher the most argillaceous materials are experimentally
physical state of associated mudstones on the deformed, they have an interesting response
basis of behaviour and mechanisms alone. that may resolve the paradox. Conventional
Boudinage, pinch-and-swell structures, and graphs of differential stress versus percentage
flattened clasts (Cowan 1978) in the study area strain (e.g. Skempton 1964) show curves stead-
all indicate ductility (or material property) ily climbing to a 'peak strength' at low total
contrasts; hence, mud flowed more readily and strain and then falling to a lower 'residual
suffered greater permanent strains on a meso- strength', or constant stress difference, at which
scopic scale than did sand and other rock types permanent deformation continues indefinitely.
(metabasalt, chert, and metamorphic rock Skempton (1964) and Tchalenko (1970) corre-
clasts). Other features pointing to mobile mud late this drop in strength with the development
are the greater degree of stratal disruption in of complex microscopic shear zones and kink
disharmonically deformed mudstone-rich inter- bands. In other words, a modest amount of
vals (Figs 3 & 6) and mudstone-filled fissures ductile strain in the Franciscan muds, accompa-
and re-entrants in sandy layers on all scales. nied by the development of some of the discon-
Although mesoscopic ductility of mudstone was tinuities observable in thin section, may have
complemented by both brittle and ductile be- fundamentally altered their properties and
haviour in sand, what I would interpret as a allowed them to achieve greater total strains
type of cataclastic flow was more widespread in than interlayered, but more 'competent', sands.
deformed mudstone on a microscopic scale. It Even if these inferences about the properties
seems everywhere to have developed crudely of Franciscan mudstones are correct, it is risky
oriented, semi-penetrative discontinuities such to assume that the microstructures described
as individual or multiple, anastomosing slip above are diagnostic of a partly consolidated
surfaces, kink-bands, and swirled zones, while physical state. Morgenstern & Tchalenko
most sand flowed homogeneously and non- (1967) and Tchalenko (1968, 1970) imply that
cataclastically. These discontinuities are actual- localized microscopic shear-zone structures and
ly expressed mesoscopically as 'scaly' foliation. induced preferred orientation develop over a
If sands were indeed water-rich as concluded wide range in water content. Although meso-
above, then it is certain muds were also water- scopic behaviour and micromechanisms could
rich when they developed the suite of structures depend on not only water content (degree of
described here. For example, Maltman (1977) consolidation) but also fluid pressure, composi-
deformed soft, ductile argillaceous sediments tion, total strain, and strain rate, there are not
with water contents up to 40% and found that yet enough experimental data to evaluate the
flow is microscopically accommodated on semi- influence of each of these variables.
penetrative shear zones and other discon-
tinuities. However, data show that gravita- Summary
tionally compacted muddy sediments dewater
rapidly and are essentially consolidated after The microscopic behaviour of sand suggests
burial of about 250-300 m (Skempton 1970; that it ranged from unconsolidated to semi-
Rieke & Chilingarian 1974). I infer that, at the lithified at the time of deformation. Even
time of deformation, mudstone had undergone though it was a water-rich, 'soft' sediment, it
a proportionally greater decrease in void ratio developed a variety of small-scale structures
and water content than interlayered sands, that may reflect local differences in water con-
which served as permeable reservoirs for water tent or fluid pressure. It commonly displays
escaping from muds. The high fluid pressures so evidence for an 'early ductility' that contrasts
generated prevented consolidation in sands and with the ductile flow and associated microstruc-
promoted intergranular flow within. ! observed tures typical of metamorphic regimes. Wood
exactly these differences in consolidation in (pers. comm. 1978), Schuster (1980), Davies &
cores from sites 496 and 497 obtained on the Cave (1976), and Lambert (1965) all have used
inner slope of the Mid-America Trench off microfabrics, in concert with other evidence, to
Guatemala on DSDP Leg 67 (see von Huene et infer that sediments were disrupted or de-
al. 1981); stiff but pliant muds are interbedded formed while soft, while Gregory (1969),
 Deformation o f Franciscan sediments, California 451

Horne (1969), and Helwig (1970) invoked dif- as a whole, I propose that it can be divided into
ferences in water content and consolidation to an accretionary substructure and an overlying
explain variations in mesoscopic structural style superstructure. Each of these large-scale do-
in semi-lithified sediments that were deformed mains in turn comprises smaller, more localized
by slumping. My conclusions about physical domains of deformation, such as fault zones
state were anticipated by Kleist (1974), whose and conceivably zones of intense flow or fold-
interpretation that Franciscan Coastal Belt ing. The key distinction between substructure
sediments in northern California (see Bach- and superstructure is that in the former, planar
man 1981) were deformed while soft was based zones of slip or flow are confined; they root
largely on the geometry of mesoscopic struc- within or beneath the wedge and need not
tures. intersect the seafloor. In contrast, the super-
structure is the realm of delapsion (Hoede-
maeker 1973) and is necessarily restricted to
Environment of deformation shallow levels of the inner trench-slope where
Modern convergent plate margins laterally unconfined deformation can be partly
accommodated on unrooted slip surfaces that
We can now try to establish the environment must intersect the seafloor. The following dis-
of deformation without having made the a cussion elaborates these concepts.
priori assumption that soft-sediment, or syn- In the substructure, most deformation occurs
consolidational, deformation is in some way as a direct consequence of the descent of
connected with slumping or sliding. Fig. 12 oceanic lithosphere; panels of sediment are
depicts a variety of hypothetical deformational scraped off and accreted at the front of the
environments that can be evaluated using the wedge or subducted beneath the inner wall of
mesoscopic structures at Piedras Blancas and the trench. Localized structures (Fig. 12) may
San Simeon. It is an idealized cross-section of a include active thrusts at and near the toe of the
simple accretionary wedge largely based on trench slope accommodating accretion (Karig et
marine geophysical and DSDP drilling data al. 1980), arcward- or trenchward-dipping faults
from active margins (e.g. Seely et al. 1974; within the wedge that might represent reacti-
Karig & Sharman 1975; Hamilton 1977; Moore vated basal thrusts (Moore & Karig 1976), and
et al. 1979; and others) seasoned with certain orderly folds (White & Ross 1979) or chaotical-
deductions about the nature and distribution of ly churned zones within inter-thrust panels. In
deformation in a dynamic subduction zone (e.g. the superstructure, deformation results directly
Moore & Karig 1976; Cowan & Silling 1978). from gravity-driven, downslope transport
Environments fall into two categories depend- accomplished by a spectrum of mass-movement
ing on scale of observation. Viewing the wedge processes (see Nardin et al. 1979 for a recent

2!M
~ ~,,~_ - - - - - ACCRETED
.,...~'~__.....-," ~ A ~ MATERIAL

Fro. 12. Diagrammatic cross-section of part of the inner wall of a trench at an active convergent plate
boundary. See text for discussion. Horizontal dashed line coincides with the approximate boundary
between layered sediments and acoustically complex (presumably accreted) material as shown on a
typical reflection profile. Wavy, anastomosing lines represent arcward dipping reflectors, interpreted
as fault zones, and a hypothetical detachment zone at the base of a slide (shaded). Deformation at
Piedras Blancas may have occurred within the shaded area as semi-consolidated sediments
gravitationally collapsed and spread laterally.
452 D . S. C o w a n

review of nomenclature). Slides (including Significance of axially symmetric extension

slumps) move en masse along localized, basal
slip zones and may suffer internal deformation The most important constraint on the defor-
as well. Slides grade into mass flows (Dott 1963; mational environment is the type of total strain
Morgenstern 1967) as sediment becomes more recorded by small-scale structures in both bed-
thoroughly dissaggregated and mixed with ded sequences and pebbly mudstones. The
water. Although many slide sheets, ranging stratal disruption in bedded sequences is due to
from 0.1-1 km in thickness, have been boudinage, pinch-and-swell structure produced
documented seismically on passive continental by necking, extension fractures, and shear frac-
margins (see numerous references in Woodcock tures (Fig. 2). All of these structures involve the
1979), it is not yet known whether they are extension of sandy layers and concomitant duc-
abundant or extensive on modern trench tile flow in associated less-competent muds or
slopes. Von Huene (1972), Coulbourn & mudstones. Unlike many examples of boudin-
Moberly (1977), and Karig et al. (1980) invoke age, however, there is no detectable preferen-
mass-transport mechanisms to explain some tial direction of extension at Piedras Blancas;
irregular topographic features on the inner extension apparently occurred with equal ease
walls of the Aleutian, Peru-Chile, and Sunda in all directions in the plane of the layering and
trenches respectively. In view of their preva- is therefore axially symmetric with respect to
lence at passive margins, it is likely that shal- the normal to layering. An imaginary ellipsoid
low-level slide and mass flow deposits will describing the total strain at the scale of hand
prove to be at least as common on relatively specimens and outcrops would have the X Y -
steep, seismically active, and topographically plane (where X I> Y i> Z are the principal
complex trench slopes. They may be confined axes) parallel to layering, and 21 = 22 > 1
to the thin (<2 km) apron of sediments draped (where 2 I, 22, 23 are maximum, intermediate,
over some trench slopes (e.g. off Peru-Chile as and minimum quadratic elongations). The bulk
shown in Kulm et al. 1977, and above the mesoscopic strain is qualitatively recorded by
Mid-America Trench as illustrated by Ibrahim the distortion of clasts in pebbly mudstones,
et al. 1979), although slide detachment surfaces both at Piedras Blancas and south of San Si-
could theoretically form at shallow depths on meon (Cowan 1978), most of which were de-
the inner wall regardless of the composition or formed into flattened oblate ellipsoids. At
consolidation history of sediments if shear stres- mesoscopic and macroscopic scales, the sedi-
ses on the potential detachment exceed the ments behaved as a continuum, but the discon-
shearing resistance of the material. Sliding tinuities accommodating the strain are obvious
would be favoured where sediments are semi- in most hand specimens and thin sections.
lithified and fluid pressures are high (e.g. Booth It is emphasized that the total strain is a true
1979). triaxial flattening (21 = 22 > 1; 23 < 1) involv-
It would be ideal if one could identify a ing finite extension in the XY-plane and not
unique site or trajectory of deformation for any 'apparent' flattening produced by a volume loss
part of the Franciscan, let alone the section accompanying compaction. Gravitational com-
exposed in the study area, in Fig. 12. However, paction is a uniaxial strain involving only verti-
this task is complicated by the probability that cal shortening in the Z-direction (Fig. 13) but
some key deformation styles are common to there are no corresponding elongations parallel
both substructure and superstructure, and the to X and Y because compacting beds are later-
smaller our scale of observation, the more ally confined within a sedimentary basin. Thus,
equivocal our conclusions regarding large-scale the oblate shape of a total strain ellipsoid is not
setting are apt to be. For example, the suite of in itself proof that extension parallel to X or Y
mesoscopic structures developed at A and B in has occurred, and Ramsay & Wood (1973) and
Fig. 12 may be identical since both sites are in Sanderson (1976) showed that an oblate shape
fault zones, but the immediate causes of dis- inherited from compaction can persist even
placements along them are utterly different. where plane tectonic strains, involving no
Woodcock (1976) also cautions that the elongation in the Y-direction, are superposed.
geometry and style of mesoscopic folds alone Dewatering and volume loss were clearly im-
cannot be used to differentiate slide-related portant at Piedras Blancas, but the structures
from more closely confined 'tectonic' deforma- prove that compaction-related changes in the
tion. Some of the following conclusions about Z-direction were either compensated or suc-
both large- and small-scale deformational en- ceeded by true extension in the XY-plane. I
vironments are necessarily speculative in view estimate that layers and clasts locally suffered
of these ambiguities. tectonic extensions as great as 75-100%.
 Deformation of Franciscan sediments, California 453
UNIAXIAL AXIALLY SYMMETRIC to impede strike elongations. The situation is
COMPACTION EXTENSION
somewhat analogous to shallow-level d6colle-
ments in on-land orogens, above which de-
tached masses can deform independently from
confined, subjacent terranes. One could argue
that the pronounced XY-elongations in the
I
I
I
I
study area are a local anomaly that developed
I
i
I in an overall field of plane 'tectonic' strain in
I I the substructure, but I discount this hypothesis
because all of the deformed Franciscan rocks
::::::::::::::::::::::::::::::::::::::::::::::::::::::
: [.. ~ exposed along the coast for 45 km SE of Pied-
ras Blancas appear to record the same type, but
different amounts of strain (Cowan 1978).
FIG. 13. Volume changes in dewatering Several small-scale deformational environ-
sediments are ordinarily accomplished by ments in the superstructure are seemingly com-
vertical shortening without corresponding
horizontal elongation, since layers are con- patible with this type of strain. In the simplest
fined in sedimentary basins. Sediments in case shown in Fig. 12, lateral spreading may:
the study area were laterally unconfined (1) accompany penecontemporaneous sliding
and able to extend in all directions parallel within a few metres of the sediment-water
to layering while shortening in a direction interface; (2) occur within the basal detachment
normal to layering. zone, which is really just a fault zone in a
special, shallow-level setting; (3) occur some-
! suggest that the only plausible large-scale where above the basal slip surface in a zone of
setting compatible with the substantial axially distributed flow accommodating gravitational
symmetric extensions documented here and in collapse. However, I believe (1) can be elimin-
Cowan (1978) is the superstructure. First, con- ated if the pervasive deformation in the Piedras
sider the alternative, using a small volume Blancas-San Simeon region is compared with
element P in the substructure, shown in cross- other examples of penecontemporaneous
section in Fig. 12. Even when the effects of 'slumping' (e.g. Gregory 1969; Corbett 1973;
volume change are taken into account, deform- Woodcock 1976), where slump folds and stratal
ing P would require an elongation of at least disruption occur in stratigraphically confined
50% normal to the plane of the figure and intervals that typically have sharply truncated
parallel to the strike of the accretionary wedge upper contacts.
(assuming Z is in the plane of the section). A choice between environments (2) and (3) is
Strains of this magnitude are doubtful, because more difficult. In a gross sense, faults are
the slice of P is effectively confined in this localized zones of slip characterized by both
direction by the wedge itself. The substructure intense strain relative to adjacent rocks and
is analogous in this regard to classical on-land non-coaxial deformation paths involving a ma-
orogenic belts, such as the Canadian Rockies jor component of progressive simple shear (e.g.
and Appalachian Valley and Ridge province, Ramsay & Graham 1970). Ramberg (1977,
where most of the tectonic deformation is consi- 1978, 1980) argues, on the basis of theory and
dered to be plane strain with negligible elonga- experiment, that horizontal lengthening in gra-
tions parallel to orogenic trends. Hossack vitationally spreading and collapsing sheets or
(1979) reviews estimates of strike elongations in nappes is compensated by vertical shortening
the Appalachians and parts of the Caledonides and that penetrative 'pure shear' (coaxial pro-
and suggests that regional values of 10% are gressive strain) is important in this regime.
typical. In contrast, the shallow-level, laterally Unfortunately, it is not possible to establish
unconfined setting of the superstructure can whether a strain path was coaxial or non-coaxial
accommodate regional, axially symmetric ex- from the shape of the total strain ellipsoid alone
tensions. Deformed sediments in the study area (Hobbs et al. 1976), but I believe that both the
can be thought of as having spread laterally, geometry of the small-scale structures in the
and to do this they must have been in or above a study area and their overall setting favour the
detachment or slip surface that intersected the 'collapse', rather than 'fault-zone' hypothesis.
ocean floor. The low average slope of the ocean First, the same style of deformation is pervasive
floor constrains spreading to be approximately at mesoscopic and larger scales over a large
horizontal and compensatory shortening to be area. The overall impression is that the hetero-
vertical. In this environment (Fig. 12), there are geneity in amount of strain reflects local differ-
no lateral restrictions, other than topography, ences in the ratio of mud to sand and perhaps in
454 D. S. C o w a n

consolidation history. It does not appear that

deformation is sharply localized in 'shear zones' COAXIAL "-' ~ J--"
t
separating thrust sheets or fault blocks that
have been juxtaposed or even modestly dis-
placed relative to one another. Secondly, the
symmetry of the strain itself makes a fault-zone
origin suspect. Faults are generally considered FIG. 14. Profile of the lower part of an
active glacier, modified from Hooke &
to result in plane strain with negligible exten- Huddleston (1978). An internal 'collapse
sions normal to slip directions, but triaxial strain' (Ramberg 1977) is distributed
flattening is theoretically possible if rocks in the through the spreading ice, but the nature of
fault zone are somehow elongaged along strike. the deformation path varies throughout as
Finally, the preponderance of structures re- shown. This model may apply to gravita-
cording layer-parallel extension and the near tionally spreading sediment on a trench
absence of folds in both incipiently and intense- slope, as discussed in the text. Non-coaxial
ly deformed beds and layers (Fig. 2) suggest shear strain near the base probably reflects
that layering was oriented at low angles, or frictional resistance to sliding.
possibly parallel, to the XY-planes of both
infinitesimal and finite strain ellipsoids through- Hooke & Huddleston (1978), summarizes the
out the deformation. For example, delicately progressive strain inside a glacier. It illustrates a
thinned layers in regions of extreme necking did two-dimensional case, but the same deforma-
not rotate with respect to either their parent tion paths would apply in ice caps or ice sheets
layers or nearby less-deformed layers. If cor- even though substantial elongations occur nor-
rect, this relationship between the XY-plane mal to the plane of the diagram. There is a
and layering would indicate a coaxial strain domain of coaxial strain ('pure shear' in Hooke
path. In summary, the simplest hypothesis per- & Huddleston 1978), accommodating vertical
mitted by these observations and interpreta- collapse and laterally unconfined strain, in the
tions is that lateral, horizontal spreading was upper part of the glacier, and the strain path
distributed through a gravitationally collapsing, progressively changes to non-coaxial in a basal
vertically shortening package of partly consoli- zone characterized by high shear strains. I
dated sediments, rather than restricted to a visualize the penetrative, mesoscopic triaxial
narrow zone of progressive simple shear at its flattening in Piedras Blancas sediments as hav-
base. If both spreading and layering were ing occurred in the domain of coaxial deforma-
approximately horizontal, then it is probable tion, perhaps in the upper part of the larger
that collapse affected a thin apron of slope slide. The inherently low strength of water-rich,
sediments mantling the inner wall of a trench, incompletely consolidated modern slope sedi-
rather than sediments that had been accreted or ments deposited on surface slopes as great as 7°
otherwise deformed and rotated from their (see seismic profiles in Hamilton 1977) would
originally horizontal orientation. invite collapse. The main reason for drawing an
analogy between slides on the inner wall and
glaciers is to raise the possibility that signficant
A model for deformation at Piedras Blancas
and distinctive deformation can occur within a
I propose that a piedmont glacier or an ice moving mass as it spreads and collapses. The
cap is an appropriate analogue for deformation model may also apply to thin submarine slides
at Piedras Blancas. Ramberg (1977, 1980) has on continental margins in general, although
stressed the geometric, kinematic, and dynamic very little is known about the internal structure
similarities between ice caps and Caledonian of deposits in this setting. The overall style of
nappes in Scandinavia. Both can be visualized deformation described here may be part of a
as flowing outward in response to the gravita- spectrum developed as slides transform into
tional potential induced by the surface slope of progressively more fluid-like sediment gravity
the moving mass. Ice and rock in effect collapse flows (see Dott 1963 and Morgenstern 1967).
and spread laterally under their own weight.
Ramberg emphasizes the internal deformation,
or 'collapse strain', distributed throughout Conclusions
these bodies and suggests that lateral displace-
ments (forward movements) are due primarily An important conclusion of this paper is that
to vertical shortening balanced by horizontal Franciscan sediments were deformed while in
elongation rather than to sliding along a basal an unconsolidated to semi-lithified state. Simi-
detachment or glide surface. Fig. 14, from lar interpretations elsewhere in the Franciscan,
 D e f o r m a t i o n o f Franciscan s e d i m e n t s , California 455

and in many other terranes as well, rely heavily the sediment-water interface from all those
on the geometry of mesoscopic structures beneath, but it does not accommodate m a n y
thought to be compatible with 'soft-sediment mass-transport processes. The third distinction
slumping', but I feel that deformational implies that there is some minimum scale
mechanisms are a less equivocal criterion. The beyond which gravity assumes an important
axially symmetric extension recorded by a vari- role. Although gravitational forces are obvious-
ety of small-scale structures is an unusual type ly required to propel thin, surficial slide sheets
of strain requiring a laterally unconfined defor- downslope, Elliott (1976) and Hamilton (1977)
mational environment. Vertical shortening and suggest that they are also directly responsible
compensatory horizontal elongation are a possi- for displacements along imbricate, arcward dip-
ble consequence of gravitational collapse and ping thrusts within accretionary wedges. 'Tecto-
spreading at a shallow level on the trench slope. nic deformation' is a useful concept if it can be
Although I am applying these conclusions only identified with the large-scale setting of a suite
to the Franciscan in the study area, my general of structures. My proposed distinction between
approach could be useful in other parts of the substructure and superstructure is influenced by
complex. The suggestion that stratal disruption Dennis's (1972) treatment of slides and thrusts,
and fragmentation occurred in response to gra- H o e d e m a e k e r ' s (1973) definitions of tectonism
vitational spreading may be controversial, since and delapsion, and Woodcock's (1976) discus-
these features have also been ascribed to fault- sion of tectonic and slump structures, all of
ing both in the Franciscan (e.g. Cowan 1974; which embody the connotation that 'tectonic'
Bachman 1981) and in other ancient subduction deformation is confined on a large scale and can
complexes (e.g. Moore & Karig 1980). All be conceptually differentiated from laterally
available structural data must be precisely de- unconfined sliding, slumping, and spreading.
scribed and evaluated in order to test these
different hypotheses.
It is appropriate to ask whether the conven-
tional distinctions 'tectonic versus soft- ACKNOWLEDGMENTS:My interest in soft-sediment de-
sediment', 'tectonic versus sedimentary', and formation was stimulated by discussions in the field
with D. S. Wood, F. A. Gibbons, and M. T. Bran-
'tectonic versus gravity-driven' have any mean- don. It was catalysed by my participation on Leg 67 of
ing in a trench slope setting. The results of this the Deep Sea Drilling Project which afforded the
study argue for abandoning the first and allow- chance to study sediment cores from the Mid-
ing 'tectonic' to denote deformational environ- America Trench off Guatemala. I am indebted to
ment rather than the physical state of sedi- M. T. Brandon for his continued suggestions and
ments. The second may separate processes at criticism.

References
AYDIN, A. 1978. Small faults formed as deformation sand: a study of cataclastic flow. In: GRIGGS, D.
bands in sandstones. Paleogeophys. 116, 913-30. & HANDIN J. (eds). Rock Deformation. Mere.
BACHMAN,S. B. 1978. A Cretaceous and early Terti- geol. Soc. Am., 79, 133--91.
ary subduction complex, Mendocino Coast, CORBErr, K. D. 1973. Open-cast slump sheets and
northern California. In: HOWELL, D. G. & their relationship to sandstone beds in an upper
McDOUGALL, K. A. (eds). Mesozoic Cambrian flysch sequence, Tasmania. J. sedi-
Paleogeography of the Western United States, ment. Petrol. 43, 147-59.
419-30. Pacif. Sect. Soc. econ. Paleontol. Miner- COULBOURN, W. T. & MOBERLY, R. 1977. Structural
al. Tulsa. evolution of fore-arc basins off South America.
1981. The Coastal Belt of the Franciscan: Can. J. Earth Sci. 14, 102-16.
youngest phase of northern California subduc- COWAN, D. S. 1974. Deformation and metamorphism
tion (this volume). of the Franciscan subduction zone complex
BOOTH, J. S. 1979. Recent history of mass-wasting on northwest of Pacheco Pass, California. Bull.
the upper continental slope, northern Gulf of geol. Soc. Am. 85, 1623-34.
Mexico, as interpreted from the consolidation -- 1978. Origin of blueschist-bearing chaotic rocks
states of the sediment. In: DOYLE, L. J. t~¢ in the Franciscan Complex, San Simeon, Califor-
PILKEY, O. H. (eds). Geology of Continental nia. Bull. geol. Soc. Am. 89, 1415-23.
Slopes. Spec. Publ. Soc. econ. Paleontol. Miner- SILLING, R. M. 1978. A dynamic, scaled
al. Tulsa, 27, 153-64. model of accretion at trenches and its implica-
BORG, I., FRIEDMAN,M., HANDIN,J. & HIGGS, D. V. tions for the tectonic evolution of subduction
1960. Experimental deformation of St. Peter complexes. J. geophys. Res. 83, 5389-96.
456 D. S. Cowan
DAVIES, W. & CAVE, R. 1976. Folding and cleavage HOSSACK, J. 1979. The use of balanced cross-sections
determined during sedimentation. Sediment. in the calculation of orogenic contraction: a
Geol. 15, 89-133. review. J. geol. Soc. London, 136, 705-11.
DENNIS, J. G. 1972. Structural Geology. New York. Hs¢), K. J. 1968. Principles of m61anges and their
DOT]', R. H., JR. 1963. Dynamics of subaqueous bearing on the Franciscan-Knoxville paradox.
gravity depositional processes. Bull. Am. Assoc. Bull. geol. Soc. Am. 79, 1063-74.
Petrol. Geol. 47, 104-28. 1969. Preliminary report and geologic guide to
DUNN, D. E., LAFOUNTAIN, L. J. & JACKSON, R. E. Franciscan m61anges of the Morro Bay--San
1973. Porosity dependence and mechanism of Simeon area, California. Spec. Publ. Calif. Div.
brittle fracture in sandstones. J. geophys. Res. Mines Geol. 35.
78, 2403-17. 1974. M61anges and olistostromes. In: D o n ,
ELLIOt, D. 1976. The motion of thrust sheets. J. R. H., JR & SHAVER, R. H. (eds). Modern and
geophys. Res. 81,949-63. Ancient Geosynclinal Sedimentation. Spec. Publ.
FRIEDMAN, M., HUGMAN III, R. H. H., & HANDIN, J. Soc. econ. Paleontol. Mineral. Tulsa, 19, 321-
1981). Experimental folding of rocks under con- 33.
fining pressure, part VIII--forced folding of IBRAHIM, A. K., LATHAM, G. V. & LADD, J. 1979.
unconsolidated sand and of lubricated layers of Seismic refraction and reflection measurements
limestone and sandstone. Bull. geol. Soc. Am. in the Middle America trench offshore Guate-
91, 307-12. mala. J. geophys. Res. 84, 5643-9.
GRANT-MACKIE, J. A. & LOWRY, D. C. 1964. Upper KARIG, D. E., MOORE, G. F., CURRAY, J. R. &
Triassic rocks of Kiritehere, SW Auckland, New LAWRENCE, M. B. 19811. Morphology and shallow
Zealand. Part I: submarine slumping of Norian structure of the trench slope off Nias Island,
strata. Sedimentology, 3, 296-317. Sunda Arc. In: HAVES, D. E. (ed.). The Tectonic
GREGORY, M. R. 1969. Sedimentary features and and Geologic Evolution of Southeast Asian Seas
penecontemporaneous slumping in the Waite- and Islands. Geophys. Monogr. Am. geophys.
mata Group, Whangaparaoa Peninsula, North Union 23, 179-208.
Auckland, New Zealand. N.Z. J. Geol. Geo- - - & SHARMAN, G. F. III 1975. Subduction and
phys. 12, 248-82. accretion in trenches. Bull. geol. Soc. Am. 86,
GRIGGS, D. & HANDIN, J. 1960. Observations on 377-89.
fracture and a hypothesis of earthquakes. In: KLEIST, J. R. 1974. Deformation by soft-sediment
GRIGGS, D. & HANDIN, J. (eds). Rock Deforma- extension in the Coastal Belt, Franciscan Com-
tion. Mem. geol. Soc. Am. 79, 347-64. plex. Geology, 2, 501-4.
HAl, L, C. A., JR 1975. San Simeon-Hosgri fault KULM, L. D., SCHWELLER, W. J. & MASIAS, A. 1977.
system, coastal California: economic and en- A preliminary analysis of the subduction proces-
vironmental implications. Science, 190, 1291-4. ses along the Andean continental margin, 6 ° to
1976. Geologic map of the San Simeon-Piedras 45°S. In: TALWANI, M. & PITMAN, W. C. III
Blancas region. San Luis Obispo County, Cali- (eds). Island Arcs. Deep Sea Trenches, and
fornia. U.S. geol. Surv. Map, MF-784. Back-arc Basins. Am. geophys. Union, M. Ew-
HAMILTON, W. 1977. Subduction in the Indonesian ing Ser. l, 285-301.
region. In: TALWANI, M. & PITMAN, W. C. III LAMBERT, J. L. M. 1965. A reinterpretation of the
(eds). Island Arcs, Deep Sea Trenches, and breccias in the Meneage crush zone of the Lizard
Back-Arc Basins. Am. geophys. Union, Ewing boundary, south-west England. Q. J. geol. Soc.
Ser. 1, 15-31. London, 121, 339-57.
HANDIN, J., HAGER, R. V., JR, FRIEDMAN, M. & MALTMAN, A. J. 1977. Some microstructures of ex-
FEATHER, J. N. 1963. Experimental deformation perimentally deformed argillaceous sediments.
of sedimentary rocks under confining pressures: Tectonophysics, 39, 417-36.
pore pressure tests. Bull. Am. Assoc. Petrol. MOORE, G. F. & KARIG, D. E. 1976. Development of
Geol. 47, 717-55. sedimentary basins on the lower trench slope.
HELWIG, J. 1970. Slump folds and early structures, Geology, 4, 693-7.
northeastern Newfoundland Appalachians. J. - - & -- 1980. Structural geology of Nias
Geol. 78, 172-87. Island, Indonesia: implications for subduction
HOBBS, B. E., MEANS, W. D. & WILLIAMS, P. F. zone tectonics. Am. J. Sci. 280, 193-223.
1976. An Outline of Structural Geology. New MOORE, J. C., WATKINS, J. S. et al. 1979. Progressive
York. accretion in the Middle America trench, south-
HOEDEMAEKER,P. J. 1973. Olisthostromes and other ern Mexico. Nature, 281,638-42.
delapsional deposits, and their occurrence in the MORGENSTERN, N. R. 1967. Submarine slimping and
region of Moratalla (Prov. of Murcia, Spain). the initiation of turbidity currents. In: RICHARDS,
Scr. Geol. 19, 1-207. A. F. (ed). Marine Geotechnique. 189-220.
HOOKE, R. L. & HUDDLESTON, P. J. 1978. Origin of Urbana.
foliation in glaciers. J. Glaciol. London, 20, -- & TCHALENKO, J. S. 1967. Microstructural
285-99. observations on shear zones from slips in natural
HORNE, G. S. 1969. Early Ordovician chaotic deposits clays. Proc. Geotechnical Conf. Oslo, 1,147-52.
in the central volcanic belt of northeastern New- NARDIN, T. R., HEIN, F. J., GORSLINE, D. S. &
foundland. Bull. geol. Soc. Am. 80, 2451-64. EDWARDS, B. D. 1979. A review of mass move-
 D e f o r m a t i o n o f F r a n c i s c a n s e d i m e n t s , California 457

merit processes, sediment and acoustic character- gins, 249-60. Springer-Verlag, New York.
istics, and contrasts in slope and base-of-slope SKEMPTON, A. W. 1964. Long-term stability of clay
systems versus canyon-fan-basin floor systems. slopes. Geotechnique. London, 14, 77-101.
In: DOYLE, L. J. & PILKEY, O. H. (eds). Geology -- 1970. The consolidation of clays by gravitation-
of Continental Slopes. Spec. Publ. Soc. econ. al compaction. Q. J. geol. Soc. London, 125,
Paleontol. Mineral. Tulsa, 27, 61-73. 373-411.
PAGE, B. M. 1978. Franciscan melanges compared SMITH, R. B. 1977. Formation of folds, boudinage,
with olistostromes of Taiwan and Italy. Tecto- and mullions in non-Newtonian materials. Bull.
nophysics, 47, 223-46. geol. Soc. Am. 88, 312-30.
POTTER, P. E. & PETTIJOHN, F. J. 1977. Paleocurrents SMITH, G. W., HOWELL, D. G. & INGERSOLL, R. V.
and Basin Analysis, 2nd edn. Berlin. 1979. Late Cretaceous trench-slope basins of
RAMBERG, H. 1977. Some remarks on the mechanism central California. Geology, 7, 303-6.
of nappe movement. Geol. FOr. Stockh. FOrh. TCHALENKO, J. S. 1968. The evolution of kink-bands
99, 110-17. and the development of compression textures in
1978. Further on stress equations. Geol. FOr. sheared clays. Tectonophysics, 6, 159-74.
Stockh. FOrh. 100, 254. 1970. Similarities between shear zones of
1980. Diapirism and gravity collapse in the different magnitudes. Bull. geol. Soc. Am. 81,
Scandinavian Caledonides. J. geol. Soc. London, 1624-40.
137, 261-70. TYLER, J. H. 1975. Fracture and rotation of brittle
RAMSAY, J. G. 1967. Folding and Fracturing of Rocks. clasts in a ductile matrix. J. Geol. 83, 501-10.
New York. VON HUENE, R. 1972. Structure of the continental
& GRAHAM, R. H. 1970. Strain variation in margin and tectonism at the eastern Aleutian
shear belts. Can. J. Earth Sci. 7, 786-813. Trench. Bull. geol. Soc. Am. 83, 3613-26.
& WOOD, D. S. 1973. The geometric effects of YON HUENE, R. et al. 1981. A summary of Deep Sea
volume change during deformation processes. Drilling Project Leg 67 shipboard results from
Tectonophysics, 16, 263-77. the Mid-America Trench transect off Guatemala
RAST, N. 1956. The origin and significance of boudin- (this volume).
age. Geol. Mag. 93, 401-8. WEISS, L. E. 1972. The Minor Structures of Deformed
RIEKE, H. H. • CHILINGARIAN,G. V. 1974. Compac- Rocks. New York.
tion of Argillaceous Sediments. Amsterdam. WHITE, R. S. & Ross, D. A. 1979. Tectonics of the
RUPKE, N. A. 1976. Large-scale slumping in a flysch western Gulf of Oman. J. geophys. Res. 84,
basin, southwestern Pyrenees. J. geol. Soc. Lon- 3479-89.
don, 132, 121-30. WHITrEN, E. H. T. 1966. Structural Geology of
SANDERSON, D. J. 1976. The superposition of com- Folded Rocks. Chicago.
paction and plane strain. Tectonophysics, 30, WILLIAMS,P. F., COLLINS,A. R. & WILTSHIRE,R. G.
35-54. 1969. Cleavage and penecontemporaneous de-
SCHUSTER, D. C. 1980. The nature and origin of the formation structures in sedimentary rocks. J.
Late Precambrian Gwna melange North Wales, geol. 77, 415-25.
United Kingdom. Thesis, PhD, Univ. Illinois WOODCOCK, N. H. 1976. Structural style in slump
(unpubl.). sheets: Ludlow Series, Powys, Wales. J. geol.
SEELY, D. R., VAIL, P. R. & WALTON,G. G. 1974. Soc. London, 132, 399-415.
Trench slope model. In: BURK, C. A. & DRAKE, 1979. Sizes of submarine slides and their signifi-
C. L. (eds). The Geology of Continental Mar- cance. J. Struct. Geol. 1, 137-42.

DARREL S. COWAN,Department of Geological Sciences, University of Washington,

Seattle, Washington 98195, U.S.A.
 Initiation and evolution of the Great Valley forearc basin of
northern and central California, U.S.A.

Raymond V. Ingersoll
SUMMARY: The late Mesozoic Great Valley forearc basin of northern and central
California evolved from a residual forearc basin formed on top of young oceanic crust to a
composite forearc basin resting on both oceanic and continental crust. Depositional
environments preserved in outcrop along the west side of the Sacramento Valley evolved
from deep oceanic floor and slope in the Late Jurassic to basin plain in the Early
Cretaceous to submarine fan complexes in the Late Cretaceous to slope and shelf in the
Palaeogene. The forearc basin widened and enlarged, and was supplied with voluminous
sediment primarily derived from the coeval magmatic arc to the east. However, significant
quantities of ophiolitic detritus (chert, and sedimentary, mafic volcanic and mafic
metavolcanic lithic fragments) occur in the lower Great Valley strata, thus indicating
erosion of 'tectonic highlands' formed during arc-arc collision immediately preceeding
initiation of the late Mesozoic subduction regime.
The Great Valley forearc owes its geometry and history to the peculiarities of
subduction initiation and to the shapes of the continental margin and magmatic arcs extant
before initiation of the Great Valley subduction phase. The residual forearc developed on
oceanic crust that had formed previously by spreading behind an east-facing intraoceanic
arc. After this east-facing arc collided with the west-facing continental-margin arc, a new
trench formed west of the suture belt. The shape of the new continental margin was
irregular, so that a wide residual forearc basin formed in the Great Valley area, whereas
no residual forearc basin formed to the north (Klamath area) or possibly, to the south
(southern California borderland area that has been dislocated by Neogene strike-slip
motions). Ttie northern coastal promontory (Klamath area) provided much of the
ophiolitic detritus within the lower part of the Great Valley strata. Sediment derived from
the Klamaths was transported southward directly into the forearc basin, as well as
westwards and then southwards into the trench west of the forearc basin. The Great
Valley forearc basin is preserved and displayed so well today due to its mode of initiation
(residual forearc formed in a previously backarc area) and its mode of termination
(northward migration of a triple junction that converted a convergent margin into a
transform margin).

The Great Valley forearc basin of northern and died petrologically, sedimentologically and
central California (Fig. 1) is probably the most stratigraphically (e.g. Dickinson & Rich 1972;
thoroughly studied and best-understood ancient Ingersoll 1978b, c, 1979a; Mansfield 1979;
basin of its kind in the world (Ingersoll 1978a, Ojakangas 1968). The batholithic roots of the
1979a). All three major components of the magmatic arc have been dated, and their petrol-
arc-trench system (Franciscan subduction com- ogy and chemistry have been studied by numer-
plex, Great Valley forearc basin and Sierra ous workers (e.g. B a t e m a n & Dodge 1970;
Nevada magmatic arc) (Dickinson 1970; Dick- Evernden & Kistler 1970; Presnall & B a t e m a n
inson & Seely 1979) are exposed on land essen- 1973). West-to-east increase of K-feldspar con-
tially in their original relationships. The sub- tent and eastward migration of the volcanic
duction complex contains the most intricate front with time are consistent with Dickinson's
relations (e.g. Aalto 1981, Bachman 1978, (1973, 1975) model of arc evolution.
1981; Cowan 1974) due to intense deformation The three components of the late Mesozoic
and metamorphism. Nonetheless, coherent pat- arc-trench system evolved concurrently from
terns of a westward decrease in age of strata, their beginnings in the Late Jurassic (Kimmerid-
degree of deformation and grade of meta- gian-Tithonian) through the Cretaceous, for
morphism are apparent (Dickinson & Seely which their relations are most clear. Subduction
1979; Dickinson et al. 1979; Ingersoll et al. that created this linked system has been termin-
1977). Sediments deposited within the forearc ated in the Neogene by the northward migra-
basin (Great Valley G r o u p - - n e w nomenclature tion of the Cape Mendocino triple junction
that replaces Great Valley Sequence of Bailey (Atwater 1970; Crowell 1979), following a com-
et al. 1964--and related strata) have been stu- plex Palaeogene history (Dickinson et al. 1979).

459
460 R. V. Ingersoll
IlOeW

\\

20 40 60 80 ~ 0 w,m
l[""I
/SAN INDAEA! F I O L ! I V I T I I d ~1(
[----]CIEII0701C IIDIMIrMT$ &ND VOLCANIC
ll~ F~IANCtlCAmm AND m l I . A T I O mOCKS
Ill,AT VALLlV GmOUP
ILAIIA'r~, -IIt~llA m l f V A O A - I I A I . PNIA
,@sloul Auo ulrAlJOm~,l,c vlRn,

S__ -
,llO

FIG. 1. Location map of northern and central California, showing principal components of the late
Mesozoic arc-trench system and geographic locations. The Sierra Nevada igneous and metamorphic
terranes represent the roots of the magmatic arc, and the Franciscan complex represents the highly
deformed subduction complex formed landward of the trench. The Great Valley Group is primarily
Upper Cretaceous in the San Joaquin Valley and is Upper Jurassic to Upper Cretaceous in the
Sacramento Valley. Small outcrops of uppermost Cretaceous that nonconformably overlies Sierra
Nevada basement occur along the east side of the Sacramento Valley, but are too small to show at
this scale. The Great Valley Group rests nonconformably on Klamath basement near Redding. Coast
Range ophiolite underlies the Great Valley Group at other surface exposures (after Ingersoll 1978a).

Thus, sedimentation and deformation in the to the west) intraoceanic arc approached the
subduction complex, sedimentation in the fore- western margin of North America where a
arc basin, and intrusion, extrusion, metamorph- west-facing (subduction down to the east) con-
ism and erosion within the magmatic arc all tinental-margin arc resided (Figs 2 & 3). Bath-
occurred concurrently and resulted in a closely olithic rocks of this continental-margin arc
linked triad of major tectonostratigraphic units occur today in the eastern Sierra Nevada in
(Dickinson 1977). California and Nevada. Intervening oceanic
The purpose of this paper is to outline in lithosphere was consumed beneath both of
broad terms the history of this elegant forearc these arc-trench systems. At the same time that
basin. The reader is referred to the numerous the east-facing intraoceanic arc consumed
references for details that are touched on only oceanic lithosphere on its east side, backarc
briefly here. spreading produced new oceanic lithosphere on
its west side. U p o n collision of the two arc-
trench systems in the Kimmeridgian (Nevadan
orogeny), a new trench formed west of the
Initiation of forearc basin
suture belt and the Franciscan-Great Valley-
Schweickert & Cowan (1975) have outlined the Sierra Nevada system initiated (Fig. 3b). In the
most likely scenario for the initiation of the process of forming a new west-facing continen-
Franciscan-Great Valley-Sierra Nevada arc- tal-margin arc superimposed on a former east-
trench system. During pre-Kimmeridgian facing intraoceanic arc, oceanic crust that had
Jurassic time, an east-facing (subduction down formed previously in a backarc setting became
 Evolution o f the Great Valley forearc basin o f California 461

¢ ,

, A

MIDjuRAS JURASSIC\
~\ LATE "¢ \\\ LATE ~ A
CRETACEOLI~A \\
N \ ( / N , I
" 0 IOO 200 ,.-t t Io \.,"

Fl6.2. Schematic palaeogeographic-palaeotectonic maps for the Great Valley forearc. The following
symbols indicate various tectonic and geographic features: dashed lines--present California state
outline and shoreline; heavy solid lines--Neogene San Andreas fault zone; dashed lines with thrust
symbols--trenches; double solid lines with diverging arrows--spreading centre; mountains--
magmatic arcs; light solid lines--palaeoshorelines; parallel light solid lines--oceanic crust; stitched
line--suture belt; stippled pattern--subduction complex; single-stemmed arrows--sediment disper-
sal directions. See text for discussion. See Ingersoll (1978a) for additional maps.

the basement (Coast Range ophiolite of Bailey able evidence (Howell et al. 1977) is consistent
et al. 1970) upon which the Great Valley forearc with forearc deposition primarily on continental
formed (Fig. 3b). crust and accreted material (constructional
The continental margin created by the com- rather than residual forearc basin) along the
bination of backarc spreading, arc-arc collision late Mesozoic continental margin south of the
and initiation of a new subduction zone was Great Valley area. Thus, the Great Valley area
irregular in shape. Trenches tend to have owes its unique characteristics (continuous re-
smooth, curving shapes in plan view, so that the cord of sedimentation from Late Jurassic
irregular continental margin formed indenta- through the entire Cretaceous) to its specific
tions (oceanic crust trapped in the forearc area) plate-tectonic history as well as inherited fea-
and promontories (continental crust in the fore- tures. The pre-subduction shape of the con-
arc area) (Fig. 2). In the Great Valley area, a tinental margin controlled the location of the
wide belt of oceanic crust (ophiolite) was trap- deep residual forearc basin (e.g. Dickinson &
ped behind the trench, so that a residual forearc Seely 1979).
basin was created (terminology of Dickinson & Depositional environments and sedimentary
Seely 1979). To the north (Klamath area), a history of the earliest stages of the Great Valley
promontory of accreted arc terranes existed forearc are obscure. The basal part of the Great
directly behind the new trench. Therefore, no Valley Group accumulated on the Coast Range
forearc basin formed and sediment derived ophiolite (Bailey et al. 1970) in a backarc
from this area was transported either directly setting. The local occurrence of andesitic volca-
into the trench and was accreted, or was trans- nics directly above the ophiolite (Evarts 1977)
ported southwards and northwards into adjoin- supports this interpretation. It is difficult to
ing forearc basins and trenches. There may also know what proportion of the lower part of the
have been a promontory at the south end of the Great Valley Group accumulated before initia-
Great Valley forearc basin, but Neogene trans- tion of the Great Valley forearc basin (Ingersol!
form-related deformation has been so extensive 1978a). Basal deposits include radiolarian
that reconstruction of the late Mesozoic con- cherts likely to have formed in the backarc area
figuration is difficult. However, the best avail- before the influx of voluminous suture- and
462 R. V. Ingersoll

arc-derived detritus (Fig. 3a) (Pessagno 1973).

Most of the contacts between ophiolite and
basal Great Valley Group are tectonic (Bailey
et al. 1970; Ingersoll 1978a), but where the
contact is depositional, deep oceanic (including
basin-plain) deposits seem to predominate.
NEOGENE f Soon after initiation of the Great Valley phase
of subduction, great thicknesses of mud with
incised conglomerate-filled channels began
accumulating. These deposits represent slope
deposition within the newly created residual
forearc basin. A wide slope existed from
shorelines along the east and north (Klamath
PALAEOGENE W" 2 e area) sides of the basin down to the trench to
TSB \i/ the west (Ingersoll 1978a). Direct access to the
trench thus was provided for suture- and arc-
derived detritus.
The Stony Creek Formation along the west
side of the Sacramento Valley includes the
Tithonian (Upper Jurassic) and Neocomian
rsa I (Lower Cretaceous) parts of the Great Valley
Group (Ingersoll et al. 1977). Locally, Kimmer-
v 1 ~/ t
.- i f/Ill// idgian (Upper Jurassic) breccias and ophiolitic
detritus occur below the Tithonian. Kimmerid-
gian conglomerates primarily consist of mafic
"~b Kim. collision metavolcanics probably derived directly from
EARLY CRETACEOUS (Nevadan)
TSB ~ \1/~ -
the underlying ophiolite (Bertucci 1980). The
majority of the Stony Creek Formation includes
/j /,-- I conglomerate and sandstone clasts consisting of
chert, metachert, argillite and other sediments,
and varieties of volcanic, plutonic and meta-
LATE JURASSIC b morphic detritus. The indicated provenance is
one of combined suture-derived detritus (up-
\1/ \1/ lifted and accreted ophiolites, deep-marine
v
sediments and arc terranes), and arc-derived
detritus (felsic, intermediate and mafic volca-
nics) (Bertucci 1980; Ingersoll 1979a, b). Poly-
crystalline quartz (including chert and
MID JURASSIC ~, o
metachert) is more prevalent to the north, and
volcanic and metavolcanic detritus increases to
FIG. 3. Schematic cross-sections of north- the south within the Sacramento Valley. The
ern California from formation of Coast ratio of polycrystalline quartz to total quartz
Range ophiolite behind an east-facing in- decreases up-section within the lower part of
traoceanic arc (a) to termination of Great the Great Valley Group. These relations sug-
Valley forearc by conversion to transform gest that the Klamath area was the source of
margin (f). Vertical arrows indicate loca-
tion of present outcrops of Great Valley most of the suture-derived material, and that
Group along west side of Sacramento Val- following erosion of the sutured terranes ('tecto-
ley. These are used as fixed reference point nic highlands' formed by the previous arc-arc
for each cross-section, Stippled pattern in- collision), detritus eroded from the newly
dicates upper-slope discontinuity. TSB is formed continental-margin arc (primarily rep-
trench-slope break. Foothill suture belt resented by the Sierra Nevada area today)
shown by diagonal and wiggly lines. Plus began accumulating within the forearc basin.
and minus symbols indicate motion on San Thus, there is a transition upward within the
Andreas fault zone (f). See text for discus- Great Valley Group from locally derived
sion (after Dickinson & Seely 1979, Ing-
ersoll 1979a, and Schweickert & Cowan ophiolitic detritus to distantly derived suture-
1975). belt detritus to distantly derived magmatic-arc
detritus.
 Evolution o f the Great Valley forearc basin o f California 463

of the forearc basin concurrently with and as a

result of the combined effects of_westward and
upward growth of the Franciscan subduction
complex, and the eastward migration of the
Sierran magmatic arc (Figs 3b-e & 4) (Dickin-
son & Seely 1979; Ingersoll 1978a, 1979a; Ing-
ersoll et al. 1977). Thus, the forearc basin (and
the arc-trench gap) widened through time as the
trench migrated westwards and the magmatic
front migrated eastwards. Intense deformation
and low-temperature, high-pressure metamorph-
ism characterized the subduction complex (e.g.
Bailey et al. 1964; Ernst 1970), whereas extru-
sion, intrusion and high-temperature, low-
pressure metamorphism characterized the
magmatic arc (e.g. Dickinson 1970; Dickinson
& Rich 1972; Evernden & Kistler 1970). Be-
tween these two terranes, the relatively undis-
turbed Great Valley Group accumulated in an
enlarging basin.

/ o__
Some form of mid-Cretaceous deformation
likely occurred within the forearc basin as evi-
denced by subtle warping of Lower Cretaceous
strata, local serpentinite protrusions on the east
side of the subduction complex, local mid-
Cretaceous disconformities, and thrust or
"L) reverse faulting within the basin (Brown &
Rich 1967; Dickinson & Seely 1979; Ingersoll et
SPAN:OF FOREARC BASIN al. 1977; Jones & Bailey 1973; Jones & Irwin
1971; Jones et al. 1969; Peterson 1967). Origi-
nally low-angle faults near the base of the Great
FIG. 4. Schematic map illustrating secular Valley Group in the northern Sacramento Val-
increase in width of Great Valley forearc ley connect with the Coast Range fault to the
basin. Approximate positions of migratory north-west and may connect in the subsurface
boundaries of the forearc basin are shown with presumed reverse faults underneath the
for times early in Early Cretaceous (125
Ma), in mid-Cretaceous (100 Ma) and late centre of the modern Great Valley (Dickinson
in the Late Cretaceous (75 Ma). Positions & Seely 1979; Ingersoll et al. 1977). This intra-
of western (outer) boundary at migratory basinal faulting may reflect increased crustal
trench-slope break marking inner limit of telescoping along the upper-slope discontinuity
active subduction are inferred from eas- (sensu Karig & Sharman 1975) during the mid-
ternmost extent of successively younger Cretaceous.
strata within the Franciscan complex. Posi- With the exception of the above-mentioned
tions of eastern (inner) boundary at migra- deformation, the forearc basin evolved syste-
tory shoreline marking easternmost extent matically as the arc-trench gap widened, and
of successively younger strata present in
subsurface of modern Great Valley. Posi- sedimentation continued at a rapid pace (aver-
tions of the magmatic front marking west- age sedimentation rate for the entire basin
ern limit of igneous belt are inferred from during the Late Cretaceous was 280 m/my)
westernmost limits of radiometric dates for (Ingersoll 1979a). By the end of the Cre-
Sierra Nevada plutons (Evernden & Kistler taceous, the forearc basin was filled to near
1970) (after Ingersoll et al. 1977). sea-level in many places (Fig. 3e) and Sierran
arc-derived sediment again was able to traverse
the arc-trench gap and be incorporated into the
subduction complex (Ingersoll 1978a). The
basin had evolved into a composite forearc
formed on top of accreted material, and both
Evolution of forearc basin
oceanic and continental crust (Dickinson &
The Cretaceous development of the Great Val- Seely 1979).
ley forearc consisted of enlargement and filling Increased rates of westward motion of North
464 R . V. I n g e r s o l l

America relative to the Pacific basin and possi- site of primarily non-marine deposition with
ble changes in the direction of spreading during local marine conditions owing to sea-level fluc-
the Palaeogene led to the rapid eastward migra- tuations. Deep basins formed in the southern
tion of the magmatic arc and possible shearing part of the San Joaquin Valley due to wrench
within the subduction complex (Coney 1976, tectonics associated with Neogene development
1978; Dickinson 1979; Dickinson & Snyder of the San Andreas fault system (Blake et al.
1978; Dickinson et al. 1979; Graham 1979; 1978; Crowell 1979; Harding 1976).
Howell et al. 1977; Nilsen & Clarke 1975). Petrographic data provide further insight re-
Throughout the Palaeogene, the Great Valley garding forearc evolution. The Great Valley
area remained a forearc basin, but with de- Group consists of suture-derived and immature
creased influence of the magmatic arc which arc-derived sedimentary and volcanic material
was far to the east (Fig. 3e) (Dickinson et al. near the base (Stony Creek Formation). The
1979). A wide shelf extended from the eroded overlying Lodoga Formation (Aptian-Albian)
roots of the Cretaceous arc in eastern California contains increased proportions of monocrystal-
and Nevada to the trench-slope break along the line quartz, potassium feldspar, mica and meta-
coast. morphic lithic fragments, suggesting that over-
Northward migration of the Mendocino triple lying volcanics, sediments and sutured material
junction during the Neogene has converted the had been eroded through to underlying plutonic
subducting margin into a transform margin terranes by the mid-Cretaceous (Dickinson &
(Atwater 1970; Crowell 1979), and has dis- Rich 1972; Ingersoll 1979b). Intrabasinal tec-
rupted palaeotectonic relations, especially tonism during the mid-Cretaceous (discussed
along the south-western margin of the former above) may have resulted in reworking of some
forearc basin (Figs 1 & 3f). The subduction previously deposited sediments, thus increasing
complex has risen above sea-level with the the quartz content of the Lodoga Formation
cessation of subduction, probably due, in part, (Dickinson & Rich 1972; Ingersoll et al. 1977).
to isostatic adjustment. In the process, the west During the Late Cretaceous, alternating epi-
side of the forearc basin has been eroded and sodes of intrusion-extrusion and erosion in the
the Great Valley Group has been exposed (Fig. magmatic arc resulted in interbedded volcanic-
3f). lithic-rich and arkosic (plutonic-derived) sand-
Depositional environments along the west stones. Sandstones in the San Joaquin Valley
side of the Sacramento Valley evolved from contain higher proportions of mica and meta-
deep oceanic before initiation of Great Valley morphic lithic fragments and more felsic volca-
subduction to basin plain and slope in the latest nics than do those of the Sacramento Valley
Jurassic to basin plain and outer fan in the Early (Ingersoll 1978c). These characteristics can be
Cretaceous to complex fan and slope in the Late explained by the fact that the magmatic arc
Cretaceous to slope, shelf and non-marine in formed on top of true continental crust (includ-
the Palaeogene (Fig. 3) (Ingersoll 1978a). ing older metamorphic terranes) to the south
These changes in depositional environment and crust with more oceanic affinities to the
were responses to the tectonic events described north.
above as well as the sequential filling of the By the end of the Cretaceous, arkosic sand-
basin, although eustatic sea-level rise in the stones (monocrystalline quartz and alkali feld-
Cretaceous and fall in the Palaeogene (e.g. spars with significant quantities of mica)
Hancock & Kauffman 1979) had modifying accumulated throughout the forearc region
effects (Ingersoll 1980). Submarine sediment- (Ingersoll 1978c). These deposits represent ero-
dispersal systems (shelf-slope-channels-fans) sion of the batholithic roots of the magmatic arc
enlarged concurrently with basin growth. Thus, (Dickinson & Rich 1972). Similar arkosic sand-
individual submarine fan deposits tend to be stones were deposited throughout the
thinner and less areally extensive in the Lower Palaeogene as the active magmatic arc re-
Cretaceous than in the Upper Cretaceous. By mained far to the east of the Great Valley area
the Late Cretaceous and Palaeogene, submarine (Dickinson et al. 1979) (Fig. 3e). The following
canyons had incised the wide shelves so that upward petrological trends within the Great Val-
gravity flows were localized and large fan sys- ley Group are the result of the arc evolution
tems developed (Ingersoll 1978a, 1979a). The described above (Ingersoll 1978c, 1979b): in-
remnant deep basin was filled during this time creasing monocrystalline quartz, decreasing
by the westward progradation of slope-shelf- polycrystalline quartz, increasing total feldspar,
shoreline (deltaic) sequences (Dickinson & increasing potassium feldspar, decreasing total
Seely 1979; Dickinson et al. 1979). During the lithics, increasing mica and decreasing
Neogene, the former forearc basin has been the sedimentary lithics.
 E v o l u t i o n o f the Great Valley f o r e a r c basin o f California 465

Conclusions the Great Valley forearc; therefore, they may

not be applicable to all other cases. Some of the
The tectonic, sedimentological and petrological variables that may require changes in these
history outlined above suggests some general models when applied to other areas include the
principles regarding the evolution of arc-trench following: m o d e of initiation of subduction,
systems. Arc-trench gaps tend to widen with geometry and composition of pre-existing con-
time by the prograde growth of the subduction tinental margin, location along that margin,
complex and the retrograde migration of the characteristics of adjoining forearcs, rate and
magmatic front (Dickinson 1973). Forearc basins orientation of subduction, intensity of volcan-
may initiate as trapped segments of oceanic ism and plutonism, and climate. The mechan-
crust behind newly formed trenches, and grow ism by which subduction is terminated (triple-
to be composite wide basins overlapping the junction migration, arc-arc collision, continen-
former continental margins as well as the grow- tal collision, etc.) determines the final charac-
ing subduction complex (Dickinson & Seely teristics of the basin. Each forearc has experi-
1979). Depositional environments tend to enced a unique history and must be u n d e r s t o o d
evolve from d e e p oceanic to base-of-slope to on its own terms. Nonetheless, knowledge of
slope and shelf to non-marine (Ingersoll 1978a). the detailed history of the Great Valley forearc
Low in the stratigraphic fill of the forearc, there provides insights to the possible history of
may be ophiolitic and suture-derived others.
sediments; these are commonly overlain by
volcaniclastic and plutonic-metamorphic detri-
ACKNOWLEDGMENTS: I thank W. R. Dickinson for
tus. Arkosic potassium-rich detritus predomin-
numerous dicussions regarding Great Valley geology
ates at the top of the section (Dickinson & Rich throughout my involvement in this study. Dickinson,
1972; Ingersoll 1978c). S. A. Graham and J. K. Leggett provided helpful
These models must be applied to other fore- reviews of the manuscript. The Research Allocations
arc basins with caution. The models owe many Committee of the University of New Mexico pro-
of their features to information obtained from vided financial support.

References
AALTO, K. R. 1981. The Franciscan Complex of Sequence conglomerate, northwestern Sacra-
northernmost California: sedimentation and tec- mento Valley, California. Thesis, MS, Univ.
tonics (this volume). Calif., Davis, 143 pp.
ATWATER,T. 1970. Implications of plate tectonics for BLAKE, M. C., JR, CAMPBELL,R. H., DIBBLEE,T. W.,
the Cenozoic tectonic evolution of western North JR, HOWELL, D. G., NILSEN, T. H., NORMARK,
America. Bull. geol. Soc. Am. 81, 3513-35. W. R., VEDDER, J. G. • SILVER, E. A. 1978.
BACHMAN,S. B. 1978. A Cretaceous and early Terti- Neogene basin formation in relation to plate-
ary subduction complex, Mendocino coast, tectonic evolution of the San Andreas fault sys-
northern California. In: HOWELL, D. G. & tem, California. Bull. Am. Assoc. Petrol. Geol.
McDOUGALL, K. (eds). Mesozoic Paleogeogra- 62, 344-72.
phy of the Western United States. Pacific Sec., BROWN, R. D., JR & RICH, E. I. 1967. Implications of
Soc. econ. Paleontol. Mineral. Pacif. Coast two Cretaceous mass transport deposits, Sac-
Paleogeogr. Syrup. 2, 419-30. ramento Valley, California: comment on a paper
--1981. The coastal belt of the Franciscan: by Gary L. Peterson. J. sediment. Petrol. 37,
youngest phase of northern California subduc- 240-8.
tion (this volume). CONEY, P. J. 1976. Plate tectonics and the Laramide
BAILEY, E. H., IRWIN, W. P. & JONES, D. L. 1964. orogeny. Spec. Publ. New Mex. geol. Soc. 6,
Franciscan and related rocks, and their signifi- 5-10.
cance in the geology of western California. Bull. 1978. Mesozoic-Cenozoic Cordilleran plate tec-
Calif. St. Min. Bur. 183, 177 pp. tonics. Mere. geol. Soc. Am. 152, 33-50.
--, BLAKE,M. C., JR & JONES, D. L. 1970. On-land COWAN, D. S. 1974. Deformation and metamorphism
Mesozoic oceanic crust in California Coast of the Franciscan subduction zone complex
Ranges. Prof. Pap. U.S. geol. Survey. 700-C, northwest of Pacheco Pass, California. Bull.
70C-81C. geol. Soc. Am. 85, 1623-34.
BATEMAN,P. C. & DODGE, F. C. W. 1970. Variations CROWELL, J. C. 1979. The San Andreas fault system
of major chemical constituents across the central through time. J. geol. Soc. London, 136, 293-
Sierra Nevada batholith. Bull. geol. Soc. Am. 81, 302.
409-20. DICKINSON, W. R. 1970. Relations of andesites, gra-
BERTUCCI, P. F. 1980. Petrology and provenance of nites, and derivative sandstones to arc-trench
Upper Jurassic-Lower Cretaceous Great Valley tectonics. Rev. Geophys. Space Phys. 8, 813-60.
 466 R. V. Ingersoll
- 1973. Widths of modern arc-trench gaps pro-
- Pacific Sec., Soc. econ. Paleontol. Mineral.
portional to past duration of igneous activity in Pacif. Coast Paleogeogr. Field Guide, 2, 109 pp.
associated magmatic arcs. J. geophys. Res. 78, INGERSOLL, R. V., 1978a. Paleogeography and
3376-89. paleotectonics of the late Mesozoic forearc basin
- 1975. Potash-depth (K-h) relations in continen-
-
of northern and central California. In: HOWELL,
tal margin and intra-oceanic magmatic arcs. D. G. & McDOUGALL, K. (eds). Mesozoic
Geology, 3, 53-6. Paleogeography of the Western United States.
- 1977. Tectono-stratigraphic evolution of sub-
- Pacific. Sec., Soc. econ. Paleontol. Mineral.
duction-controlled sedimentary assemblages. In: Pacif. Coast Paleogeogr. Symp. 2, 471-82.
TALWANI, M. & PITMAN, W. C. III (eds). Island 1978b. Submarine fan facies of the Upper Cre-
Arcs, Deep Sea Trenches and Back-arc Basins. taceous Great Valley Sequence, northern and
Am. geophys. Union, M. Ewing Ser. 1, 33-40. central California. Sediment. Geol. 21,203-30.
1979. Cenozoic plate tectonic setting of the
- -
1978c. Petrofacies and petrologic evolution of
Cordilleran region in the United States. In: the Late Cretaceous fore-arc basin, northern and
ARMENTROUT, J. M., COLE, M. R. & TERBEST, H., central California. J. Geol. 86, 335-53.
JR (eds). Cenozoic Paleogeography of the West- 1979a. Evolution of the Late Cretaceous fore-
ern United States. Pacific Sec., Soc. econ. arc basin, northern and central California. Bull.
Paleontol. Mineral. Pacif. Coast Paleogeogr. geol. Soc. Am. 90, pt. I, 813-26.
Syrup. 3, 1-13. - - 1979b. Petrofacies and provenance of the lower
-- , INGERSOLL, R. V. & GRAHAM, S. A. 1979. part of the Great Valley Sequence, Sacramento
Paleogene sediment dispersal and paleotectonics Valley, California. Abstr. Progr. geol. Soc. Am.
in northern California. Bull. geol. Soc. Am. 90, I1, 85-6.
pt. I, 897-8, pt. If, 1458-528. - - 1980. Eustatic, tectonic and sedimentary con-
& RICH, E. I. 1972. Petrologic intervals and
- -
trois of Cretaceous transgressive-regressive sequ-
petrofacies in the Great Valley Sequence, Sac- ences, western, central and eastern USA. Abstr.
ramento Valley, California. Bull. geol. Soc. Am. Progr. geol. Soc. Am. 12, 453.
83, 3007-24. -- , RICH, E. I. & DICKINSON, W. R. 1977. Great
& SEELY, D. R. 1979. Structure and stratigraphy of
- -
Valley Sequence, Sacramento Valley. Cord.
forearc regions. Bull. Am. Assoc. Petrol. Geol. 63, Sect. , geol. Soc. Am. Fieldtrip Guidebook,72 pp.
2-31. JONES, D. L. & BAILFY, E. H. 1973. Preliminary
-- • SNYDER, W. S. 1978. Plate tectonics of the biostratigraphic map, Colyear Springs quad-
Laramide orogeny. Mere. geol. Soc. Am. 151, rangle, California. U.S. geol. Surv. Map, MF-
355-66. 517.
ERNST, W. G. 197{). Tectonic contact between the - - , BAILEY, E. H. & IMI,AY, R. W. 1969. Struc-
Franciscan melange and the Great Valley sequ- tural and stratigraphic significance of the Buchia
ence, crustal expression of a late Mesozoic Be- zones in the Colyear Springs-Paskenta area,
nioff zone. J. geophys. Res. 75, 886-901. California. Prof. Pap. U.S. geol, Surv. 647A, 24
EVARTS, R. C. 1977. The geology and petrology of the PP.
Del Puerto ophiolite, Diablo Range, central - - & IRWIN, W. P. 1971. Structural implications of
California Coast Ranges. Oregon Bull. Dep. an offset Early Cretaceous shoreline in northern
Geol. Miner. Ind. 95, 121-39. California, Bull. geol. Soc. Am. 82, 815-22.
EVERNDEN, J. F. & KISTLER, R, W. 1970. Chronology KARIG, D. E. & SHARMAN, G. F. 1975. Subduction
of emplacement of Mesozoic batholithic com- and accretion in trenches. Bull. geol. Soc. Am.
plexes in California and western Nevada. Prof. 86, 377-89.
Pap. U.S. geol. Surv. 623, 42 pp. MANSFIELD, C. F. 1979. Upper Mesozoic subsea fan
GRAHAM, S. A. 1979. Tertiary paleotectonics and deposits in the southern Diablo Range, Califor-
paleogeography of the Salinian block. In: nia: record of the Sierra Nevada magmatic arc.
ARMENTROUT, J. M., COLE, M. R. & TERBEST, H. Bull. geol. Soc. Am. 90, pt I, 1025-46.
JR (eds). Cenozoic Paleogeography of the West- NILSEN, T. H. & CLARKE, S. H., JR. 1975. Sedimenta-
ern United States. Pacif. Sec., Soc. Econ. Paleon- tion and tectonics in the early Tertiary continen-
tol. Mineral. Pacif. Coast Paleogeogr. Symp. 3, tal borderland of central California. Prof. Pap
45-51. U.S. geol. Surv. 925, 64 pp.
HANCOCK, J. M. & KAUFFMAN, E. G. 1979. The great OJAKANGAS, R. W. 1968. Cretaceous sedimentation,
transgressions of the Late Cretaceous. J. geol. Sacramento Valley, California. Bull. geol. Soc.
Soc. London, 136, 175-86. Am. 79, 973-1008.
HARDIN(;, T. P. 1976. Tectonic significance and hyd- PESSAGNO, E. A., JR 1973. Age and geologic signifi-
rocarbon trapping consequences of sequential cance of radiolarian cherts in the California
folding synchronous with San Andreas faulting, Coast Ranges. Geology, 1, 153-6.
San Joaquin Valley, California. Bull. Am. PETERSON, G. L. 1967. Implications of two Cre-
Assoc. Petrol. Geol. 60, 356-78. taceous mass transport deposits, Sacramento
HOWELL, D. G., VEDDER, J. G. & McDOUGALL, K. Valley, California: reply to comment by Brown
(eds) 1977. Cretaceous geology of the California and Rich. J. sediment. Petrol. 37, 248-57.
Coast Ranges, west of the San Andreas fault. PRESNALL, D. S. & BATEMAN, P. C. 1973. Fusion
 Evolution o f the Great Valley forearc basin o f California 467

relations in the system NaAISi308-CaA12Si208- SCHWEICKERT, R. A. & COWAN, D. S. 1975. Early

KA1Si30s-Si02-H20 and generation of granitic Mesozoic tectonic evolution of the western
magmas in the Sierra Nevada batholith. Bull. Sierra Nevada, California. Bull. geol. Soc. Am.
geol. Soc. Am. 84, 3181-202. 86, 1329-36.

RAYMONDV. INGERSOLL,Department of Geology, University of New Mexico, Albuquerque,

New Mexico 87131, U.S.A.
 Cretaceous-Palaeogene Flysch Zone of the East Alps and
Carpathians: identification and plate-tectonic significance of
'dormant' and 'active' deep-sea trenches in the
Alpine-Carpathian Arc

Reinhard Hesse
SUMMARY: The Flysch Zone of the East Alps is interpreted as a Cretaceous-Palaeogene
deep-sea trench fill that accumulated over a period of about 70 Ma when the subduction
zone associated with the trench was dormant. In Aptian-Albian time the trench had an
elongate, at least 100 km (possibly 200 or 300 km) long basin plain located below the
calcite compensation level; sediment input was from two lateral sources (small acute-
angled fan or broad feeder channel exposed in the Falknis- and Tasna Nappe; submarine
canyon facies in the area north of Salzburg). The existence of a basin plain has also been
proved for the Campanian by the recognition of sheet-like turbidites that are continuous
for 50 km. The small number of lateral sediment sources is attributed to the existence of
slope basins that intercepted the downslope movement of turbidity currents. The lack of
syndepositional deformation and of volcanism and volcaniclastic detritus show that
sediment accumulation in the trench was not accompanied by active subduction.
This 500 km long and 10 (in the east perhaps 80 or 100) km wide trench of the East Alps
was connected with the 1000 km long Carpathian trench whose western segment was also
not associated with subduction until Oligocene time. The East Carpathian flysch,
however, experienced continuing deformation and (relative) eastward migration of the
depositional trough from Albian to Oligocene time. Huge Neogene volcanic centres of the
Inner Carpathians are associated with the final diastrophic phases both in the West and
East Carpathians, whereas in the East Alps there is no volcanism because subduction does
not seem to have reached a great enough depth. The East Alpine-West Carpathian
segment of this trench system may represent a major Cretaceous-Palaeogene transform
fault zone separating Europe and its bordering seas in the north from the Cretaceous
Penninic-Pannonian plate in the south. The latter has been eliminated almost entirely
from the surface due to Late Cretaceous-Palaeogene subduction at its southern boundary
nearly juxtaposing the European plate (Helvetic and Uitrahelvetic domains) with the
Austro-Alpine/South Alpine (plus Adriatic?) plate. Postulated strike-slip motion in the
west-east trending segment of the trench was followed by convergent motion, when
southward subduction (of the European plate) resumed in late Eocene-Oiigocene time.
Combination of an eastward facing trench with active subduction and a dormant trench
along a transform fault zone suggests an analogy between the Alpine-Carpathian Arc and
the modern West Atlantic loop arcs in the Caribbean and South Scotia Seas. The
geodynamic link between these loop-arc structures could be the westward drift of major
lithospheric plates between which smaller plates remain relatively stationary because they
are 'sheltered' behind westward-dipping subduction zones that are more or less anchored
in the asthenosphere.

O n e of the m o r e c h a l l e n g i n g tasks awaiting (tilting, s y n d e p o s i t i o n a l versus post-deposi-

sedimentologists in flysch t e r r a n e s is to d e t e r - tional d e f o r m a t i o n ) of the basin in g e n e r a l .
m i n e the t e c t o n i c setting of ancient flysch O n this basis the Flysch Z o n e will be inter-
basins. This is a difficult task and can only be p r e t e d as an a n c i e n t d e e p - s e a t r e n c h , w h i c h
d o n e by c o m p a r i s o n with m o d e r n basins that t o g e t h e r with its c o n t i n u a t i o n in the W e s t Car-
we think r e p r e s e n t flysch e n v i r o n m e n t s . pathians was d o r m a n t (i.e. not associated with
In this p a p e r t h e t e c t o n i c setting of the Flysch active s u b d u c t i o n ) d u r i n g m o s t of its existence,
Z o n e of the East Alps and C a r p a t h i a n s will be w h e r e a s the East C a r p a t h i a n s reflect active
analysed (with e m p h a s i s on the East Alps) Cretaceous-Tertiary subduction.
using: (1) p a l a e o c u r r e n t analysis and location
and c o m p o s i t i o n of the s o u r c e areas; (2) ana-
t o m y and facies t r e n d s of individual flysch Geological setting and previous studies
formations; (3) i n f e r r e d d e p o s i t i o n a l environ-
ments; (4) r e c o n s t r u c t e d basin g e o m e t r y includ- T h e Flysch Z o n e of the East Alps ( R h e n o d a n u -
ing p a l a e o d e p t h ; a n d (5) the tectonic b e h a v i o u r bian Flysch of O b e r h a u s e r 1968) flanks the

471
472 R. H e s s e

Northern Calcareous Alps as an approximately Mediterranean Tethys (Alps, Apennines, Car-

500 km long, 5-20 km wide, almost straight pathians, etc.) and collectively known under
fold belt forming the forested northern foothills the name 'helminthoid flysch' (Caron et al.
of the East Alps (Fig. 1). It consists of a 1979, 1981) after the meandering trace fossil
moderately thick sequence of lithologically dis- Helminthoidea.
tinct turbidite formations whose cumulative The middle to upper Cenomanian-lower
thickness reaches a maximum of 1.5-2 km Turonian Reiselsberg Sandstone represents
(Table 1). Carbonate flysch formations alter- the only major influx of massive terrigenous
nate with terrigenous flysch (Fig. 2). This flyschsands into the flysch basin. Lithologically very
was laid down during a 70 Ma period of similar poorly sorted, often mica-rich, poorly
apparently continuous deposition (Barremian- cemented sandstones occur in the upper Maes-
Aptian to Palaeogene). trichtian-Palaeocene Bleicherhorn Formation.
Sedimentological studies in the Flysch Zone These terrigenous flysch formations have an
(Butt 1977; Faupl 1975, 1976; Freimoser 1973; easterly source. Intercalated carbonate turbi-
Hesse 1973a,b, 1974, 1975; Hesse & Butt 1976; dites in these two formations, however, display
von Rad 1973) have revealed three significant palaeocurrent directions from the west. The
findings: (1) Palaeocurrent directions are pre- only terrigenous flysch formation with a wester-
dominantly parallel or subparallel to strike of ly source is the Aptian-Albian Gault Formation
the flysch belt but reverse orientation several whose sand turbidites are glauconitic quartz
times (Fig. 3). (2) At two stratigraphic levels greywackes with a significant amount (25-30%)
sequences of individual turbidites comprising of biogenic and terrigenous carbonate frag-
either an entire formation or significant parts ofments.
a formation have been traced along strike for The abundance of calcarenitic and calcilutitic
distances up to 115 and 50 km, respectively. (3) turbidites of biogenic origin in the Upper Cre-
Green hemipelagic and red pelagic claystones taceous helminthoid flysch facies but also in
are free of calcium carbonate although the the Lower Cretaceous Tristel Beds suggests
turbidites with which they alternate may be widespread carbonate-producing shelf and
carbonate-turbidites. Deposition was below the slope areas which were tapped by submarine
calcite compensation level (CCL). canyons. The fine-grained, quartz-rich
glauconitic turbidites of the Gault Formation
Palaeocurrent analysis, composition and with their substantial biogenic carbonate con-
location of the source areas tent probably represent shelf sands that moved
for relatively long distances by (long-shore)
Reversal of the palaeocurrent directions at currents before they became trapped in canyon
certain formation boundaries is accompanied heads incising the continental shelf. High mat-
by a change in lithology from one formation to rix content (particularly mica content) and poor
the next (Fig. 4) indicating that the palaeocur- sorting of the Reiselsberg Sandstone and part of
rent reversals are related to switching to petro- the Bleicherhorn Formation turbidites, on the
graphically different source areas. Carbonate other hand, suggest a different origin: these
turbidite formations (Barremian-Aptian Tristel sands were probably funneled through delta-
Beds, the thin-bedded Upper Turonian to systems during flood stages of rivers without
Lower Campanian Piesenkopf Beds, the much winnowing, reworking and redeposition.
Campanian Zementmergel Formation and the For the Reiselsberg Sandstone this is surprising,
Upper Campanian-Lower Maestrichtian H~il- because the late Cenomanian-Turonian was
lritzer Formation) have a westerly source probably the time of fastest sea-level rise during
although in the Ofterschwang Beds (Albian- the Cretaceous (Watts & Steckler 1979) where
Lower Cenomanian) some easterly derived one would expect little terrigenous detritus to
beds occur that are lithologically quite similar reach the outer shelf (as evidenced by the wide-
to the westerly derived beds. This is the only spread occurrence of chalk on the European
formation where turbidites of similar petro- and North American continental shelves). The
graphic composition but with opposing conclusion to be drawn from this is that tectonic
palaeocurrent directions alternate with one activity caused uplift and steepening of the
another (yon Rad 1964). The carbonate turbi- relief on land and narrowing of the shelves at
dites consist of mostly bioclastic calcarenites sea thus facilitating the supply of coarse ter-
and calcilutites (nannofossil limestones), the rigenous sands to the deep sea. The source
latter prevailing in the Piesenkopf Beds. Car- terranes for the Upper Cretaceous terrigenous
bonate turbidite facies of this kind are very sands must have been composed predominantly
widespread in the Upper Cretaceous of the of continental crust (Faupl et al. 1970;
 .o-.

~ o o
oI I I I I
,oo, I I
~O0.m

/,,'~ o o o o o
o o o ~. ~
~o$$~ o
o o
B o
S, o
o
o o n

o ~,~ O5~,,
• o o ~
o
U o o!
r~
CFLYSCH

o o /I
o o 9'
o o o o//
o o o o o o o o o //
o o o o o o o g ~
o o o_--o, o o ~
o o c~/I ---~
s
o ,~
o c/
I

F L Y S C H OF E A S T A L P S
AND CARPATHIANS

FIG. 1. Flysch Zone of East Alps and Carpathians and adjacent structural units mentioned in text (modified after Tollmann 1969).

4~
"--.I
L~
474 R. Hesse

phyl s= i o

ra n

44

terrigenous carbonate turbidites

FIG. 2. Average mineralogical composition of turbiditic sandstones and calcarenites of various flysch
formations in Bavaria. (Data from Hesse 1973b, p. 175 and von Rad 1973, p. 127; number of samples
analysed for each formation in centre of diagram.) Analysis are for basal portions of beds. Formation
names: see Fig. 4.

Freimoser 1973) because pebbles and rock frag- et al. (1972), Hesse (1973a,b), Niedermayr
ments of the sands include granite, gneiss, (1966), von Rad (1973) and Wieseneder (1967)
schist and argillite besides Jurassic and earliest have produced nothing but the occasional
Cretaceous sediments (calcarenitic reef- basaltic or diabasic rock fragment. This has
limestone, echinoderm breccia but also pelagic important repercussions on the tectonic inter-
filament limestone, saccocoma limestone, cal- pretation for the Flysch Zone of the East Alps.
pionellid limestone, tintinnid and aptychi limes- It is clear that active margin environments with
tone). Although Faupl et al. (1970) and subduction related volcanism are not suitable as
Freimoser (1973) seem to favour a northern tectonic setting for the Flysch Zone of the East
source, no conclusive evidence exists as to Alps and also of the West Carpathians (before
whether this material was supplied from the Miocene time).
north or the south and palaeogeographic con- Components derived from basic and ultraba-
siderations make a southern source at least as sic rocks indicative of exposed oceanic crust are
likely as a northern one (see later section). scarce but not entirely lacking in sediments of
One of the puzzles of Alpine orogeny is the the Flysch Zone. Chromspinel has been re-
scarcity of Mesozoic and Cenozoic volcanism. ported by Faupl (1975) from the Kaumberger
This is also manifest in the paucity of volca- Beds in the Vienna Woods (eastern part of the
nogenic rock fragments in the terrigenous sedi- Flysch Zone) and by Prey and Schnabel from
ments of the Cretaceous-Palaeogene flysch. sandstones of Reiselsberg type in the southern
More than 1000 petrographic analyses reported Vienna Woods and near Ybbsitz (P. Faupl,
by Faupl et al. (1970), Freimoser (1973), Grfin written comm. 1980). Pebbles of ultramafic
 Flysch Zone of East Alps and Carpathians 475
Munich

Middle Cr.

w 4nmm E

W ~ E

! I
50 lOOkm

Measurements by v. RAD
a n d HESSE

F°lk nl $ tt'~:~'

Fl6.3. Reversal of longitudinal palaeocurrent directions for three different times in the Flysch Zone
of the East Alps. Current directions for the Falknis and Tasna Nappes are shown only for the Lower
Cretaceous because Upper Cretaceous deposits of these nappes are largely pelagic. Note that in
middle Cretaceous time (Ofterschwang Beds, Reiselsberg Sandstone) some current directions
oppose the general trend from east to west by 180°, a few are at 90°. East-to-west directions in Upper
Cretaceous deposits are largely confined to the Bleicherhorn Formation. Length of arrow prop-
ortional to number of measurements.

composition, however, are more abundant in Composition of the source areas which sup-
the Ultrahelvetic Zone to the north of the plied detritus to the Flysch may thus be estab-
Flysch Zone (Dietrich & Franz 1976). In South lished with a reasonable degree of confidence
Penninic units (Verspala Flysch, Arosa Zone) from sandstone and conglomerate mineralogy
and the U p p e r Austro-Alpine deposits of the and petrography if one keeps in mind the
Gosau Basins to the south of the Flysch Zone possible introduction of bias due to intrastratal
chromspinel occurs in strata as young as early solution and selective removal of components
Campanian according to Woletz (1967). during transport and weathering. Locating
However, Faupl (written comm.) and the these source areas, however, is more difficult.
author (unpubl. data) found it also in Maes- Palaeocurrent indicators may be useful to iden-
trichtian-Danian strata in the Weyerer B6gen- tify the general transport directions, but are not
Gosau and late Campanian strata of the K6ssen diagnostic to pin-point the exact locations
Gosau basin, respectively. where the detritus was introduced into the
 ",.3

TABLE 1. Comparison of stratigraphic columns forflysch zones of East Alps and West Carpathians
(after Dzulynski et al. 1959; Hesse 1974; Oberbauser 1980)
FLYSCH ZONE Carpathians
i

Vorar Iberg North I South Greifenstein Nappe Vienna Woods Laober Noppe Silesian Noppe
(sigiswang) Facies f (Obersfdor f) Facies Kohlenberg Noppe
.... J
Krosno Beds BOO--> 2000 m

Menilite Beds 150-5CX~ m

Greifenstein Beds --500 m Loober Beds 2000-3000 Hieroglyphic B e d s 500-80 m
Potoeocene
(Danian) Trotenbach Formation Ciezkowice S~ndstone fSO m
Bleicher horn Formation ~ . . . . . . . Upper Istebno Beds
mica-rich greywocke, shale
Moestrichtian Fano]oFm 500-700 m
mica-rich greywocke, Hallritzer Formation
calcorenite and shale Uppermost VorJegated Clayst. Altlengboch Beds > I000 m Slavering Beds -700 m
] Companion I PlonknerbriJcke Fm --550 m carbonate greywacke + shale --400 m J Uppermost Variegated Cloystone
o / Sonfontan t Corb°note greywockeand sb°le , r Zementmergel Formahon
calcorenite, corbonotic siltstone Zementmergel Formation KohlenbergBeds -- 500m Kaumberg Beds )O0-300m
P,leset~kopfBeds ~ ~ _ thick marl and green shale
Lower Istebna Beds 1000 m
~ F..n_ iOcion : co[cflutite and shale /

I Turonion I Schwobbrunnen Prn ;> 600 m 600-200 m ReiselsbergSandstone 10-50 m Reiselsberg Sandstone
poorly sorted, mica-rich greywacke poorly sorted, rnico-rich greywacke 0-500 m
Godula Beds 2000 m
Cenomonion Basis Fro, ~ Lower
Ofterschwong Bed
, , ~ Variegated Cleyst, -200 m Vorlegated Cloystone
--IO0 m siltstone and shale /

Goult Formation 200 m Goult Formation - I 0 0 m Mikuszowice Rodiolorite <50 m

t~ ion quartz-greywocke, black and green shale (in port Bortberg Beds) Lgoto Beds 500 m
Trlstel Beds 150 m
graded colccfenite, colcilutite Wernsdorf Shale 200m
LBorremion green and block shale
substratum unknown
~ l Houterivien Grodischt Sandstone 50 m

Valonginion
Cteszyn Shale and Limestone
Berriosion 700 m
 Flysch Zone of East Alps and Carpathians 477
TASLE 2. Characteristics of turbidites on laterally confined, elongate basin plains
Environments with
similar characteristics
(1) Palaeocurrent directions:
Predominantly longitudinal currents, i.e. Large deep-sea
current directions parallel to long axis channels (mid-
of basin (or parallel to strike of structural belt). ocean channels)
Azimuths show bimodal distributions 180° apart, Perhaps elongate
i.e. current reversals are common, between abyssal plains (e.g.
successive formations as well as within Horse-Shoe abyssal
individual formations plain), otherwise non-
existent
Deposits with opposing current directions are
often petrographically distinctly different
(different sediment source)
(2) Continuity of individual turbidites:
up to 100 km or more Deep-sea channels
(East Alps: 115 km, (i.e. Cascadia Channel)
Apennines: 190 km, Ricci-Lucchi 1975a;
Caucasus: 200 km, Grossheim 1961)
(3) Bed thickness of turbidites:
longitudinally continuous layers: cm to more than
10 m (Apennines, Ricci-Lucchi 1975a, Ricci-Lucchi &
Valmori 1980)
(4) Sedimentary structures:
base-cut out sequences often starting with b-division
(lower parallel lamination), a-divisions (graded or
massive) occur occasionally (e.g. felspar-rich marker bed
of Gault Fm.). Repeated alternation of b and
c-divisions within individual turbidite (possibly due
to current reflection from the basin walls). Thickness
of e-division dependant on availability of pelitic material,
e.g. Gault Fm.: subordinate or absent, Zementmergel
Fm: dominant sediment type
(5) Maximum grain size:
typically medium to fine sand. Rarely coarse sand
(felspar-rich marker bed of Gault Fm.) or gravel
(pebbly mudstone, Gault Fm.; Reiseisberg Sandstone)
(6) Down-current lithological variations (bed thickness, grain-size,
mineral composition, sedimentary structures):
minimal. Not detectable on outcrop-scale.
Recognizable only within continuous turbidites
if traced for long distances (see Fig. 7)

basin. Ever since palaeocurrent measurements lished, lateral sediment sources cannot have
have been used there has been a need to verify existed during the time of deposition, because
by detailed facies analysis the locations of the all beds are continuous or pinch out (i.e. the
point sources of sediment supply. If the mea- thinner ones) in an easterly direction and no
sured directions are parallel to strike of the new beds appear (Fig. 5). The same applies to
structural belt, does this mean that the sources the Zementmergel Formation in which a 190 m
were far away at the basin ends; if they are thick sequence of carbonate turbidites has been
perpendicular to strike, does this mean we see a traced bed-by-bed for a 50 km distance (Hesse
nearby lateral source? For the Flysch Zone of 1979). Since the two formations have
the East Alps the absence of sediment sources palaeocurrent directions from the west, the
on parts of the basin margins has been proved point sources where the material was intro-
by the striking continuity of individual turbidite duced into the flysch basin, therefore, was to
beds of the Gault Formation, in connection the west of the correlated sections.
with longitudinal palaeocurrents. Along that This, however, should not be misinterpreted
115 km long segment of the Flysch Zone for to mean that lateral sediment sources did gener-
which the continuity of beds has been estab- ally not exist along the margins of the flysch
478 R. Hesse

N
, z

Maestr.

• ~ • t • i • I , I • L~ I • t ~ : u L , ~ . I [ • I ' z ~ _ ~ _ ~ 2 - ~ Z Z = : ~ z ~ ~ ........

- ~ _ . . . . . ._ _~,_.~. . . . . . . . . . . .~ . ,. ', ;_+~_+~. ~ .....

. . . . . . . . . . '~:~/U 0 ~::= . . . . . ............ -
° ' • , i I ~ - i • i, , - - - 000
~ u _ l . . t - i i . f i , i. I , ~ i: i { ~ ] ~ + 'T i - --
Camp. - . • /,+ + ~ ~ . + .+ i . l . ~ m + L ~ .

~ ~ L ~ ~ ~ . ~ ~ . _

Sant.
C oniac.__:, -500
Turon.
Cenom. i
Albian
Albian
' Aptian
++- " • . . . . ' " " f't'--~

Om
+10 km ,*
FI(;. 4. Palinspastic cross-section for central portion of the Flysch Zone of East Alps in Bavaria
(between rivers Lech and Inn). ft: Tristel Beds: fg: Gault Formation: fo: Ofterschwang Beds; fs:
Reisclsberg Sandstone: fp: Picsenkopf Beds: fz: Zementmergcl Formation: fh: Hfillritzer
Formation: fb: Bleicherhorn Formation. (In south-eastern Bawlria the latter occurs also in the
southern half of the Flysch Zone according to Freimoser 1973.) Circled dots: formations with
palaeocurrent directions from the west. Circled crosses: formations with palaeocurrent directions
predominantly from east.

basin. They did exist, e.g. for the Gault Forma- fill may be unravelled: lithological variations
tion in the area north of Salzburg, where coarse (bed thickness, grain-size, mineralogical com-
to very coarse detritus suddenly appears that position) may be evaluated quantitatively and
has no upcurrent equivalent farther west (Hesse gradients be established• Since outcrops of the
1973a). A n o t h e r lateral source was established in Gault Formation occur along a narrow belt less
the west in GraubOuden and Liechtenstein in than 2 km wide, only longitudinal, downcurrent
the Gault Formation of the Falknis and Tasna gradients have been established, whereas for
Nappes (see section on depositional environ- the Z e m e n t m e r g e l Formation gradients per-
ments). Such lateral sources can have existed in pendicular to palaeocurrent direction are also
any area at any time, but their existence has to available (unpubl. data). The large number of
be d o c u m e n t e d by appropriate facies evidence available thickness data required statistical
(occurrence of submarine canyons or deep-sea treatment by a m e t h o d described by Hesse
fans) in addition to palaeocurrents perpendicu- (1973b). Thickness gradients for the Gault
lar to strike. Formation are extremely low and yield an
average of - 6 mm km -1 for 11 beds decreasing
downcurrent in thickness and of 4 mm k m - i for
Anatomy of individual turbidite formations
12 beds increasing downcurrent (Fig. 7). Gra-
If the turbidites of individual formations can dients for thin-bedded basin-plain turbidites of
be traced as continuous layers over major parts the A p e n n i n e s (Ricci-Lucchi & Valmori 1980,
of the basin (by 'fingerprinting' them using bed table 3) are of the same order of magnitude,
thickness (Fig. 5) and mineralogical composi- whereas thick-bedded basin-plain turbidites
tion (Fig. 6)), the anatomy of parts of the basin show somewhat higher gradients. This com-

Fw,. 5. Example of correlation of turbidite sections (Gault Formation, Aptian-Albian, simplified

from Hesse 1974). Only the lower half of the formation (i.e. Lower Claystone Member, Lower
Greywacke Member, Middle Claystone Member) is shown. Note that bed-by-bed correlation of the
sections shown is nearly perfect (except with the most proximal section 225, which contains a larger
number of additional thin beds). Bed F2 is a relatively felspar-rich petrographic marker bed
('felspar-bed') which, in a number of outcrops, is amalgamated with underlying bed F1.
 section no. 56 64 70
Middle
Clayst°ne Member

27km
25 J32 section no. 41
4 appr. 90 km IL
Lowe r Ctayst°ne Member
Downcurrent direction of ~'~
m25 sediment transport /

claystone M ember
i.o~er
-15
Tristel Beds
lr~stel Seds

~r~stel Beds
480 R. Hesse
Gault Formation, Eastern Alpine Flysch
Standard i i i I I 1
section I
m
30o/o Q u a r t z 60 o 10o/0Feld s p a r ~18
-30 Q

O
N
M
-20
L
K
J
H
G ........2 $
-,0 F]
i
E
D
J
C
,o"0 I

F[(;. 6. Bed-by-bed correlation of selected sections (225, 232, 241--shortened to last two digits on
figure) of the Gault Formation by mineral composition of basal laminae of every bed, (Simplified
from Hesse 1973a, fig. 3: only turbidites A - Q of the Lower Greywacke Member are shown.) Quartz
and feldspar contents are plotted versus standard section. Beds that deviate significantly are denoted
by full circles (e.g. bed L in section 232, bed N in section 225).

THICKNESS X
m
GAULT FM, (Beds no. 12-21))

3.0"

( 2 0 ) = 0 2 21 0 . 0 9
-
~ "- - " " (2l) : P 2.05 0.76
2.0

(17') =L 1.49 0.01

J (16) --K 1.4.2 0.20

1.0-

(18) =M 0 . 8 2 - 0 . 6 5
(15) =d 0.64. 0 . 0 5
(19) =N 0 . 6 9 - 0 . 3 2

(12) =H 0 . 2 7 - 0 . 7 4
, z~-
I0 50 [00 I{0 Km
DISTANCE

FIG. 7. Vertical and horizontal bed thickness variation of beds no. 12 ( - H on Fig. 5) to 21 ( - P on
Fig. 5) over a downcurrent distance of 106 kin. Beds 15-17 and 20, 21 slightly increase downward in
thickness with an average gradient of 2.7 m km -1. Average gradient for beds thinning downcurrent
is - 4 . 2 mm km -1. Gradients obtained by linear regression analysis. Column to the right: average
bed thickness. Note general upward thickening trend of sequence (beds with higher numbers are
generally thicker). On the basin plain such a trend reflects a systematic increase in availability of the
fine sand-sized material, which may or may not be inherited from a prograding depositional lobe on
an adjacent deep-sea fan.
 Flysch Z o n e o f East A l p s and Carpathians 481

pares well with an average figure of tion in the south. Going from north to south
- 7 mm km -I for three continuous layers in individual turbidites lose their basal structural
recent turbidites of the Puerto Rico Trench divisions (mostly b-divisions, some a-divisions)
(gradient determined for a correlation distance but at the same time gain thickness in the
of 81 km, Conolly & Ewing 1967) and - 3 and e-division of homogeneous grey marlstone.
2 mm km-~ for three downcurrent thinning and This suggests that the competence of individual
three downcurrent thickening layers, respec- flows to carry sand-sized sediment decreased
tively, of the Cascadia Deep-Sea Channel (cor- very slowly in the downcurrent direction but
relation distance 83 km, Griggs et al. 1969). relatively rapidly in the direction perpendicular
Modern deep-sea trenches and deep-sea chan- to flow. The two stratigraphic columns that
nels are clearly environments with very low have been established for the north and south
horizontal gradients for lithological variations facies in the central segment of the Flysch Zone
in turbidites. Deep-sea channel deposits, howev- (in Bavaria) can therefore be applied to areas
er, will show unidirectional palaeocurrents and far to the east in Austria and the west in Allg/iu
an association of channel-fill with spill-over and Vorarlberg (for names of age-equivalent
facies; they are therefore not possible ana- formations see Table 1). Some of the forma-
logues for the Flysch Zone of the East Alps. tions can even be recognized in the West Car-
Downcurrent variations in grain-size and pathians in Czechoslovakia (Prey 1960, 1965)
mineralogical composition of individual beds this being the basis for treating the East Alpine-
display equally low gradients (Hesse 1973a, Carpathian Flysch as one continuous belt.
1974). For example, the median grain-size at
the base of marker bed F2 (characterized by its Inferred flepositionai e n v i r o n m e n t s
high detrital felspar content) decreases from
more than 500 pm to less than 200 pm over a
Elongate basin plain with limited number of
distance of 90 km in the downcurrent direction. point sources
Simultaneously its felspar content decreases The long-distance continuity of layers
from 18 to 11% as a result of hydrodynamic together with the longitudinal palaeocurrent
sorting accompanying the grain-size decrease. directions and the reversals in current azimuth
(The felspar, which is exclusively plagioclase, is conclusive evidence that the depositional
was enriched in the coarser grain-size fractions environment representing a major part of the
of the starting material.) flysch basin was a flat, nearly horizontal basin
These low but detectable gradients for down- plain. In addition, this was an elongate basin
current lithological variations are reflected in plain. The suggestion of a nearly horizontal
the remarkable facies constancy along strike. basin plain is inescapable--how else could the
Across strike, i.e. perpendicular to the preva- currents have reversed flow direction? Tilting
lent palaeocurrent directions, on the other hand of the basin floor would be a possibility.
facies varies relatively rapidly. Consequently, However, if current reversals occur repeatedly
two separate stratigraphic columns are in use, and even between successive turbidites as in the
one for the southern 'Oberstdorf Facies' and Ofterschwang Beds, tilting is not a realistic
one for the northern 'Sigiswang Facies' (Table alternative. The long-distance continuity of
1). Tristel Beds and Gault Formation are res- layers is also best explained by a basin plain
tricted to the southern facies (except in the area environment--by analogy: elongate basin
near Salzburg and at one western locality in the plains and deep-sea channels are the only mod-
Allg/iu, yon Rad 1964) either because they ern environments for which continuous turbi-
have been sheared off the base of the Flysch dite layers have been reported (e.g. Conolly &
Nappe during tectonic transport or because the Ewing 1967; Griggs et al. 1969; Chough &
basin was much narrower in Early Cretaceous Hesse 1980; Pilkey et al. 1980).
time, not comprising that part which was later Basin plains may develop in various kinds of
to become the northern facies realm. In the turbidite basins including deep-sea trenches.
latter case deposition of the Reiselsberg Sand- For modern trenches Piper et al. (1973) and von
stone (and the Ofterschwang Beds) in the Huene (1974) suggested that their trench plains
northern facies realm would indicate a mid- are characterized by an axial leveed channel
Cretaceous widening of the basin and north- based on findings in the Aleutian and Peru-
ward overstepping of the turbidite facies by Chile trenches. Obviously, in the Gault Forma-
2-5 km. In the Upper Cretaceous the facies tion and in the Zementmergel Formation of the
change from north to south is best observed Flysch Zone such a channel is not present. It is
between Hfillritzer Formation (in the north) questionable, however, whether an axial chan-
and coeval parts of the Zementmergel Forma- nel is a necessary element of a deep-sea trench
482 R. H e s s e

m 14

1 -lit
E-7.t~
3
k m ~ ~

~ E ,
lli . ;
~:~ , _~

J
-'l • ~1 ~ ~k "---~H
: f, ]-
-'1 =
d; ~ ~ " D

- ¢ B

3_

I---2.5 km =I" ,rkm .~ ,o 5 - ~ - - - - o 25 ~ . , - - - - - - - - - ~ ~ /

~km i
2 6 8 i5 ,6 17 18 ~-~ ,/'3z

i 23
~--- 50 km --"P~ I 10 km-"~

FIG. 8. Correlation of sections, Gault Formation, Falknis Nappe (sections 202,205 to 208,212,214 to
218), Tasna Nappe (223), and, for comparison, Flysch Zone of East Alps (232) (modified from
Hesse 1973a, plate 3---first digit of section no. omitted on figure). The Lower thin-bedded Greywacke
Member (I, Upper part), the Lower Greywacke Member (II), the Middle thin-bedded Greywacke
Member (III) and lower part of Upper thick-bedded Greywacke Member (IV) are shown. The
depositional environment of the Falknis and Tasna Nappes is interpreted as an acute-angled
deep-sea fan or broad feeder channel. Two groups of turbidites (f2 and n-o-p) are tentatively traced
as continuous layers through most sections of the Falknis and Tasna Nappes and are probably also
present in section 232 of the Flysch Zone.
Note the general thickening-upward trend from beds 'c' to 'p' and a second similar trend at the top
of the sections shown. Similar trends are inherited by the basin plain facies (section 232) of the Flysch
Zone (see Figs 5 & 8).

plain, because few trenches have been studied in (1980) to infer a basin plain environment for
this regard. In the Mid-America trench no much of the Miocene Marnoso-Arenacea
well-defined axial channel was detected during Formation of the Northern Apennines. It
recent deep-sea drilling off Guatemala (von should be noted that in vertical bed sequences
Huene et al. 1980, fig. 2). One also can imagine of the Gault Formation certain trends such as a
that in narrow trenches the trench walls them- thickening-upward succession can be de-
selves may act as channel walls. ciphered (Fig. 5) which might be misinterpreted
The characteristics of basin plain turbidites as a 'prograding fan-lobe trend' if taken out of
are listed in Table 2. The same kind of evidence context. This is to be kept in mind in trend
(i.e. long distance continuity of layers, analysis of turbidite sequences: there may be
palaeocurrent reversals) was used by Ricci- factors other than the geomorphology of a
Lucchi (1975 a, b) and Ricci-Lucchi & Valmori specific depositional environment (such as a
 Flysch Zone of East Alps and Carpathians 483

Fx6. 9. (A) North wall of Falkinshorn, Liechtenstein, displaying thickening upward sequence of
proximal turbidites (lithological members III and IV, see profile 5 on Fig. 8) interpreted as
prograding fan lobe. Exposed thickness (between X-X 1) approximately 100 m. Lower 20 m of
sequence (below X) separated from overlying sequence by nearly bedding-parallel fault. Detailed
bed-by-bed section (profile 5, Fig. 8) measured approximately at centre of picture with the aid of
mountain guides J. Gassner and O. Lampert, Triesen, Liechtenstein.
(B) Parallel-laminated and partly wavy-laminated, 80 cm thick turbidite with upper layer of oblique
lamination (left side of picture) characteristic of bypassing without deposition of major portion of
turbidity current (turbidite facies B2 of Mutti & Ricci-Lucchi 1975). Gault Formation, Tasna Nappe,
Breite Krone, Grisons.
(C) Internal reworking (reactivation) surfaces in thick, coarse-grained turbidite, deep-sea-fan facies.
Base of bed to the right. Falknis Nappe, Naafkopf, 2360 m, Liechtenstein.
(D) Protrusions of convolute bedding on upper bedding surface current-remolded into 'false flute
casts'. Tasna Nappe, Breite Krone, 2740 m, Siivretta Ranges, Grisons.

deep-sea fan) which control vertical thickness model. Although these sections occur in iso-
trends ('megarhythms' of Lajoie 1979), etc. lated tectonic units of the Falknis and Tasna
These may include climate, sea-level changes Nappes separated 50 km from the closest Gault
and the tectonic behaviour of the source area. sections in the Flysch Zone, it has been known
for some time that the rocks of the three nappes
were palaeogeographically linked (e.g. Richter
Deep-sea fan in the Gault Formation of the
Falknis and Tasna Nappes 1933). Palaeocurrent directions (Fig. 3) as well
as detailed facies analysis prove that the Falknis
In the westernmost outcrops of the Gault and Tasna Gault Formations are the upcurrent
Formation in the West-lEast Alpine border re- equivalents of the Gault Formation in the
gion in Liechtenstein and Graubiinden Flysch Zone.
(Switzerland) bed-sequences occur that bear In the Falknis and Tasna Gault bed-by-bed
some distinct petrographic and lithological correlation of sections is generally not possible,
relations with those farther N E in the Flysch although at certain stratigraphic intervals sec-
Zone but also display some characteristic differ- tions have been tied together bed-by-bed over a
ences and do not fit a basin-plain depositional limited distance (Fig. 8). The formation has
484 R. Hesse
TABLE 3. Greywacke/claystone thickness ratio, Gault Formation
Lithological Falknis and Tasna Nappes Flysch Zone
member Min. Max. Av. Min. Max. Av.
IV 11.8 42.7 25.4 0.9 12.5 3.1
II and III 4.7 13.6 7.4 0.6 4.2 1.9
I 1.7 3.8 2.8 0.1

been subdivided into five lithostratigraphic coarse-grained felspar-bearing sandstone

members that are homogeneous lithologically, beds, e.g. five in Member II of the Falknis-
have limited thickness, can be identified at Tasna Nappe compared to one in Member II of
almost any locality if exposed and correspond the Flysch Zone (two in the westernmost out-
to five lithological members of the Gault crops in the Allg~iu area). In Member IV this
Formation in the Flysch Zone. For the Falknis effect is even more pronounced, where a large
and Tasna Nappes these are (I) a Lower thin- number of felspar-bearing coarse beds occurs
bedded Turbidite Member containing a pebbly in the Falknis-Tasna area but none further east
mudstone and a conglomerate layer, (II) a in the Flysch Zone: i.e. the turbidity currents
Lower Turbidite Member of intermediate had deposited their coarser load before
thickness, (III) a Middle thin-bedded Turbidite reaching the basin plain of the Flysch Zone.
Member, (IV) an Upper thick-bedded Turbidite This is an excellent example of horizontal
Member (Fig. 9a) and (V) an Upper Clay- gradation occurring over a distance of about
stone Member. The overall thickness of the 50 km. The finer-grained sand fractions which
formation is about the same in the three nap- make up the turbidites in the Flysch Zone are
pes, except that the Tasna outcrops are strati- present in the Falknis-Tasna area as upper
graphically incomplete due to faulting. In each portions of the coarser beds or as individual
member the greywacke-claystone thickness thin layers. The same effect of horizontal grada-
ratio for sections of the Falknis-Tasna Nappe is tion is displayed by the greater abundance and
an order of magnitude higher than for corres- larger diameter of resedimented claystone
ponding members in the Flysch Zone (Table 3). clasts (intraclasts) in the west.
This increase in sand-turbidite thickness at the The number of amalgamated beds is con-
expense of turbiditic and hemipelagic shale or siderably higher in the Falknis-Tasna area than
claystone is due to the fact that much of the in the Flysch Zone--this being one reason why
fine-grained turbiditic sediment bypassed the bed-by-bed correlations were not possible in
Falknis-Tasna area and most of the hemipelagic the former. Lower structure divisions (a-
clay deposited was eroded by subsequent tur- divisions) are much more abundant in the for-
bidity currents. Many of the thinner turbidites mer than in the latter, although this has not
of the Falknis-Tasna Nappes pinch out before been quantified. Similarly, there are
reaching the Flysch Zone. Therefore, the thin- sedimentary structures such as low-angle cross-
bedded turbidite members of the Fatknis-Tasna bedding (dune and antidune cross-bedding,
Nappes correspond to claystone members in the Hesse 1973a, plate 9, fig. 2), internal reactiva-
Flysch Zone, the thick-bedded turbidite mem- tion (reworking) surfaces (Fig. 9c) and small-
bers to turbidite members of intermediate scale channels (less than 1 m deep, Hesse 1973a,
thickness. This change in character is gradual. plate 8, fig. 3) in the Falknis and Tasna areas
In the westernmost section of the Gault Forma- that have been described elsewhere from
tion in the Flysch Zone (in the Allgfiu area) the proximal environments such as deep-sea fan
middle member is still a thin-bedded turbidite valleys; these are lacking in the Flysch Zone.
member, as in the Falknis-Tasna Nappes. Rippled surfaces, thin cross-bedded layers at
Further east it loses its turbidites and becomes a bed-tops (Fig. 9b) and remoulded, stream-lined
claystone member. convolute bedding surfaces ('false flute casts',
For the Falknis-Tasna Gault, bed thickness, Fig. 9d), indicative of bypassing have also been
grain-size and felspar content of sand- observed in the Falknis and Tasna Nappes.
turbidites show a significant increase compared In summary the Gault Formation of the
to corresponding members of the Flysch Gault. Falknis and Tasna Nappes is distinctly more
This is a continuation of the slow gradual proximal in character than the Gault Formation
upcurrent changes observed within the Flysch of the Flysch Nappe and contains the upcurrent
Zone. The increase in felspar content and portions of the turbidites whose distal ends on
grain-size is reflected in a large number of the basin plain of the Flysch Zone extend at
 Flysch Zone of East Alps and Carpathians 485

least 150 km farther east, perhaps as much as and (3) the presence of thickening upward
270 km to the area north of Salzburg. The cycles (Fig. 8). The main difference in compari-
observed facies associations are similar to Mutti son with the Mutti & Ricchi-Lucchi fan model is
& Ricci-Lucchi's (1975) middle and outer fan the presence of only two thickening upward
facies, characterized by (1) their facies associa- sequences that may be interpreted as prograd-
tions B1 (thick, coarse to medium grained sand- ing fan lobes (members I + II, and members III
stone with occasional granules and pebbles, + IV). This might suggest that the fan was
showing wedge-shaped sets of thick, coarse- relatively small and consisted of a single chan-
grained inclined laminae, plane-parallel and nel and depositional lobe probably confined by
wavy laminations, dish structures and fluid a funnel-shaped incision at the base-of-slope.
escape pipes, common claystone chips, small- 'Feeder-cone' is perhaps a better term for this
scale cross-lamination and/or convolute lamina- feature if one considers its function as the main
tion), B 2 and E (both similar to B I but thinner supplier of sediment for the adjacent basin
bedded), C1 and C2 ('classical' turbidites with plain. Although it may have existed for 5-8 Ma
more or less complete turbidite structure- the rate at which turbidity currents passed
sequences), (2) the absence of larger channels, through this system was probably not very high

o ,oo zoo k~ K R AIKOW

GENEVA ZURICH MUNICH VIENNA /~" ~ = . -'~'~N~ BUCHAREST

~o

FIG. 10. Inferred schematic arrangement of palaeogeographical zones of Alpine Tethys in Albian
time and position of point sources (arrows) for the flysch trench (after Faupl 1978; Gwinner 1971;
Hesse 1973a; Hesse & Butt 1976; Lemcke 1970; Sandulescu 1973; Tollmann 1969).

I
OUTER RIDGE

,~sooo '~Z73/z'~'=z'~"

s • ~ X~Z E~/2~,, -
-~ S O o °

_,.... \ ... ,.._

I 1 I I I

FIG. 11. Bathymetry of the Puerto Rico Trench showing main tributary channel (Mona Canyon) on
the south wall (simplified after Conolly & Ewing 1967 and Ewing & Heezen 1955). The Puerto Rico
Trench is considered a suitable modern analogue for the Cretaceous deep-sea trench of the Flysch
Zone of the East Alps. Isobaths in fathoms.
486 R. Hesse
(up to 50 Ma -1, 10 of which reached the basin conglomerate in the Gault Formation north of
plain). The deep-sea fan or feeder cone of the Salzburg (Aberer & Braumfiller 1958; Hesse
Falknis and Tasna Nappes in Albian time was in 1973a), whose significance as a possible sub-
all likelihood located on the south side of the marine canyon deposit has already been men-
SW-NE to W-E oriented fan-basin plain array tioned.
(Fig. 10), because palaeocurrents are generally
toward NE and grain-size decreases in the same Basin geometry and palaeodepth
direction. A northward decrease in grain-size is
especially obvious for the Upper Jurassic Falk- The deep-sea fan/basin plain association of
nis Breccia (Allemann 1956) which may repre- depositional environments that has been estab-
sent an inner fan valley deposit of the same fan. lished repeatedly in time appears to be repre-
If we look for a modern equivalent of this sentative for the Flysch Zone, although other
fan-basin plain assemblage, the modern Puerto environments may have existed at times (e.g.
Rico Trench seems to provide a close example. during Reiselsberg time, see below). This basin
It has a main tributary on the south wall (Mona plain was relatively narrow compared to its
Canyon) which joins with the trench plain (Fig. length judging from: (1) the present shape of
11). the flysch belt; (2) the relatively low degree of
The proposed correlation between the Gault scatter in the predominantly longitudinal
Formations of the Falknis and Tasna Nappes palaeocurrent directions; (3) the successful bed-
and the Flysch Zone was the first example of by-bed correlation in two turbidite formations.
ancient deep-sea fan deposits that have actually If the flysch belt as we see it today was merely a
been traced laterally into their basin-plain stripe of ocean floor cut randomly out of a
equivalents (Hesse 1973a). Similar results have broad abyssal plain it would be a miraculous
been obtained recently by Ricci-Lucchi & Val- coincidence that its structural orientation is
mori (1980) in Miocene flysch of the Apen- parallel to the sedimentological trends. If the
nines. Contrary to some current models (e.g. basin plain was part of an originally broad
Walker 1978), basin-plain turbidites are not abyssal plain, it would be much more likely to
restricted to thin-bedded, c-d-e-division distal see more scatter and a different pattern in the
varieties but include many beds that start with palaeocurrent directions. The currents not
b-divisions and the occasional a-division bed. being confined by the basin walls would shift
Some of these may be very thick. This may be a laterally in a similar fashion to fan-lobes on a
special characteristic of an elongate basin plain deep-sea fan and it would be most unlikely to
fed by a small narrow fan or wide canyon or find 50 consecutive beds that are stacked up in
feeder channel which did not trap major por- an orderly fashion so that each one can be
tions of the sediment during downslope trans- traced for more than 100 km. Sediment
port. volumes contained in individual beds would
Discussion of the depositional environments become unreasonably large, if the shape of the
in this section has been restricted to two basin plain was square rather than narrow
selected formations, for which detailed data are rectangular--to compare two geometric situa-
available. The deep-sea fan/basin-plain associa- tions that are obviously simplifications but not
tion of environments may be characteristic for unrealistic ones, from what we know of modern
certain other formations of the Flysch Zone, basin plains. Estimates for individual beds of
particularly carbonate flysch formations. It may the Gault Formation range from 1 to 25 km 3 (of
not be applicable, however, to formations such solid particles), the lower value applying to an
as the Reiselsberg Sandstone which displays approximately 1 m thick bed (of 2% porosity)
invariably proximal characteristics over a dis- extending for 100 km (proved length) times
tance of at least 200 km along strike. Petro- 10 km (assumed basin width), the maximum
graphically similar beds of the Bleicherhorn applying to a bed of 2.5 m average thickness
Formation also display distinctly proximal fea- (bed 'AJ', Hesse 1974, fig. 3) extending for
tures. These are the only two formations which 300 km (possible length) times 30 km (possible
also contain significant amounts of fine-grained basin width). These figures compare well with
conglomerates: elsewhere conglomeratic de- estimates of Ricci-Lucchi & Valmori (1980) for
posits occur only as isolated layers (i.e. pebbly basin-wide, continuous turbidites in the Apen-
mudstones in the Gault Formation (Hesse nines. If basin width was assumed to be an
1973b); conglomerates in the Ofterschwang order of magnitude larger, the thickest beds
Beds (Stephan & Hesse 1966) and Zement- would contain sediment volumes in the order of
mergel and H~illritzer Formations (Freimoser 100 km3--volumes comparable to estimated
1973)). An exception is the coarse boulder slump volumes that result from open-slope
 Flysch Zone of East Alps and Carpathians 487

failure (e.g. 70 km 3 for the slump following the in Ofterschwang and Reiselsberg time. If so,
Kwanto earthquake of 1923, Menard 1964) but the widening was accompanied by at least 150-
unlikely for flows that were funnelled through 550 m deepening (compared to the southern
submarine canyon-fan valley systems. Such sys- part), this amount being the difference in thick-
tems must be capable of containing the entire ness of the Reiselsberg Sandstone (in the com-
flow more or less at a time--i.e, the sediment pacted state) between the southern facies
volume plus perhaps 40 times the amount of (usually less than 50 m) and the northern facies
water in order to make the slurry move. (200-600 m, von Rad 1973). This deepening
Palaeodepth of the basin plain was below the due to down-faulting (Fig. 4) is the only evi-
calcite compensation level as shown by the dence for possible syndepositional tectonic
absence of calcium carbonate in the green activity in the Flysch Zone.
hemipelagic claystone which occurs throughout
the stratigraphic column of the Flysch Zone
often alternating with carbonate turbidites
(Hesse & Butt 1976). Carbonate-free red clay- Tectonic setting of the Flysch Zone of
stone which occurs at three stratigraphic levels the East Alps: 'dormant'
(Albian--lower Cenomanian lower Varie- Cretaceous-Palaeogene deep-sea trench
gated Claystone, upper Turonian--lower Con- associated with transform faulting
iacian upper Variegated Claystone and upper
Campanian Uppermost Variegated Claystone) Discussion in previous sections has shown that
has been interpreted as equivalent to modern none of the following types of turbidite basins
brown abyssal clay on the basis of chemical and are adequate modern analogues for the Flysch
clay mineralogical analysis (Faupl 1976). Zone of the East Alps: (1) active margin basins
(i.e. upper slope, mid-slope or lower slope
Tectonic behaviour of the basin basins, deep-sea trench associated with active
subduction); (2) continental rise apron; (3)
There is little if any evidence for syndeposi- abyssal cone; (4) abyssal plain; (5) mid-ocean
tional deformation. If the absence of the Lower channel; (6) ponded basin on mid-ocean ridge;
Cretaceous Tristel Beds and Gault Formation (7) back-arc basin plain. Among the remaining
in the northern facies realm is not due to possible basin types the deep-sea trench with
tectonic erosion at the base of the Flysch Nap- dormant subduction is the tectonic environment
pc, it may reflect widening of the basin and best suited to fit the observations (Table 4).
northward overstepping of the turbidite facies Basin geometry (great length compared to nar-

TABL~ 4. Flysch Zone of East Alps: Cretaceous-Palaeocene deep sea trench

Evidence for trench:
Dimensions: Length: 500 km (without Carpathians)
Width: 10-50 km (widening toward E: possibly > 100 km)
Depth: below palaeo-calcite compensation level.
Red claystone as equivalent of brown abyssal clay
Depositional environments:
(a) Elongate, nearly horizontal basin plain:
Palaeocurrent directions predominantly parallel
to strike (basin axis)
Repeated reversal of current directions (change
in azimuth by 180° in successive turbidites)
Long-distance continuity of individual turbidites
Gault Formation: 115 km
Zementmergel Formation: 50 km
Low gradients of downcurrent change in bed thickness,
grain-size, etc. along individual beds (high gradients
perpendicular to flow)
(b) Small to intermediate size deep-sea fan (Falknis-Tasna Nappes)
Limited number of lateral sediment sources suggests existence of slope basins
that intercept access routes of turbidity currents
Lack of contemporaneous subduction:
Lack of volcanism and volcaniclastic detritus
Continuous sedimentation without major diastrophism for 70 Ma
Occurrence of slivers of basic and ultrabasic rocks (ophiolite suite) N and S of Flysch Zone
488 R. Hesse
row width, water depth below the calcite com- cided with tectonic activity in the flysch trench
pensation level), association of depositional proposed by the possible widening and partial
environments (basin plain plus relatively small downfaulting in Reiselsberg time. This might
deep-sea fans) and limited number of lateral reflect a pulse of subduction--the only one
sediment sources, tectonic behaviour (lack of detected for the 50-70 Ma life-span of the
any significant syndepositional deformation), flysch trench. At the same time the postulated
lack of contemporaneous volcanism, and Falknis fan ceased to issue sediment to the
occurrence of possible relics of oceanic crust in basin plain. It was apparently cut off from its
adjacent tectonic units are in accordance with a detrital sediment source either due to uplift or
deep-sea trench not associated with subduction. due to formation of a barrier blocking sediment
This analogy first suggested by Hsfi (1972) and supply or (most likely) due to submergence of
analysed here in detail implies Late Jurassic- the shelf during the Cenomanian transgression.
earliest Cretaceous formation by subduction of In Late Cretaceous time the Falknis area re-
a trench which then became dormant for a ceived pelagic sediment. The coarse clastic sedi-
period of 50-70 Ma from Barremian-Aptian to ments of the Reiselsberg Sandstone did not
Palaeocene or mid-Eocene time. reach the Falknis area because of its elevated
Although this type of deep-sea trench is not position (as a fan). Mass flows transporting
very common, the Puerto Rico Trench provides these sediments westward bypassed the Falknis
a modern example. It is 8000 m deep, has a area and followed a more northern route sup-
more than 400 km long and up to 30 km wide plying Reiselsberg Sandstone as far west as
basin plain with turbidite layers that are con- the Wildhaus Syncline in eastern Switzerland
tinuous for 200 km (Conolly & Ewing 1967). (Forrer 1946).
The trench fill of horizontally stratified sedi- The uniform proximal character of the
ments is up to 1.7 km thick (Ewing & Ewing Reiselsberg Sandstone over a distance of at
1962). In the east half of the trench the basin least 200 km parallel strike could be due to the
plain branches into several arms with a com- presence of several lateral sediment sources
bined width of 60 km including the intervening (von Rad 1973). Alternatively, it might be due
ridge (Fig. 11). At the point of branching, to a single stationary source on the trench wall
turbidity currents coming from the principal whic.h issued sediment to the trench floor while
source (Mona Canyon) at the western end of this was moving past the source like a conveyor
the main plain might theoretically turn south- belt--a situation which has been implied for
ward toward the Puerto Rican shelf, i.e. toward fans located on the San Andreas Fault in Cali-
the source area. A similar situation could be fornia (Nelson & Nilsen 1974). This would be
envisaged to interpret Faupl's (1975) north-to- an attractive possibility, since it would imply that
south palaeocurrent directions in the Kaumber- the flysch trench was in fact the site of (large-
ger Beds of the Vienna Woods without having scale) strike-slip motion. In this case the
to resort to an internal source area within the Reiselsberg Sandstone would consist of imbri-
Flysch Zone. The limited number of sources cate diachronous sediment lobes rather than
supplying sediment to the main plain of the synchronous lobes that existed separately side-
Puerto Rico Trench (Fig. 11) is typical for by-side but coalesced with time. With the avail-
modern trenches, which usually have very few able stratigraphic tools at present it will hardly
access routes for turbidity currents to reach the be possible to differentiate between the two
trench bottom. This is the result of the develop- possibilities.
ment of tectonic ridges and ramps on modern
trench slopes, which intercept the downslope
movement of gravity driven sediment flows Cretaceous palaeogeography of the
trapping them in slope basins. The Puerto Rico northern margin of the Tethyan Sea in
Trench was in all likelihood created by pre- the Alpine-Carpathian sector and
Oligocene subduction, but present-day seismic subduction models
activity indicates sinistral strike-slip motion
(Molnar & Sykes 1969) along a major transform Superficially the East Alps and the Carpathians
fault zone running through the trench slope and are separated by the depression of the Vienna
forming the northern boundary of the Carib- Basin, but this is a young (Miocene) feature.
bean plate. Geologically the Flysch Zone of the East Alps
It is of interest that tectonic activity in the and the Carpathian Flysch form a continuous
source area postulated to derive the terrigenous belt as shown by stratigraphic and facies rela-
detritus of the Reiselsberg Sandstone (at a time tionships (Prey 1960, 1965; Tollmann 1969).
of rapid global sea-level rise) may have coin- The Carpathian flysch consists of two seg-
 Flysch Zone o f East Alps and Carpathians 489
ments--the West Carpathian segment compris- 150 km by a broad zone of Neogene calcalka-
ing three separate troughs filled with more or line volcanic rocks which represent one of the
less synchronous turbidite sequences of Upper largest volcanic accumulations in Cenozoic
Jurassic-Lower Cretaceous to Oligocene age, Europe. Extrusion of the volcanics coincides
and the East Carpathian segment representing with the closure of the flysch troughs in the
a diachronous succession of troughs which de- West Carpathians and with diastrophism in the
crease in age from west to east (Contescu 1974). youngest, outermost depositional trough of
In particular, the middle of the three West the East Carpathians.
Carpathian flysch troughs, the Silesian Trough It appears justified, therefore, to assume that
(Table 1), is linked to the Flysch Zone of the the deep-sea trench, which received the flysch
East Alps, whereas the two others, the Skole sediments of the East Alps, continued as a deep
Trough in the north and the Magura Trough in furrow around the Carpathian arc dating at
the south do not seem to have direct equiva- least from the Early Cretaceous, changing in
lents in the East Alps. However, the Pieniny character from a conservative to a convergent
Klippen Belt which borders the Magura Trough plate boundary at the West-East Carpathian
(and thus the Carpathian flysch as a whole) in transition. It was only in Miocene time that the
the south, might be traced laterally into the West Carpathian segment became convergent
Tiergarten Klippen-Zone of the Vienna Woods as well, whereas in the East Alps this change
(Faupl, pers. comm.). More southerly tectonic occurred after the Palaeocene (probably in late
units of the Carpathians such as the Kri~na- Eocene---early Oligocene time). In the West
Nappe, the Veporids, the Cho6-Nappe and the Carpathians the Miocene subduction zone
Gemerids, which comprise older Mesozoic sedi- apparently reached depths sufficient for magma
ments and metamorphic rocks, have been generation by partial melting as evidenced by the
correlated with the Lower, Middle and Upper vast amounts of Neogene calcalkaline volcanics
Austroalpine Nappes, respectively, of the Alps (Bleahu et al. 1973). The deep-sea trench inter-
(e.g. Tollmann 1969; Stegena et al. 1975). Of pretation for the East Alpine--West Car-
the structural and palaeogeographic elements pathian flysch given on the basis of sedimento-
north of and under the Carpathian Flysch the logical evidence from the Alps and the island-
Subsilesian Nappe is considered an equivalent arc interpretation for the Carpathians by
of the Ultrahelvetic Buntmergel Formation in Bleahu et al. (1973) are mutually supporting if
Austria (e.g. Ksiazkiewicz 1956). one keeps in mind that the trench became
Like the Cretaceous-Palaeogene flysch of the eliminated as a topographic feature in the East
East Alps the West Carpathian flysch sequ- Alps--West Carpathians at the time when is-
ences are not interrupted by major diastrophic land arc volcanism started in the West Car-
events or hiatuses except one at the Cretaceous- pathians. In the East Alps evidence for a volca-
Tertiary boundary in the Magura Trough whose nic arc accompanying the post-mid-Eocene sub-
southernmost portion was stripped off its base- duction zone is lacking, except for mid-Tertiary
ment and was displaced southward into the volcanism in the South Alps (Dietrich 1976a,b).
Pieniny Klippen Belt (Birkenmajer 1970). Subduction may not have been deep enough or,
Quite unlike the East Alps and West Car- alternatively, the volcanic products may have
pathians the East Carpathian flysch experi- been lost in tectonic sinks without leaving a
enced tectonic deformation from Albian to trace. In the East Carpathians the subduction
Miocene time. Both the development of flysch zone seems to have been active since the later
troughs and subsequent diastrophism migrate part of the Early Cretaceous as suggested by the
eastward toward the convex side of the arc ages of diastrophism, but active volcanism is
(Contescu 1974); a well-expressed diastrophic not documented until Neogene time either (ex-
polarity which is clear evidence for subduction. cept for some Upper Jurassic volcanism).
(For a similar diachronous sequence of flysch Again, this must not necessarily mean that
deposits in North Africa and Sicily see Wezel subduction did not generate volcanism as the
1973.) Thus the Baraolt Nappe is affected in example of the Upper Eocene-Lower Oli-
Aptian time; the Ceahlau Nappe in Albian- gocene Tavayannaz Sandstone in the
Cenomanian (-Turonian) time; the Bobu, Tele- Ultrahelvetic of the West Alps shows--a volca-
ajen, Macla and Audia Nappes after the end of nogenic turbidite sequence composed of andesi-
the Turonian; and the Tarcau and Marginal tic detritus for which not the slightest trace of a
Folds Nappes in Late Oligocene and Early volcanic source area exists. Similarly,
Miocene times, respectively (Stefanescu 1980). Palaeogene and older volcanic arcs of the East
The Carpathian flysch belt is accompanied on Carpathians may have been lost in tectonic
its inner, concave side at a distance of up to sinks.
490 R. Hesse

Existence of this Cretaceous-Palaeogene eastern exposed end of this trough in the Lower
East Alpine-Carpathian deep-sea trench with Engadine Window ophiolites occur between the
an actively subducting eastern segment and Biindner Schiefer complex (North Penninic
western segment lacking subduction poses the Valais Trough) and the Tasna Nappe (Middle
question of how close to the European conti- Penninic). Thus the Valais Trough to the north
nent this feature was in Cretaceous time, in of the Flysch trench appears to have been
other words--how wide were the intervening foored, at least in part, by oceanic crust.
oceanic areas? Although one can only speculate Realizing that the entire Pr~itigau Flysch is part
on this question, intuitively I would expect they of this trough, 50 km is an absolute minimum
were very broad. This is based on the fact that for the width of the area between the Flysch
we do not see continental margin deposits of trench and the southern continental margin of
the southern European margin of any reason- Europe in the Rhine Valley region. Further
able thickness exposed in the East Alps (except east, i.e. east of the Engadine Window, no
in the Allga6i) or the Carpathians. We have to equivalents of the North Penninic Valais
go to the West Alps to see these sequences in Trough are known. It is possible that the post-
the Helvetic Nappes and the Ultrahelveticum in ulated oceanic area between the southern Euro-
Switzerland or in the Zone Dauphinoise and pean continental margin and the East Alpine-
the Zone Ultradauphinoise in France. Sequ- Carpathian deep-sea trench, of which almost all
ences of similar thickness therefore should be traces have been lost east of the Engadine
present in the subsurface in the east, under the Window (except the above-mentioned pebbles
cover of thrust sheets and nappes, unless the and boulders in the Ultrahelveticum), may have
margin was starved. been wider in the Alpine sector than in the
It has been variously suggested that northern Carpathian sector, because the Carpathian
source areas supplied detritus to the Alpine- flysch belt is several times as wide as the East
Carpathian Flysch trench (e.g. Dzulynski et al. Alpine flysch belt, perhaps at the expense of
1959; Freimoser 1973, a.o.). In the West Car- this postulated oceanic area north of the
pathians submarine canyons have been de- trench--but this is mere speculation.
tected in drill holes beneath the Carpathian South of the Cretaceous-Palaeogene flysch
Flysch nappes (Picha 1974); these undoubtedly trench a zone of islands and shallow platforms
supplied clastic sediment southward. But has it existed which includes rocks of the Sulzfluh-
ever been proved that this material reached the Nappe, possibly the Jurassic Hochstegen Facies
trench? There is evidence it may not have. in the Tauern Window (Tollmann 1964) and the
Why, for example did none of the coarse ultra- Brianqonnais Platform. Those parts of the plat-
mafic and mafic detritus observed in the form that are preserved today seemed to be
Ultrahelveticum in Bavaria (Dietrich & Franz submerged in Early Cretaceous time. However,
1976) find its way into the Flysch trench? Its the abundance of detri~al quartz (and felspar) in
occurrence suggests exposure of (at least) sliv- the Gault Formation indicates that the granitic-
ers of oceanic crust north of the flysch trench, gneissic basement of the platform was ex-
but this distinctive material does not appear in posed somewhere. The exposure of silicic base-
the trench fill sediments. Where some chrom- ment of the Middle Penninic platform is already
spinel appears in the Vienna Woods Flysch visible in the coarse granitic detritus of the
(Faupl 1975), it can probably be derived from a Upper Jurassic breccia formations of the Falk-
submarine ridge within the Flysch belt itself or nis Nappe.
may even have a southern source, despite the South of this Middle Penninic belt of plat-
palaeocurrent directions which point southward forms and islands that were floored by con-
(as discussed elsewhere). tinental crust the South Penninic Piemontais
Correlation of the Gault Formation between trough (or Schistes lustr6s trough) followed:
the Falknis and Tasna Nappes and the Flysch this is the main oceanic realm of the Alpine
Zone established for the Gault Formation Tethys. This oceanic area was successively
places the Flysch Zone in a palaeogeographic being eliminated from the surface by southward
position north of the Middle Penninic Falknis subduction at the South Alpine-Austro-
and Sulzfluh Nappes and their western equiva- Alpine/South Penninic convergent plate bound-
lent, the Brian~:onnais Platform (Triimpy & ary which acted as the main subduction zone in
Haccard 1969), but south of the North Penninic Cretaceous time. It has long been argued that
Valais Trough, which at its western end (near closure of the South Penninic oceanic area in
Bourg St Marucie in the French Savois Alps) the East Alps was complete by late Campanian
comprises an incomplete Late Cretaceous time when chromspinel disappears as a heavy
ophiolite sequence (Antoine 1971). Also at the mineral from the turbidite sands in the Gosau
 Flysch Zone of East Alps and Carpathians 49:
Uttra-
Helvetic I. heir. Z. ,North Penninic Z, Hiddte Penninic Zone Austroatpine Zone
i i Ftysch Zone
.kin ~ ~ Pr~tigau FI. I of East Atps x , _ I(~ssen lleichlnhall

4".-. = ~ , /i .i - -1
I i
Upper Campani!n , ,;
)i It 50km I

i i i
-, iuc
' f i South Pen ninic I Piemontaisel Zorte t ~/
,
i i
i
i
i

J(m I;
Nt i
'
f
r
. . j ,..oo....... ..s
U -- ...... ~ - - --i

Jill ,,:'¢d.'@). L \ "-

Lower Cretaceous - ~7 ";;..¢'.g \@',,

Cenomanian

Fro. 12. Hypothetical palinspastic cross-sections (north-south) through the northern Tethys in the
western East Alps for Early Cretaceous-Cenomanian and late Campanian times (from Hesse & Butt
1976). Depth ranking of the K6ssen and Reichenhall Gosau Basins, the Flysch Zone of the East
Aips, and the Ultrahelvetic flysch relative to the CCL based on chemical and mircopalaeontological
data. No data were available for the North Penninic Pr~itigau Flysch. Horizontal distances are
absolute minimum figures. Note vertical exaggeration. Continental crust: crosses. Oceanic crust:
vertical banding. Mid-slope basins on the southern wall of the flysch trench may be represented by
the Triesen Flysch of Leichtenstein, the Tratenbach Beds of Bavaria or perhaps the Laaber Nappe of
the Vienna Woods.

Basins (e.g. Tollmann 1978), but--as men- tains or Translyvanian Basin, to include major
tioned--chromspinel is still present in Upper parts of the predecessor of the present Panno-
Campanian and Maestrichtian-Danian strata nian Basin before emplacement of the Kri~na,
showing that ultramafic rocks, a l t h o u g h of un- Veporid, Cho~ and Gemerid Nappes. By the
known origin, were still exposed at that time. end of the Cretaceous this former Penninic-
At the same time that subduction proceeded plate sensu lato had been eliminated from the
in the South Penninic realm a second plate surface and the Austro-Alpine/South Alpine
boundary marked by the flysch trench existed realm had been juxtaposed with the southern
between the North Penninic and the Middle upper slope (or forearc) of the trench (Fig. 12).
Penninic realm. It must have been generated as
a topographic feature by an initial phase of
subduction in latest Jurassic-earliest Cretaceous
time. During most of the Cretaceous and The Alpine-Carpathian A r c m a
Palaeogene, however, it acted as a transform Tethyan analogue of the modern West
fault zone, to become the site of subduction Atlantic loop arcs in the Caribbean and
again only after mid-Eocene time. South Scotia Seas
In Cretaceous time the Alpine Tethys is thus
encompassed between two more or less east- If deep-sea trenches are anchored in the asthe-
west trending plate boundaries (not one, as nosphere to some extent by their subduction
Dietrich 1976a, b suggested) which enclosed a zones (Kaula 1975; Tullis 1972)--then the
small plate with both continental (Middle Pen- Alpine-Carpathian Arc might r e p r e s e n t - - i n a
ninic) and oceanic (South Penninic) crustal broad s e n s e - - a n analogue of the present-day
domains between them. During this time the Caribbean or South Scotia Arcs. In these mod-
northern boundary acted as a transform fault, ern loop arcs active volcanism is (more or less)
the southern boundary as a subduction zone. restricted to the eastward convex, west dipping
This plate may have extended eastward as far as subduction zone complexes (Lesser Antilles,
the western boundary of the Apuseni Moun- South Sandwich Islands) whereas the flanks of
492 R. Hesse

the arcs appear to be large transform fault other hand would be 'sheltered' behind the East
zones marked by islands and, notably in the Carpathian subduction zone and would there-
case of the Caribbean, by the Puerto Rico fore be relatively stationary.
deep-sea trench which has been inactive as far The analogy is of course a very general one.
as subduction is concerned probably at least In particular the former southern plate bound-
since the beginning of the Miocene. At these ary is problematical because it is now juxta-
flanks motion is largely strike-slip between the posed with the northern boundary due to the
westward moving American plates (Molnar & almost complete elimination of the intervening
Sykes 1969) and the relatively stationary Carib- plate, which except in the West Alps, is only
bean (Jordan 1975) and (?) South Sandwich exposed in tectonic windows. The evolution of
plates (Forsyth 1975). On the southern flanks of this southern boundary in the Alps was discus-
both arcs slow subduction seems to occur, sed by Trfimpy (1975), and in the Eastern
however, as evidenced by the Tertiary to Re- European Alpine system by Burchfiel (1980).
cent calc-alkaline volcanism in the South Shet- Laubscher (1971) envisaged 300km of right
land Islands and structural deformation in the lateral movement which would have occurred
South Caribbean Basin (Case 1974). along a transform fault zone in the area of the
The analogy implies strike-slip motion be- southern plate boundary. Envisaging the Cre-
tween Europe and the East Alpine-West Car- taceous palaeogeography of the Alpine-
pathian deep-sea trench, which would be sinis- Carpathian region in terms of a reasonable
tral because in Cretaceous time the Eurasian modern analogue may not solve all the prob-
plate can hardly have moved eastward at a time lems but it may help to spur our imagination
when opening in the North Atlantic had not yet and provide some constraints for a more realis-
started. Eastward (absolute) plate motion of tic picture. The geodynamic link between the
the Eurasian plate in the Tertiary is also unlike- three loop-arc structures could be the westward
ly because of the constraints placed on it by the drift of major lithospheric plates between which
opening of the marginal seas in the West Pacific smaller plates remain relatively stationary be-
and the present-day motions of the plates adja- cause they are trapped behind westward dip-
cent to West Pacific subduction zones (Chase ping subduction zones that are anchored in the
1978). The Penninic-plate sensu lato on the asthenosphere.

References
ABERER, F. & BRAUMULER,E. 1958. 0ber Helveti- the north part of the Eastern Alps. Doctorate
kum und Flysch in Raum nOrdlich Salzburg. Thesis. Univ. Ttibingen, Germany. 129 pp (un-.
Mitt. Geol. Ges. Wien, 49, 1-59. pubi.).
ALLEMANN, F. 1956. Geologie des F(irstentums CARON, C., HESSE, R. & KERCKHOVE,C. 1979. Alpine
Liechtenstein. IIl. Teil-Selbstverl. Hist. Ver. helminthoid flysch sequences. Intn. Geol. Cor-
Fiirstentum Liechtenstein, Vaduz, 224 pp. rel. Proj. 105: Continental margins in the Alps.
ANTOINE, P. 1971. La zone de Br~ches de Tarentaise 4th workshop. F%ld guide. Fribourg, Switzer-
entre Bourg-Saint Maurice (Vall~ de l'ls~re) et land. 86 pp.
la frontibre italo-suisse. These, Univ. Grenoble. CARON, C., HESSE, R., KERCKHOVE, C. VAN
367 pp. STUIJVENBERG, J. & TASSI~, N. 1981. Comparison
BLEAHU, M., BOCCALEITI,M., MANEITi • PELTZ, S. preliminaire des flysches ~ helminthoides sur
1973. Neogene Carpathian arc: a continental arc trois transversales des Alpes. Eclog. geol. Helv.
displaying the features of an 'island arc'. J. 74 (in press).
geophys. Res. 78, 5025-32. CASE, J. E. 1974. Major basins along the continental
BIRKENMAJER,K. 1970. Pre-Eocene fold structures in margin of northern South America. In: BURK,
the Pieniny Klippen Belt (Carpathians) of C. A. t~ DRAKE, C. L. (eds). The Geology of
Poland. Stud. geol. Poln. 31, 1-77. Continental Margins, 733-41. Springer-Verlag,
1976. The Carpathian orogen and plate tecto- New York.
nics. Publ. Inst. Geophys. Pol. Acad. Sci. A-2, CHASE, C. G. 1978. Extension behind island arcs and
101, 43-53. motions relative to hot spots. J. geophys. Res.
BURCHFIEL, B. C. 1980. Eastern European Alpine 83, 5385-7.
system and the Carpathian orocline as an exam- CHOUGH, S. K. & HESSE, R. 1980. The Northwest
ple of collision tectonics. Tectonophysics, 63, Atlantic mid-ocean channel of the Labrador Sea:
31-61. III. Head spill vs. body spill deposits from tur-
Burr, A. A. 1977. A comparative study of the deposi- bidity currents on natural levees. J. sediment.
tional environments of Upper Cretaceous rocks in Petrol. 50, 227-34.
 Flysch Z o n e o f E a s t A l p s a n d C a r p a t h i a n s 493

CONNOLLY, J. R. & EWIN~, M. 1967. Sedimentation GROSSHEIM, V. A. 1961. The feasiblity of long dis-
in the Puerto Rico Trench. J. sediment. Petrol. tance, horizon-by-horizon correlation of flysch
37, 44-59. sections ('teleconnection'). Izvestiya Akad.
CONTESCU, L. R. 1974. Geologic history and Nauk. SSSR, (1962), 41-8.
palaeogeography of Eastern Carpathians: Exam- GRON, W., KITTLER, G., LAUER, G., PAPP, A. &
ple of Alpine geosynclinal evolution. Bull. Am. SCHNABEL, W. 1972. Studien in der Unterkreide
Assoc. Petrol. Geol. 58, 2436-76. des Wienerwaldes. Jahrb. Geol. B.-A. Wien.
DEWEY, J. F., PITMAN,W. C., III, RYAN, W. B. F. & 115, 103-85.
BONN1N,J. 1973. Plate tectonics and evolution of GWINNER, M. P. 1972. Geologie der Alpen.
the Alpine system. Bull. geol. Soc. Am. 84, Schweizerbart, Stuttgart. 477 pp.
3137-80. HESSE, R. 1973a. Flysch-Gault und Falknis-Tasna-
DIETRICH,V. J. 1976a. Evolution of the Eastern Alps: Gault (Unterkreide): Schrittweiser Ubergang
a plate tectonics working hypothesis. Geology, 4, v o n d e r distalen zur proximalen Flyschfazies auf
147-52. einer penninischen Trogebene der Alpen.
- - 1976b. Plattentektonik in den Ostalpen--eine Geologica et Palaeontologica, Sonderband, 2, 55
Arbeitshypothese. Geotekton. Forsch. Stuttgart, pp., 12 plates. Elwert-Verlag, Marburg.
5 0 , 1-84. 1973b. Lithostratigraphie, Petrographie und
DIETRICH, V. & FRANZ, U. 1976. Ophiolith-Detritus Entstehungsbedingungen des bayerischen Flys-
in den santonen Gosau-Schichten (N6rdliche ches: Unterkreide. Geologica Bay. 66, 148-222.
Kalkalpen). Geotekton. Forsch. Stuttgart, 50, 1974. Long-distance continuity of turbidites:
85-109. Possible evidence for an Early Cretaceous trench
DZULYNSKI, S., KSIAZKIEWICZ,M. & KUENEN, PH.H. abyssal plain in the East Alps. Bull. geol. Soc.
1959. Turbidites in flysch of the Polish Car- Am. 85, 859-70.
pathians. Bull. geol. Soc. Am. 70, 1089-118. 1975. Turbiditic and non-turbiditic mudstone of
EWING, J. I. & EWING, M. 1962. Reflection profiling Cretaceous flysch sections of the East Alps and
in and around the Puerto Rico Trench. J. geo- other turbidite basins. Sedimentology, 22, 387-
phys. Res. 67, 4729-39. 416
EWING, M. & HEEZEN B. C. 1955. Puerto Rico 1979. Bavarian Alps (Trauchgau-Ammergau).
Trench geophysical data. In: POEDERVAART,A. In: CARON, C., HESSE, R. & KERCKHOVE, C.
(ed.). Crust of the Earth. Spec. Pap. geol. Soc. (eds). Alpine helminthoid flysch sequences. Intn.
Am. 62, 255-68. Geol. Correl. Pro]. 105. Continental margins in
FAUPL, P. 1975. Schwermineralien und Str6mungs- the Alps. 4th workshop, field guide, 65-86,
richtungen aus den Kaumberger Schichten Fribourg, Switzerland.
(Oberkreide) des Wienerwald-Flysches, Nieder- --, & Burr, A. A. 1976. Paleobathymetry of
6sterreich. Neves Jahrb. Geol. Paleontol. Mont- Cretaceous turbidite basins of the East Alps
shofte. 9, 528-40. relative to the calcite compensation level.
1976. Vorkommen und Bedeutung roter Pelite J. Geol. 84, 505-33.
in den Kaumberger Schichten (Oberkreide) des Hs0, J. K. 1972. Alpine flysch in a Mediterranean
Wienerwald-Flysches. Nieder6sterreich. Neves setting. Repts 24th Int. Geol. Congr. 6, 67-74,
Jahrb, Geol. Paleontol. Montshofte. 8, 449-70. Montreal.
1978. Zur rfiumlichen und zeitlichen Entwick- HUENE, R. VON 1974. Modern trench sediments, In:
lung von Breccien-und Turbiditserien in den BURK, C. A. & DRAKE, C. L. (eds). The Geology
Ostalpen. Mitt. Ges. Geol. Bergbaustud. Osterr. of Continental Margins, 207-11, Springer-
2 5 , 81-110. Verlag, New York.
, GRUN, W., LAUER, G., PAPP, A., SCHNABEL, , AUBOUIN, J., AZEMA, J., BLACKINTON, G.,
W. & STURM, M. 1970. Zur Typisierung der CARTER, J. A., COULBOURN, W. T., COWAN,
Sieveringer Schichten im Flysch des Wiener D. S., CURIALE,J. A., DENGO, C. A., FAAS, R.
Waldes. Jahrb. geol. B.-A. Wien, 113, 73-158. W., HARRISON,W., HESSE, R., HUSSONG,D. M.,
FORRER, M. 1946. Uber den Flysch der 6stlichen LADD, J. W., MUZYLOV, N., SHIKI, T.,
Wildhausermulde. (Vorlaufige Mitteilung). THOMPSON, P. R. d~ WESTBERG, J. 1980. Leg 67:
Eclog. geol. Helv. 39, 132-40. The Deep Sea Drilling Project Mid-America
FORSYTH, D. W. 1975. Fault plane solutions and Trench transect off Guatemala. Bull. geol. Soc.
tectonics of the South Atlantic and Scotia Sea. Am. 91,421-32.
J. geophys. Res. 80, 1429-43. JORDAN, T. H. 1975. The present-day motions of the
FREIMOSER, M. 1973. Zur Stratigraphie, Sediment- Caribbean Plate. J. geophys. Res. 80, 4433-9.
petrographie und Faziesentwicklung der Siidost- KAULA, W. M. 1975. Absolute plate motions by
bayerischen Flyschzone und des Ultrahelveti- boundary velocity minimizations. J. geophys.
kums zwischen Bergen/Obb. und Salzburg. Res. 80, 244-8.
Geologica Bavar. 66, 7-91. KSIAZKIEWICZ, M. 1956. Geology of the Northern
GRIGGS, G. B., CAREY, A. G., JR. • KULM, L. D. Carpathians. Geol. Rdsch. 45, 369-411.
1969. Deep-sea sedimentation and sediment- LAJOIE, J. 1979. Origin of megarhythms in flysch
fauna interaction in Cascadia Channel and on sequences of the Ou6bec Appalachians. Can. J.
Cascadia Abyssal Plain. Deep Sea Res. 16, 157- Earth Sci. 16, 1518-23.
70. LAUBSCHER, H. P. 1971. Das Alpen-Dinariden-
494 R. Hesse

Problem und die Palinspastik der sfidlichen , 1975b. Miocene paleogeography and basin
Tethys. Geol. Rdsch. 60, 813-33. analysis in the Periadriatic Apennines. In:
LEMCKE, K. 1970. Epirogenetische Tendenzen im SQUYRES, C. (ed.). Geology of Italy, 5-11. Pet-
Untergrund und in der Ffillung des Molassebeck- rol. Explor. Soc. Libya.
ens n6rdlich der Alpen. Bull. Ver. Schweiz. - - & VALMORI, E. 1980. Basin-wide turbidites in a
PetroL-Geol. u. Ing. 37, 25-34. Miocene, over-supplied deep-sea plain: a
MENARD, H. W. 1964. Marine Geology of the Pacific. geometrical analysis. Sedimentology, 27, 241-70.
McGraw-Hill, New York. 271 pp. RICHXER, M. 1933. Alter and Stellung der siidbayeris-
MOLNAR, P. & SYKES, L. R. 1969. Tectonics of the chen Flyschzone. Centralbl. Mineral. Geol.
Caribbean and Middle America regions from Palaeontol, 496-508.
focal mechanisms and seismicity. Bull. geol. Soc. SANDULESCU, M. 1973. Essai de reconstitution des
Am. 80, 1639-84. 6.16ments preparoxismaux alpins des Dacides (in-
M u l l , E. & RICcI-LuccHI, F. 1975. Turbidite facies ternides) orientales. Rev. Roumain GOol. GOo-
and facies associations. In: M u l l , E. et al. (eds). phys., G(ograph, Ser. GOolog. 17, 145-56.
Examples of Turbidite Facies and Facies Associa- STEFANESCU, M. 1980. Time of flysch deposition in
tions from Selected Formations of the Northern the Eastern Carpathians. Ann. Inst. G(ol. GOo-
Appenines. 9th Int. Sedimentol. Congress Nice, phys. 56, 133-42.
Field trip Guide All, 21-36. STEGENA, L., GECZY, B. & HORVAXH, F. 1975. Late
NELSON, C. H. & NILSEN, T. H. 1974. Depositional Cenozoic evolution of the Pannonian basin. Tec-
trends of modern and ancient deep-sea fans. In: tonophysics, 26, 71-90.
DoTr, R. H. JR. & SHAVER, R. H. (eds). Modern STEPHAN, W. & HESSE, R. 1966. Erliiuterungen zur
and Ancient Geosynclinal Sedimentation. Spec. Geologischen Karte von Bayern 1:25,000, Blatt
Publ. Soc. econ. Palaeontol. Mineral. Tulsa, 19, Nr. 8236 Tegernsee. Bayerisches Geologisches
69-91. Landesamt Mfinchen, 304 pp.
NIEDERMAYR, G. v. 1966. Beitr~ge zur Sediment- TOLLMANN, A. 1964. Zur Fortsetzung des Briancon-
petrographie des Wienerwald-Flysches. Verh. nais in den Ostalpen. Mitt. geol. Ges. Wien, 57,
geol. Bundesanst. 1/2, 106-41. 469-78.
OBERHAUSER, R. 1968. Beitr~ige zur Kenntnis der - - , 1969. Die tektonische Gliederung des Alpen-
Tektonik und Palaogeographie wfihrend der Karpaten-Bogens. Geologie, 18, 1131-55.
Oberkreide und des Pal~ogens im Ostalpen- , 1978. Plattentektonische Fragen in den Ostal-
raum. Jahrb. geol. Bundesanst. 111, 115-45. pen und der plattentektonische Mechanismus
OBERHAUSER,R. (ed.) 1980. Der Geologische Aufbau des mediterranen Orogens. Mitt. Osterr. geol.
Osterreichs. Springer-Verlag, Berlin. 70(1 pp. Ges. 69, 291-351.
PICHA, F. 1974. Ancient submarine canyons of the TROMPY, R. 1975. Penninic-Austroalpine boundary in
Carpathian miogeosyncline. In: D o n , R. H. JR the Swiss Alps: a presumed former continental
& SHAVER, R. H. (eds). Modern and Ancient margin and its problems. Am. J. Sci. 275A,
Geosynclinal Sedimentation. Spec. Publ. Soc. 209-38.
econ. Paleontol. Mineral. Tulsa, 19, 126-7. TROMPY & HACCARD, D. 1969. R6union extraordi-
PILKEY, 0 . H., LOCKER, S. D. & CLEARY, W. J. 1980. naire de la Soci6t6 G~ologique de France: Les
Comparison of sand-layer geometry on flat floors Grisons. 14 au 21 Septembre 1969. C. r. Somm.
of l0 modern depositional basins. Bull. Am. S(ances Soc. geol. France, 9, 329-96.
Assoc. Petrol. Geol. 64, 841-56. TULLIS, T. E. 1972. Evidence that lithospheric plates
PIPER, D. J. W., VON HUENE, R. & DUNCAN, J. R. act as anchors. Los, 53, 522.
1973. Late Quaternary sedimentation in the ac- WALKER, R. G. 1978. Deep-water sandstone facies
tive eastern Aleutian trench. Geology, 1, 19-22. and ancient submarine fans: Models for explora-
PREY, S. 1960. Gedanken fiber Flysch und Klippen- tion for stratigraphic traps. Bull. Am. Assoc.
zonen in Osterreich anl~if$1ich einer Exkursion in Petrol. Geol. 62, 933-66.
die polnischen Karpathen. Verh. geol. Bunde- WArtS, A. B. & STECKLER, M. S. 1979. Subsidence
sanst. 19%214. and eustasy at the continental margin of eastern
PREY, S. 1965. Vergleichende Betrachtungen fiber North America. In: TALWAN1, M., HAY, W. &
Westkarpaten und Ostalpen im Anschlul3 an RYAN, W. B. F. (eds). Deep Drilling Results
Exkursionen in die Westkarpaten. Verh. geol. in the Atlantic Ocean: continental margins
Bundesanst. 1/2, 69-107. and paleoenvironment. Am. geophys. Union,
RAD, U. VON 1964. Mineralbestand und Ablagerungs- M. Ewing, Ser. 3, 218-34.
bedingungen der Flyschsedimente im Allgiiu. WEZEL, F. C. 1973. Diacronismo degli eventi geologi-
Doctoral Thesis, Technical Univ. Mfinchen. 131 ci Oligo-Miocenici nelle Maghrebidi. Riv. Miner-
pp. (plus appendix) (unpubl.). al. Sicil. 24, 219-32.
--, 1973. Zur Sedimentologie und Fazies des WIESENEDER, H. 1967. Zur Petrologie der ostalpinen
Allg~iuer Flysches. Geologica Bay. 66, 92-147. Flyschzone. Geol. Rdsch. 56, 227-40.
RICCI-LUCCHI, F. 1975a. Depositional cycles in two WOLEXZ, G. 1967. Schwermineralvergesellschaf-
turbidite formations o f northern Apennines tungen aus ostalpinen Sedimentationsbecken der
(Italy). J. sediment. Petrol. 45, 3-43. Kreidezeit. Geol. Rdsch. 5 6 , 308-20.

REINHARD HESSE, McGill University, Department of Geological sciences, 3450

University Street, Montreal, Quebec H3A 2A7, Canada and Technische
Universit~it, Miinchen, Federal Republic of Germany.
 The anatomy of a Lower Palaeozoic accretionary forearc: the
Southern Uplands of Scotland

J. K. Leggett, W. S. McKerrow & D. M. Casey

SUMMARY: The Southern Uplands of Scotland consists of an accretionary prism which
had a prolonged (over 45 Ma) development off the SE coast of ancient North America
(Laurentia) during the late Ordovician and Silurian. To the north, a forearc basin
occupied the Midland Valley. An arc massif of older metamorphic rocks in the Grampian
Highlands, and capping calc-alkaline volcanics, supplied much of the sediment to the
trench. In its early development (late Ordovician) the accretionary complex incorporated
slivers of ocean-floor material as well as thick turbidites into a steep lower trench slope.
Later, in the Silurian, a trench slope break emerged shedding sediment north into
the Midland Valley forearc basin. By that time only turbidites and black shales were
being accreted. The Southern Uplands is dominated by coherent strata, and, despite
intense imbrication, is devoid of m61ange. We conclude that the accretionary prism was
the result of high sediment input to the trench and very slow, oblique, subduction of the
Iapetus Ocean eastwards below Scotland. D6collement surfaces developed during accre-
tion preferentially utilised a highly incompetent black shale layer near the base of the
subducting sequence. Prior to the inception of accretion (in the Caradoc), we postulate
sediment subduction, and possibly subduction erosion, during the Cambrian and early
Ordovician.

The Southern Uplands is a c. 12,000 km 2 area distribution of arc-type igneous rocks and ocean
of southern Scotland, comprising imbricated floor-type rocks (such as ophiolites) along the
and folded Ordovician and Silurian rocks of orogen. The suture along which the ocean
deep-water origin. Recent reconstructions sug- closed is recognised chiefly by contrasts on
gest that the area, and its continuation in either side in the faunas, volcanic polarities,
Ireland, is a fossil subduction complex pre- structural styles, sedimentary histories and
served on the southern margin of Laurentia basement characteristics. In the British Isles the
(Early Palaeozoic North America) which suture runs SW to N E from the Shannon estu-
resulted from accretion above an oceanic ary in western Ireland through the Solway Firth
plate subducting to the N W (Mitchell & south of the Southern Uplands, probably curv-
McKervow 1975; McKerrow et al. 1977; Leggett ing north under the present North Sea to run
et al. 1979a, b; Weir 1979). In this paper we between Norway and Shetland (Phillips et al.
reconstruct the sedimentary and tectonic his- 1976; Thirlwall 1981a). Early Palaeozoic Bri-
tory of this fossil forearc from Lower tain N W of this line formed part of Laurentia
Palaeozoic regional geology; we then review (Fig. 1). Differences between individual tran-
the evidence for an accretionary origin for the sects across the Appalachian-Caledonian
Southern Uplands, extending our comparisons orogen suggest that the continental margins of
with modern accretionary forearcs more than the Iapetus were irregular and that individual
we have hitherto. portion of the northwestern Iapetus margin.
Dewey & Kidd 1974). In Fig. 1 we show an
interpretation of the late Caradoc palaeo-
geography of the New England to G r e e n l a n d
The reconstructed Northern Iapetus portion of the north-western Iapetus margin.
active margin in the British Isles Evidence of a magmatic arc in Scotland indi-
(mid-Ordovician to Early Devonian) cates that subduction was occurring during (also
before and after) the Ordovician under the
Deformation and metamorphism within the British sector of the Laurentian margin (Phillips
Appalachian-Caledonian orogen are attributed et al. 1976)). There is no evidence for subduc-
by most geologists to tectonic events associated tion under the Appalachian or Greenland sec-
with the closure of an Early Palaeozoic ocean, tors (e.g. Henriksen & Higgins 1976; Williams
the Iapetus (e.g. numerous references in Harris 1978). The British sector may have been sepa-
et al. 1979). Evidence for the existence of this rated from these other sectors by transform
ocean comes from faunal provinciality, and the faults (Fig. 1; McKerrow, in press).

495
496 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey

i;,~i
!ii ~r"'
ill ili ['
I [ i! Carbonates
'~ ~ ~- ~ ~ 111 [l 't'-i~.~'/ [~] Clastics

~ L./" .... l @ ' ~ i !i !' , . ' ~ / ~ Land

...... -

,,u, i.,,j~okm -~

FIG. !. Possible configurations of the southeastern Laurentian margin from Long Island (LI) to
Greenland in late Ordovician times (late Caradoc) after McKerrow (in press). The trench along the
NW subducting British sector is shown with a thick line, and is probably separated from the inactive
Greenland and North American sectors by transform faults. Carbonates (including shallow marine
siliciclastic strata) demark the stable Laurentian shelf. Land areas along the margin represent
deformed terranes produced during the Taconic and Grampian orogenies. It has been suggested that
Taconic deformation was produced in the mid-Ordovocian by impingement of east-facing intra-
oceanic arcs (BH: Bronson Hill, T: Tetagouche, LB: Lushs Bight) on the Laurentian margin
(McKerrow in press, and references therein). Grampian deformation may also have been produced
by impingement of a volcanic arc on the Laurentian margin, in latest Cambrian times (McKerrow in
press, and see section on pre-middle Ordovician history of British sector). Configuration of
southeastern lapetus margin also shown; scale does not refer to width of the ocean, which is
unknown. TM: Tactonic Mountains, MB: Matepedia Basin, LR: Long Range, Newfoundland, GH:
GrampianHighlands.

Fl(;. 2. Caledonian (pre-Early Devonian) geological provinces of northern British Isles, and
interpretation in terms of northwards subduction frorn line of lapetus suture. Outcrops too small to
show ornament clearly are generalized. Major faults are dashed where continuation is inferred.
AccreAonarv complex: NB: Northern Belt, CB: Central Belt, SB: Southern Belt. Faults: KF:
Kingledores Fault, OBF: Orlock Bridge Fault, SGSZ: Slieve Glah shear zone. Places: SA: Slieve
Aughty: SBh: Slieve Bernagh: SE: Shannon estuary.
Upper slope/forearc basin: basernent only exposed in the SW Midland Valley, consisting of the
Arenig Balhmtrae ophiolite complex (BO), which is probably allochthonous on Precambrian
crystalline crystalline basement (see text). A slope/shelf succession (Llanvirn to late Silurian)
overlies the ophiolite uncomfortably, this succession and inliers to the NE (containing Llandovery to
Early Dewmian strata) record the history of a forearc basin formed behind the growing accretionary
complex south of the Southern Upland Fault.
Volcanic arc." dashed ornament is mid-Ordovician basalt, andesite and rhyolite of the Tyrone igneous
complex (T) and Charlestown inlier (C): more diffuse dashed ornament is the South Mayo Trough, a
thick Ordovician-Silurian sequence including early Ordovician basalts, interpreted as an intra-arc rift
zone (Ryan & Archer 1977). Solid ornament is Early Dewmian f?in part late Silurian-Thirlwall
1981a) andesite. OH: Ochil Hills: OF: Oehil Fault. Possible subduction-related granitoid intrusions
in metamorphic frontal arc/arc massif not shown.
Frontal arc~arc rnasstJ: most rocks herein were deformed prior to the accretionary phase in latest
Cambrian; structural grain shown (highly illustrative). Relation of Great Glen Fault, Leannan Fault
(LnF) and Leek Fault (LkF) and of Moine Thrust and Loch Skerrols Thrust (LST), from Johnson et
al. (1979). Note position of (,'onnemara Dalradian (CD) with respect to putative continuation of
Southern Upland Fault, and possible equivalents of basalts, cherts, slates and greywackes of the
South Connemara Group (SCG) to Ordovician of Northern Bclt ,of Southern Uplands. For
discussion of ophiolitic slivers of the Highland Boundary Complex (HBC) see text.
 NORTH WEST
Wenlock CALEDONIAN N O R T H - WEST
4'
Llandovery ACCRETIONARY C O M P L E X FORELAND " HIGHLANDS
Ordovician

Ordovician & Silurian FOREARC BASIN /

~'~ Ophioliticrocks
L. Devonian
MAGMATIC ARC
Ordovician
M e t a m o r p h i c frontal arc and arc massif
(Dalradian S u p e r g r o u p etc.)
GRAMPIAN HIGHLANDS

0 100 K m
I I

South Mayo Lk
Trough
hland Boundar
z
t'%

SCG/ M I D L A N D VALLEY
t~
a
Upland Fault
: ~ z - ~
IRELAND
INLIERS~A LONGFORD DOWN • o °, .,/ SOUTHERN UPLANDS
MASSIF

)SBh
IA PETU'-S' SUTURE

4~
"-4
 ...~J~v
i MIDLAND VALLEY l"
/
!
i ir;
:i IDa-'" El ,~ " ~ X
!i
GAF

,?.L~o .~ /

/ f
! ! '
I TL L
--K

i ~ Granites and Granodiorites

-i] upper Palaeozoic and Triassic

F4 ~,y/ '~ Upper Slope / ]
O ~¢ F...... B.... r
i: i Lower
1 Accretionary
~-~ Complex ~ Palaeozoic
::::i:iCG!!!i!iiiiii~i~il Ballanttae Ophiolite i
[] Sequence r~
]1
, 10 15km
lbm,L~s
F
 Anatomy of a Lower Palaeozoic accretionary forearc 499

Since Dewey's (1969, 1971) formative plate cian), County Tyrone (Caradoc), the Midland
tectonic reconstruction numerous authors have Valley and parts of the Grampians (?late Silu-
supported northwards subduction under Scot- rian and Early Devonian), and by certain con-
land (Mitchell & McKerrow 1975 and refer- t e m p o r a n e o u s granitoids in the G r a m p i a n
ences therein), and we can divide northern Highlands (early Ordovician to Early D e v o -
Britain into structural regions (Fig. 2) using the nian). The British portion of the Appalachian-
terminology of Dickinson & Seely (1979). An Caledonian orogen should be viewed as just
accretionary complex (mid-Ordovician-late one transect across the Laurentian margin;
Silurian) occupies the Southern U p l a n d s - - even within the small area of the British Isles
Longford D o w n Massif--Central Ireland In- some geological provinces are discontinuous
liers. A forearc basin (late Ordovician--Early (e.g. South Mayo Trough and Midland Valley
Devonian) occupies the southern part of the forearc basin, Fig. 2).
Midland Valley. A metamorphic frontal arc/arc
massif (pre-early Ordovician) occurs in the Accretionary complex
Grampian Highlands and analogous areas in
Ireland. A magmatic arc is represented by Interpretation of the Southern Uplands as an
volcanic rocks in South Mayo (early Ordovi- accretionary complex (Leggett et al. 1979a)

FIG. 3. Geological map of the Southern Uplands, updated from Leggett et al. (1979a, b).
Major faults: shown as thick lines where mapped (sources below), dashed where continuations
are inferred from Geological Survey Maps, dashed with a question mark where continuation of strike
fault is uncertain or where a wrench fault has been deduced from displacements in outcrops of
basalt-chert-graptolitic shale (basal lithologies). Black areas are imbricate zones of basal lithologies;
almost all of these come from mapped areas in source references below, but some inferred from belts
of outcrop of basal lithologies on Geological Survey maps.
Stratigraphic sequences shown in Fig. 4 are located by numbers 1 to 10 and I to X. Two traverses are
shown, across the central Southern Uplands and the west coast. Correlative sequences shown using
roman numerals on west coast; these can be used to trace some of the major reverse faults between
the two traverses. For over 80 years the Southern Uplands has been divided into three belts (Peach &
Home 1899). The Northern Belt corresponds to our tracts 1,2, and 3; the Central Belt to tracts 4--8;
and the Southern Belt to tracts 9-10.
Structuralprofiles shown in Figs 8-11 are located by double lines and bold letters. Composite profiles
are labelled at both ends.
Place names (towns and villages shown with an asterisk): Af Glen Afton dam, B Biggar, Ba
Ballantrae, BH Bennane Head, BHd Burrow Head, C. Coulter, CB Clanyard Bay, CI Craighead
Inlier, Cm Crossmichael, CmI Carmichael Inlier, Co B Coldingham Bay, CP Corsewall Point, Cr
Craigmichan, D. Dumfries, DB Drumbreddan Bay, DL Dobbs Linn, E. Ettrickbridgend, Ed
Eddleston. EI Eastfield Inlier, Ey Eyemouth, FB Float Bay, G Girvan, Gk Glenkiln, GI Glenluce,
GoF Gatehouse of Fleet, GPt Gipsy Point, H Hawick, Ha Hartfell, HHI Hagshaw Hills Inlier, I
Innerleithen, K Kirkcudbright, L Leadburn, La Lauder, LH Lammermuir Hills, Lm Langholm, Lw
Lesmahagow, LwI Lesmahagow Inlier, M. Moffat, MB Morroch Bay, Me Melrose, MG Mull of
Galloway, MH Moorfoot Hills, Mo Moniaive, N Noblehouse, ND Nithsdale, NS Newton Stewart,
P Portayew, Pe Peebles, Phi Pentland Hills Inlier, Pp Portpatrick, Ps Portslogan, R. Riccarton
Junction, S Stobo, SP Siccar Point, Sq Sanquhar, Sn Straiton, St Stranraer, SV Stinchar Valley, T
Foot of Talla Water (Tweedsmuir, Talla region), Th Thornhill, W Whithorn, WH Wrae Hill.
Faults: BF Bush Fault, B1F Berryfell Fault, BL Blairbuies Line, CF Carcow Fault, CPF
Carmichael-Pentland Fault, CT Culcronchie Thrust, DCF Dove Cove Fault, DF Drumblair Fault,
EVF Ettrick Valley Fault, FL Fardingmullach Line, GAF Glen App Fault, GBF Gillespie Burn
Fault, GF Grassfield Fault, GRSB Glan Reif Shatter Belt, HBF Harwood Burn Fault, HF Hyndlee
Fault, HL Hawick Line, HtF Hermitage Fault, HtL Hartfell Line, KF Kingledores Faults, KT
Killantringan Thrust, LF Langlee Fault, LKF Loch Ken Fault, LL Leadhills Line, LRF Loch Ryan
Fault, MSB Mosspaul Shatter Belt, MVF Moffat Valley Fault, PNT Pool Ness Thrust, PT Pibble
Thrust, RL Riccarton Line, SF Stinchar Fault, SUF Southern Upland Fault, TdF Talmond Fault, TF
Tudhope Fault, TL Talnotry Line, WF Wolfhopelee Fault, YLF Yad Linn Fault.
Devonian granites: CCG Cairnsmore of Carsphairn, CFG Cairnsmore of Fleet, CG Criffel, LDG
Loch Doon, SG Spango.
Sources: Clarkson et al. (1975); Cook & Weir (1979); Craig & Walton (1959); Floyd (1976); Fyfe &
Weir (1976); Gordon (1962); Kelling (1961); Leggett et al. (1979a, b); Lumsden et al. (1967); Peach
&Horne (1899); Rust (1965); Toghill (1970b); Walton (1955, 1956, 1961); Warren (1964); Weir
(1968, 1977); Welsh (1964); Williams (1962).
 SOUTHERN BELT - - - ~ L~
NORTHERN BELT - CENTRAL BELT
RL lOa lOb
S Kirk- Riccarton
EVFI F
Ma 1 ~ 7 .~--~ :'~PPBB~
WENLOCK t ...... WENLOCK
riccartonensis ,
--'-I1,L 6 M VIiF Craig- III Ettnck-
brtdg
8 end =II cudbrignt
.E-'-~ .[_.~ RGq~'~-~
B~'-~PBB murchisoni
I centrifugus
426 I tartfell I ~ I ~,>~__L ~----~t~( ~ ~?~[-- ~(~B4 ( ~ i crenulata
_,ojiestoniensis
crlspus
Talla ~ ,~/,~ ~ ~ turriculatus
LLANDOVERY sedgwickii
convolutus
gregarius
OF 2 I Twe3d. * , cyphus
vesiculosus
Ab,ngtonl da !~.~ . ~ ~ ~---- _~ ~ ~ ~ acurninatus
persculptus
438 anceps
ASHGILL complanatus '
445 linearis
clingani
wilsoni
CARADOC peltifer
cjracilis
467
LLANDEILO "X" rn
479 -- -- "f3r . S.Whitho
LLANVIRN ,? __L_ = I~> lundgreni
489 ~ N Whlthorn ~_R,~ I ell . . . . I

GBF "V centrlfugus WENLOCK ~ U S q

IS Glenlu~e c renulata J LLANDOVERY l crenulata I
~r)es_,_o~,e~
c r,spus
SUF turrlculatus
U,,E, /\ N Glenluce _/2~J LEGEND
sedgwJck,
convOIUTUS j Upper slope sediments
gregarlu~
.~'~ 'Sheet-sand' turbidite facies with 1
Girvan/ / \ KT cypnus 1 ~k.~ predominant palaeocurrent direction, s/ c~
13allantrae / \ I Port veSlCUl~ /
~="" ~ ~, I Morl acumma~ ~'.~'.~'.~'.~'.~'.~/--~'Flysch rudite' turbidite faciesr with r.~
L L A N D O V E R Y I. . . . I persculptu~ ~(,~ predominant palaeocurrent dl ect ioq/sr Greywacke
IMHF~i // '~~' "IT I
- - I - ~ % DCF T GAF I - - anceps j _~ Thin-bedded turbidite facies J
ASHGILL ~ I 1 I I compla nalus .J
- - ~,,%~ I • I Portslogan linearis____~_j Grey mudstones } Moffaf Shale Group
[L,%'~~, I Corsewall/ I Black shales
5~ I GlenApp I F"~,~ i cl,ngam .~ Radtolarian cherts
CARAOOCDM~'~ I ~D~'~ ~-~ I L-~/,A, ) * wllsonl '/
peltrfer ~ Basalt
LLAN"E'LOl:'-'-'k"L
~ I ~"~; l ~ Obduchon of Ballantrae ophiolite tnorth of Stinchar Fault
LLANVIRN~BO"~ I graptolite control
conodont control
ARENIG ~"~""
 Anatomy of a Lower Palaeozoic accretionary forearc 501

hinges on four lines of evidence: sediments and Wenlock terrestrial sediments.

(i) The Iapetus suture immediately to the Rocks older than Silurian are only seen at
south of the area and the contempor- Girvan, where an Ordovician marine succes-
aneous magmatic arc to the north indi- sion, interpreted as an upper slope and shelf
cate a forearc setting in mid-Ordovician sequence, rests unconformably on the early
through late Silurian times. Ordovician Ballantrae ophiolite complex (Fig.
(ii) Rocks preserved in the area are inter- 4).
preted as ocean floor and trench de- Sediments deposited during the latest Cara-
posits (basalts, metalliferous sediments, doc indicate a NW-inclined palaeoslope (Ing-
cherts, black shales, and greywackes). ham 1978, p. 170). This may be related to the
(iii) The structural arrangement of the rocks first e m e r g e n c e of a trench slope break in the
preserved resembles the accretionary Southern Uplands accretionary complex. En-
prism m o d e l of Seely et al. (1974). suing sediments in the Girvan area are varied
(iv) Deformation appears to have occurred Llandovery marine strata up to c. 3 km thick
progressively during the evolution of the (Fig. 4). These are largely d e e p e r water shelf
area, rather than in a climactic event. deposits having both southerly and northerly
provenance (Cocks & Toghill 1973).
Foreare basin
Llandovery strata in inliers to the east of
The Silurian succession in the Girvan area Girvan are turbiditic, and derived almost en-
(Fig. 3) is marine and all inliers NE of Girvan tirely from the south (summary in Leggett
have Silurian successions of Llandovery marine 1980a). Wenlock sediments include clasts of

F1G. 4. Time-stratigraphic diagram showing representative stratigraphic sequences for fault slices in
two traverses across the Southern Uplands from the Southern Upland Fault in the north to the
unconformable Upper Palaeozoic cover in the south. Adapted from Leggett (1980a), and Leggett
et al. (1979b). Time-scale after McKerrow et al. (1980). Numbers in circles and boxes refer to
locations in Fig. 3. Iapetus suture lies c. 20 km to SE of youngest fault slice. Greywackes divided into
three broad facies types (see text), with representative predominant palaeocurrent directions beside
columns. Details of grey mudstones in pelagic facies simplified: thin black seams occur in the Ashgill
grey mudstones, and grey mudstones are intercalated in the upper division of the Moffat Shale (see
Leggett 1978, fig. 3). Graptolite control shown only in turbidite facies: correlations in pelagic facies
accurate to within one graptolite zone unless a question mark is present, in which case horizon is
known to within + 2 or 3 graptolite zones. Otherwise no exact temporal correlations intended. Solid
vertical lines represent faults labelled as in Fig. 3. We follow Clarkson et al. (1975) in grouping the
symmetricus (now termed rigidus), linnarssoni and ellesae (formerly termed rigidus) zones together
as the 'C. ellesae' zone. Lumsden et al. (1967) grouped the same three middle Wenlock zones
together under 'C. linnarssoni" (see Fig. 9). Conodonts in cherts closely associated with some of the
basalts in the Coulter-Noblehouse sequence suggest a Llanvirn-Llandeilo age for the volcanicity
(Lamont & Lindstr6m 1957; Bergstr6m 1971). We have shown the possible presence of Arenig
basalts at the base of the sequence because basaits of this age occur both to the north of the
Coulter-Noblehousetract (in the Ballantrae ophiolite sequence) and to the south (in the Afton-Abington
sequence). Biostratigraphic data within Ballantrae ophiolite sequence and in unconformable cover
indicate emplacement within Arenig or Llanvirn (summary in Bluck 1978).
Lithostratigraphic units: AD Albany division, AF Abbotsford Flags, AG Ardwell Group, BB
Balmae Beds, BaG Barr Group, BG Balclatchie Group, BO Ballantrae ophiolite, CBB Caddroun
Burn Beds, CG Corsewall Group, DG Drummock Group, DMS Downan Mafic Sequence, GaG
Galdenoch Group, GAG Glen App Group, GF Garheugh Formation, GG Gala Group, GPB Gipsy
Point Beds, HR Hawick Rocks, IG Intermediate Group, KF Kilfillan Formation, KG Kirkcolm
Group, LB Lowther Beds, MHF Mulloch Hill Formation, PBB Penchrise Burn Beds, PG
Portpatrick Group, PyG Pyroxenous Group, RCB Raeberry Castle Beds, RF Ross Formation, RG
Riccarton Group, SB Shankend Beds, SCB Stobs Castle Beds, SF Shalloch Formation, TG Tappins
Group, WG Whitehouse Group, WL Wrae Limestone.
Sources for stratigraphic and sedimentological data: upper columns: 1, 2, 3, Peach &Horne (1899);
Eckford & Ritchie (1931); Lamont & Lindstr6m (1957); McKerrow et al. (1977); Leggett et al.
(1979a, b), Leggett (1980b) 4, 5, 6, D. Casey (unpublished data); Peach &Horne (1899); Walton
(1955), 7, 8, Peach & Horne (1899); Toghill (1970a). 9, Peach & Horne (1899); Warren (1963, 1964).
10a, Craig & Walton (1959, 1962); Walton (1968); Clarkson et al. (1975). 10b, Warren (1963, 1964).
Lower columns: 1, Walton (1956, 1961); Kelling (1961, 1962); Williams (1962). 2, 3, Kelling (1961,
1962); Welsh (1964); Kelling & Welsh (1970). 4, 5, Gordon (1962). 9b, 10c, Rust (1965);
Girvan-Ballantrae sequence from Williams (1962); Lamont & Lindstr6m (1957); Bergstr6m (1971);
Ingham (1978).
 502 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey
cleaved greywacke (McGiven 1968), clearly de- in Scotland and along strike in Ireland (Fig. 2)
formed long before any supposed end-Silurian thick late Proterozoic and Cambrian sediments
Caledonian orogeny. The emerging trench (Dalradian Supergroup) were deformed and
slope break at this time must have been fringed metamorphosed in the early Ordovician Gram-
by shallow-water deposits, because commi- pian orogeny about 500 Ma ago (summary in
nuted shelly debris occurs in the Llandovery Johnson et al. 1979). Plate tectonic explanations
turbidites (Rolfe 1961). for the Grampian orogeny are many and varied
Uplift caused the forearc basin to become but almost all relate the orogeny to subduction
emergent in late Llandovery or early Wenlock along the northwestern Iapetus margin. The
times. Terrestrial red beds, including thick allu- earliest stratigraphic sequence in the Southern
vial fan conglomerates, overlie the turbidites Uplands, immediately south of the Southern
disconformably. East of the Girvan area the Upland Fault, contains an N. gracilis Zone
only sedimentary transition between marine fauna (Fig. 4). Accretion must have started
and terrestrial strata occurs in the Lesmahagow after this time. Therefore the Grampian events
inlier (Fig. 3). During the Wenlock, movements were clearly over before accretion commenced
on the ancestral Southern Upland Fault con- in the Southern Uplands. In relation to the
trolled sedimentation in the forearc basin. Allu- mid Ordovician-late Silurian construction of
vial fan complexes prograded northwards into the Southern Uplands accretionary complex the
the basin three times, reflecting movements on Grampian belt forms a joint frontal arc and arc
the Southern Upland Fault or other faults in the massif.
emergent northern part of the accretionary
complex (McGiven 1968; Fig. 6). These may be
Magmatic arc
related to rapid uplift associated with discrete
phases of offscraping in the south. North of the Southern Upland Fault and its
The Ballantrae ophiolite is a complex faulted hidden continuation in Ireland (Fig. 2) arc-type
terrane exposed in the SW corner of the Mid- magmas were erupted at intervals from late
land Valley (Fig. 3). It largely comprises mafic Cambrian through to Early Devonian (summary
volcanic and volcaniclastic rocks; serpentinite in Stillman & Francis 1979). Contemporaneous
belts and local bodies of ultramafic rock, gab- granitoids intrude the late Precambrian--Cam-
bro, dolerite, trondjemite and amphibolite are brian (Grampian) metamorphic rocks and Pro-
also present. Radiolarian cherts, black shales terozoic high-grade gneissic basement of the arc
and debris-flow conglomerates occur above and massif (summary in Brown 1979).
within the mafic sequences. No arc volcanics of definite Silurian age are
Bamford (1979 and references therein) re- exposed, but indirect evidence for arc eruptions
cords a 6-14 km thick layer of rocks with seismic throughout the Ordovician and Silurian comes
velocities of 5 . 8 - 6 . 0 k m s -~, interpreted as from the Southern Uplands: abundant andesitic
Lower Palaeozoic, across the Midland Valley. detritus occurs in Ordovician and Silurian
These are underlain by > 6 . 4 k m s -~ rocks, greywackes (Kelling 1962; Sanders & Morris
interpreted as pre-Caledonian high-grade Pre- 1978; Walton 1956), and numerous
cambrian metamorphic rocks (Bamford 1979). metabentonitic seams occur in pelagic and
This interpretation is strongly supported by hemipelagic sediments (Cameron & Anderson
granulite xenoliths discovered in Upper 1979; Leggett 1979).
Palaeozoic volcanic vents across the Midland By far the most voluminous volcanic rocks
Valley (Graham & Upton 1978). Laboratory exposed in the magmatic arc are calc-alkaline
measurements of velocities do not appear to be extrusives and pyroclastics, mostly andesites
adequate at present to discriminate ophiolitic but also including basalts and rhyolites, inter-
lithologies from Lower Palaeozoic sediments in bedded in Lower Old Red Sandstone (ORS)
the 5.8-6.0 km s -j layer; high proportions of terrestrial deposits of possible Early Devonian
serpentinite in the ophiolites are likely to pro- age (Thirlwall 1981a). In the Ochil Hills (Fig. 2)
vide a further complication (e.g. Powell 1978). these lavas are up to 3 km thick. Systematic
Therefore, we do not know what proportion of changes in Sr, K and other elements in the lavas
the basement of the Midland Valley forearc across the region are similar to trends observed
basin comprises ophiolitic rocks of the kind across modern magmatic arcs (Thirlwall 1981a).
exposed at Ballantrae. These ORS volcanics pose a problem in dating
the closure of the Iapetus Ocean (McKerrow &
Frontal arc/arc massif Cocks 1976; Thirwall 1981a). Either the sector
of Iapetus between Scotland and England was
North-west of the Highland Boundary Fault still open in the Early Devonian, subducting to
 A n a t o m y o f a L o w e r Palaeozoic accretionary forearc 503

the NW under Scotland, or the Lower ORS is in for a distance of c. 30 km in the region east of
part of late Silurian age. Lower ORS dating on Moffat (Fig. 3).
palaeontological grounds is uncertain and K-Ar Following the accretionary prism model we
ages of Lower ORS lavas are either Silurian or propose that each fault-bounded tract repre-
Early Devonian depended on the time-scale sents a discrete packet of accreted ocean floor
chosen (Thirlwall 1981a). and/or trench sediment offscraped above the
northwards subducting Iapetus ocean crust.
The Southern Uplands accretionary The differences between stratigraphic sequ-
complex ences contained in these tracts (e.g. sequences 3
& 4, Fig. 4) suggest a large geographic separa-
Reconstruction of the evolution of the complex tion between some of the sequences prior to
accretion (McKerrow et al. 1977). The apparent
Stratigraphy lack of continuity of some of the tracts implies
In this section we review the geological evi- original offscraping of lenticular packets during
dence used to reconstruct the Southern Up- accretion.
lands accretionary complex. Fig. 3 shows the Study of Fig. 3 reveals that there are rather
main reverse faults recognised to date in the more major reverse faults than there are de-
Southern Uplands, and Fig. 4 shows repre- fined stratigraphic sequences (e.g. faults west of
sentative stratigraphic sequences from the the type areas for sequences 2 and 3 in the
tracts defined by these strike faults. We have central part of the Northern Belt and faults
described the stratigraphic sequences at length, traversing the area of sequence III on the
and discussed their correlation across the Galloway Coast). Clearly, therefore, the diffe-
Southern Uplands, in an appendix to this paper rent stratigraphic sequences should be used as
(Supplementary Publication No. 18038 of the only a general guide to the configuration of
Geological Society of London, copies obtain- original accretionary tracts; the areas of out-
able from the Geological Society Library and crop of sequences in some cases contain more
the British Lending Library at Boston Spa). than one original sliver of off-scraped ocean
Our intention here is to draw attention only to floor and/or trench sediment. This conclusion
the main points. The diagrams show that each has been confirmed wherever areas have been
tract is characterized by a different sequence, mapped in detail. The best example is in the
and that progressively younger strata are found Langholm region, where Lumsden et al. (1967)
in sequences towards the SE (site of the Iape- have mapped several major strike faults which
tus). The accretionary prism model of Seely separate greywacke sequences of different age
et al. (1974) predicts such a configuration for (progressively younger towards the SE) within
subduction complexes. The viability of this tract 10 (see Figs. 3, 4 & 10Q).
interpretation in the case of the Southern Up-
lands is strengthened by the predominant Ordovician reconstruction
younging of strata NW (continentwards) within Fig. 5 shows a reconstruction of the Southern
each tract, by the oceanic/trench affinities of Uplands accretionary complex in the late
strata in the sequences (see later) and by evi- Ordovician. The evolutionary history can be
dence for a contemporaneous magmatic arc followed using stratigraphic evidence summa-
NW of the region. rized in Figs 3 and 4. In the following account
Some of the reverse faults can be traced we refer to lettered labels on Figs 3-6.
along the entire outcrop (e.g. the Riccarton The Ballantrae ophiolite (BO) north of the
Line); one, the Kingledores Fault, can be traced Stinchar Fault (SF) was emplaced on the
into the Irish continuation of the belt (Fig. 2). Laurentian continental margin in Arenig or
Most, however, are of uncertain continuity, and early Llanvirn times (e.g. Dewey 1974) and is
several cannot be traced more than a few tens overlain unconformably by a Llanvirn to
of kilometres, suggesting that some tracts may Ashgill upper slope succession (Ingham 1978).
be discontinuous (Fig. 3). Similarly, stratig- In the southern part of the ophiolite the Dow-
raphic sequences can be traced within tracts for nan mafic sequence (DMS Fig. 4) between the
varying distances. For example, sequence 5 at Stinchar Fault and Dove Cove Fault (DCF) was
Hartfell (Fig. 4) can be traced west across the probably emplaced in the late Llandeilo or
ground south of Moniaive through the aureole early Caradoc (gracilis Zone, Barrett et al. in
of the Cairnsmore of Fleet granite, to the press and unpublished fieldwork).
Wigtownshire coast south of Glenluce, a dis- Basement below the ophiolite is not exposed,
tance of 100 km (Fig. 3). In contrast sequence 7 but boulders in the upper slope succession
at Craigmichan (Fig. 4) is only known to extend (Benan Conglomerate) suggest a Proterozoic/
504 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey
V.E.2x
O_

::~7 ~': . •

f--~, ( ." . . . . . . . .

.:.....:.........
E5_ OAXACA

\ i f i i ~ "" \~

'','" ~.-. -~ ~-~/~-i ~ , / "I--i', ~t~s --~\

- ' ."',',-, - ~ . . . . . . ,, ..-'

..... -' ,. M V
I

.... :' :-~:,' BO

V.E. 2x DCI ~ SF "

r NBof SU -- -- - -.

~ ~.~.'~,~, ~ SOUTHERN UPLANDS

:,.Y,ii~ I ";:-'z-:--',:_','-'-,-,-,. (Late Ordovician)
z~ / x . - \ i /~/l'l\l

- I ~-i.. ~,./. /\1.~1. \ t ,-, /

) ', 7,'f-" U,, I I I
' " ' -"- .' "'.'-.-. . . . . , -." ..... " - k m

Fnc;. 5. Reconstruction of Southern Uplands accretionary prism in late Ordovician, compared with
modern Middle American accretionary complex off Mexico at same scale. Southern Uplands
reconstructed on a line along west coast, Oaxaca (Mexico) profile from Moore et al. (1979).
Continental basement dotted; subducting oceanic crust and ophiolitic rocks in oblique dashed
ornament. For further explanation see text. BO Ballantrae ophiolite, DCF Dove Cove Fault, MV
Midland Valley, NB of SU Northern Belt of Southern Uplands, SF Stinchar Fault,

Cambrian metamorphic/granitic basement (dot- by d6collement surfaces which are the major
ted ornament in Fig. 5) under the SW Midland strike faults of Fig. 3; along these faults basal
Valley (MV) (Longman et el. 1979). Seismic lithologies (basalt, chert, black shale) have
refraction evidence indicates that this type of been imbricated (dashed ornament in Fig. 5).
basement may terminate at the Southern Up- The positions of sequences I-III (Figs 3 & 4)
land Fault (Bamford 1979). are shown in Fig. 5; note that there are more
South of the Southern Upland Fault, of fault-bounded tracts (probable discrete
which the Dove Cove and Stinchar faults are accreted slices) than there are defined sequ-
braiding strands (Fig. 3), greywacke tracts of ences (see faults mapped on Fig. 3 both on the
the Southern Uplands began to be accreted in west coast and in the area east of the Loch Ryan
the late Llandeilo or Caradoc. This we know Fault)•
because all graptolites recorded in the Coulter/ The Oaxaca profile in Fig. 5 makes an in-
Noblehouse and Corsewall/Glen App sequ- teresting comparison. Continental basement is
ences (sequences 1 and I, Figs 3 & 4) are from overlain by an upper slope succession lower
the gracilis Zone. The remaining tracts of the Miocene at the base. Below the thin blanket of
Ordovician Northern Belt of the Southern Up- sediment on the lower slope, landward-dipping
lands (NB of SU) were accreted by late Ordovi- reflections are accreted trench deposits. Deep-
cian time. Accretionary slices are well-defined sea drilling shows that the earliest accretion was
 Anatomy of a Lower Palaeozoic accretionary forearc 505

late Miocene (Moore et al. 1979, 1981). The succession is not exposed east of the Girvan-
main points to note in comparing the two Ballantrae area (Fig. 3). The thickest exposed
profiles in Fig. 5 are: forearc basin succession (at Girvan) is c. 3 km.
(i) Clear reverse strike faults in the South- Basement may comprise Proterozoic granulites
ern Uplands (Fig. 3) suggest that indi- at depth, as previously discussed. The forearc
vidual packets of accreted strata were succession in the north of the Midland Valley is
morphologically well defined (cf. Nank- buried under Upper Palaeozoic cover.
ai Trough profiles, Moore & Karig Ophiolitic rocks of the Highland Boundary
1976). Seismic records suggest less de- Complex (HBC) crop out at intervals along the
finition of discrete accreted slices under Highland Boundary Fault (HBF) on the north
the Oaxaca lower slope (Shipley et al. border of the Midland Valley (Fig. 2). These
1981). relate to the earlier history of the margin prior
(ii) On both margins the earliest upper to the onset of accretion in the Southern Up-
slope sediments are more than 10 Ma lands. They date from the Cambrian (Hender-
older than the first accretion of trench son & Robertson 1981) or early Ordovican
clastics. (Bluck et al. 1980). The frontal arc/arc massif
(iii) The Ordovician part of the Southern was exposed in the Grampian Highlands (GH).
Uplands accretionary complex was con- In the Sumatra forearc the accretionary com-
structed in c. 20-30Ma (from gracilis plex is also emergent. The trench slope break
times to sometime between latest Cara- forms the Mentawai Islands, one of which is
doc and end Ordovician: see constraints Nias (Fig. 5), on which both accretionary com-
of graptolite biostratigraphy in Fig. 4, plex and lower slope basin sediments have been
sequences 1-3). The Oaxaca accretion- mapped (Moore & Karig 1980). In the forearc
ary complex was constructed in c. 10 Ma basin up to c. 6 km of sediment have accumu-
(since late Miocene: Moore et al. 1979). lated behind the growing accretionary complex
(iv) Ordovician accretion in the Southern (G. F. Moore et al. 1981, and references there-
Uplands included ocean floor pelagic in). Basement under the basin may be ophiolitic
sediments and possibly slivers of the (Curray et al. 1979) or older accreted material
underlying oceanic basement: pelagic- (Karig et al. 1980).
hemipelagic sediments and basalts are The main points to note in comparing the two
preserved in all the Ordovician sequ- profiles in Fig. 6 are:
ences (see Fig. 4). Compared to this (i) Accretion of thick clastics from the
'efficient' accretion in the Ordovician
Southern Uplands, the Oaxaca accretion
VE 5x Nias
seems to be 'inefficient': pelagic- 0~ - • _

hemipelagic ocean floor deposits and E5i NF

much of the trench sediment seems to SUMATRA

have been subducted below the accre-
tionary complex (Watkins et al. 1981). ..... -_ - SUF HBF

VE 5x c[ MV ~ GH

Silurian reconstruction ....

By late Silurian times all tracts exposed today ~ 5~ HBC

(Figs 3 & 4) had been accreted. Later Silurian

sequences (?Ludlow) may lie between the Up-
~0 I ---- ~ : + ~ - -
~ ' ~
SOUTHERN UPLANDS
(Late Silurian)

per Palaeozoic overstep and the Iapetus suture FIG. 6. Reconstruction of Southern Up-
(Figs 2 & 3). lands accretionary prism in late Silurian,
Uplift caused by the prolonged accretion had compared with modern Sundra arc accre-
produced during latest Ordovician and early tionary complex off Sumatra at same scale.
Silurian a trench slope break, which shed sedi- Southern Uplands reconstructed on a cen-
ment north to the Midland Valley (MV) forearc tral section, Sumatra profile from Karig et
basin (Leggett 1980a). The trench slope break, al. (1980). Continental basement dotted;
known as Cockburnland (C), was bounded by subducting oceanic crust in oblique dashed
ornament. For further explanation see
the Southern Upland Fault (SUF), which was text.
active during Silurian sedimentation in the fore- C Cockburnland, GH Grampian High-
arc basin (McGiven 1968). The horizontal lands, HBC Highland Boundary Complex,
arrow north of the fault in Fig. 6 shows the area HBF Highland Boundary Fault, MV Mid-
in which inliers preserving the succession in this land Valley, NF Nicobar Fan, SUF South-
basin are exposed. The base of the forearc basin ern Upland Fault.
506 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey

Nicobar Fan (NF, Fig. 5) caused the also occur. Most of the basalts are non-vesicular
rapid post-Miocene growth of the or sparsely vesicular, suggesting eruption under
Sumatran accretionary complex. at least several kilometres of water (e.g. Jones
Greywacke stratigraphy in the Central 1969).
and Southern Belts of the Southern Sequence 1 contains alkali basalts and sequ-
Uplands suggests that a similar thick ence 2 predominantly tholeiites. The least-
ocean floor clastic sediment pile was mobile trace elements, Ti, Zr, Y and Nb,
being accreted in t h e Silurian along the indicate that the basalts are of oceanic deriva-
northwestern Iapetus margin in Britain tion (Lambert et al. 1981). If so, they must have
(Leggett 1980a and see later). been slivered off the top of subducting oceanic
(ii) The similar tectonic/palaeogeographic basement during accretion. This process has
role of the Southern Upland Fault been described in modern trenches (see, e.g.
(SUF) and the reverse fault/flexure be- Kulm et al. 1981). However, in some cases
hind the Sumatran trench slope break: where contacts are not exposed in the field we
both controlled forearc basin sedi- cannot be sure that the Southern Uplands
mentation (Karig et al. 1980: McGiven basalt bodies are not flow units intercalated
1968). within a pelagic sediment section.
(iii) The Sumatran accretionary complex de- Minor Caradoc volcanic rocks in the North-
veloped in 2()Ma since early Miocene ern Belt crop out in the Tweeddale area and east
(Karig et al. 1980); the Southern Up- of Sanquhar in the Bail Hill area (Fig. 3). The
lands accretionary complex developed in Tweeddale lavas are blocks of peralkaline
45-50 Ma from mid-Ordovician to late rhyolite (Thirlwall 1981b) in a debris flow de-
Silurian (Leggett et al. 1979a). Plate posit. Intercalated ash beds and blocks of lime-
convergence along the Sunda arc off stone with shallow water fossils led Leggett
Sumatra is relatively fast (Karig et al. (1980b) to interpret this debris flow unit (the
1980), so the implication of the pro- Wrae Limestone: sequence 3, Fig. 4) as an
longed development of the Southern olistostrome derived by submarine sliding from
Uplands accretionary complex is that a carbonate-capped oceanic volcano. Part of
convergence between the northern Iape- such a volcano may be preserved in the Bail Hill
tus oceanic plate and the Laurentian region, where several square kilometres of
continent was slow. varied igneous rocks with broadly dioritic affini-
ties have been interpreted as an accreted
Lithological and palaeogeographic analysis oceanic seamount (Hepworth et al. 1981).
Analysis of the Southern Uplands strati-
(2) Ocean-floor pelagic-hemipelagic sediments
graphic sequences (Fig. 4) allows reconstruc-
The earliest pelagic sediments are Fe-rich mud-
tion of environments on the Iapetus ocean floor
stones. These have intimate field relations with
and along its northwestern margin. The
the basalts, occurring in cross-cutting veins,
basalts, cherts and black shales have been inter-
interstitial pods and lenses, and beds between
preted as ocean floor lithologies (Leggett 1978,
and above flows. Fe contents of up to 40%, Mn
1979) and the greywackes as both ocean floor
contents up to 3%, low AI and Ti contents, and
and trench deposits (Leggett 1980a, b). In this
Ce-depletion in a R E E pattern from a supra-
section we briefly describe the lithologies in
basaltic sample indicate a syn-magmatic origin
upward succession, stressing the reasons for
from hydrothermal-exhalative solutions (Leg-
these interpretations. Further details can be
gett & Smith 1980). These deposits, apparent
found in the source references.
analogues of hydrothermal metalliferous sedi-
(1) Oceanic volcanics ments from modern spreading centres and
Basalts exposed at the base of sequences l basal ocean-floor sediment sections, support the
and 2 (Fig. 4) occupy strip-like outcrops up to conclusion that slivers of oceanic basement
200 m thick and 4 km long, fault-bounded at have been preserved in the Northern Belt.
the base. There is only minor interdigitation of Radiolarian cherts and siliceous mudstones
basalts in overlying pelagic sediments. Further overlie the basalts. These are red, grey and
small basalt outcrops are found in the southern green microcrystalline deposits, commonly bed-
part of the Northern Belt along strike from the ded on a centimetre-to decimetre-scale, but
type area of sequence 3. There are no basalts locally massive. Radiolarite layers are common,
in the Central or Southern Belts. and sedimentary structures include crude grad-
Massive basalt dominates outcrops, though ing (suggesting redeposition), parallel lamina-
pillow lavas, thin lava flows and basalt breccias tion, syn-sedimentary faulting and chaotic
 Anatomy o f a L o w e r P a l a e o z o i c accretionary f o r e a r c 507
NW SE

~
~ ~ .... - ~
~
4
.4~
~
m
"~
WLO __
~ Trench wedge facies
A ..... .~ . ~ , ,: . . ~ ; _. ~: : . . :%
. . : : : 2. . ~ 2 ~ ~-]Terrigenous plate facies
~e,agic °,ate facies

Continental crust

[ ~ ..... C,R,Doc
km I ~
-- AP I[I I1
TSB . . . . 1
j/J- ~ ?

(pre-turriculatus)

i - -',-,:::":":-'i':'"'~,i~",:i",;:i.;";',
!~''~}i'"i-'/'~

(post-turriculatus)

Fi6. 7. Schematic cross-sections across the forearc region of the north Iapetus margin in southern
Scotland at various times in the Ordovician and Silurian (after Leggett 1980a). Representative
sediment dispersal directions shown (see references, Fig. 4). Major accreted slices numbered as in
upper columns of Fig. 4.
(A) Stage I: Caradoc. Trench wedge facies overlie pelagic plate facies; nascent accretionary prism
(AP) oceanwards of obducted ophiolite on Laurentian continental crust. Source of Wrae Limestone
olistostrome (WLO), thought to be an oceanic island (Leggett 1980b), not shown.
(B) Stage 2: Early and middle Llandovery. Emergent trench slope break ('FSB) delimits two basinal
provinces: the lower trench slope and trench to the south, and forearc basin (FAB) to the north.
Elevated area on subducting oceanic plate (?outer trench high: see text) delimits two turbidite facies
provinces, one in the trench and another on the ocean floor. Question mark in one of the accreted
slices represents the probable existence of a fourth Ordovician slice between sequences 2 and 3
(Leggett et al. 1979a).
(C) Stage 3: Late Llandovery. Oceanic basement high overwhelmed, uniform trench-ocean floor
flysch blanket ensues.

slumped bedding. Radiolarians are primitive perhaps as low as several m Ma-1 and certainly
spumellarians, and are preserved within diagene- equally as slow as siliceous pelagic deposits in
tic balls of silica 300-400 Hm in diameter (Leg- modern ocean basins.
gett 1978, 1979). The cherts are overlain by the pelagic-
The radiolaria are of no biostratigraphic hemipelagic Moffat Shale Group, comprising
value, and dates are provided by graptolites black shales and grey mudstones, which
(Peach & H o m e 1899) and conodonts (Lamont accumulated from latest Llandeilo/early Cara-
& Lindstr6m 1957). These extend from mid- doc to late Llandovery (Fig. 4). Biostrati-
Arenig to late Llandeilo, about 25 Ma. The graphic control is excellent in these deposits,
thickest continuous chert section known is only which contain abundant graptolites. The maxi-
c. 60 m. Notwithstanding the intense imbrica- mum exposed thickness is c. 100m (Fig. 4,
tion in the outcrops of basal lithologies (see sequence 7), and the duration of sedimentation
later) and the limited number of faunas on c. 40Ma. Most of the Moffat Shale is finely
which the dating is based, it appears that the laminated, carbon-rich shale devoid of benthos,
cherts accumulated very slowly, at rates which clearly accumulated under euxinic condi-
508 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey

tions. Interbedded grey mudstones, especially deposits SE (i.e. outboard or oceanwards) of

common in the Ashgill (Fig. 4), reflect more Caradoc greywackes of the Northern Belt are
oxic conditions. Euxinification has been entirely pelagic (sequences 4-8, Fig. 4). The
ascribed to productivity fluctuations arising same applies to Ashgill strata, except that one
from eustatic transgressions (Leggett 1978, as yet unstudied greywacke unit occurs ocean-
1980c; Leggett et al. 1981). The most significant wards of the Northern Belt greywackes in
aspect of the Southern Uplands pelagic deposits sequence 8 (see discussion in Leggett 1980a).
is that they represent some 60 Ma of geological Biostratigraphic control in the Ordovician
time in which less than 200 m (after compac- greywackes is poor. Graptolite faunas have
tion) of sediment was deposited. Such pro- been recorded from sequences 1 and 3 (see Fig.
longed slow accumulation can only have occur- 4 and caption): in both cases only one graptolite
red in a wide ocean basin. zone is represented. If the Ordovician
greywackes are trench deposits the greywackes
(3) Trench and ocean floor turbidite facies in each sequence are not likely to extend any
Greywacke turbidites in the Southern Up- higher than these zones. The residence time of
lands are dominated by monotonous medium- turbidite sections in modern trenches on accre-
to thick-bedded, usually acyclic, sequences and tionary margins (e.g. Moore 1979) is not likely
massive sandstones. These are distinguished in to be longer than the 3-5 Ma average duration
Fig. 4 as the 'sheet sand' facies class. Other, less of an upper Ordovician graptolite zone
common, facies include rudites (disorganized (McKerrow et al. 198(I).
and organized) and thin-bedded turbidites, In the mid-Llandovery, greywackes appear
usually with low sandstone-shale ratios. The in sequences both 'outboard' (sequences 8, l/)a)
distribution of these facies classes is depicted and 'inboard' (sequences 4, 5, 6) of pelagic
in Fig. 4. This can only be done in an illustra- facies. This suggests the presence of an elevated
tive way because biostratigraphic control is area slowly accumulating pelagic sediment
generally poor in the greywackes, and detailed (black shale), separating two turbidite pro-
lithostratigraphy is generally lacking. vinces (Fig. 7B). Leggett (1980a) interpreted
Fig. 7 is a reconstruction of the history of the elevated area as an outer trench high. By
clastic sedimentation along the northwestern this interpretation greywackes in the 'inboard'
lapetus margin using stratigraphic information sequences are trench deposits and greywackes
summarized in Fig. 4 and previously published in the 'outboard" sequences are abyssal plain
facies analysis (Leggett 1980a). Ordovician (ocean floor) turbidites. Perhaps more likely, in
greywackes are interpreted as trench deposits view of the duration (several graptolite zones)
on the basis of: of the strata in sequences 4 and 5 (Fig. 4), is
(i) common rudites (Fig. 4, sequences 1-3 that the high represents an intra-oceanic base-
and I-III) which map out as lenticular ment feature, such as the Ninetyeast Ridge in
bodies along the length of the Northern the Indian Ocean, separating an 'inner' trench-
Belt and have predominantly lateral ocean floor turbidite accumulation and an "outer'
(NW to SE) palaeocurrents, suggesting abyssal plain accumulation.
localized point sources (Kelling & Hol- After turriculatus times (sequence 7, Fig. 4)
royd 1978; Fig. 7A); the high was submerged (sequences 5, 8, 9, 10a
(ii) immature petrography (e.g. Kelling and b in Fig. 4) and was covered by a uniform
1962; Walton 1965), suggesting direct sediment blanket (Fig. 7C). The thick
input of arc- and ophiolite-derived greywackes of these later sequences contain
detritus without trapping in a forearc interbedded graptolitic horizons and accumu-
basin; lated for long periods. Clarkson et al. (1975)
(iii) evidence for flow confinement and low have proposed a true thickness of c. 5 km for
depositional gradients, provided by axial the strata in sequence 9 (Fig. 4) in the Kirkcud-
palaeocurrents most commonly flowing bright area (Fig. 3). This supports the inter-
NE to SW but with some also flowing SW pretation of later Silurian greywackes as a
to NE (Fig. 4, sequences 1-3 and I-III). Nicobar Fan-type accumulation (Fig. 6). Facies
Rare palaeocurrent indicators suggest- considerations also support this view: most beds
ing currents from the SE (Kelling 1962) are derived axially, are relatively quartzose,
perhaps represent flows which rebound- and devoid of appreciable rudite bodies (sum-
ed off the outer trench slope. mary in Walton 1965).
The Ordovician trench interpretation is sup- A final category of trench and ocean floor
ported by the stratigraphy of the Ordovician clastics comprises the debris flow/olistostrome
strata preserved in the Central Belt. Caradoc deposits. The Wrae Limestone olistostrome has
 A n a t o m y o f a L o w e r Palaeozoic accretionary forearc 509

GAF KF
- " ~ P. MB . . . . . PTZ.

--~ ~J~r PGr

[]
GAF GM

CR LR I GR BR

SUF CF LF LL FL KF

H.
,Z,
,, /
:::::::::::::::::::::::::::::::
2

3 Correlatlltlve
SUF GF LIZ LL ? ~V KF _1~ greywacl:ke
I~ ~--, DVI ..... GV ~- / ~ ~KR I EV ~ ~' ~ 1 ~~o: 4It .......

I
I C- N T l ? A-AT TT
® ® ® :)::.!::
; ; "": 7 J
~ Basalt_*Chert_"
_¢km ~ Black shale
i

FIG. 8. Structural profiles in the Northern Belt of the Southern Uplands. Lines of section are on Fig.
3. Major reverse faults, ie. those bounding tracts containing considerably different stratigraphic
sequences and considered to be ddcollement surfaces dating from initial offscraping during
accretion, are shown extending above and below line of section. Illustrative bedding attitudes shown
with thin lines in top two-thirds of section; arrows above section are representative sedimentary
younging data. North-west on the sections is on the left. F. the Rhinns of Galloway (from Kelling
1961). G. the region east of Loch Ryan (from Welsh 1964). H. the region west of Nithsdale (from
Floyd 1976). I. the Clyde valley region (from Leggett et al. 1979a).
Faults: labelled as in Fig. 3, plus KTZ--Killantringan Thrust Zone; LIZ--Leadhills Imbricate Zone;
PTZ--Portayew Thrust Zone.
Greywacke units (lettering below sections: units named by authors in papers from which sections are
taken). For details of correlations see Figs 3, 4 and text).
1. Corsewall Group (CGr); Marchburn Formation (MF); Coulter-Noblehouse Tract (C-NT),
2, 3, 4. Kirkcolm Group (KGr)--divided into Lower Barren (ii), Upper Barren (iv), and intervening
Meta-clast (iii) divisions (profile F); Lochryan Rocks (LR)---divided into Lower (iii) and Upper (iv)
divisions (profile G); Afton Formation (AF); Afton-Abington tract (A-AT).
5. Galdenoch Group (GCr); Cairnerzean Rocks (CR).
6, 7 Portpatrick Group (PGr)---divided into a lower Acid-clast division (6) and an upper Basic-clast
division (7); Glenwhan Rocks (GR); Boreland Rocks (BR); Scar Formation (ScF): Shinnei
Formation (ShF); Tweeddale tract (T-F).
Locations: BP Broad Port; COP Cave Ochtree Point; CP Corsewall Point; CV Clyde Valley; DB
Dally Bay; DV Duneaton Valley; EV Elvanwater valley; GM Glenwhan Moor; GV Glengonnar
Valley; KR Kirkton Rig; MB Morroch Bay; P Portpatrick, Po Portobello; SPB Salt Pans Bay.

already been mentioned. Other such Caradoc in the Southern Uplands is probably too deep to
carbonate debris flows occur in the Northern have preserved extensive lower slope sedi-
Belt, though volumetrically of minor import- ments. Mapping in more recent accretionary
ance and poorly exposed (summary in Leggett complexes suggests that large differences in
1980b). These may represent olistostromes de- style of deformation and age between adjacent
rived from the upper slope, where Caradoc terranes are characteristic of lower slope sedi-
limestones are known (e.g. Craighead Lime- ment sections resting on accretionary basement
stone in the Ardwell Group, Williams 1962; (e.g. Bachman 1981; Moore & Karig 1976; see
A G in Fig. 4). They may equally well derive also J. C. Moore et al. 1981). There is no
from carbonate-capped oceanic islands, an in- evidence in the Southern Uplands for signifi-
terpretation favoured by the volcanic blocks cant age differences between contiguous
and tuffaceous interbeds in the Wrae Limes- greywacke sections, as would perhaps be ex-
tone olistostrome. pected between accretionary basement and a
younger mantling slope sediment pile. Possible
(4) Lower trench slope sediments minor exceptions are an area of Wenlock rocks
We believe that the present level of erosion juxtaposed with Llandovery greywackes in the
510 J. K. Leggett, W. S. McKerrow & D. M. Casey
GBF

Correlative
greywacke
DF terr0nes

Black shale
KF / / Sknl
GF i

BF CT PT

,
~,~;5 .~~::: ....
.~--/.-~ ~ ~
,
...~"~~'~!:~'~.'~
..

? HtL

~......._._._._._. lt':-:..-.lt::.:~7:
lt..l.,- :.::~,~-~,o-:~m.:;: :: :: f':5 :..:::. :: .:::~:

FIG. 9. Structural profiles of the Central Belt of the Southern Uplands. Details as in Fig. 8.
North-west on the sections is to the left. J. west Wigtownshire coast, north (from Gordon 1962).
K. west Wigtownshire coast, south (from Gordon 1962). L. region west of Cairnsmore of Fleet
granite (from Cook & Weir 1979). M. region north of Moffat (from unpublished Institute of
Geological Sciences field logs of the 1976 British Gas pipeline section).
Faults: labelled as in Fig. 3, plus BF Blairbuies Fault, T-I'Z Talnotry Thrust Zone.
Greywacke units (lettering below sections: units named by authors in papers from which sections are
taken). For details of correlation see Figs 3, 4 and text.
1. Kilfillan Formation (KF); Craignell Formation (CF) in part. 2. Garheugh Formation (GF);
Craignell Formation (CF) in part. 3. Knockeans Formation (KsF). Correlation between profiles K
and L is tentative (see Cook & Weir 1979); correlation between profile L and unnamed greywacke
terranes in profile M is extremely tentative.

RL

CB terranes
3

O. : : : ; : : : : : : : : : : , : , : , : , : . : . : . . . : ..:.:.._,.. ,., ::

HR I RB o 5kin

1 RL

Hr |

RL BF
- it . . . . . BA - - - HS . . . . . . .
• : : .': : ": .':!~ ; .:]:: ; ::;:7!::
o ~\~ ,~,
HR "l--mu ri "~ . . . . . . . . . . . . 5 .............
BLF
RL SB

GG I HR ~~o ~ . o i' SB ~' . . .S. .C B

HF

......"7 : . . . ~ ~z-.. :~,:~.,. :~ :~ ~-

I . . . . . . . . . . . . . . . . . . .
PBB SB PBB CBB
 Anatomy o f a L o w e r Palaeozoic accretionary forearc 511

continuation of the Central Belt in the Long- faces during accretion of the tracts, which infers
ford-Down Inlier of Ireland (Leggett et al. that they were initiated as low-angle thrusts.
1979b), and the anomalous inlier at Colding- This view is supported by parallelism or near-
ham on the NE coast in Scotland. At Colding- parallelism of bedding and most of the major
ham abundant flat-lying structures of possible strike faults (Figs 8-10). Rotation within a
submarine slide origin contrast markedly with growing accretionary complex, coupled with
the steeply dipping greywacke terranes to the probable oversteepening resulting from even-
north and south. Other such candiates for lower tual impingement against the European margin,
slope sediment remnants may come to light may explain the steep dips of these faults today
with future mapping, but we do not anticipate (McKerrow et al. 1977).
that any will be large. Basal lithologies are imbricated along the
major faults in irregular zones (Fig. 3) from
(5) Irish sequences
several tens of metres to more than 1 km wide
Stratigraphy of sequences in the Irish con-
(Leggett et al. 1979a; Weir 1979). 'Line', a term
tinuation of the Southern Uplands, summarized
used by Floyd (1976) and Weir (1979) for some
by Leggett et al. (1979b), is broadly similar to
of the major faults, derives from the promin-
that of sequences described here. A major
ence of linear outcrops of basal lithologies
difference is greater interbedding of basaltic
(basalt, chert and graptolitic shale) marking the
volcanics, both in pelagic and clastic succes-
traces of imbricate zones on Geological Survey
sions, in the equivalent of the Northern Belt in
maps (Fig. 3).
Cavan (NW corner of the Longford-Down mas-
Within tracts, steeply dipping homoclinal
sif, Fig. 2). Greywacke facies and petrology are
zones in which beds young NW are predomi-
broadly analogous, with more rudites and more
nant. F 1 folds, both megascopic and mesosco-
immature, volcaniclastic detritus in Ordovician
pic, are chiefly upright, asymmetric, open to
than in Silurian rocks (Leggett et al. 1979b). In
isoclinal, and SE-verging, with variably plung-
detail, however, sequences in individual tracts
ing NE-SW axes (summary in Stringer &
do not correlate between mapped areas in the
Treagus 1980). Axial planes are predominantly
west and east of the Longford-Down massif. As
inclined steeply NW or SE (e.g. Anderson &
in the Southern Uplands, it appears that most
Cameron 1979). Widespread slump folds occur
tracts do not extend along the belt for more
in the western part of the Longford-Down inlier
than c. 100 km. (Morris 1979).
Cleavage transects F 1 folds for a distance of
Structure
up to 50 km north of the Iapetus suture, and is
Structural profiles of the Northern, Central rotated clockwise with respect to F~ axial traces
and Southern Belts (Figs 8-10 respectively) by up to 20 ° (Phillips et al. 1979). Phillips et al.
illustrate the megascopic structure of the South- (1979) suggest that the non-axial planar cleav-
ern Uplands. Reverse strike faults and belts of age results from dextral shear parallel to the
homoclinal, steeply-dipping NW-younging regional strike during brittle deformation.
greywackes dominate. Major strike faults are Stringer & Treagus (1980) following Borradaile
those which bound tracts of differing stratigra- (1978) suggest that it is a result of folding of
phy (Figs 3 & 4) and have indeterminable planes non-orthogonal with respect to bulk
though presumably very large displacements. strain axes. In non-coaxial strain histories cleav-
Major faults so defined we interpret for reasons age formation can be delayed relative to fold
already outlined as original d6collement sur- formation as a result of intergranular move-

FIG. 10. Structural profiles of the Southern Belt of the Southern Uplands. Details as in Fig. 8.
North-west on the sections is to the left. N. coast section east of Whithorn (from Rust 1965). O.
Kirkcudbrightshire coast (from Craig & Walton 1959). P. Region south of Moffat (a composite
section from unpublished Institute of Geological Sciences field logs of the 1975 and 1976 British Gas
pipeline sections. Q. Langholm region (from Lumsden et al. 1967). R. Hawick region (from Warren
1964). S. Riccarton Junction region (from Warren 1964).
Faults: labelled as in Fig. 3.
Greywacke units (lettering below sections: named by authors in papers from which sections are
taken; for details of correlation see Figs 3, 4 and text).
1. Gala Greywackes (Llandovery) (GG). 2. Kirmaiden Beds (KB); Carghidown Beds (CB); Hawick
Rocks (HR). 3. Riccarton Beds (RB); Stobs Castle Beds (SCB); Shankend Beds (SB); Penchrise
Burn Beds (PBB); Caddroun Burn Beds (CBB). In the Riccarton Beds, and subdivisions theoreof,
units containing diagnostic graptolite zones are shown: mu, murchisoni (solid line); ri, riccartonensis
(long dashes); li, linnarssoni (short dashes).
 512 J. K. Leggett, W. S. McKerrow & D. M. Casey

LEGEND

~ UppM Palaeozoic
(ORS-Did Red SandstoneC-Cml=oniferous)
I~ 5kin ~ post-Llandove~y
I SUudanred beds I MidlandValleyforeafc succession
Llandovery j (Silurian)
turbidlte sequences
~ post-Benansediments
~ Benanconglomerate 1 Girvanforearc succession
~i s ~,;,?,? ,-,(~s s
A. " s s ,?'~-~',',7\ ~- s s ~ Kirklandconglomerate ((~rdovician)
! I Superstesmudstones
~ G l e n App Greywackes
~ Serpentinitebelts I
Baiian~rae ophiolite
Mafic vok:anics

I
ORS HwB LW -- :~ ORS

GL P HB NR ORS

D. E.
FiG. 11. Structural profiles of the Midland Valley. Lines of section are on Fig. 3. Illustrative bedding
attitudes shown with thin lines, faults with thick lines. Arrows are representative sedimentary
younging data. North-west on the sections is to the left. (A) The Ballantrae ophiolite (from Dewey
1974). (B) The Stinchar valley region (from Williams 1959--vertical scale exaggerated several
times). (C) The Lesmahagow inlier (from Geological Survey one inch sheet 23). (D) The Hagshaw
Hills inlier (from Rolfe 1961). (E) The Pentland Hills inliers (from Tipper 1974).
Faults: AF Assel Fault, DT Doune Thrust, DnT Dupin Thrust, KF Knocklauch Fault, SF Stinchar
Fault, TI" Tormitchell Thrust.
Localities: BH Benan Hill, DH Dinvin Hill, GL Glenbuck Loch, HB Henshaw Burn, HwB
Hareshaw Burn, LW Logan Water, NR North Esk Reservoir, P Parisholm, PH Parisholm Hill.
ment within beds (Borradaile 1978). Such Cameron 1979). In the southern part of the
movements are highly likely in the early phases Southern Uplands similar F 2 folds are concen-
of deformation in subduction complexes, where trated in a linear belt 2-3 km wide (Stringer ,8,:
interstitial fluid pressures are likely to be high Treagus 1980). Phillips et al. (1979) ascribe
(e.g. Cowan 1981). these relatively minor post-DI deformation
The transecting SI cleavage, and probable effects to final closure of the Iapetus.
primary variation in Fl fold axis plunges (Strin- Wrench faults which displace D 2 structures
ger & Treagus 1980) has caused confusion are common. These are orientated NNW, N
leading to publication of overly complex and NE (Fig. 3) and are predominantly sinistral
polyphase deformation histories (Anderson with displacements of up to 1 km (summary in
1980, commenting on Weir 1979 and references Weir 1979). Anderson & Cameron (1979) re-
therein). Most recent structural interpretations late these to closure of Iapetus. Intrusive rocks
have related all main folds and strike faults to a of probable Early Devonian (post-closure) age
compound, diachronous D~ event produced have exploited faults in the Southern Uplands.
during accretion (Anderson & Cameron 1979; Porphyrites and lamprophyres commonly in-
Eales 1979; Leggett et al. 1979a, b). trude wrench faults; some of these are sheared,
Post-D 1 folding becomes progressively more indicating continuing fault movements (Ander-
common towards the suture (Phillips et al. son & Cameron 1979; Weir 1979). Small linear
1979). In the eastern part of the Longford- granite stocks have the same trends as reverse
Down massif profile (Fig. 2) F 2 folds are gentle, faults and wrench faults (Geological Survey
south-verging flexures and F 3 folds are north- one-inch sheets 15 and 24).
verging monoclines; both are of Caledonide Structural geology appears more complex in
trend (NW-SE axes), have associated crenula- the western than in the eastern part of the
tion cleavages, short limbs of < 20 m, and do Longford Down massif and the Southern Up-
not affect the regional outcrop (Anderson & lands (Morris, unpublished data). In particular
 Anatomy of a Lower Palaeozoic accretionary forearc 513

there is common pre-cleavage overturning of sure of the Iapetus ocean during deposition of
strata (Morris 1979). the ORS (?latest Silurian or Early D e v o n i a n - -
Prehnite-pumpellyite facies metamorphism see earlier discussion) is the simplest explana-
occurs in basalts and basic-clast greywackes in tion for this deformation.
Longford-Down (Oliver 1978) and the South-
ern Uplands (Oliver & Leggett 1980). Assemb-
lages of the greenschist and zeolite facies have Discussion
not been found. Detrital prehnite indicates Mechanism of offscraping
sedimentary recycling of metamorphic minerals
and suggests that the burial metamorphism The level of initial d6collement during accre-
derives from 'tectonic' burial associated with tion of individual tracts in the Southern Up-
accretion (Oliver & Leggett 1980). lands was fundamentally influenced by the
Detrital glaucophane in late Llandeilo or black shales of the Moffat Shale Group, which
Caradoc greywackes from Cavan (Sanders & provided an incompetent level in the subduct-
Morris 1978) and the west coast of the Southern ing ocean floor/trench sequences. Black shales
Uplands (Kelling 1962; Walton 1956) may indi- are thin or are absent in the entirely Ordovician
cate pre-existing (?Cambrian or early Ordovi- sequences of the Northern Belt (Fig. 4). D6col-
cian) subduction terranes, eroded to deep lement occurred at the top of, or within, chert
levels, exposed along the Laurentian margin. sections, or locally within basalts in sequence 1
Glaucophane-bearing clasts are exposed in where no black shales are developed. Cherts
local m61ange outcrops in the Ballantrae and basalts are also preserved locally below thin
ophiolite, but their geological significance is as black shales in the other Ordovician tracts. In
yet unclear (summary in Bluck 1978). the Central Belt the thickness of the Moffat
Shale Group increases due to the diachronism
of the greywackes, and all sequences were de-
Structure of the Midland Valley tached at the top, within, or at the base of the
The structure of the post-ophiolite Ordovi- black shales (Fig. 4). Underlying ocean floor
cian and Silurian strata of the Midland Valley is lithologies were presumably subducted or
varied, though much less complicated than accreted at deeper levels: an example of selec-
structure in the Southern Uplands. North- tive subduction (Moore 1975). D6collement
wards-directed, SE-inclined thrusts in the Gir- occurred within greywacke sequences in the
van succession (Williams 1962, Fig. l lb) can southern part of the Central Belt and the
only be dated as 'post-Ordovician'. Weir (1979) Southern Belt. The ocean-floor clastic pile sub-
relates this thrusting to northwards obduction ducting at that time was probably too thick to
of the whole ophiolite plus sedimentary cover enable offscraping of complete sections.
on to Laurentian continental basement some- During detachment of each packet of strata
time towards the end of the Silurian. (tract) folding and low-angle thrusting were
Folding varies in intensity along the Midland essentially synchronous. Folding was probably
Valley. Strata in the Lesmahagow inlier are accomplished mostly during the early stages of
folded into a gentle open asymmetric anticline accretion and uplift. The intensity of imbrica-
(Fig. l lC). The Hagshaw Hills rocks form a tion, both in zones along tract-bounding major
tight asymmetric anticline with a reverse- faults and within greywacke terranes, indicates
faulted axis (Fig. l i D ) . In the Pentland Hills that deformation within tracts was accom-
and in much of the Girvan area, strata are plished by continued strike faulting (reverse or
steeply inclined without obvious fold axes. thrust motion) after strain-hardening and dur-
Cleavage is absent. Ingham (1978) ascribes ing rotation of tracts within the prism. We
local intense folding in thin-bedded sandstones analyse the relationship between folding and
of the Ardwell Group (Fig. 4) near Girvan (Fig. faulting in more detail elsewhere (Leggett &
3) to slumping. Casey, in press).
Folding of forearc strata occurred at some
stage during deposition of the Lower Old Red Controls on structural style
Sandstone (ORS), because ORS beds are
folded with Silurian in the Lesmahagow and Accretionary complexes dominated by co-
Hagshaw Hills inliers, but are unconformable herent strata tend to develop above modern
on Silurian in the Pentland Hills inlier (Fig. 11). subduction zones along which sediment input is
Southwards vergence of the folds and north- high and convergence is fast (Cowan & Silling
wards tectonic transport on the thrusts (Fig. 11) 1978; Moore 1979). The Southern Uplands is
suggest compression from the south. Final clo- devoid of m61ange, despite intense imbrication.
514 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey
Sediment input to the subduction zone during the Midland Valley forearc basin may have
the accretionary phase was undoubtedly high. been bypassed further to the NE, in the portion
Stratigraphy of greywacke sequences in the of the margin now buried under the North Sea.
Central and Southern Belts shows that for much Most Silurian greywackes show a dominance of
of Silurian time thick ocean-floor clastics were currents from the NE (Fig. 4; Walton 1965)
being fed into the Scottish subduction zone. i.e. parallel to the trench axis (Fig. 7),
Rates calculated from published estimates of
thickness in greywacke units (excluding com- Pre-middle Ordovician history of the
paction effects) are c. 165-185 m Ma -1 (Leggett northwestern lapetus margin in Britain
1980a). Slow convergence is indicated by the
relatively slow development of the Southern Volcanic and plutonic evidence already out-
Uplands accretionary complex compared with lined indicates that Iapetus oceanic crust was
the Sumatran margin (Fig. 5 and previous sec- subducting under the British sector of the
tion). So the essentially coherent nature of Laurentian margin in late Cambrian and early
accreted strata in the Southern Uplands can Ordovician times. By implication, the require-
readily be explained in terms of what Moore ment for a subduction-related mechanism for
(1979) considers the two most important con- the Grampian orogeny (Dewey 1971; Phillips
trois in structural style in modern accretionary et al. 1976) supports this. Accretion did not
complexes. begin in the Southern Uplands until mid-
Other parameters were undoubtedly impor- Ordovician, and there is no evidence for an
tant. The effect of stratigraphy in the subduct- appreciable older accretionary terrane under
ing sequences has already been described. Obli- the Midland Valley. What happened in the
que convergence (Thirlwall 1981a) may further forearc before onset of accretion in the South-
have facilitated offscraping of coherent strata. ern Uplands?
By analogy with processes observed on mod-
Controls on sedimentation
ern active margins (summary in Scholl et al.
1981) there are three possible explanations.
Most sediment in the early Silurian forearc First, the margin may have been non-
basin was derived from the south, off the rising accretionary; any sediment accumulating on the
trench slope break of the Southern Uplands, trench or coming into the trench on the down-
accretionary complex. Yet sand finding its way going ocean plate may have been subducted.
to the trench and ocean floor at the same time Secondly, the subducting ocean plate may have
compositionally reflects the arc/orogen hinter- mechanically eroded the over-riding Laurentian
land to the north (e.g. Walton 1965). This margin prior to the Southern Uplands accre-
introduces two problems: why does the frontal tionary phase. Thirdly, strike-slip movements
arc/magmatic arc not contribute more sediment on major faults sub-parallel to the margin may
to the Midland Valley forearc basin, and how have removed older (Cambrian and early
does the sediment bypass the forearc basin to Ordovician) forearc elements.
reach the ocean floor? Unfortunately, Lower There is evidence for missing early Ordovi-
Palaeozoic strata under the northern part of the clan stratigraphy. As already outlined, the up-
Midland Valley are buried beneath Upper per slope succession on the Ballantrae ophiolite
Palaeozoic cover. ORS sediment in the Strath- includes thick northerly derived conglomerates
more syncline along the northern border of of Llanvirn and Llandeilo age. According to
the Midland Valley are some 10 km thick Williams (1962) some of these are of shelf- and
(Armstrong & Paterson 1970). This immense some of slope-origin. Age-equivalent sediments
thickness indicates a substantial depression, to the south, in the early accreted tracts in the
possibly inherited from Silurian time and in Southern Uplands, are pelagic cherts, siliceous
some way linked to the lack of northerly- mudstones and thin metabentonites. Facies in-
derived material in the southern part of the termediate between these conglomerates and
Silurian forearc basin. cherts are not preserved. Furthermore, the
The volcaniclastic nature of Silurian flysch in earliest accreted slice immediately south of the
sequences accreted in the Llandovery is not ophiolite was not emplaced until gracilis times,
such a puzzle. Much of the sediment may have more than 10Ma after deposition of upper
been derived from previously accreted Ordovi- slope sediment began on top of the ophiolite
cian flysch (Fig. 4, currents from the NW). (Figs 4 & 5). The missing stratigraphy may be
Additionally, forearc basins along modern ac- explained by strike-slip removal of Llanvirn-
tive margins are rarely continuous for long Llandeilo accreted tracts and slope sediment,
distances (e.g. Dickinson & Seely 1979), and but sediment subduction is equally likely.
 Anatomy o f a L o w e r P a l a e o z o i c accretionary f o r e a r c 515

Subduction erosion may also have occurred. normal ocean crust (Lewis 1976). Voluminous
Longman et al. (1979) cite dated granitic and monomict basaltic breccias are intercalated in
metamorphic boulders in a Llanvirn conglomer- the marie sequences (Lewis & Bloxam 1977);
ate from the supra-ophiolitic succession as evi- rounded pebbles therein and intense vesicular-
dence for a Cambrian-early Ordovician magma- ity in pillows indicate eruption in relatively
tic arc under the SW Midland Valley. No scale shallow water. There is no appreciable sheeted
is shown on their diagrammatic reconstruction, dyke complex.
but clearly the arc-trench gap would be far too Bluck et al. (1980), following Lewis & Blox-
narrow (a few tens of kilometres) to allow am (1977), use these features to interpret the
subduction from th" line of the Southern Up- marie sequences in the ophiolite as part of an
land Fault, no matter how steep the subduction island arc assemblage. Dating construction of
zone. Although strike-slip removal of a pre- the ophiolite (from gabbro and trondjemite)
middle Ordovician forearc terrane is again a and emplacement (from amphibolite and an
possibility, mechanical erosion of the overrid- associated middle Arenig olistrostrome), Bluck
ing Laurentian plate during Llanvirn-Llandeilo et al. (1980) show only a limited time gap
times may also be responsible. Lower ORS between the two. This, they argue, indicates
andesites are also anomalously close to the that elements of a short-lived marginal basin,
suture in the southern Midland Valley, giving closure of which emplaced the ophiolite on the
an arc-trench gap of only c. 70 km (Fig. 2). Laurentian margin, are also included. Barrett
Thirlwall (1981a) invokes a steep subduction et al. (in press) stress the lack of evidence for
zone associated with the late stages of closure to what should have been an appreciable magma-
explain this. Subduction erosion during latest tic arc terrane south of the ophiolite, such as
Silurian and/or Early Devonian may equally those associated with Ordovician ophiolites in
well be responsible, and would perhaps explain Newfoundland (Fig. 1). They prefer to inter-
the lack of Early Devonian marine strata in the pret the Ballantrae ophiolite as accreted slivers
southern part of the accretionary complex if the of ocean floor volcanic edifices. This explana-
Lower ORS is indeed of Early Devonian age. tion is consistent with the atypical features of
There is no evidence for the missing Cam- the ophiolite and explains its forearc setting
brian-early Ordovician forearc terrane in Ire- (Laurentian ophiolites in Newfoundland are
land, and total strike-slip displacement of sever- demonstrably of back-arc origin: see Dewey
al 100 km would be necessary to remove it to 1974 and Williams 1978). It also explains the
the submerged extension of the Caledonides on intense disruption of the ophiolite by high-angle
the British continental shelf (assuming oblique reverse faults. Intermittent accretion of high-
ESE-WNW convergence at this time--Fig. 1 standing ocean crust slivers, possibly in a reg-
and Thirwall 1981a). We suspect that strati- ime of overall sediment subduction, would also
graphic links across the Southern Upland Fault explain the division of the ophiolite into Arenig
do not allow this amount of movement along marie sequences north of the Stinchar Fault and
the south border of the Midland Valley after a late Llandeilo or early Caradoc sequence
the mid-Ordovician. The unique late Llandeilo- (Downan marie sequence, Fig. 4) to the south.
early Caradoc flysch rudites at Corsewall Point Volcanism on the ocean plate just prior to
and in Glen App (sequence I, Fig. 4) occur accretion in the Arenig would explain the nar-
immediately south of the Ballantrae ophiolite row time gap between construction of the
and contain clasts 90% of which can be matched ophiolite and emplacement.
with lithologies exposed in the ophiolite Geochemistry of marie rocks does not help
(Walton 1956; Kelling & Holroyd 1978). Petro- with the debate as to whether the ophiolite was
graphy can similarly be linked between boul- generated in an arc, composite arc/marginal
ders in Wenlock conglomerates of the Midland basin, or ocean floor environment. On balance,
Valley forearc succession and lithologies ex- arguments from regional geology best support
posed in the Northern Belt, south of the South- an open-ocean origin for the ophiolite.
ern Upland Fault. So any strike-slip movements Finally, how is the latest Cambrian/early
must have occurred in Cambrian and early Ordovician Grampian orogeny explained? Evi-
Ordovician times. dence for a Cambrian/early Ordovician arc
Clearly the geological significance of the Bal- under the SW Midland Valley makes an arc/
lantrae ophiolite bears heavily on this problem. continent collision one of the more attractive
The ophiolite has several atypical features. theories for the Grampian Orogeny in Scotland
Differentiated lavas, as well as tholeiites, occur (Longman et al. 1979). However, this does not
in the pillow lava and massive basalt sequences, readily explain the Grampian deformation in
which are anomalously thick compared with the west of Ireland (Connemara Dalradian, Fig.
516 J. K. Leggett, W. S. M c K e r r o w & D. M. Casey

2) where the Midland Valley-type basement logy either side of the major faults exposed
rocks appear to be missing. It is possible that (Fig. 2) to allow strike-slip displacements on
they have been removed by strike-slip faulting. the scale of those in Cenozoic western North
In summary, pre-middle Ordovician rocks in America. Across the Great Glen Fault, not-
southern Scotland and northern Ireland give a withstanding limited post-Early Devonian pre-
tantalizingly incomplete picture of subduction- Carboniferous strike-slip displacement (sum-
related events along the Laurentian margin. mary in Johnson et al. 1979) stratigraphy, struc-
Sediment subduction, erosion, and strike-slip ture and metamorphism in Precambrian rocks
translation by their very nature inevitably result are very similar. The Highland Boundary Fault
in 'lost' stratigraphy and other complexities. is not traceable west of the Midland Valley, and
For this reason we do not expect to be able to seems to be a local feature. Early Palaeozoic
reconstruct pre-middle Ordovician events as strike-slip movements may well have occurred
confidently as we do mid-Ordovician to late along the Laurentian margin in Britain, but the
Silurian events. fault or faults responsible are not obvious.
Further assessment of this possibility must
Effects of closure of the Iapetus await palaeomagnetic work.

Compared with the intense deformation Basement below southern Scotland

which produced nappes in Norway, Greenland,
and Shetland, the Southern Uplands seems to Seismic velocities indicate that the basement
have escaped closure-related tectonism when below the Midland Valley comprises 6-14 km
the Laurentian and European continents col- of Lower Palaeozoic rocks, already discussed,
lided. All major (Dl) structures (reverse faults and a layer of pre-Caledonian high-grade meta-
and associated folds) feasibly relate to accretion morphic rocks extending to 20 km depth (Barn-
(Anderson & Cameron 1979; Eales 1979: Leg- ford et al. 1977, Bamford 1979). This metamor-
gett et al. 1979a; Weir 1979). Anderson & phic layer, with seismic velocities of 6.4 km s -I
Cameron (1979) ascribe the D2 deformation, abuts against a 6.3 km s -~ layer at the Southern
which is mesocopic only, to closure. Other Upland Fault. This 6.3 km s -~ layer under the
possible effects, summarized by Johnson et al. Southern Uplands is overlain by c. 14-16 km of
(1979), are thrust motion during initial move- 5.8-6.0 km s-1 rocks interpreted as greywackes,
ments on the Highland Boundary Fault and and may be pre-Caledonian basement but this
along the Moine Thrust, transcurrent move- is not certain (Bamford 1979, p. 94). Gneissic
ment on faults between these two, and folding xenoliths in Upper Palaeozoic vents confirm
in the Midland Valley. crystalline basement below the Midland Valley,
Relative lack of closure-related deformation but have not as yet been found in Carboniferous
in the Southern Uplands could be due to dia- vents in the Southern Uplands. Strogen (1974)
chronous collision, which according to Phillips has found such xenoliths in Carboniferous vents
et al. (1976) occurred earlier in the NE and immediately along strike from the Southern
produced considerable dextral strike-slip along Uplands in central Ireland. This might indicate
the suture. Thirlwall (1981a), on the other a local prong of continental basement at depth,
hand, uses geochemical data in Lower ORS perhaps explaining the greater overturning of
volcanics to argue for oblique ESE-WSW con- strata in the western part of Longford-Down
vergence and for a swing in the suture line to massif (Leggett et al. 1979b). It could equally
north-south under the North Sea. This is an well indicate continental basement below the
attractive hypothesis, allowing head-on colli- whole accretionary terrane.
sion, and associated nappe formation, in Nor- Evidence already outlined indicates that the
way, Greenland and Shetland, and producing Southern Uplands rocks derive originally from
plate motion almost parallel to the northern an ocean-floor/trench environment and cannot
British portion of the margin. Consequent slow have been deposited in situ on a continental
subduction under the Southern Uplands would substrate. If the 6 . 3 k m s -1 layer below the
allow preservation of primary structures in the Southern Uplands is older continental material,
accretionary complex. Vein arrays near the we face the intriguing possibility that the entire
suture indicate both dextral and sinistral move- fossil accretionary complex is allochthonous. If
ments (Anderson, pers. comm. 1981) and this that did prove to be the case, the most likely
topic requires further investigation. explanation would be obduction of the accre-
Despite this evidence for oblique relative tionary terrane, possibly including the Ballan-
plate motion we suspect that, except along the trae ophiolite and its Ordovician cover (cf.
suture, there are too many similarities in geo- Weir 1979), over thinned continental basement,
 Anatomy of a Lower Palaeozoic accretionary forearc 517

as a result of closure of Iapetus. Such a process to have h a p p e n e d b e l o w the N e w f o u n d l a n d

n e e d involve n o r t h w a r d s translation of only sector of the o r o g e n ( S h e r i d a n 1974, figs 3 &
several tens of k i l o m e t r e s . R e c e n t seismic re- 5). This d o e s not n e g a t e the s u b d u c t i o n -
flection w o r k in the S o u t h e r n A p p a l a c h i a n s accretion e x p l a n a t i o n for s o u t h e r n Scotland.
shows that a l r e a d y - d e f o r m e d t e r r a n e s can be W e suggest that it is entirely possible for an
t r a n s l a t e d several h u n d r e d s of k i l o m e t r e s over a c c r e t i o n a r y c o m p l e x to be g e n e r a t e d by sub-
crystalline b a s e m e n t d u r i n g collision ( C o o k d u c t i o n of an o c e a n of which all o t h e r traces
et al. 1979). w e r e o b l i t e r a t e d on closure.
F u r t h e r indications of a n o m a l o u s crustal F u r t h e r r e s o l u t i o n of crustal s t r u c t u r e b e l o w
structures b e l o w the S o u t h e r n U p l a n d s are the s o u t h e r n S c o t l a n d awaits C O C O R P - t y p e d e e p
lack of a clearly d e f i n e d M o h o ( B a m f o r d seismic reflection profiling.
1979) and the o c c u r r e n c e of a block of rock of
ACKNOWLEDGMENTS:We are grateful to Mike Audley-
high electical c o n d u c t i v i t y e x t e n d i n g well into
Charles for reviewing the manuscript, and to Bernard
the m a n t l e ( H u t t o n et al. 1980). H o w e v e r , t h e r e Anderson, Alistair Robertson, and the participants
is no e v i d e n c e in seismic w o r k ( B a m f o r d 1979) on the 1980 Geological Society field trip to the
or gravity w o r k (Powell 1971) that a thick relic Southern Uplands for many helpful discussions. JKL
of Iapetus o c e a n crust has b e e n p r e s e r v e d in the & DMC thank NERC for financial assistance during
p r e s e n t crustal profile ( D e w e y 1971), as s e e m s fieldwork.

References
ANDERSON, T. B. 1980. Deformation sequences in the Caledonian Orogeny and Ordovician time scale.
Southern Uplands. Scott. J. Geol. 17, 78-80. Geology, 8, 492-5.
ANDERSON, T. B. & CAMERON, T. D. J. 1979. A BORRADAILE, G. J. 1978. Transected folds: a study
structural profile of Caledonian deformation in illustrated with examples from Canada and Scot-
Down. In: HARRIS, A. L., HOLLAND, C. H. & land. Bull. geol. Soc. Am. 89, 481-93.
LEAKE, B. E. (eds). The Caledonides of the BROWN, G. C. 1979. Geochemical and geophysical
British Isles--reviewed. Spec. Publ. geol. Soc. constraints on the origin and evolution of
London, 263-7. Caledonian granites. In: HARRIS, A. L.,
ARMSTRONG, M. & PATERSON, I. B. 1970. The Lower HOLLAND, C. H. & LEAKE, B. E. (eds). The
Old Red Sandstone of the Strathmore region. Caledonides of the British Isles--reviewed. Spec.
Rep. No. 70/12, Inst. Geol. Sci. Publ. geol. Soc. London, 645-51.
BACHMAN, S. B. 1981. The Coastal Belt of the CAMERON, T. D. J. & ANDERSON,T. B. 1979. Silurian
Franciscan: youngest phase of northern Califor- metabentonites in County Down, Northern Ire-
nia subduction (this volume). land. Geol. J. 15, 5%75.
BAMFORD, D. 1979. Seismic constraints on the deep CLARKSON,C. M., CRAIG,G. Y. & WALTON,E. K.
geology of the Caledonides of northern Britain. 1975. The Silurian rocks bordering Kirkcud-
In: HARRIS, A. L., HOLLAND, C. H. & LEAKE, bright Bay, South Scotland. Trans. R. Soc. Edin-
B. E. (eds). The Caledonides of the British Isles-- burgh, 69, 313-25.
reviewed. Spec. Publ. geol. Soc. London, 93--6. COCKS, L. R. M. ~ TOGHILL, P. 1973. The biostra-
-- , NUNN, K., PRODEHL, C. & JACOB, B. 1977. tigraphy of the Silurian rocks of the Girvan
LISPB-III. Upper crustal structure of Northern district, Scotland. J. geol. Soc. London, 129,
Britain. J. geol. Soc. London, 133, 481-8. 209-43.
BARRErf, T. J., JENKYNS, H. C., LEGGETT, J. K. & COOK, D. R. & WEIR, J. A. 1979. Structure of the
ROBERTSON, A. H. F. 1981. Age and origin of Lower Palaeozoic rocks around Cairnsmore of
Ballantrae ophiolite and its significance to the Fleet, Galloway. Scott. J. Geol. 15, 187-202.
Caledonian orogeny and Ordovician time scale: COOK, F. A., ALBAUGH, D. S., BROWN, L. D.,
Comment. Geology, in press. KAUFMAN, S., OLIVER, J. E. ~ HATCHER, R. D.,
BERGSTROM, S. M. 1971. Conodont biostratigraphy of JR. 1979. Thin-skinned tectonics of the crystal-
the Middle and Upper Ordovician of Europe and line Appalachians: COCORP seismic reflection
eastern North America. In: SWEET, W. C. t~ profiling of the Blue Ridge and Piedmont. Geol-
BERGSTROM, S. M. (eds). Symposium on Con- ogy, 7, 563-7.
odont Biostratigraphy. Mere. geol. Soc. Am. COWAN, D. S. 1981. Deformation of partly dewatered
127, 83-162. and consolidated Franciscan sediments near
BLUCK, B. J. 1978. Geology of a continental margin Piedras Blancas Point, California (this volume).
1: The Ballantrae Complex. In: BOWES, D. R. & & SILLING, R. M. 1978. A dynamic, scaled
LEAKE, B. E. (eds). Crustal Evolution in North- model of accretion at trenches and its implica-
western Britain and Adjacent regions. Geol. J. tions for the tectonic evolution of subduction
Spec. Issue No. 10. 151-62. complexes. J. geophys. Res. 83, 5389-96.
, HALLIDAY, A. N. AFTALION, M., & CRAIG, G. Y. & WALTON, E. K. 1959. Sequence and
MACINTYRE, R. M. 1980. Age and origin of structure in the Silurian rocks of Kirkcudbright-
Ballantrae ophiolite and its significance to the shire. Geol. Mag. 96, 209-20.
518 J. K. Leggett, W. S. McKerrow & D. M. Casey
- - & 1962. Sedimentary structures and HEPWORTH, B. C., OLIVER, G. J. H. & MCMURTRY,
palaeocurrent directions from the Silurian rocks M. J. 1981. Sedimentology, volcanism, structure
of Kircudbrightshire. Trans. Edinb. geol. Soc. and metamorphism of the Lower Palaeozoic
19, 100-19. accretionary complex; Bail Hill-Abington area
CURRAY, J. R., MOORE, D. G., LAWVER, L. A., of the Southern Uplands of Scotland (this
EMMEL, F. J., RAITr, R. W., HENRY, M. & volume).
KIECKHEFER, R. 1979. Tectonics of the Andaman HUTrON, V. R. S., INGHAM, M. R. & MBIPOM, E. W.
Sea and Burma. Mem. Am. Assoc. Petrol. Geol. 1980. An electrical model of the crust and upper
29, 189-98. mantle in Scotland. Nature, London, 287, 30-2.
DEWEY, J. F. 1969. Evolution of the Appalachian/ INGHAM, J. K. 1978. Geology of a continental margin
Caledonian orogen. Nature, London, 222, 124- 2: middle and late Ordovician transgression,
9. Girvan. In: BOWLS, D. R. & LEAKE, B. E. (eds).
- - 1971. A model for the Lower Palaeozoic evolu- Crustal Evolution in Northwestern Britain and
tion of the early Caledonides of Scotland and Adjacent Regions. Geol. J. Spec. Issue No. 10,
Ireland. Scott. J. Geol. 7, 219-40. 163-76.
1974. Continental margins and ophiolite obduc- JOHNSON, M. R. W . , SANDERSON, D. J. • SOPER,
tion: Appalachian Caledonian System. In: BURK, N. J. 1979. Deformation in the Caledonides of
C. A. & DRAKE, C. L. (eds). The Geology of England, Ireland and Scotland. In: HARRIS,
Continental Margins, 933-50. Springer-Verlag, A. L., HOLLAND, C. H. & LEAKE, I . E. (eds).
New York. The Caledonides of the British Isles--reviewed.
& KIDD, W. S. F. 1974. Continental collisions Spec. Publ. geol. Soc. London, 165-86.
in the Appalachian Caledonian orogenic belt. JONES, J. G. 1969. Pillow lavas as depth indicators.
Variations related to complete and incomplete Am. J. Sci. 267, 181-95.
suturing. Geology, 2, 543-6. KARIG, D. E., LAWRENCE, M. B., MOORE, G. F. &
DICKINSON, W. R. t~L SEELY, D. R. 1979. Structure CURRAY, J. R. 1980. Structural framework of the
and stratigraphy of forearc regions. Bull. Am. fore-arc basin, NW Sumatra. J. geol. Soc. Lon-
Assoc. Petrol. Geol. 63, 2-31. don, 137, 77-91.
EALES, M. U. 1979. Structure of the Southern Up- KEELING, G. 1961. The stratigraphy and structure of
lands. In: HARRIS, A. L., HOLLAND, C. H. & the Ordovician rocks of the Rhinns of Galloway.
LEAKE, B. E. (eds). The Caledonides of the Q. J. geol. Soc. London, 117, 37-75.
British Isles--reviewed. Spec. Publ. geol. Soc. 1962. The petrology and sedimentation of the
London, 269-73. Upper Ordovician rocks of the Rhinns of Gallo-
ECKFORD, R. J. A. & RITCHIE, M. 1931. The lavas of way, south-west Scotland. Trans. R. Soc. Edin-
Tweeddale and their position in the Caradocian burgh, 65, 107-37.
sequence. Suture. Prog. geol, Surv. G.B. Part 3, 81, HOEROYD, J. 1978. Clast size, shape, and
46-57. composition in some ancient and modern fan
FLOYD, J. D. 1976. The Ordovician rocks of West gravels. In: STANLEY, D. J. & KEELING, G. (eds).
Nithsdale. Thesis, PhD. Univ. St Andrews (un- Sedimentation in Submarine Canyons, Fans and
publ.). Trenches, 138-59. Dowden, Hutchinson & Ross,
FY~E, T. B. & WEre, J. A. 1976. The Ettrick Valley Stroudsburg, Pennsylvania.
Thrust and the upper limit of the Moffat Shales --& WELSH, W. 1970. The Loch Ryan Fault.
in Craigmichan Scaurs (Dumfries and Galloway Scott. J. Geol. 6, 266-71.
Region: Annandale and Eskdale District). Scott. KULM, L. D., RESIG, J. M., THORNBURG, T. M. &
J. Geol. 12, 93-102. SCHRADER, H.-J. 1981. Cenozoic structure, stra-
GORDON, A. J. 1962. The Lower Palaeozoic rocks tigraphy and tectonics of the central Peru forearc
around Glenluce, Wigtonshire. Thesis, PhD. (this volume).
Univ. Edinburgh (unpubl.). LAMBERT, R. ST. J., HOLLAND, J. G. & LEGGETT,J. K.
GRAHAM, A. M. & UPTON, B. G. J. 1978. Gneisses in 1981. Petrology and tectonic setting of some
diatremes, Scottish Midland Valley: petrology Ordovician volcanic rocks from the Southern
and tectonic implications. J. Geol. Soc. London, Uplands of Scotland. J. geol. Soc. London, 138,
135, 219-28. 421-36.
HARRIS, A. L., HOLLAND, C. H. & LEAKE, B. E. LAMONT, A. & LINDSTROM, M. 1957. Arenigian and
(eds). 1979. The Caledonides of the British Isles-- Llandeilian cherts identified in the Southern
reviewed. Spec. Publ. geol. Soc. London. Scot- Uplands of Scotland by means of conodonts etc.
tish Academic Press, Edinburgh. 768 pp. Trans. geol. Soc. Edinburgh, 17, 60-70.
HENDERSON, W. G. t~ ROBERTSON, A. H. F. 1981. LEGGETT, J. K. 1978. Eustacy and pelagic regimes
The Highland Border rocks and their relation to in the Iapetus Ocean during the Ordovician and
marginal basin development in the Scottish Cale- Silurian. Earth planet. Sci. Lett. 41, 163-9.
donides. Geol. Soc. Newsl. London, 10, (2), -- 1979. Oceanic sediments from the Ordovician
23-4. of the Southern Uplands. In: HARRIS, A. L.,
HENRIKSEN, N. & HIGGINS, A. K. 1976. East Green- HOLLAND, C. H. & LEAKE, B. E. (eds). The
land Caledonian fold belt. In: ESCHER, A. & Caledonides of the British Isles--reviewed. Spec.
WATT, W. S. (eds). Geology of Greenland, 182- Publ. geol. Soc. London, 495-8.
246. Geological Survey of Greenland. --1980a. The sedimentological evolution of a
 Anatomy of a Lower Palaeozoic accretionary forearc 519
Lower Palaeozoic accretionary fore-arc in the - -& COCKS, L. R. M. 1976. Progressive faunal
Southern Uplands of Scotland. Sedimentology, migration across the Iapetus Ocean. Nature,
27, 401-17. London, 263, 304-6.
1980b. Palaeogeographic setting of the Wrae , LAMBERT, R. ST.J. • CHAMBERLAIN, V. E.
Limestone: an Ordovician submarine slide de- 1980. The Ordovician, Silurian and Devonian
posit in Tweeddale. Scott. J. Geol. 16, 91-104. time-scale. Earth planet. Sci. Lett. 51, 1-8.
1980c. British Lower Palaeozoic black shales , LEGGErr, J. K. & EALES, M. H. 1977. Imbri-
and their palaeo-oceanographic significance. cate thrust model of the Southern Uplands of
J. geol. Soc. London, 137, 139-56. Scotland. Nature, London, 267, 237-9.
-- & CASEY, D. M. 1981. The Southern Uplands MITCHELL, A. H. G. ~ MCKERROW, W. S. 1975.
accretionary prism: implications for controls on Analogous evolution of the Burma orogen and
structural development of subduction com- the Scottish Caledonides. Bull. geol. Soc. Am.
plexes. In: DRAKE, C. L. & WATKINS,J. S. (eds). 86, 305-15.
Continental Margin Processes. Spec. Publ. Am. MOORE, G. F., CURRAY, J. R. ~ EMMEL, F. J. 1981.
Assoc. Petrol. Geol. (in press). Sedimentation in the Sunda Trench and Forearc
--, MCKERROW, W. S., COCKS, L. R. M. & region (this volume).
RICKARDS, R. B. 1981. Periodicity in the early & KARIG, D. E. 1980. Structural geology of Nias
Palaeozoic marine realm. J. geol. Soc. London, Island, Indonesia: implications for subduction
138, 167-76. zone tectonics. Am. J. Sci. 280, 193-223.
- - , MCKERROW, W. S. & EALES, M. H. 1979a. MOORE, J. C. 1975. Selective subduction. Geology, 3,
The Southern Uplands of Scotland; a Lower 530--2.
Palaezoic accretionary prism. J. geol. Soc. Lon- 1979. Variation in strain and strain rate during
don, 136, 755-70. underthrusting of trench deposits. Geology, 7,
- - , MCKERROW, W. S.., MORRIS, J. H., OLIVER, 185-8.
G. J. H. & PHILLIPS, W. E. A. 1979b. The -- & KARm, D. E. 1976. Sedimentology, structural
north-western margin of the Iapetus Ocean. In: geology, and tectonics of the Shikoku subduction
HARRIS, A. L., HOLLAND, C. H. & LEAKE, B. E. zone, southwestern Japan. Bull. geol. Soc. Am.
(eds). The Caledonides of the British Isles-- 87, 1259-68.
reviewed. Spec. PuN. geol. Soc. London, 499- -- & WATKINS, J. S. et al. 1979. Progressive accre-
511. tion in the Middle America Trench, Southern
-- t~ SMITH, T. K. 1980. Fe-rich deposits associ- Mexico. Nature, London, 281,638--42.
ated with Ordovician basalts in the Southern & 1981. Tectonic processes along the
Uplands of Scotland: possible Lower Palaeozoic Middle America Trench inner slope (this
equivalents of modern active ridge sediments. volume).
Earth planet. Sci. Lett. 47, 431-40. MORRIS, J. H. 1979. Lower Palaeozoic soft-sediment
LEWIS, A. D. 1976. The geochemistry and geology of deformation structures in the western end of the
Girvan-Ballantrae ophiolite and related Ordovi- Longford-Down inlier, Ireland. In: HARRIS,
cian volcanics in the Southern Uplands of Scot- A. L., HOLLAND, C. H. & LEAKE, B. E. (eds).
land. Thesis, PhD. Univ. Wales (unpubl.). The Caledonides of the British Isles--reviewed.
LEWIS, A. D. & BLOXAM, T. W. 1977. Petrotectonic Spec. Publ. Geol. Soc. London, 513-6.
environments of the Girvan-Ballantrae lavas OLIVER, G. J. H. 1978. Prehnite--pumpellyite facies
from rare earth element distributions. Scott. J. metamorphism in County Cavan, Ireland. Na-
Geol. 13, 211-22. ture, London, 274, 242-3.
LONGMAN, C. D., BLUCK, B. J. ff~;VAN BREEMEN, O. - - & L E ~ E r r , J. K. 1980. Metamorphism in
1979. Ordovician conglomerates and evolution an accretionary prism: prehnite-pumpellyite
of Midland Valley. Nature, London, 280, 578-81. facies metamorphism of the Southern Uplands
LUMSDEN, G. I., TULLOCH, W., HOWELLS, M. F. & of Scotland. Trans. R. Soc. Edinburgh, 71,
DAVIES, A. 1967. The geology of the neighbour- 235-46.
hood of Langholm. Mere. geol. Surv. Scotland. PEACH, B. N. & HORNE, J. 1899. The Silurian rocks of
Sheet 11. 225 pp. Britain, 1, Scotland. Mere. geol. Surv. Scotland,
McGIVEN, A. 1968. Sedimentation and provenance of 749 pp.
post-Valentian conglomerates up to and including PHILLIPS, W. E. A., STILLMAN, C. J. & MURPHY, T.
the basal conglomerate of the Lower Old Red 1976. A Caledonian plate tectonic model.
Sandstone in the southern part of the Midland J. geol. Soc. London, 132, 579-609.
Valley of Scotland. Thesis, PhD. Univ. Glasgow -- , FLEGG, A. M. & ANDERSON, T. B. 1979. Strain
(unpubl.). adjacent to the Iapetus suture in Ireland. In:
MCKERROW, W. S. 1979. Ordovician and Silurian HARRIS, A. L., HOLLAND, C. H. & LEAKE, B. E.
changes in sea level. J. geol. Soc. London, 136, (eds). The Caledonides of the British Isles--
137-45. reviewed. Spec. Publ. geol. Soc. London, 257-
- - 1 9 8 1 . The Northwest margin of the Iapetus 62.
Ocean during the early Palaeozoic. In: DRAKE, POWELL, D. W. 1971. A model for the Lower
C. L. & WATKINS, J. S. (eds). Continental Mar- Palaeozoic evolution of the southern margin of
gin Processes. Spec. Publ. Am. Assoc. Petrol. the early Caledonides of Scotland and Ireland.
Geol. (in press). Scott. J. Geol. 7, 369-72.
520 J. K . L e g g e t t , W . S. M c K e r r o w & D. M. Casey

- - 1978. Geology of a continental margin 3: grav- TOGHILL, P. 1970a. Highest Ordovician (Hartfell
ity and magnetic anomaly interpretation of the Shales) graptolite faunas from the Moffat area,
Girvan-Ballantrae district. In: BOWES, D. R. & south Scotland. Bull. Br. Mus. nat. Hist., Geol.
LEAKE, B. E. (eds). Crustal Evolution in North- 1-26. 1 9 ,

western Britain and Adjacent Regions. Geol. J. 1970b. The south-east limit of the Moffat Shales
Spec. Issue No. 10, 177-82. in the upper Ettrick valley region, Selkirkshire.
ROLEE, W. D. I. 1961. The geology of the Hagshaw Scott. J. Geol. 6, 233-42.
Hills Silurian Inlier. Trans. Edinb. geol. Soc. 18, WALTON, E. K. 1955. Silurian greywackes in Peeble-
240-69. shire. Proc. R. Soc. Edinb. B65, 327-57.
RusT, B. R. 1965. The stratigraphy and structure of 1956. Two Ordovician conglomerates in South
the Whithorn area of Wigtonshire, Scotland. Ayrshire. Trans. geol. Soc. Glasgow, 22, 133-
Scott. J. Geol. 1, 101-33. 56.
RYAN, P. D. & ARCHER, J. B. 1977. The South Mayo - 1961. Some aspects of the succession and struc-
-

Trough: a possible Ordovician Gulf of Califor- ture in the Lower Palaeozoic rocks of the South-
nia-type marginal basin in the west of Ireland, ern Uplands of Scotland. Geol. Rdsch. 50, 63-
Can. J. Earth Sci. 14, 2453-6l. 77.
SANDERS, I. S. & MORRIS, J. H. 1978. Evidence for 1965. Lower Palaeozoic rocks: stratigraphy,
Caledonian subduction from greywacke detritus palaeogeography and structure. In: CRAIG, G. Y.
in the Longford-Down inlier. J. Earth Sci. R. (ed.) The Geology of Scotland, 161-227. Oliver
Dubl. Soc. l, 53-62. & Boyd, Edinburgh.
SCIlOLL, D. W., WON HUENE, R., VALLIER,T, L. & 1968. Some rare sedimentary structures in the
HOWEIJ,, D. G. 1981. Sedimentary masses and Silurian rocks of Kirkcudbrightshire. Scott. J.
concepts about tectonic processes at underthrust Geol. 4, 335-69.
ocean margins. Geology, 8, 564-8. WARREN, P. T. 1963. The petrography, sedimentation
SEEI,Y, D. R., VAIL, P. R. & WALTON, G. G. 1974. and provenance of the Wenlock rocks near
Trench Slope Model. In: BURK, C. A. & DRAKE, Hawick, Roxburghshire. Trans. geol. Soc. Edin-
C. L. (eds). The Geology of Continental Mar- burgh, 19, 225-55.
gins, 249-60. Springer-Verlag, New York. - 1964. The stratigraphy and structure of the
-

SltEmDAN, R. E. 1974. Atlantic continental margin of Silurian rocks south-east of Hawick, Roxburg-
North America. In: BURK, C. A. & DRAKE, C. L. shire. Q. J. geol. Soc. London, 120, 192-222.
(cds). The Geology of Continental Margins. 391- WATKINS, J. S., MOORE, J. C. et al. 1981. Accretion
407. Springer-Verlag, New York. underplating, subduction and tectonic evolu-
SttlPI.EY, T. H., LADD, J. W., BUFEI.ER, R. T. & t i o n - M i d d l e America Trench, southern Mex-
WATKINS, J. S. 1981. Facies belts of the Middle ico: results from Leg 66 DSDP. In: Oceanologica
America Trench and forearc region, southern Acta, in press.
Mexico: results from Leg 66 DSDP (this WEIR. J. A. 1968. Structural history of the Silurian
volume). rocks of the coast west of Gatehouse, Kirkcud-
STIt,LMAN, C. J. & FRANOS, E. H. 1979. Caledonide brightshire. Scott. J. Geol. 4, 31-52.
volcanism in Britain and Ireland. In: HARRIS, -- 1977. The Ettrick Valley Thrust and the upper
A. L., HOLLAND, C. H. & LEAKE, B. E. (eds). limit of the Moffat Shales in Craigmichan Scaurs:
The Caledonides of the British Isles--reviewed. a reply. Scott. J. Geol. 13, 75-7.
Spec. Publ. geol. Soc. London, 557-77. 1979. Tectonic contrasts in the Southern Up-
STmNCER, P. & TREAGUS,J. E. 1980. Non-axial planar lands. Scott. J. Geol. 15, 169-86.
cleavage in the Hawick Rocks of the Galloway WELSH, W. 1964. The Ordovician rocks of North-
area, Southern Uplands, Scotland. J. struct. West Wigtonshire. Thesis, PhD. Univ. Edin-
Geol. 2, 129-48. burgh (unpubl.).
STROGEN, P. 1974. The sub-Palaeozoic basement in WILLIAMS, A. 1959. A structural history of the Girvan
central Ireland. Nature, Phys. Sci. 250, 262-3. district, S.W. Ayrshire. Trans. R. Soc. Edin-
TmRLWALL, M. F. 1981a. Implications for Caledo- burgh, 63, 629-97.
nian plate tectonic models of chemical data from -- 1962. The Barr and Lower Ardmillan Series
volcanic rocks of the British Old Red Sandstone. (Caradoc) of the Girvan district, southwest Ayr-
J. geol. Soc. London, 138, 123-38. shire. Mem. geol. Soc. London, 3, 267 pp.
- - 1981b. Peralkaline rhyolites from the Ordovi- WILLIAMS, H. 1978. Geological development of the
cian Tweeddale lavas, Peebleshire. Geol. J. 16, northern Appalachians: its bearing on the evolu-
41-4. tion of the British Isles. In: BOWES, D. R. &
TIPPER, J. C. 1974. The Marine Silurian of the North LEAKE, B. E. (eds). Crustal Evolution in North-
Esk lnlier. Thesis, PhD. Univ. Edinburgh (un- western Britain and Adjacent Regions. Geol. J.
publ.). Spec. Issue No. 10, 1-22.

JEREMY K. LEGGETT, Department of Geology, Royal School of Mines, Imperial College,

London SW7 2BP.
W. STUARTMCKERROW, DAVID M. CASEY,Department of Geology and Mineralogy, Parks
Road, Oxford OX1 3PR.
 Sedimentology, volcanism, structure and metamorphism of the
northern margin of a Lower Palaeozoic accretionary complex;
Bail Hill-Abington area of the Southern Uplands of Scotland

B. C. Hepworth, G. J. H. Oliver & M. J. McMurtry

SUMMARY: NE-SW faults in the Bail Hill-Abington area of the Northern Belt of the
Southern Uplands define blocks u p to 3.2 km wide. The strata, folded and locally
overturned, young predominantly to the NW but successive blocks to the SW contain
progressively younger sequences. Analogous configurations occur in modern accretionary
margins. The oldest rocks are Arenig basalts, dolerites, cherts and brown mudstones
underlying red shales, possibly Llanvirn, and black fossiliferous shales and cherts of
Llandeilo and Caradoc age.
Trench sediments overlying pelagic sequences represent a range of depositional
mechanisms. Rudites and associated fine-grained lithologies of lateral origin relate to a
lower trench slope canyon system, whilst axially transported sands, originating on the
lower trench slope, were deposited by turbidity currents and related flows.
The Bail Hill Volcanic Group (Upper Llandeilo) represents a mildly alkaline seamount
in the Iapetus Ocean, with volcanic activity spanning the transition from pelagic plate to
trench sedimentation before accretion.
Faults, initially low-angle thrusts, and bedding were rotated through the vertical within
the accretionary complex, pre-dating or accompanying slaty cleavage development. Soft
sediment deformation, two fold phases and a kink-band set are recognized. Imbricate fault
zones located in incompetent pelagic sequences are tentatively equated with tectonic
m61anges of other accretionary complexes.
Index minerals, illite crystallinity and 'vitrinite' reflectance establish metamorphic grade
as a zeolite to prehnite-pumpellyite facies.

The Southern Uplands of Scotland are familiar Raven Gill Formation (70 m)
as a Lower Palaeozoic accretionary complex, This lowest unit in the Northern Belt consists
based on evidence summarized in Leggett et al. of massive basalts, dolerites, blue-grey radio-
(1981). The Northern Belt (Peach & Horne larian cherts and brown mudstones (type sec-
1899) consists entirely of Ordovician sediments t i o n - R a v e n Gill NS 920199). The mudstones
and volcanics and forms the northernmost zone contain conodonts (Lamont & Lindstr6m 1957)
of the complex. Caledonian deformation pro- and graptolites (Peach & H o r n e 1899) dated as
duced a series of fault-bounded blocks (Fig. 1) Lower Arenig, and inarticulate brachiopods
of steeply dipping strata (Mitchell 1974; less specific in age. The igneous rocks include
McKerrow et al. 1977). spilitic lavas displaying vesicular and trachytic
textures. A 0.4 m wide shear zone separates
this formation from tectonically lower but stra-
tigraphically higher turbiditic sandstones to the
Stratigraphy south.
Detailed stratigraphical sequences for the area
are published elsewhere (McMurtry 1980a; Hep- Post-Lower Arenig, Pre-Moffat Shale strata
worth, in preparation). Formation distribution (minimum thickness 20 m)
and age relationships are shown schematically Conodonts in red shales from the Leadhills
in Fig. 2. New geological names are introduced Imbricate Zone 4 km E N E of Abington, the
in italics. The pelagic succession is described in Fardingmullach Line in west Nithsdale (Floyd
this section in a conventional manner, starting 1975) and at Morroch Bay in the Rhinns of
with the oldest lithologies. Trench sediments Galloway, have been dated as Llandeilo
are classified into formations based on lithofa- (Lamont & Lindstr0m 1957). Faunas of upper
cies and are described in the next section, Arenig or Llanvirn age (D. hirundo to D.
followed by a description of the Bail Hill Volca- murchisoni zones) have not been found in the
nic Group (McMurtry 1980a). The stratig- Southern Uplands. Black shales exposed along
raphical correlations of Williams et al. (1972) the northern sides of the Eller and Howcon
are used throughout. Faults are underlain by green and grey radiola-

521
522 B. C. Hepworth, G. J. H. Oliver & M. J. McMurtry
...'~:.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'..'.'.'.'.'.'.'.'.'." / .'.'.'.'.-.'.'.'.. ~ ," ".'.'.'.'.'.'.'.-.'.'.'.'.'.-.-.'.'.'.'.'.'.'.'.'.'." I .'.'.'.'.'.'.
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
,,,~.::::.:.:.:::.:.:.:.:.::.:...,-o..::.:.:.:.:.:.:.:.:.::::
:.:.:.:.:.:.:.:.:.:.:..# ..:.:.:.:.:.:.:
:'::::::':::::: SUF ":::.... :':'.":.'":i~>-':'5-->""~:'"'i':':':':':':':':':'""" .... '.'.'.'.'.'.'.'.'.:.:.:.:.'.'.'.'.'.'.'.'.'.'.'.'...'.'.'.'.v.v.
......- ..... = ..... ~ i ;! . " ~: .....
, : ;,-~_~7;~,..--.
!' i ~, i "~ -••:--~---47
;~-ii;:i
•>Z-X
.... (>;'~/,-','-
\; ;,y :"~,-,,'/--,,,,
L, ',-',,,,-'--,"
.... -,,'_-,~
,~, . .~..',~. '-'-',',/,-~_;
;, ~ > : ,~'~ ~-'A
' <,'-:~} ~.'.;~
"-:'~ x~.-,.,.-'.-7", :2--:" ,";'4 ',,_';'-'c-~ -,-,,.,~<:"-':;F_,'/ .
• Study

:::::::::::~:::i:: :':':':'I':- :-,. .......... -~y--.-. ".,.- -"- i

~']'~'i']'i'~'i'~'~'i'~'['i'i'. "Z~ Bail Hill "~ . . . . . N ................. -- ~"
"'"'"""'"'"'"'"'"'" ' " " 75 ~ ~ SP "-- . . . . . . . . . ~_
.... Fault
"'LL". . . . . . . . . . :':';~---'-" ~ ° ? . C~W,CK B OCK _~---~
~_) K i r k c o n n e l ~ ( ~ p ~ KI _,._~'-~"- ~-.~.-.----~-~.-':-'4
Geological boundary
• EF ....... _-t'-.. . . . . . . . . . . . . . . . . . . . i
-.'.....-.'.'..... . . . . . ~ . . . . . . . . . . "-- -x. 50
--'-- Bedding • v.'.v.'.'.'X'X'>.'b,. --,., 7.
•,'.'.-..............:.:.:.. ,'~, ! _.j~
--,- Bedding (vertical) ".'.'.'.°o',',°°',°,°°'.'o°°, |

"':':':-:':':':[][:::.~!?~..,~. ABINGTON B L O C K - i
[] Kaolinite

Prehnite
• ,o "':':':->:-X'X'X"
•.::................ _ _ _ ~ ~ -__--_------_~_------_--J---z-!~--
~ ; - ~ - ~ , .--""- --~-
Prehnite + Pumpellyite •.'.'.'.'.'.'.'.'. ~. - - ~ ~ - ~ - - - - - - - - - _ . - - _ . . - ' 7 . ~ 7
Prehnite + pumpellyite
÷ t homsonlte ", . . . . . . . . . .
"':'X'.'.'.'.'.'.'.'.'.'.v •
"
" : :.~ ~ 70 ~
_-----7--~-
~

. ~ \. ~=-~-~.CON~ BLOCK~ ¢ 1

s
0 1, 2kms , ~ \
~M w
' ~ {~ Menn \ \ '~
\ ~ ~.~
Outline of a r e a s mapped by a u t h o r s ;
rest a d a p t e d from Geol. Surv. S c o t - o~
~and Sheet 1 5 Villages s h o w n on
main map indicated by capital letters.

FxG. 1.(a) Map of the Ordovician rocks of the Bail HilI-Abington area. The Leadhills Imbricate
Z o n e extends from the Abington area south-westwards into the Bail Hill district where it is inferred
structurally to underlie the Conrig Block. The Duntercleuch Fault is interpreted as a splay from the
Leadhills Imbricate Zone. The Spango Granite is Devonian in age.

Formations Structural abbreviations

AB Abington Formation CJB Crawfordjohn Block
CF Crawfordjohn Formation FIZ Fardingmullach Imbricate Z o n e
CR Conrig Formation GF Glencaple fold belt
EV Elvan Formation LIZ Leadhills Imbricate Z o n e
GC Glencaple Formation MBB Mill Burn Block
GU Guffock Formation SBIZ Shield Burn Imbricate Z o n e
KI Kiln Formation
MB Mill Burn M e m b e r
RG Raven Gill Formation
SP Spothfore Formation
Faults
CF Crawfordjohn Fault
DF Duntercleuch Fault
EF Eller Fault
FF Fardingmullach Fault
HF H o w c o n Fault
SBF Shield Burn Fault
SUF Southern Uplands Fault
WDF Wellgrain D o d Fault
 Northern margin of the Southern Uplands accretionary complex 523

................................-...........-.............-.....- / ......~'-~ ....-.....-.......................-.............-.......-.'.'.'." I .'.'.'.'.'.'.'.'.-.-.'~

................................. ~ .....,. v .............................. .'" (b .'.'.°'". . . . . . ."
.1.'.'.'.'.'.'.'.','.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.*.'.'.','."
• % .'.'...'.'." . . . . ..-.-,....-.-........,,...-,....-.,...-,'.'.•.'.'.'°'.'.'... o> . . . . . . . . . . . . .
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
A :::::::::::::::2::::::::::::::::::::::::::::::::::::::
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 4 "-":: :-" =:- ~ >~
":':':':" "~ '. . . . . . . . . . . . . n 50 \
I:!:.'.'.'.! 70 M LL BURN BLOCK Crawfordloh CF .......
.'lv'.'.:.:~
• .. -(~ ......................
.:.:.:.....~ ~.~
•:.:.:.:.~ .,.--."" 67 CRAWFORDJOHN BLOCK
-~:':':':"~
~ - ~
................. ..~ - _ -_
~~ s el. . Z. . . . . .
-:.:
~ .~
-

I . . . . . -_---2-2-

I ~ ~ : -
~Z.-..~-~-.~- ............
.:!i!::iiiiiiiiii!iiiii!i!".'~'-
I~i.iii~ j --
"5r2
ABINGTON B L O C K - e a s t
Abington
I~)
I / Ordovician

:,-r~. i Imbricate zone

• --,- ~.~--.:.: :) ..~ [ ~ . L. .I Z. .
Bail Hill Volcanic
Group ( B H V G )
O u t l y i n g tuffs

....
i I. . . .
i-----~:÷J. . . . .
! - -~:~---CC-~~~----C--CC-C:~
E::-_--y:::_:~:----_-~: Havengill Dod A -C-C--1
- GC " II from B H V G

Granite

Post Ordovician

~-E--C- CC~-C:--C7~-! . . . . i . • ELVANFOOT BLOCK

C--~=:-:-J ~ • • •
I (~) Leadhills
I t ~I~ 88 FIZ E Ivanfoot j

I _ - - ~ - - C - - - - - - _ - . ~ '~ '

VV.n,oc..ead i~~:-----~o-?~ ..... I"

i 65 B ~

~ G ~ 7 ~ ' - ~ i Green Lowther -*-- US

~o / / ..,

SUF CF SBF WDF FF

,.
i
i MBB I CJB I Abington Block GF Elvanfoot Block
i -<..<
:::::::1
I ~
::i::!::~::i
:!:! !l
!:!:!:!!
i
i<xx
i
i
I
i

v~
Ally/~,,iXAN< >- ~'< /
I
i
I
i
SB~Zi LIZ
i

-< Younging direction

-- Bedding
--- C l e a v a g e (S I)

FIG. 1.(b) Schematicstructural profile acrossthe Abingtondistrict. Stratigraphicalsequencesare cut

up into a series of NE-SW orientated blocks by major reverse faults. Line of section and scale given
on main diagram.
524 B. C. Hepworth, G. J. H. Oliver & M. J. McMurtry
NW SE
ElVaBfool

~
Abmgton
Ablngton B - east
Cr a ~ i c k B- west GC
Crawfotdlohn
Mdl Burn B
B
CR-
AB EV

GU-
MB CF
KI ~> 0

(,AlgAl)()(; C;A,qAL~)C;
0 0
t LAND[ I£() Fg i t ANDEILL)

[ [ ANVU4N ............ HF DF [ [ ANVIIIN

CF '~

,RG
............
EF

San£Jston~ ~ Brown .shale and Over!

.........
I Black s h a m and chetl ~ Chert ,qGF

a n O t:or~cilomerat e

Thin - bedded ,~, Fossil control

sill - ~hale

Doh~rlt~; ,tnci bas, dl .... fault

Ball Hill Volcanic

Group

FI(~. 2. Stratigraphical sequences of the structural blocks in the Bail HilI-Abington area. For
abbreviations see Fig. 1. Based on McMurtry (1980a) and Hepworth (unpublished data).

rian cherts which have an estimated minimum 1980a) to ferromagnesian-rich (e.g. Elvan
thickness of a few tens of metres; Leggett et al. Formation). Graptolite evidence from the
(1979) described a 100 m thick continuous sec- Moffat Shales (Fig. 2) shows a progressive
tion in the Coulter-Noblehouse area (equiva- south-eastward delay in the onset of clastic
lent to the Mill Burn Block, Fig. 2), consisting sedimentation, as estabished for the Northern
of red cherts, siliceous mudstones and turbiditic and Central Belts (e.g. Lapworth 1878).
sandstones. These unfossiliferous chert sequ- Sandstone sequences attain a maximum
ences may represent the Upper Arenig- thickness of 3200 m in the Elvanfoot Block; this
Llanvirn time span (e.g. Peach & Horne 1899). value probably reflects tectonic thickening by
faults rather than the original depositional
Moffat Shales (c. 20 m) thickness. Assuming a moderate sedimentation
Across the Northern and Central Belts rate for trench sediments of 1000 m Ma -1 (Pip-
(Peach & Home 1899) black shale sedimenta- er et al. 1973) and a post-depositional compac-
tion commenced during N. gracilis times (Fig. tion of 50% (cf. Moore et al. 1980), the clastic
2). The black shales, with interbedded radiola- sediments in the Elvanfoot Block might have
rian cherts, are typically exposed to the north of been deposited in less than 6.5 Ma. This figure
faults, with younger graptolite zones and/or is compatible with the time range of 6.5 Ma
faunas cropping out in successive fault blocks calculated by Carter et al. (1980) for the D.
to the SE (Lapworth 1878; Walton 1965; clingani zone. It is possible that the other,
McKerrow et al. 1977). The shales pass con- thinner turbidite formations were likewise de-
formably up into turbidites at the base of the posited within the space of a single graptolite
Elvan Formation. zone.

Clastic sediments
Turbiditic sandstones conformably overlie Sedimentology
the Moffat Shales and have been subdivided
into formations on the basis of sedimentology Sedimentary facies within the clastic formations
and petrography. Modal analysis (Fig. 3) shows were produced by sedimentary processes oper-
petrographic compositions ranging from quartz- ating within a trench environment (Piper 1972;
rich (e.g. Glenflosh Formation, McMurtry Leggett 1980). The sedimentary facies have
 Northern margin of the Southern Uplands accretionary complex 525

Laterally transported sediments

The Crawick Block (Fig. 2) includes four
facies associations which define a large-scale
thickening and coarsening-upward cycle (Ricci-
Lucchi 1975). Pelagic black Moffat Shales at the
40 base of the block pass upwards into the thin-
60
bedded Kiln Formation (McMurtry 1980a), in
which two facies associations define upper and
60 lower (Sheil Rig and Butt Hill) members of the
40
formation. The Kiln Formation is succeeded
conformably by the rudaceous Spothfore
8O Formation (McMurtry 1980a). Both formations
are laterally impersistent (Fig. 1).

M ' ' ' ' ' ' ' F

Spothfore Formation (300-800 m)
20 40 60 8O The Spothfore Formation (type s e c t i o n - -
upper Spothfore Burn 767138, Fig. la) consists
? of rudites, coarse- to fine-grained sandstones
and siltstones. Clasts range in size up to several
metres and are all of intrabasinal origin, with
sediments constituting 95% and volcanic rocks
5% of the total. The rudites are generally
disorganised, suggesting emplacement by deb-
ris flows (Middleton & H a m p t o n 1973),
although local lamination and imbrication indi-
cate currents flowing towards the SE, trans-
verse to the postulated trench (McMurtry
1980a). Thin-bedded sandstones and siltstones
interbedded with the rudites were also depos-
ited by currents flowing towards the SE.
Laterally discontinuous sequences of red and
green chert with dimensions of several tens of
metres both across and along strike crop out
20 40 60 80 within the Spothfore rudites. The cherts show
randomly orientated folds (782153) in contrast
FIG. 3.(A) Modal analyses of sandstones; to nearby rudites which are unfolded; bedding
triangles = Abington Formation; squares is oblique to that in rudites (780153) and the
= Glencaple Formation; inverted triangles contacts are sharp and irregular. The cherts are
= Elvan Formation. Q = quartz + acid + interpreted as large gravity-emplaced (exotic)
metamorphic fragments. F = feldspar + blocks within the rudites.
amphibole + pyroxene + basic fragments.
M = matrix + mica + sedimentary frag- Graptolite faunas in shale clasts in the rudites
ments. Analysed by B.C.H. are from low in the N. gracilis zone and are
(B) Modal analyses of sandstones; filled older than in situ faunas from the Crawick
circles = Glenflosh Formation; open cir- Block. This suggests the clasts (including the
cles = Kiln Formation; squares = Guffock large chert blocks) were derived from the
Formation; open triangles = Spothfore north, originating from earlier accreted sedi-
Formation (matrix); filled triangles = ments and/or lower slope deposits.
Spothfore Formation (clasts). Analysed by The Spothfore Formation was formed in an
M.J.M. inner fan environment, related to a canyon
system incised across the lower slope of the
accretionary complex.
been separated into those of lateral and those of
axial origin (Table 1 and Fig. 4). Stratigraphy Kiln Formation (700-900 m)
and field relations are not implied in the discus- The thin-bedded Kiln Formation (type sec-
sion, unless stated, and the lithologies are de- tion in the Kiln Burn between 768142 and
scribed in a progressively downcurrent direc- 773133) comprises six members. The Butt
tion (i.e. becoming more distal). Hill and Sheil Rig Members represent distinct
526 B. C. Hepworth, G. J. H. Oliver & M. J. McMurtry
TABLE 1. Summary of sedimentological characteristics of selected facies from the Bail Hill-Abington
area. 1 = Spothfore Formation; 2 = Shell Rig Member; 3 = Butt Hill Member; 4 = Elvan
Formation; 5 = Abington Formation thick-bedded facies; 6 -- Abington Formation medium-bedded
facies; 7 = Guffock Formation; 8 = Glencaple Formation. Superscripted ~umber (1) Describes sand
layer only. (2) D = distribution grading, I = inverse grading. (3) After Walker & Mutti (1973). (4)
After Mutti & Ricci-Lucchi (1975). (5) Direction given from source to depositional site

Ch;ff~i(:terlS| ic f a c i e s 1 2 3 4 5 6 7 8

Bed thickness ( i ) range 06-5m n/m-era mm-cm 04-4m 02-4m 15-120cr'n 03-O 6m 10- 8Ocm

{ l i ) average - - 105cm 110m 32 cm 30cm 28cm

L a t e r a l w(:dgiIl~4 yes no rare no no no

A nlalgamahon y~s y(~s yes ra re rare

Sharp fops 1 yes yes yes yes no no no no

Sharp bases 1 yes y~,s yes yes yeS y(~s yes yes

Sol{t m a r k s no ,io no r,o no no yes no

G r a n l size7 ( i ) f y p u ; a l CS- boulder MS FS-MS VCS VCS CS CS CS

h~) m a x i m u m 5 m - gr~c~ule pebble VCS - pebble

Graded 2 no no r a t (?ly no D I D D D

Bouma sequ~ll(:~s no no no no ae ae ae a-e

Mudst I sandst (:lasts < 2m no no < 2cm < 0 3m yes yes yes

S o i l s~(J d e f o r m a t i o n 11o no y~s i are no no no yes

F a c i(.,s 3 A1 E - B C C C C4

CtJr r~?llts s SE SE - - - SW SW SW

A ssoch~tt~cJ J a c l e s 3 A2E F A1 C 2 4 C24 C2aIL B2D D.E DIE C14 8 2 , D . G

Mod(~ of deOOSdlOtl D#brls Reworked "Ow~rbank" Hiqh conc khgh cone Turbldde Turbldlte LOW c o n c
flow t ur bldd~ turbldlte qrav 1 l o w ~ur b l d l l e turb~dite

sedimentary facies within which four tuffaceous graded sandstones which show transport to-
members crop out (see below). wards the SW (i.e. parallel to the trench),
The Kiln Formation occurs between the indicating that deposition of the thin-bedded
underlying Moffat Shales and the overlying lithologies took place within an open trench
Spothfore Formation; deposition of these sedi- environment, rather than in a canyon or chan-
ments thus spans the transition from an area of nel incised into the lower slope.
pelagic to a more proximal environment of
channel-mouth sedimentation. Axially transported sediments
Sheil Rig Member. This upper member under-
lies and is interbedded with rudites of the Although palaeocurrent data are generally
Spothfore Formation. It is dominated by len- sparse the dominant transport direction was
ticular, cross-laminated, fine-grained sand- towards the SW, the thick sedimentary sequ-
stones and siltstones and is considered to be of a ences retaining their distinctive sedimentologi-
proximal nature. The sandy laminae are less cal dispersal patterns parallel to the present-day
than 3 cm thick, are moderately sorted and strike.
contain abundant matrix (15%). Palaeocurrents
indicate transport to the SE (i.e. transverse to Elvan Formation (2500 m)
the trench). Similar lithologies have been de- The Elvan Formation (type section in Long-
scribed from the Hecho Group in the Pyrenees cleuch from 913175 to 921172) is a thick-bedded
and ascribed to a proximal environment (Mutti facies rich in basic igneous clasts and ferro-
1977). magnesian minerals (Fig. 3a). Massive and peb-
Butt Hill Member. This lower member overlies bly sandstones constitute over 60% of the
pelagic black Moffat Shales and grey and green formation. Beds are characteristically thick and
radiolarian cherts at the base of the Kiln do not show Bouma sequences. Deposition
Formation. Units of fine-grained, parallel- took place from high-concentration sediment
laminated sandstone or siltstone alternate gravity flows.
every few centimetres with units of fine-grained There is a general consensus that similar
siltstone or shale. Occasional well-sorted sands deposits formed in proximal fan environments
and silts with sharp tops and bases are inter- (e.g. Walker 1966; Stanley et al. 1978). The lack
preted as turbidites reworked by bottom cur- of associated canyon-wall deposits (slump
rents. Both the Sheil Rig and Butt Hill Mem- facies, etc.), channelling and material coarser
bers have intercalations of thicker-bedded than granule size, together with the lateral and
 Northern margin of the Southern Uplands accretionary complex 527
OR Outer rise NE
TW Trench wedge
B Pelagic facies
3
Basement 41 2
6,7 5 ~ ~ . . ~ . tW
8
sw ":
'l

Flo. 4. Schematic diagram of the distribution of sedimentary facies shown by the sandstone
sequences. No stratigraphical correlations are implicit unless stated in the text. Details of individuaJ
facies are shown in Table 1 using the same numbering system.

vertical continuity of the lithology, precludes massive sandstones (Fig. 4) within a broadly
deposition of the Elvan Formation in an inner channellized distributary system.
fan environment. Deposition in the trench near Medium-bedded sandstones and shales. The
a canyon mouth is considered likely, but the dominant lithology of the Abington Formation
lack of knowledge of sand body geometry consists of interbedded sandstone/shale coup-
precludes discrimination between a broadly lets, 10-100 cm thick (typically 30 cm), which
channellized distributary system and a non- are plane-parallel at outcrop scale. Distribution
channellized sheet-flow accumulation. grading (Tae Bouma intervals) is common and
T b, T c and T d divisions are occasionally
Abington and Glenflosh Formations (2800 m) observed in the upper parts of units.
The quartz-rich (Fig. 3) Abington and Glen- Tae sequences are generally considered typi-
flosh Formations are lateral equivalents (Fig. cal of a 'proximal' environment (Walker 1976).
2), and in the following descriptions will be Similar sequences have been described as pass-
referred to as the Abington Formation. The ing downcurrent and laterally into more com-
Glenflosh Formation has a type section in road plete Bouma sequences (Elmore et al. 1979).
and stream cuttings flanking Glenflosh Hill The lack of conspicuous channels in this facies
(795140); the Abington Formation has type suggests deposition as sheet flows, analogous to
sections at Craighead Quarry (919238) and road a mid-fan, non-channellized area.
cuttings north of Abington (930243) and has
been subdivided into a thick-bedded and Glencaple Formation (c. 1100 m)
medium-bedded sandstone facies. This formation conformably overlies the
more thickly bedded Elvan Formation (Fig. 2)
Thick-bedded sandstones. These are of local and together the two formations define a first
development within the Abington Formation order thinning-upward cycle (Ricci-Lucchi
and display thinning and fining-upward sequ- 1975). The type section is in Glencaple Burn
ences on a scale of a few metres, a feature (92419%924193). Frequent incursions of the
typical of distributary channels in a mid-fan massive Elvan facies and similar petrographies
environment (Walker & Mutti 1973). Massive, indicate a conformable transition between the
thick-bedded and very coarse sandstones occur, two formations.
similar to lithologies of the Elvan Formation. Characteristically Ta_e Bouma divisions are
However, the dominant lithology of the Abing- present. Most beds begin with a graded T a
ton Formation contrasts with the Elvan Forma- division of the distribution variety, although
tion in showing internal sedimentary structures, delayed, symmetrical and inverse coarse-tail
including distribution and inverse grading. It is grading are observed in the thicker beds. Simi-
suggested that this facies represents deposition lar complete vertical sequences of Bouma divi-
from more 'mature' flows (turbulence being the sions have been described from outer fan non-
major sediment support mechanism) than the channellized depositional lobes (Mutti & Ricci-
Elvan Formation lithologies. Hence these de- Lucchi 1975) and mid-fan sparsely channelled
posits are inferred to occur downcurrent from depositional lobes (Walker & Mutti 1973).
528 B. C. Hepworth, G. J. H. Oliver & M. J. McMurtry

Medium-bedded turbidites showing relatively suggests a single stable source for each of them,
complete Bouma sequences have been operational over a considerable time span.
documented to pass upcurrent into both mas- Markedly different petrographies for sand-
sive, coarse-grained, thick-bedded, channel- stone sequences of identical age (i.e. Abington
lized deposits (e.g. Walker 1966; Stanley et al. and Elvan Formations) indicate the simul-
1978) and Wae sequences (Elmore et al. 1979). taneous operation of two different point
This facies is suggested to be the most distal sources transporting sediments into two distinct
of all the sandstone lithologies described, and geographically separated basins.
formed from low-concentration flows in the Nowhere in the area can a lateral transition
trench. The presence of thick-bedded Elvan be observed from proximal to distal facies, as
lithologies indicates sporadic incursions of high- described from deep-sea fans and basins (e.g.
concentration flows, resulting from local pro- Stanley et al. 1978) and established for turbidite
gradations of the channellized mid-fan in a models (Mutti & Ricci-Lucchi 1975). Contrast-
down-fan direction. The transition from the ing lithofacies can be recognized at separated
Elvan Formation to the Glencaple Formation is localities, which individually are comparable
attributed either to a lateral shift in the axis of with components of the fan model. These form
deposition of the coarse, thick-bedded facies, thick sequences, suggesting that sediment
or to a general decrease in the volume of supply originated from widely separated points,
turbidite flows into the trench. as in the Chile Trench (Schweller & Kulm
1978).

Discussion Volcanic rocks

The clastic sediments have been interpreted as The Bail Hill Volcanic Group crops out at the
trench deposits principally because of the domi- western edge of the area and constitutes a range
nance of linear sediment dispersal patterns of mildly alkaline lavas and pyroclastics
(NE-SW) and the lack of a source to the SE (McMurtry 1980a,b). Early basalts were
(Piper 1972). In the Northern Belt of the South- erupted on to pelagic black shales of N. gracilis
ern Uplands axial sediment transport and de- age. The basalts contain phenocyrsts of clinopy-
position was principally by turbidity currents. roxene and bytownite, are locally vesicular and
No evidence has yet been found to suggest that amygdaloidal, although more commonly show
the deposition of the bulk of the sandstones was autobrecciated textures indicative of eruption
related directly to an axial channel, such as in under water. These basalts are succeeded by
the Chile Trench (Schweller & Kulm 1978). In middle-stage hawaiite and mugearite lavas and
contrast, apart from the channel-related thick- pyroclastics containing phenocrysts of oligo-
bedded facies of the Abington Formation, the clase/andesine, pargasitic amphibole, biotite/
majority of the sediments were deposited by phlogopite and apatite.
sheet flows. Retention of distinctive sedimento- The presence of gabbroic and dioritic xeno-
logical characteristics over 20 km down-basin liths in some lithologies and the relationships of
and parallel-bounding surfaces to most sand- mineral compositions between the coarse-
stone units suggest individual flows of large grained xenoliths and volcanic phenocrysts
volume occupying the whole width of the would suggest that a large intrusive body under-
trench, and extending for considerable dis- lay the volcano in Ordovician times and
tances over the flat trench-floor. However, the periodically extruded the material that now
sequences of massive coarse sandstones in the constitutes the Bail Hill Volcanic Group. This
Elvan Formation indicate that other mechan- body was subducted along with layers 2 and 3 of
isms such as high-concentration sediment grav- the Iapetus Ocean crust. The presence of extru-
ity flow can produce thick sedimentary bodies. sive and shallow intrusive material (not found
The transition from lateral to axial flow in the at outcrop) as clasts in an infilled volcanic neck
trench is thought to have resulted from deflec- would suggest that extrusive activity was litho-
tion of the turbidity currents by the outer slope logically more diverse than present outcrops
(Piper 1972). This transition was apparently indicate.
accomplished at the depositional site of the Four outlying pyroclastic members are inter-
Elvan Formation, which is the coarsest-grained bedded with turbiditic sediments of the Kiln
and thickest-bedded of the laterally persistent Formation (Fig. 1). The thickest, the Stoodfold
lithologies. The lateral consistency of sedi- Member (McMurtry 1980a) is petrographically
mentological characteristics of most formations similar to the middle-stage activity of the Bail
 Northern margin of the Southern Uplands accretionary complex 529

Hill volcano and comprises lithic and crystal nantly sandstones) up to 3.2 km thick. Litholo-
tufts, volcaniclastics and agglomerates inter- gies above the faults are strongly imbricated
bedded with siltstones containing flattened into linear outcrop belts, surrounded to the NW
pumiceous material. and SE by younger sandstones, thereby defin-
The volcano is believed to have originated on ing inliers.
the outer rise as a mildly alkaline seamount, its One such inlier, the Leadhills Imbricate
activity spanning the transition from pelagic Zone (Leggett et al. 1979), displays internal
plate to trench environment, and to have be- organization of the fault slices at the base of the
come extinct prior to burial by rudites of the Abington Block. Progressively older rocks crop
Spothfore Formation. The volcano was de- out towards the centre of the inlier, in the
tached from the downgoing Iapetus Ocean sequence C. peltifer black shales, Llandeilo red
plate and accreted (McMurtry 1980a). shales and Arenig mudstones and cherts (Fig.
5). To the NW of the Arenig strata imbricate
thrusts of approximately equal displacement
Structure bring to the surface fault slices consisting of
black shales, cherts and thin wedges of Abing-
The Bail Hill-Abington area consists largely of ton Formation turbidites. Imbrication is in-
NW younging, typically vertical beds with little tense, with a minimum of 11 fault slices recog-
evidence of the 'flat belts' of Craig & Walton
nized in a 700 m poorly exposed transect
(1959), although megascopic folds are found across strike. Towards the NW margin of the
infrequently in all blocks. Overturning of bed-
inlier the fault slices die out, and black shales
ding occurs close to the Southern Uplands are inferred to pass up into turbiditic sand-
Fault.
stones of the Abington Formation.
Faulting Both the Leadhills and the Fardingmullach
Imbricate Zones can be traced SW for 15 km
Outcrop distribution in the Northern Belt into Nithsdale (Floyd 1975) and for 80 km
is primarily controlled by large strike faults, thence to the coast in the Rhinns of Galloway
recognized on the basis of structural and (Kelling 1961, 1962) where they are exposed at
palaeontological data, trending ENE-WSW and Killintringan and Morroch Bay respectively.
separating sedimentary sequences (predomi- Other faults such as the Howcon and Eller

RGF CBF WDF

\ I

t/,',
\ I
\ -< l

iiI l !lvrIl\\\ , ii!iiiiiiii~ i-ii

I
\
\ Pel

Abington Fm -- Fault
0 300m
Raven Gill Fm - - - - Inferred fault

m Black shales and cherts -< Younging direction

Red shales (Llandedo) cl clingani

Green shales (age ?} pe pettffer graptofite zone

RGF Raven Gill Fault gr gracilis

CBF Coom Burn Fault -- Bedding

WDF Wellgrain Dod Fault Areas not exposed left blank

FI~. 5. Structural profile through the Leadhills Imbricate Zone; location of line of section is shown
on Fig. I(A). The sandstones of the Abington Formation indicated on the extreme left of the
diagram are continuous for 2.8 km north-westwards up to the Shield Burn Fault. Strata range in age
from Arenig to upper Caradoc. Faults are recognized on the basis of structural and palaeontological
data.
530 B. C. H e p w o r t h , G. J. H. Oliver & M. J. M c M u r t r y
Faults (McMurtry 1980a) and the Duntercleuch folds are rarely seen in the field, possibly due to
Fault and Leadhills Imbricate Zone merge to- the formation of axial-planar strike faults
wards the NE, thereby reducing the number of associated with inlier development. Where un-
structural segments recognized in sections faulted, fold pairs verge towards the SE.
across strike (Fig. 1). Weir (1979) considered
the faults to show a listric geometry, with Post-D 1deformation
translation towards the SE. Later folding (F2) has only been recognized
It is envisaged that movement initially occur- in the Mill Burn Block, immediately south of
red along low-angle thrusts, thereby juxtapos- the Southern Uplands Fault. Inclined plunging
ing lithologies originally deposited tens of folds of sinistral vergence show a transecting $1
kilometres apart (McKerrow et al. 1977). Sub- cleavage, and possibly represent D 0 soft sedi-
sequent rotation of bedding and thrusts through ment deformation. This cleavage is locally de-
the vertical took place within the accretionary formed by F 2 folds of dextral vergence which
complex (Eales 1979). The imbricate zones are plunge southwards at angles of about 40 ° .
envisaged as laterally extensive fault zones A conjugate kink-band set has been identi-
underlying largely coherent blocks several fied from the Leadhills Imbricate Zone. Single
kilometres in thickness. Innumerable brittle sets of kink-bands have been found in the
fractures and faults sub-parallel to strike served Abington Block.
as planes along which displacement occurred;
thus a zone of intense tectonic dislocation facili-
Discussion
tated underthrusting of structural blocks
towards the NW. An implication of the accretionary complex
The amount of thickening of the strata by model is that deformation is likely to be di-
faulting can be estimated from the 1700 m wide achronous across the width of the Southern
Leadhills Imbricate Zone which consists of Uplands (McKerrow et al. 1977). Hence D1, D2
repetitions of the 180 m thick pre-upper etc. defined in any one block are not necessarily
Llandeilo lithologies plus slivers of turbiditic contemporaneous. South-easterly vergence of
sandstones from the overlying Abington early fold phases across the Southern Uplands
Formation. The presence of Llandeilo and Are- is consistent with underthrusting towards the
nig lithologies in the Leadhills Imbricate Zone NW (Anderson & Cameron 1979). This uni-
suggests that ddcollement occurred within Iape- formity of fold vergence direction is seen in
tus Ocean Layer I in the Northern Belt at a modern accretionary complexes, and is domi-
lower horizon than in the Central Belt where no nantly oceanward (J. C. Moore & Karig 1976).
units older than Caradoc (N. gracilis zone) are The major fault-bounded blocks are interpre-
found. ted as individual accreted packets (McKerrow
et al. 1977). The thickness of each fault block
Folding falls within the values of 2-8 km inferred for the
accretionary packages of the Shikoku subduc-
Main deformation (D1) tion zone (J. C. Moore & Karig 1976, fig. 11).
The recognition of a deformational chron- Eales (1979) considered the main fold phase to
ology is hindered by the variable and limited be partly coeval with thrusting.
development of slaty cleavage ($1). In the Mill In contrast to many other accretionary com-
Burn and Abington Blocks, $1 maintains a plexes, tectonic m61anges are absent in the
statistically vertical attitude, despite variations Southern Uplands. We consider that their place
in bedding orientation ranging from dipping is taken in the Northern Belt by imbricate zones
steeply to the north, to SE (overturned) dipping located in the incompetent pelagic sequences.
strata. Rotation of bedding and the major faults Both tectonic m61anges and imbricate zones are
from the horizontal to the vertical is considered associated with strong fault movements (Cowan
to be either a pre- or syn-S~ event. 1974; Connelly 1978).
Folds, generally without the development of
a slaty cleavage, are found in all blocks and are Metamorphism
preferentially developed in thin- to medium-
Index Minerals
bedded lithologies (including cherts). Axial sur-
faces are parallel to the general strike; fold axes Spilites from the Leadhills Imbricate Zone
plunge towards the SW at angles between 16°- Spilitized lavas, dolerites and breccias from
90 °. Wavelength of folds varies from a few the Leadhills Imbricate Zone (Fig. 1) show
centimetres up to metres. Folds are generally recognizable igneous textures but the rocks are
isolated in development. Hinges of major F1 altered although unfoliated. Calcic feldspar has
 Northern margin of the Southern Uplands accretionary complex 531

altered to albite which itself is partially replaced CHLORITE~

by mixtures of chlorite (Fig. 6a), carbonate,
prehnite and green pumpellyite (Fig. 6b). Phe-
nocrysts of clinopyroxene are unaltered or are R
partly replaced by chlorite along cracks and
edges. Sphene has largely replaced opaque • v
I
'i 6
oxides. Amygdales contain variable mixtures of
radiating sheaves of chlorite, quartz, carbonate, Fe2+ tot
albite, sericite, pumpellyite and prehnite. None
of the fine-grained basalts have recognizable rA t
4
prehnite or pumpellyite, though they are com-
monly riddled with veins of carbonate + chlor-
ite + albite + quartz + prehnite + pumpellyite.
The secondary minerals indicate prehnite- f
!
pumpellyite facies metamorphism. Comparison i

of these pumpellyites with those from else- 5 6 7

Sb
where in the world (Fig. 6b) suggests that the
metamorphism is of a grade similar to Van- 16
couver Island (field b of Kuniyoshi & Liou [-'..\
1976), with temperatures between 300 and
• ",,4•
'.~
360°C and pressures exceeding 1-3 kbar (Nitsch
'~ ', •
1971). 12

Veins in the Elvan Formation

O
Feldspar detritus in these sandstones is
11 • ~ ........
albite. An abundance of prehnite + calcite + .<8
quartz veins indicates appreciable mobility of AA•
Si, Ca and A1, supporting the metamorphic
A A o
origin of the albite by the breakdown of detrital
calcic plagioclase. These veins may have
formed in hydraulic fractures of metamorphic
origin (cf. Norris & Henley 1976). Pumpellyite,
however, has not been identified in prehnite
veins--this assemblage does not define a par- PUMPELLYITE 44
ticular metamorphic grade but does indicate a
grade lower than the greenschist facies and, 18 2~2 2~6 30
because zeolites have not been found, presum- Wt ?. AI203
ably prehnite-pumpellyite facies.
FIG. 6.(A) Chlorite probe analyses. Filled
Lavas of the Bail Hill Volcanic Group triangles = Leadhills Imbricate Zone
The in situ rocks of the Bail Hill Volcanic spilite; diamonds = Bail Hill Volcanic
Group contain original calcic plagioclase, occa- Group; inverted triangles = Elvan Forma-
sionally altered to K-feldspar (possibly adu- tion. R = ripidolite; B = brunsvigite; Py =
pycnochlorite; D = diabantite; P = penni-
laria) or, more frequently, partially replaced by nite (after Hey 1954). Analysed by B.C.H.
carbonate, chlorite and white mica. and G.J.H.O. at Edinburgh and Glasgow
Amygdales in the lavas contain mixtures of Universities.
carbonate, thomsonite, prehnite, pumpellyite (B) Pumpellyite composition in terms of wt
and albite. The lava matrix sometimes contains % FeO t°t and wt % AI203. Diamonds =
pumpellyite and always chlorite. Veins of car- Bail Hill Volcanic Group; triangles =
bonate and analcite are common. The presence Leadhills Imbricate Zone spilite. Dot-dash
of thomsonite and analcite indicates that the = composition field Vancouver Island (a),
Bail Hill Bolcanic Group has been metamorph- after Kuniyoshi & Liou (1976); dotted line
= composition field Vancouver Island (b),
osed under conditions of the zeolite facies. The after Surdam (1969); solid line = composi-
non-albitization of the calcic feldspar corrobo- tion field, Wakatipu I and II (Kawachi
rates this. 1976); dash-dot-dot line = composition
The presence of prehnite and pumpellyite in field, Wakatipu III (Kawachi 1976). Analy-
the same rocks suggests that pressures of over sed by B.C.H. and G.J.H.O. at Edinburgh
1 kbar may have been reached (Ernst 1974). and Glasgow Universities.
532 B. C. Hepworth, G. J. H. Oliver & M. J. McMurtry
The presence of analcite implies that metamor- block in the direction of younging (i.e. towards
phic temperatures did not exceed 200°C, so that the NW). No significant patterns of IC distribu-
the analcite + quartz = albite + water reaction tion emerge across the Abington and Elvanfoot
did not occur (Liou 1971). The thomsonite + blocks, so it is inferred that metamorphism
quartz = wairakite reaction at 280-316°C continued after the succession had been cut up
(Coombs et al. 1959) has not taken place, into fault-bounded blocks and rotated to verti-
although it is to be remarked that the Bail Hill cality. Illite crystallinity has not yet been meas-
Volcanic Group is quartz-free. The zeolite ured in the blocks north of the Abington Block.
metamorphic minerals may have formed soon
after extrusion of the volcanics as a result of
deuteric alteration, or at some time later under Reflectance
conditions of burial metamorphism.
The mean maximum reflectance (&o max) of
Siltstones from the Stoodfold Member 'vitrinite' (graptolite) fragments in black shales
These siltstones cor~tain small white rounded has been measured for the Abington and Elvan-
foot Blocks. Values range from 3.91 to 8.22%,
kaolinite patches up to 2 mm in diameter,
which are distributed through the bedding lami- equivalent to meta-anthracite coal rank and the
nae. The patches formed after the sediment prehnite-pumpellyite facies (Kisch 1974). No
had been partially or wholly compacted, hence significant variation occurs in the area. On the
their non-flattened nature and random distribu- basis of one set of determinations from one
locality giving /~0 max = 5.75%, shales at the
tion.
Volcaniclastics associated with the siltstones base of the Elvan Formation are considered to
show fresh calcic plagioclase, although some be of the same metamorphic grade as those at
have been partially altered to carbonate and the base of the stratigraphically lower Abington
white mica; amphiboles have altered to mix- Formation to the NW. This finding agrees with
tures of chlorite, carbonate and iron oxides. conclusions drawn on the basis of IC.
The lack of albitization and presence of kao-
linite in sediments in the same block as the Discussion
zeolite-bearing Bail Hill Volcanic Group con-
firms the lower grade of metamorphism in the The Bail Hill Volcanic Group has been meta-
Crawick Block compared with the Elvanfoot morphosed to a lower grade than the spilites of
Block. A stability maximum between 115 and the Leadhills Imbricate Zone. It is not known
200°C has been suggested for kaolinite whether the change in grade from zeolite to
(Dunoyer de Segozac 1970) and similar condi- prehnite-pumpellyite facies is coincident with
tions are inferred for the Crawick Block. the fault separating the Crawick Block from the
Crawfordjohn Block (Figs 1 & 2), or whether it
occurs within one of the blocks. On this point, it
lllite crystallinity
is significant that illite crystallinity values do not
Turbiditic sandstones throughout the area vary in sections across the Leadhills Imbricate
show considerably more than 10% matrix (Fig. Zone to the top of the underlying Glencaple
3) suggesting that they have undergone recrys- Formation. Given the present-day thickness of
tallization with unstable (volcanigenic) detritus the turbidite formations, in excess of 3 km,
breaking down to fm:m more matrix (Cummins burial metamorphism must have occurred at the
1962). X-ray diffraction analyses of the <2 Bm base of the sedimentary pile in the trench whilst
fraction of shales, siltstones and sandstones the top of the sequence was still unlithified.
show quartz + feldspar + chlorite + illite Metamorphism throughout a sandstone succes-
mineralogies which are present in all specimens sion would only have been possible after thrust-
examined from the Crawfordjohn, Abington ing under the previously accreted wedge, within
and Elvanfoot Blocks. Probe analyses of detri- the accretionary complex. It is considered
tal and metamorphic chlorites show a range of that metamorphism was a continuous process,
compositions, including ripidolite, brunsvigite, initiated in the trench environment and
pycnochlorite and diabantite (Fig. 6a). All illite progressively increased in grade as thrusting
crystallinity (IC) values of < 2 / ~ m grain-size occurred and bedding rotated towards the
fractions fall within the anchizone (Kiibler vertical.
1968) which has been equated with the The low grade of metamorphism in the Bail
prehnite-pumpellyite facies (Kisch 1974). Hill-Abington area is similar to that found in
Under burial metamorphic conditions IC the matrix of m61anges and sandstones of other
might be expected to decrease within a fault subduction complexes such as those on Kodiak
 Northern margin of the Southern Uplands accretionary complex 533
Island ( C o n n e l l y 1978), t h e Franciscan C o m - ton and Dr. J. A. Weir for much helpful discussion
plex of California (Bailey et al. 1964; C o w a n and for critically reading the manuscript. We are also
1978) and the O y o C o m p l e x of Nias Island grateful to the Grant Institute of Geology, University
( M o o r e & Karig 1980). F u r t h e r i n t e r p e t a t i o n of of Edinburgh, the Organic Geochemistry Unit of the
University of Newcastle upon Tyne and the Depart-
the area awaits m o r e t h o r o u g h investigations of
ment of Geology, University of Glasgow for use of
y o u n g e r and tectonically active a c c r e t i o n a r y analytical facilities; and to Mrs S. Johnson, Mrs J.
c o m p l e x e s . T h e i n f e r r e d tectonic history pre- Galloway and J. Jukes for typing the manuscripts.
s e n t e d here m a y be useful in d e v e l o p i n g a m o r e Photographic work was produced by Mr J. Allen.
d e t a i l e d m o d e l to explain the c o m p l i c a t e d rela- We gratefully acknowledge financial support from
tionships b e t w e e n d e f o r m a t i o n a n d m e t a - the N.E.R.C.; G.J.H.O. was financed in addition by
m o r p h i s m at o t h e r a c c r e t i o n a r y margins. the University of St. Andrews. Special thanks are
given to the reviewers for improvements to the manu-
ACKNOWLEDGMENTS:We thank Professor E. K. Wal- script.

References
ANDERSON, T. B. & CAMERON, T. D. J. 1979. A ERNST, W. G. 1974. Metamorphism and ancient
structural profile of Caledonian deformation in continental margins. In: BURK, C. A. & DRAKE,
Down. In: HARRIS, A. L., HOLLAND, C. H. & C. L. (eds). The Geology of Continental Mar-
LEAKE, I . E. (eds). The Caledonides of the gins, 933-50. Springer-Verlag, New York.
British Isles--reviewed. Spec. Publ. geol. Soc. FLOYD, J. D. 1975. The Ordovician rocks of West
London, 8, 263-7. Nithsdale. Thesis, Ph.D. Univ. St. Andrews (un-
BAILEY, E. H., IRWIN, W. P. & JONES, D. L. 1964. publ.).
Franciscan and related rocks, and their signifi- HEY, M. H. 1954. A new review of the chlorites.
cance in the geology of western California. Bull. Mineralog. Mag. 30, 277-92.
Calif. Div. Mines Geol. 183, 89-112. KAWACHI, Y. 1975. Pumpellyite-actinolite and con-
CARTER, C., TREXLER, J. H., JR. & CHURKIN, M., JR. tiguous facies metamorphism in part of Upper
1980. Dating of graptolite zones by sedimenta- Wakatipu district, South Island, New Zealand.
tion rates: implications for rates of evolution. N.Z. J. Geol. Geophys. 18, 401-41.
Lethaia, 13, 279-87. KEELING, G. 1961. The stratigraphy and structure of
CONNELEY,W. 1978. Uyak Complex, Kodiak Islands, the Ordovician rocks of the Rhinns of Galloway.
Alaska: A Cretaceous subduction complex. Bull. O.J. geol. Soc. London, 117, 37-75.
geol. Soc. Am. 89, 755-69. -- 1962. The petrology and sedimentation of Up-
COOMBS, D. S., ELLIS, A. J., FYFE, W. S. & TAYLOR, per Ordovician rocks of the Rhinns of Galloway.
A. M. 1959. The zeolite facies, with comments Trans. R. Soc. Edinburgh, 65, 107-37.
on the interpretation of hydrothermal syntheses. KISCH, H. J. 1974. Anthracite and meta-anthracite
Geochim. cosmochim. Acta, 17, 53-107. coal ranks associated with 'anchimetamorphism'
COWAN, D. S. 1974. Deformation and metamorphism and 'very-low-stage' metamorphism, I, II, and
of the Franciscan subduction zone complex III. Proc. Kon. Ned. Akad. Wet. Amsterdam,
northwest of Pacheco Pass, California. Bull. Set. B, 77, 81-118.
geol. Soc. Am. 85, 1623-34. KOBLER, B. 1968. La cristallinit6 de l'illite et les zones
- - 1978. Origin of blueschist-bearing chaotic rocks tout h fait supGrieures du m6tamorphisme. Et-
in the Franciscan Complex, San Simeon, Califor- ages Tectoniques, 105-22. A la Baconni~re,
nia. Bull. geol. Soc. Am. 89, 1415-23. Neuchfitel, Switzerland.
CRAIG, G. Y. & WALTON,E. K. 1959. Sequence and KUNIYOSHI, S. & LIOU, J. G. 1976. Burial meta-
structure in the Silurian rocks of Kirkcudbright- morphism 6f the Karmutsen volcanic rocks,
shire. Geol. Mag. 96, 209-20. northeastern Vancouver Island, British Col-
CUMMINS, W. A. 1962. The greywacke problem. umbia. Am. J. Sci. 276, 1096-119.
Liverpool Manchester geol. J. 3, 51-72. LAMONT, A. & LINDSTROM, M. 1957. Arenigian and
DUNOYERDE SEGONZAC,G. 1970. The transformation Llandeilian cherts identified in the Southern
of clay minerals during diagenesis and low-grade Uplands of Scotland by means of conodonts, etc.
metamorphism: A review. Sedimentology, 15, Trans. geol. Soc. Edinburgh, 17, 60--70.
281-346. LAPWORTH, C. 1878. The Moffat Series. Q.J. geol.
EALES, M. H. 1979. Structure of the Southern Up- Soc. London, 34, 240--346.
lands. In: HARRIS, A. L., HOLLAND, C. H. & LEGGEd, J. K. 1980. The sedimentological evolution
LEAKE, B. E. (eds). The Caledonides of the of a Lower Palaeozoic accretionary fore-arc in
British Isles--reviewed, Spec. Pubi. geol. Soc. the Southern Uplands of Scotland. Sedimentolo-
London, 8, 269-73. gy, 27, 401-17.
ELMORE, R. D., PILKEY, O. H., CLEARY, W. J. --, MCKERROW, W. S. & EALES, M. H. 1979. The
CURRAN, H. A. 1979.Black Shell turbidite, Hat- Southern Uplands of Scotland: A Lower
teras Abyssal Plain, western Atlantic Ocean. Palaeozoic accretionary prism. J. geol. Soc. Lon-
Bull. geol. Soc. Am. 90, 1165-76. don, 136, 755-70.
534 B. C. H e p w o r t h , G. J. H. O l i v e r & M. J. M c M u r t r y

-- , MCKERROW, W. S. & CASEY,D. M. 1981. The NORRIS, R. J. & HENLEY, R. W. 1976. Dewatering of
anatomy of a Lower Palaeozoic accretionary a metamorphic pile. Geology, 4, 333-6.
prism: The Southern Uplands of Scotland (this PEACH, B. N. 8z HORNE, J. 1899. The Silurian rocks of
volume). Britain, 1, Scotland. Mere. geol. Surv. Scotland,
Llou, J. G. 1971. Synthesis and stability relations of 749 pp.
prehnite, Ca2Al2Si3010(OH)2. Am. Mineral. 56, PIPER, D. J. W. 1972. Trench sedimentation in the
507-31. Lower Palaeozoic of the Southern Uplands.
MCKERROW, W. S., LEGGErr, J. K. & EALES, M. H. Scott. J. Geol. 8, 289-91.
1977. Imbricate thrust model of the Southern , VON HUENE, R. d~ DUNCAN, J. R. 1973. Late
Uplands of Scotland. Nature, London, 267, 237- Quaternary sedimentation in the active eastern
9. Aleutian Trench. Geology, 1, 19-22.
MCMURTRY, M. J. 1980a. The Ordovician rocks of the Rlccl-Luccnl, F. 1975. Depositional cycles in two
Bail Hill area, Sanquhar, South Scotland: Vol- turbidite formations of northern Apennines
canism and sedimentation in the lapetus Ocean. (Italy). J. sediment. Petrol. 45, 3-43.
Thesis, Ph,D. Univ. St. Andrews (unpubl.). SCHWEELER, W. J. & KULM, L. D. 1978. Depositional
- - 1980b. Discussion of: Evidence for Caledonian patterns and channelized sedimentation in active
subduction from greywacke detritus in the Long- eastern Pacific trenches. In: STANLEY,D. J. ~,z
ford-Down inlier. J. Earth Sci. R. Dubl. Soc. 2, KEELING, G. (eds). Sedimentation in Submarine
209-12. Canyons, Fans, and Trenches, 311-24. Dowden,
MIDDLETON, G. V. & HAMPTON, M. A. 1973. Sedi- Hutchinson & Ross, Stroudsburg, Pennsylvania.
ment gravity flows: mechanics of flow and de- STANLEY, D. J., PALMER, H. D., d~. DILL, R. F. 1978.
position. In: MIDDLETON, G. V. & BOUMA, A. H. Coarse sediment transport by mass flow and
(eds). Turbidites and Deep Water Sedimentation. turbidity current processes and downslope trans-
Pacif. Sec. Short Course, Soc. Econ. Paleontol. formations in Annot Sandstone canyon-fan val-
Mineral. 1-38. ley systems. In: STANLEY,D. J. & KEELING,G.
MITCm-:LE, A. H. G. 1974. Flysch-ophiolite sequ- (eds). Sedimentation in Submarine Canyons,
ences: polarity indicators in arc and collision- Fans, and Trenches, 85-115. Dowden, Hutch-
type orogens. Nature, London, 248, 747-9. inson & Ross, Stroudsburg, Pennsylvania.
MOORE, G. F., BILLMAN, H. G., HEHANUSSA, P. E. & SURDAM, R. C. 1969. Electron microprobe study of
KARIG, D. E. 1980. Sedimentology and paleo- prehnite and pumpellyite from the Karmutsen
bathymetry of Neogene trench-slope deposits. Group, Vancouver Island, British Columbia.
Nips Island, Indonesia. J. Geol. Chicago, 88, Am. Mineral. 54, 256-66.
1 6 1 - 8 0 . WAI.KER, R. G. 1966. Shale Grit and Grindslow
MOORE, G. F. & KARIG, D. E. 1980. Structural Shales: transition from turbidite to shallow water
geology of Nips Island, Indonesia: Implications sediments in the Upper Carboniferous of north-
for subduction zone tectonics. Am. J. Sci. 280, ern England. J. sediment. Petrol. 36, 90-114.
193-223. - - 1976. Facies Models 2. Turbidites and associ-
MOORE, J. C. & KARIG, D. E. 1976. Sedimentology, ated coarse clastic deposits. Geoscience Can. 3,
structural geology and tectonics of the Shikoku 25-36.
subduction zone, southwestern Japan. Bull. geol. -- & Murrl, E. 1973. Turbidite facies and facies
Soc. Am. 87, 1259-68. associations. In: MIDDLETON, G. V. 8,: BOUMA,
M u l l , E. 1977. Distinctive thin-bedded turbidite A. H. (eds). Turbidites and Deep Water Sedi-
facies and related depositional environments in mentation. Pacif. Sec., Short Course, Soc. econ.
the Eocene Hecho Group (South-central Paleontol. Mineral. 119-57.
Pyrenees, Spain). Sedimentology, 24, 107-31. WALTON, E. K. 1965. Lower Palaeozoic rocks: stra-
-- & RICCI-LuCCHI, F. 1975. Turbidite facies and tigraphy, palaeogeography and structure. In:
facies associations. In: Muvrl, E., PAREA, G. C., CRAIG, G. Y. (ed). The Geology of Scotland,
Rlccl-LuccHl, F., SAGRI, M., ZANZUCCHI, G., 161-227. Oliver & Boyd, Edinburgh.
GHIBAUDO, G. & JACCARINO, S. (eds). Examples WEIR, J. A. 1979. Tectonic contrasts in the Southern
of Turbidite Facies and Facies Associations from Uplands. Scott, J. Geol. 15, 169-86.
Selected Formations of the Northern Apennines. WILLIAMS, A., STRACHAN, I., BASSETT, D. A., DEAN,
IX Int. Congr. Sedim. (Nice), Field Trip A-II, W. T., INGHAM, J. K., WRIGHT, A. D. &
21-36. WHITTINGTON, H. B. 1972. A correlation of
NITSCH, K. H. t971. Stabilitfitsbeziehungen von Pre- Ordovician rocks in the British Isles. Spec. Rep.
hnit- und Pumpellyithaltigen Paragenesen. Con- geol. Soc. London, 3, 74 pp.
trib. Mineral. Petrol. 30, 240-60.

BARRY C. HEPWORTH, GRAHAMEJ. H. OLIVER • MICHAEL J. MCMURTRY, Depart-

ment of Geology, University of St. Andrews, Fife, Scotland.
 Sedimentary facies associations within subduction complexes

Michael B. Underwood & Steven B. Bachman

S U M M A R Y : Sedimentation patterns within modern subduction zones are complex and
variable, and do not necessarily follow models of submarine fan sedimentation. Environ-
mental reconstructions within ancient subduction complexes should follow modern
analogues as closely as possible and consider several criteria, including turbidite facies
associations, vertical depositional cycles, regional palaeocurrent patterns, and in many
cases, structural style and sandstone petrology.
Important variables in trench sedimentation include the volume and texture of sediment
entering the trench and the distribution of major sediment sources, especially large
submarine canyons. Sediment transport in the trench is commonly longitudinal, although
locally, such as at the mouth of a submarine canyon, flow may be at a high angle to the
continental margin.
Patterns of trench-slope sedimentation depend largely upon the topography of the
slope. In general, coarse sediment is either trapped behind tectonic ridges (within slope
basins) or bypasses the slope via submarine canyons. Background sedimentation is
dominated by hemipelagic settling. Current directions are commonly at a high angle to the
margin, but longitudinal flow may occur within large elongate slope basins. Major facies
associations include submarine-canyon, slope, mature-slope-basin, and immature-slope-
basin.

Classification schemes for turbidites and related trench and trench-slope deposits are also consi-
resedimented deposits (e.g. Walker & Mutti dered. We then model turbidite facies associa-
1973; Mutti & Ricci-Lucchi 1975; Ricci-Lucchi tions for the major depositional settings. Be-
1975) are applicable to deep-sea deposits re- cause many trench and trench-slope environ-
gardless of the geometry of the depositional ments have not been adequately sampled, our
system. Turbidite facies are commonly used to facies associations in some cases are largely
describe specific facies associations, such as inferential, and designations are based upon the
slope, submarine fan, and basin plain (e.g. types of deposits logically expected for a given
Walker & Mutti 1973; Bouma & Nilsen 1978; topography and sediment input. In this paper,
Ingersoll 1978a; Walker 1978; Pescadore 1978). turbidite facies terminology is used in only the
In these models, analogies are drawn between most general sense (Table 1), following the
stratigraphic sequences of turbidites and classifications of Walker & M u t t i (1973) and
modern depositional geometries. However, Ricci-Lucchi (1975).
submarine fan models (Normark 1970, 1978; The deformation associated with subduction
Nelson & Kulm 1973; Nelson & Nilsen 1974; and/or accretion commonly hampers the en-
Nilsen 1980) are not applicable to all modern vironmental reconstruction of ancient trench
sedimentary basins, particularly elongate and trench-slope deposits. However, the scale
troughs (e.g. Hsii et al. 1980; Pilkey et al. 1980). of structural dismemberment, in many cases, is
Moreover, turbidite facies associations, deposi- such that identifiable depositional cycles and
tional cycles, and submarine-fan facies associa- facies changes are preserved. Nevertheless,
tions are not necessarily equivalent or inter- several criteria must generally be considered
changeable. before contrasting sedimentary environments
Meaningful environmental reconstructions (e.g. trench floor versus slope basin) can be
within ancient subduction complexes depend recognized within structurally complex terranes
upon well-established modern analogues, and (e.g. Bachman 1978, 1981; Moore 1979; Moore
should, as closely as possible, be patterned after & Karig 1980; Moore et al. 1980).
the geometry of those analogues. For this
reason, we first discuss the diversity of
sedimentary processes and depositional systems Trench-floor deposits
within modern subduction zones, where several
types of sediment bodies, including submarine Surface samples and Deep Sea Drilling Project
fans, are developed. The sedimentological (DSDP) cores show that modern trench sedi-
characteristics of several examples of ancient ments generally consist of varying proportions

537
538 M. B. U n d e r w o o d & S. B. B a c h m a n

TABLE 1. Classification o f turbidites and associated deposits, after Walker & Mutti (1973) and
Ricci-Lucchi (1975)
Facies
A - - thick to massive, lenticular beds of coarse sandstone, pebbly sandstone, and conglomerate
B - massive to thick-bedded, medium- to coarse-grained sandstone
C - - thick-bedded, classical sandy turbidites, with graded beds at base
D - - thinner-bedded and finer-grained turbidites, including lutite turbidites; basal graded beds absent
E - - t h i n - to medium-bedded, coarse- to fine-grained sandstone, with irregular wedge-shaped beds and
cross-stratification
F - - chaotic deposits formed by slumps and slides
G - - hemipelagic mudstone and shale

of sandy turbidites and interbeds of mud (Ross Schweller & Kulm 1978; Moore, Watkins et al.
1971; Kulm, von H u e n e et al. 1973; Piper et al. 1979). This sequence would not be expected,
1973; Kulm & Fowler 1974; von H u e n e 1974; however, if large volumes of terrigenous turbi-
Karig, Ingle et al. 1975; J. C. Moore & Karig dites are deposited on the abyssal floor seaward
1976; Moore, Watkins et al. 1979; von Huene, of the trench, as with the Bengal Fan (Curray &
Aubouin et al. 1980; McMillen & Haines 1981; Moore 1974), or if the trench floor is starved of
McMillen et al. 1981). Sedimentary sequences coarse terrigenous material, as is the Japan
underlying the trench turbidites commonly in- Trench (Langseth, O k a d a et al. 1978; Arthur,
clude abyssal pelagic clay or biogenic ooze and von H u e n e et al. 1981).
hemipelagic sediments (Piper et al. 1973; The geometry of trench fill is dependent

I06" 22 ° 105 ° 21 ~ 104 ° 20° 103 ° 19°iI02°-~

~,;o

~'\ i

8ta~derc
Monzanillo

~4000~

4500If
500C
5500
6000

FIc. 1. Bathymetric map and trench-axis profile of the Middle America Trench from Banderas Bay
to Rio Balsas, Mexico. General bathymetry (in metres) is modified from Fisher (1961). Axial profile
incorporates data from Fisher (1961), Ross & Shor (1965), and unpublished reflection records from
the following sources: Scripps Institution of Oceanography cruises Scan-11, Cocotow-4, Iguana-I,
Iguana-5, Lamont-Doherty Geological Observatory cruise Vema-28, and Glomar Challenger transit
to Deep Sea Drilling Project Leg 66 drilling sites. Axial-depth values are in corrected metres.
Sediment thickness was calculated using V = 2.0 km s -1, such that 0.5 s penetration = 500 m of
sediment. See Fig. 2 for explanation of symbols. Modified from Underwood & Karig (1980).
 Sedimentary facies associations within subduction complexes 539

upon the interaction between sediment supply 1981) as well as the Oregon-Washington con-
and convergence rate, the distribution and rela- tinental margin (Carlson & Nelson 1969; Bar-
tive size of submarine canyons along the trench nard 1978; Kulm & Scheidegger 1979) and the
slope, and the bathymetric relief on the down- southern Chile Trench (Scholl et al. 1970;
going oceanic plate (Schweller & Kulm 1978; Hayes 1974; Prince et al. 1980). These canyons
Underwood et aI. 1980). The great bulk of have been responsible for funnelling ter-
terrigenous material reaches the trench floor rigenous sediment from the continental shelf
via large submarine canyons that are deeply directly to the trench floor, thereby bypassing
incised into the trench slope (Underwood & the trench slope. Sediment bypassing was parti-
Karig 1980). The sand/mud ratio of trench cularly effective during low stands of sea-level
deposits, in many cases, is thus dependent upon associated with Pleistocene glacial epochs (e.g.
the proximity of submarine canyons to the site Scholl et al. 1977; Barnard 1978). Data from
of deposition. After sediment reaches the floor current-meter studies indicate that at least one
of the trench, axial transport involving dis- of the canyons off Mexico (Rio Balsas Canyon)
tances of hundreds of kilometres is indicated is still active (Reimnitz & Gutierrez-Estrada
(Scholl 1974; Kulm et al. 1977; Schweller & 1970; Shepard et al. 1979); by analogy, other
Kulm 1978). prominent canyon systems that head on the
narrow continental shelf (Figs 1 & 2) are prob-
ably active as well.
Sediment bypassing of the slfpe off Mexico is
Submarine canyons and sediment bypassing
also supported by an axial profile of the Middle
Large submarine canyons are prominent fea- America Trench, which shows a strong correla-
tures along the landward slope of the northern tion between the thickness of sediment and the
Middle America Trench (Figs 1 & 2; Fisher location of large submarine canyons (Figs 1 &
1961; Underwood & Karig 1980; Shipley et al. 2; Underwood & Karig 1980). These data sug-
1980; McMillen & Haines 1981; McMillen et al. gest that the submarine canyons act as the

102 ° "~ 1(31o I00 ° 99 ° 98 ° 97 °

~ ~ EXPLANATION
. . . . . . 170-"
.- 19o ---'~A /i) ~SUBMARINE CANYON .
~-- ---TRENCH AXIS (
103~ X)..~" o DATA POINT FOR ~ a- (
L /~ AXIAL PROFILE ~ , "~ ~ ' ~
(,'~,o E] SEDIMENT THICKNESS __/.f~'~ ~.J,,~,ff "-~
/ . ~ - IN METERS ~ ~ /~,~; (,, ~ -

~ ~ . .............. -";,~opu~o . . . . . _ . . . . . ~ e : '~ (.~-

~_-#£,~-o5.R'/c~: ~__ o oooo~ o-

I - ,5 r
I ~ ~/ ~ Cocotow=4 Survey ~ t' ~r I
/ ~ DSDP Leg 66 /

,000 r " q

F
,ooo 1-:, 11

F]o. 2. Bathymetric map and trench-axis profile of the Middle America Trench from Rio Balsas to
Tehuantepec Bay, Mexico. Depth calculations and bathymetry are the same as for Fig. 1. Data
sources, in addition to those cited in Fig. 1, include Karig et al. (1978), Shipley et al. (1980), and
unpublished reflection profiles from the University of Texas Marine Lab (DSDP Leg 66 site survey).
Modified from Underwood & Karig (1980).
540 M. B. Underwood & S. B . B a c h m a n

primary conduits for sediment entering the redistribution of the sediment by longitudinal
trench. Sediment thickness decreases in both flow mechanisms may be expected (Piper et al.
directions away from point sources at the 1973).
canyon mouths, indicating that bi-directional
transport probably occurs on the trench floor. Axial transport
Seismic reflection data at the mouth of Rio
Balsas Canyon show a thick wedge of sediment Axial channels are prominent features in the
that is channellized along its upper surface; eastern Aleutian Trench (Piper et al. 1973; von
these appear to be distributary channels of a Huene 1974) and the southern Chile Trench
submarine fan (Underwood & Karig 1980). (Scholl et al. 1970; von Huene 1974; Schweller
Rates of sedimentation in the Middle Amer- & Kulm 1978; Underwood et al. 1980), and are
ica Trench (Ross 1971; Moore, Watkins et al. locally developed in the Japan Trench (Arthur,
1979) are not high enough to allow seaward von Huene et al. 1981), the northern Sunda
progradation of submarine fans over the outer Trench (G. F. Moore, pers. comm. 1980), and
trench slope (Underwood et al. 1980). As a the Middle America Trench off Oaxaca (Ship-
result, individual channels initially orientated ley et al. 1980; McMillen et al. 1981). The
perpendicular to the trench axis either shift spectacular channel off southern Chile is con-
their courses in response to the axial depth tinuous for over 1000 km (Schweller & Kulm
gradient or are deflected by the outer trench 1978); however, the extent of individual flow
slope. Channellized transport on the trench units has not been documented. The longitudi-
floor appears to be very localized, as channels nal distance and continuity of axial transport is
are no longer evident within a distance of thus poorly known.
approximately 30 km from the mouth of Rio Non-channellized longitudinal flow almost
Balsas Canyon (Underwood & Karig 1980; certainly occurs along trench floors in many
Underwood et al. 1980). cases. Axial channels are absent in the Middle
In trench systems with higher rates of sedi- America Trench north of Acapulco, for exam-
mentation and lower convergence rates (e.g. ple (Ross & Shor 1965; Karig et al. 1978), but
the Oregon-Washington margin), submarine textural data and sedimentary structures sug-
fans have built out over the seaward trench gest that turbidity currents transport sediment
slope, masking the trench as a bathymetric away from the mouths of submarine canyons
feature (Kulm & Fowler 1974; Schweller & (Ross 1971). Moreover, the continuity of the
Kulm 1978). The Astoria Fan off Oregon, for trench wedge, which maintains a thickness of
example, has been used as a model of sub- over 300 m north of Rio Balsas Canyon (Figs 1
marine-fan sedimentation (Nelson & Kulm & 2), suggests that axial transport occurs for
1973; Nelson & Nilsen 1974); in this case, the distances of 200-300 km between major point
effects of tectonics on trench sedimentation sources. Other trenches that do not display
appear to be minimal. axial channels include the central Aleutian
Large submarine canyons do not provide the Trench (Scholl 1974), the Kuril-Kamchatka
only source of sediment to reach the trench Trench (Scholl 1974), northern California
floor. Secondary sources include background (Silver 1971), Oregon-Washington margin
settling of hemipelagic debris and locally-de- (Kulm & Fowler 1974; Barnard 1978), the Gulf
rived mass flows, including slumps and slides of Oman (White & Klitgord 1976), the Peru
(e.g. Piper et al. 1973; von Huene 1974; Moore Trench (Schweller & Kulm 1978), the Ecuador
et al. 1976; Karig et al. 1981). High pore water Trench (Lonsdate 1978), and the Middle Amer-
pressures resulting from tectonically induced ica Trench off Guatemala (Seely et al. 1974;
dewatering at the base of the trench slope, Ladd et al. 1978; Ibrahim et al. 1979).
combined in some cases with oversteepened The presence or absence of an axial channel
slopes, provides a mechanism for recurrent is probably dependent upon several variables,
slope failure (e.g. Carson 1980). Mass flows such as the axial depth gradient, the texture of
thus initiated may transport reworked slope sediment reaching the trench floor, the rate of
sediments or accreted trench deposits to the sedimentation, and the type of mass flow
trench floor, commonly via small submarine mechanism involved in axial transport. Where
canyons that head on the lower slope (e.g. channels are lacking, coarse sediment is prob-
Karig et al. 1981). In general, hemipelagic ably transported as sheet-like turbidites, in a
deposition and locally-derived mass Ilows pro- manner analogous to sand-layer deposition
vide a significant source of trench sediment only within flat-floored basins along the Atlantic
if the terrigenous supply is low. Moreover, once margin (e.g. Bennetts & Pilkey 1976; Pilkey et
slumps and debris flows reach the trench floor, al. 1980). Individual sand layers in these abys-
 Sedimentary facies associations within s u b d u c t i o n c o m p l e x e s 541

sal-plain basins are continuous for distances of and apparently isolated ponds of sediment (Fig.
up to 500 km (Elmore et al. 1979). Sand-layer 2).
thickness and the percentage of sand gradually
decreases away from the source (Pilkey et al. Ancient analogues
1980). Similar sheet-like flows in trenches are
generally confined by the seaward and land- In spite of post-depositional deformation,
ward trench slopes, but in some circumstances, sedimentary sequences up to hundreds of
sand bodies may migrate up or overtop the metres in thickness are preserved within ancient
outer trench slope (Damuth 1979; J. C. Moore subduction complexes, such as the Coastal Belt
et al. 1981). Franciscan Complex of northern California
Most trenches appear to contain a continuous (Bachman 1978, 1981). Inferred trench sedi-
wedge of turbidites. In some cases, however, ments within the Coastal Belt include all turbi-
sediment supplies are not large enough for dite facies (Table 1). However, massive chan-
sediment bodies to prograde over basement nellized sandstone (facies B) is locally dominant
highs on the downgoing oceanic plate, such as and closely associated with thin-bedded de-
seamounts, aseismic ridges, and fault blocks posits of probable channel overspill origin.
(e.g. Scholl 1974; Kulm et al. 1977). In the Chaotic fine-grained deposits (facies F) may
northern Middle America Trench, basement represent hemipelagic material slumped off the
topography has created silled basins south of base of the trench slope. Turbidite facies asso-
Rio Balsas Canyon (Fig. 3). Depth to the ciations are consistent with upper- and mid-fan
trench floor increases by over 200 m from the depositional settings, but outer-fan facies asso-
north side of an exposed basement ridge to the ciations are generally lacking. Palaeocurrent
south side (Figs 2 & 3). South-directed turbidity data indicate that flow directions were parallel
currents emanating from Rio Balsas Canyon to the inferred continental margin; deposition
apparently filled the trench virtually to the crest may therefore have occurred within an axial
of the basement ridge; subsequent flows could channel on the trench floor. Elsewhere within
then spill over the ridge and continue down the the Franciscan Complex, mid- to outer-fan
trench axis. Farther south, near the DSDP Leg facies associations, consisting mostly of facies C
66 drilling sites, closely spaced reflection pro- and D turbidites, occur together with inner-fan
files show that axial transport is locally blocked deposits (Aalto 1976). These data suggest that
by basement highs (Shipley et al. 1980); low trench fans were at least locally developed
rates of sedimentation result in discontinuous within the Franciscan trench.

I-- c~
z ~:
ii I 0 I ~ -I"

,,~ w z 0 tmZ
~o w 0
c'r" - i
w oF)
(.9<i[
w
rr"
0
._I
Z W o
,~rr ~
~rF
0 Go ~- It. _I ~-iz)
4500 1 --6.0 U
W
O9
:2
ILl
--6.5~
F-
;$L'
d
ILl
>
--7.0 <
~,~:: ~'~Q%-/~ !i • , " ~'~ ~ ' ~ " ~ " # ~ :~t',',- "~ - ! ~ ~,-~~ ' , ~ : ~ - ' . ~ . . ~C=~ ` :~: ::::: : S-,,: ~:

>-
5500 <
-- 7.,5 ~
I' I I I I

0 I0 20
KM
(VE : 12X)

FIG. 3. Single-channel seismic-reflection (air-gun) profile across the floor of the Middle America
Trench south of Rio Balsas Canyon. Location of trackline B-B' is shown in Fig. 2. Seismic profile is
from DSDP Leg 66 transit.
542 M . B. U n d e r w o o d & S. B. B a c h m a n

Sedimentological data from other ancient axial transport and distal sedimentation within
subduction complexes demonstrate the diversi- a major trench-floor system in SE Alaska
ty of possible transport processes and deposi- (Nilsen & Bouma 1977; Nilsen & Moore 1979;
tional geometries within trenches. Ordovician Nilsen & Zuffa 1981).
greywackes in the Southern Uplands of Scot-
land, for example, are interpreted as accreted Sedimentary facies model
trench deposits and consist primarily of facies B
and C sandstones (Leggett et al. 1979; Leggett Based upon the available data from both
1980). Localized conglomeratic units may rep- modern and ancient trench deposits, we pro-
resent point sources within a largely non- pose a conceptual model for trench sedimen-
channellized trench-floor transport system tation which includes four major types of
(Leggett 1980). Permian to Jurassic strata in sediment bodies and corresponding facies
New Zealand include both m61anges and cohe- associations. The four trench-floor facies asso-
rent packets of strongly channellized thick- ciations are trench-fan, axial-channel, non-
bedded sandstone and conglomerate (facies A channellized, and starved-trench (Fig. 4).
& B) with associated levee deposits and over- Trench fans, in most cases, are not large
bank turbidites (Carter et al. 1978). Some of the enough to prograde over the outer trench slope,
channel-fill sequences probably accumulated in and the transition from more proximal to more
channels orientated parallel to the trench axis; distal facies associations takes place in a direc-
localized point sources are also likely (Carter et tion that is parallel to the trench axis. Sub-
al. 1978). Tectonically juxtaposed packets of marine fans distorted by restricted basin
accreted trench turbidites on Barbados are geometries have been recognized within the
laterally continuous and up to 500 m thick; geological record (e.g. Pescadore 1978).
facies associations are consistent with deposi- Trench-fan deposits include all turbidite facies
tion within middle and outer portions of one or and submarine-fan facies associations, and
more submarine fans (Speed 1981; Pudsey & palaeocurrent patterns range from radial to
Reading 1981). Inferred trench sediments on longitudinal (Fig. 4). Vertical cycles should
Kodiak Island, Alaska, were probably depo- include typical thinning- and fining-upward
sited within a basin-plain environment; domi- sequences associated with channel migration
nant facies include fine-grained facies D turbi- and abandonment, as well as thickening- and
dites and interbeds of facies G hemipelagic coarsening-upward cycles associated with pro-
shale (Nilsen & Moore 1979). Regional facies gradation of outer-fan depositional lobes (e.g.
relations and palaeocurrent data suggest that Ricci-Lucchi 1975; Mutti et al. 1978; Walker
the basin-plain deposits represent long-distance 1978).

,'
, J/Y
TRENCH " , , : : ? / ¢ / . so
TE.., NOOS
~"%~,i,~,F A N ' , '>Y, ~:: . . . . . .
" - ~ " - - "~ -'~/<x..\~\ , C H A N N E L"' :, "'(".,"~

FACIES'- ('A),B, " ~ ~ J

C, D, E, (F), 6 : - - ~
Associo[ions :
INNER TO OUTER
FAN
Associations :
Paleocorrents: INNER TO MID C, D, (F), G - \ ..... : ~ ~
RADIAL ; MAY FAN
BE DOMINANTLY Poleocurrents :
Associations: F A C I E S: (D],(F), G
L OMG/TUDINAL OUTER FAN TO
LONG/TUDINAL BASIN PLAIN Associations =
Poleocurrents: SLOPE OR B A S I N
PLAIN
L ONGI TUD/NA L

FIG. 4. Conceptual diagram showing the turbidite facies and palaeocurrent patterns predicted for
sediment bodies on the trench floor. See Table 1 for definition of facies terminology. Minor turbidite
facies are shown in parentheses, and equivalent and/or similar associations, including submarine-fan
facies associations, are also indicated. Adapted from Underwood et al. (1980).
 Sedimentary facies associations within subduction complexes 543

Trenches containing large axial channels do shelf deposits to large complexes of coalescing
not fit well into a model of submarine-fan submarine fans (Dickinson 1971, 1974; Ing-
sedimentation. Thick sequences of channellized ersoll 1978b; Dickinson & Seely 1979). In this
sandstone (facies B & C) are likely in such paper, forearc basins are considered only in so
cases; these deposits would pass sharply into far as they affect sedimentation along the lower
associated fine-grained levee and overbank de- trench slope.
posits (facies E, with D & G). Palaeocurrents Sedimentary basins along the lower slope are
reflect longitudinal flow (Fig. 4). Facies associa- generally elongate and bounded by tectonically
tions are similar to those of inner- to mid-fan active ridges (e.g. von Huene 1972, 1979; Kulm
facies, but more distal fan facies associations & Fowler 1974; Carson et al. 1974; G. F. Moore
are notably lacking. Basin-floor channels of this & Karig 1976; White & Klitgord 1976; Karig et
type have been described in the Alps and al. 1979; Arthur, von Huene et al. 1981). Local-
Apennines by Sagri (1979). ly, slope basins may contain: (1) hemipelagic
Evidence from seismic reflection data sug- deposits or remobilized material slumped off
gests that non-channellized sheet flow is com- adjacent bathymetric highs; (2) turbidites of
mon in modern trenches. Facies associations shallow-water origin transported beyond the
related to this style of sedimentation are prob- trench-slope break via submarine canyons; or
ably similar to non-channellized outer-fan and (3) debris-flow deposits containing blocks of the
basin-plain facies associations, although underlying accretionary complex (G. F. Moore
palaeocurrent patterns would be dominantly & Karig 1976). Any small basin or terrace is a
longitudinal (Fig. 4). Facies C and D turbidites potential site for the accumulation of sand-sized
and interlayered hemipelagic mudstones (facies detritus, and sand layers have been cored in
G) are expected, but well-developed vertical lower-slope basins off Kodiak Island (Kulm,
cycles are unlikely. Analogies with modern yon Huene et al. 1973; Howell & v o n Huene
abyssal-floor basins suggest a decrease in both 1978; M. Hampton, pers. comm. 1979), New
sandstone/shale ratios and sand-layer thickness Zealand (Lewis 1981), and Oregon-Washington
with increasing distance from the sediment (Kulm, von Huene et al. 1973; Kulm & Fowler
source (Pilkey et al. 1980). 1974; Barnard 1978; Kulm & Scheidegger
Trenches may be starved of coarse clastic 1979). Because of upslope blockage of sediment
material either because no terrigenous sands transport by tectonic ridges and deposition
reach the trench or axial transport on the trench within basins located higher on the slope, the
floor is blocked. The sedimentary facies chances of coarse sediment reaching a lower-
associations expected with a 'starved' trench are slope basin decreases with increasing distance
similar to those of the slope or basin plain (Fig. down the trench slope (Underwood et al. 1980).
4); deposits are dominated by facies G Where basins are not present, the apron of
hemipelagic muds and fine-grained facies D sediments covering the lower trench slope
turbidites, along with possible facies F slump generally consists of hemipelagic muds and
bodies. occasional thin beds of fine sand and silt (Ross
1971; Kulm, von Huene et al. 1973; Barnard
1978; Langseth, Okada et al. 1978; Moore,
Trench-slope deposits Watkins et al. 1979; Kulm & Scheidegger 1979;
Krissek et al. 1980; von Huene, Aubouin et al.
The structure and morphology of forearc re- 1980; McMillen & Haines 1981 ; McMillen et al.
gions are diverse and variable (Seely 1979; 1981; Arthur, von Huene et al. 1981). In many
Dickinson & Seely 1979). Large sedimentary cases, the slope muds overlie much coarser
basins such as the Aleutian Terrace (Marlow et trench sediments that were accreted at the base
al. 1973; Scholl 1974) and the Mentawai of the lower slope ( e . g . J . C . Moore & Karig
Trough, off Sumatra (Karig et al. 1980) are 1976; Moore, Watkins et al. 1979). Tectonic
commonly located between the active magmatic loading and oversteepened slopes may cause
arc and the trench-slope break. Several smaller sediment failure (slumps and slides) within the
basins or benches may develop on the upper apron of slope muds (e.g. Hampton et al. 1978).
trench slope, such as off Oregon-Washington
(Barnard 1978; Kulm & Scheidegger 1979) and Upslopetrapping
Peru-Chile (Coulbourn & Moberly 1977;
Moberly et al. 1981; Kulm et al. 1977; Kulm et Small submarine canyons play an important
al. 1981). Models of forearc-basin sedimenta- role in the distribution of sediment along the
tion are well-established. Sedimentary trench slope by providing an effective transport
sequences range from non-marine, deltaic, and route for coarse detritus. Most of the canyons
544 M. B. Underwood & S. B . B a c h m a n

on the landward slope of the northern Middle distinguishing between these two types of units,
America Trench head downslope of the con- as trench deposits are generally more highly
tinental shelf (Figs 1 & 2; Fisher 1961), and, as deformed than associated slope sediments (e.g.
a result, they probably do not receive a direct Bachman 1978; Moore & Karig 1980).
supply of terrigenous detritus with the present Perhaps the most complete record of lower-
position of sea-level. It is likely, however, that slope sedimentation is exposed on Nias Island,
sand-sized material entered the canyons during off Sumatra (Moore 1979; Moore & Karig 1976,
global low stands associated with Pleistocene 1980; Moore et al. 1980). Stratigraphic se-
glacial epochs. Moreover, unconfined mass quences exposed on Nias define an overall
movements (sediment creep, slumps, slides, coarsening- and thickening-upward trend
debris flows) may carry coarse debris beyond associated with basin uplift and the prograda-
the shelf break (Field & Clarke 1979; Nardin et tion of a submarine fan complex (Moore et al.
al. 1979). Small submarine canyons can then 1980). Basal slope strata overlie m61ange
intercept and funnel downslope the turbidity (accreted trench deposits) and consist of
currents generated by mass movements occur- hemipelagic marls and lutite turbidites (facies D
ring upslope or on the outer shelf. & G). There is a pronounced increase upsection
Tectonic ridges and other bathymetric highs in coarser facies A, B, and C deposits, and
on the trench slope commonly cut off the paths small-scale vertical cycles suggest that deposi-
of submarine canyons and block the transport tion of coarser facies occurred within outer- to
of turbidites and other mass flows. The mecha- inner-fan environments (Moore et al. 1980).
nics of upslope trapping are well-documented in The stratigraphic record on Nias demonstrates
the northern Middle America Trench, where the influence of tectonic processes on lower-
detailed local bathymetry shows that the small slope sedimentation. Apparently, individual
canyons typically coalesce downslope into basins, initially isolated from sources of coarse
fewer and larger canyons (Karig et al. 1978; detritus, were uplifted and eventually linked
Underwood et al. 1980). As an individual with downcutting submarine canyons which
canyon approaches a ridge, such as the trench- headed off the coast of Sumatra (Moore 1979;
slope break, the canyon channel is blocked, the Moore et al. 1980). Continued uplift and de-
canyon gradient decreases, and sediment be- position allowed progradation of submarine
comes ponded behind the ridge in a slope or fans and the infilling of the slope basins.
forearc basin. Reflection profiles display far Several examples of possible slope-basin de-
fewer submarine canyons below the trench- posits occur within the Franciscan Complex of
slope break (Underwood et al. 1980). In gener- California (Howell et al. 1977; Underwood
al, only the largest canyons continue across the 1977; Bachman 1978, 1981; Smith et al. 1979).
active ridges of the trench slope (Underwood & These strata are relatively undeformed and
Karig 1980). Approximately 80% of the sub- generally dominated by thick sequences of
marine canyons off Mexico end without channellized facies B sandstone, with associ-
reaching the trench floor; most terminate above ated deposits of levee/overbank origin (facies
the trench-slope break (Figs 1 & 2). Where a E, with minor D & G). Some sections in the
greater number of active ridges are present, Coastal Belt contain abundant facies C & D
such as the Sunda Trench (G. F. Moore & turbidites, but distal fan-facies associations are
Karig 1976; Karig et al. 1979), the effects of generally lacking (Bachman 1978, 1981). Facies
structural blockage of sediment transport are associations suggest deposition within mid- to
more pronounced, and the chances of sub- outer-fan channels and depositional lobes, or
marine canyons maintaining their channels their facies equivalents (Howell et al. 1977;
greatly diminishes downslope. Bachman 1978; Smith et al. 1979). The pre-
dominance of thick-bedded sandstone can be
Ancient analogues attributed to selective trapping of the coarse-
grained parts of gravity flows within restricted
In general, differentiation of trench-floor and slope basins, while more turbulent fine-grained
trench-slope deposits within ancient subduction fractions were able to bypass the basins and
complexes is difficult on the basis of sedi- settle farther down the trench slope (Bachman
mentary facies analysis alone, except in cases 1978; Smith et al. 1979).
such as Barbados, where a clear lithological Ancient trench-slope deposits on Kodiak Is-
distinction exists between thin-bedded biogenic land consist primarily of thick sequences of
marls (slope deposits) and accreted trench facies G mudstone (Nilsen & Moore 1979).
turbidites (e.g. Speed 1981). Structural style is Chaotic deposits (facies F) and thick beds of
perhaps the most commonly used criterion for channellized sandstone and conglomerate
 Sedimentary facies associations within subduction complexes 545

(facies A & B) crop out locally and probably C deposits; in large canyons, levee and over-
represent slump deposits and canyon/channel bank deposits (facies D, E, & G) may represent
fill, respectively (Nilsen & Moore 1979). local deposition outside the channel thalweg.
Trench-slope deposits on New Zealand contain Slumps off canyon walls (facies F) may also
deposits of low energy bottom currents (con- become interstratified with canyon fill. Inactive
tourites) in addition to the dominant hemipela- canyons are filled largely with hemipelagic
gic mudstones and lutite turbidites (Carter et al. slope deposits and fine-grained turbidites
1978). (facies D & G). Palaeocurrents from canyon fill
should be dominantly at a high angle to the
margin (Fig. 5).
Sedimentary facies model
Because of upslope trapping, trench-slope
In our facies model for trench-slope deposits basins near the base of the landward slope
(Fig. 5), submarine-canyon and slope facies generally receive only fine-grained sediments
associations are generally the same as for tecto- derived from hemipelagic settling and dilute
nically inactive environments (e.g. Whitaker density flows (facies D & G). Slumps and slides
1974; Kelling & Stanley 1976). Slope deposits derived from tectonically active bounding
are dominated by facies G hemipelagic mud- ridges are also likely, but such chaotic deposits
stones and thin-bedded, fine-grained facies D would consist primarily of remobilized fine-
turbidites (Fig. 5). The occurrence of chaotic grained slope sediments. In our facies model,
deposits associated with slope failure (facies F) basins that receive only fine-grained material
may be more common within subduction com- are classified as 'immature' slope basins
plexes, however, because of the related tectonic (Fig. 5).
activity. Channel lag deposits from submarine Progressive growth of the accretionary prism
canyons consist largely of coarse facies A, B, & generally causes uplift of trench-slope basins

SUBMARINE cANYoN ~/. ' : / / ~ TERRG i ENOUS,L

FACIES~XI, B,(C),(DJ, E, ';,)~// . SOURCE~,.
~Po,.ocu.,n,,, ~ ~I..~..<<"~/>. : / / ~/~//, /
/ . ,/,, HIGH A N G L E TO .~\',~\'i~/-~ . _~,////. J'/- // / //
/ i V/ MARGIN "~..~.~'.-~-----~" "/'// FACIES= / / /I

/,,,, ~~ ; ~ ..j / . \~,:~/./

//
~ V ~ -:_~.~
c " j ~ r f i ~,i,} i
i,J/'/1, J,
Poleocurrents:
, I //
/ .'---(D),(F),6 /'~//~'~ / i/I~ ;. . " ' ,~ ~ !
/,/,/ ," / , , ~ :-\ . / .' ., , I III/ , , I

/ .~. . .'. . .\. . , , ,1 /J,//

~ / // ; ,.
- -L:~ -_ i ~
, S. . .L. . O P E BASIN '. ¢
£'/I - ~ ~
..... .FLOOR ?_L./t.',// /./,/~7" ---~~~ I,l,17 s o u R c E
>__--J~_ -.. ~- \ < < ~ ~ , ,I</Is ~ ~ ~ j , , .
_ -/_-.-- ~ - - - - - --~-..-__ ~ Associations, / / /1 , ~ 7- -?~
......... ~. ..-U~ ; i ~'~ ~==--- SLOPE OR BASIN PLAIN/ " I ~- J /
~- ...... i- i / ~ i-- . . . . __. - ~ "~-~--'-~- ~ r ~ .~. ' '~; . / h i -- " "- "
----__J_.___-<_._:::~ . . . . 11 :~:- __: ___~---~-.._~_..\7'"~.~ ~ ~ J

FIG. 5. Conceptual diagram showing the turbidite facies and palaeocurrent patterns predicted for
trench-slope settings. The terminology follows that of Fig. 4 (see Table 1). The terms 'mature' and
'immature' slope basin refer to the presence or absence, respectively, of a source of coarse
terrigenous sediment, and are not necessarily indicative of any particular bathymetric position on the
trench slope. Adapted from Underwood et al. (1980).
546 M. B. U n d e r w o o d & S. B. B a c h m a n

through time; this is associated with a decrease gies is clearly not appropriate in all cases. The
in tectonic activity with increasing distance identification of submarine-fan deposits within
from the active basal slope (Seely et al. 1974; subduction complexes should depend upon the
Karig & Sharman 1975). As a result, bounding documentation of appropriate palaeocurrent
thrust ridges may become inactive during the patterns, vertical cycles, and fan facies associa-
uplift history of individual slope basins, thereby tions that are consistent with submarine-fan
increasing the width of the basins through time progradation or regression. In the absence of
(G. F. Moore & Karig 1976). The combined such documentation, facies association termi-
effect of these processes may be the sudden nology should be linked with the geometries of
influx of coarse terrigenous detritus into pre- the other types of sediment bodies observed
viously isolated slope basins, as basin uplift is within modern trenches, such as axial channels.
accompanied by progressive downcutting of Our facies models have other implications for
submarine canyons (e.g. Moore et al. 1980). environmental reconstructions within accre-
In our facies model, we define any such basin tionary margins, including the identification
that receives a direct supply of coarse clastic and differentiation of separate tectono-
material as 'mature', regardless of its actual stratigraphic units. Because of the complexity
position on the trench slope (Fig. 5). and variability of sedimentation that is possible
The facies associations for mature slope within a single trench, general lithologies and
basins are varied and depend largely upon the turbidite facies associations can change drama-
geometry of the basin. All of the submarine-fan tically within a relatively short distance. For
facies associations may be present in large slope example, structural blockage of axial transport
basins. In contrast, small narrow basins may paths by a seamount on the downgoing oceanic
selectively trap coarse material and allow fine- plate could allow thick sequences of sandy
grained sediment to bypass the basin. turbidites to accumulate on one side of the
Palaeocurrent patterns may also vary from ra- blockage, with hemipelagic muds dominating
dial to longitudinal (Fig. 5). the other side. Subsequent accretion and uplift
of the seamount and trench fill could expose
sections of strata that might appear to represent
Discussion completely unrelated geological histories. The
resulting juxtaposition of sandy turbidites,
The sedimentary facies models presented in this basalts, and mudstones could be erroneously
paper are somewhat speculative, due largely to attributed to large-scale tectonic events associ-
the general lack of detailed sampling within ated with accretion, rather than the original
modern subduction zones. The available data depositional geometry that existed on the
clearly indicate, however, that over-simplified trench floor. In other cases, it may be nearly
models cannot account for the observed com- impossible to differentiate between trench fill
plexity and diversity of sediment bodies and and trench-slope deposits. For example, abyssal
patterns of sedimentation. We have attempted sediments, trench sediments, and slope sedi-
to stress the variety of possible facies associa- ments could all be very similar if the trench is
tions and identify some of the variables that starved of coarse clastic material. If tectono-
affect sedimentation within trench-floor and stratigraphic designations are based primarily
trench-slope environments. upon sedimentary facies relations, then all of
Turbidite facies terminology is an important the resulting strata would be included in the
tool in the analysis of sedimentary strata, same unit, and the designations would have
including strata exposed within ancient sub- little genetic meaning.
duction complexes. However, turbidite facies In addition to the predicted lithological varia-
associations within subduction complexes may tions and complexities, significant differences in
be other than those of a submarine fan. The sandstone petrology could result for trench-
facies associations that do not fit well into the floor and slope-basin deposits that accumulate
slope/fan/basin-plain model of deep-basin sedi- in different sites along the length of an accretion-
mentation include axial-channel, non- ary margin. For example, both regional and
channellized (sheet-flow) trench-floor, starved- local changes in source rocks and onland drain-
trench, immature-slope-basin, and certain asso- age patterns could introduce sands of differing
ciations within mature slope basins, such as composition into the submarine canyons that
small basins that selectively trap coarse detri- funnel detritus into different sedimentary
tus. Submarine fans also occur within subduc- basins. Moreover, long-distance axial transport
tion zones, but we believe that the use of in the trench could juxtapose detritus of one
terminology linked to submarine-fan morpholo- composition against sediments of a different
 Sedimentary facies associations within subduction complexes 547

provenance that were transported via local sub- cannot be ignored. The separation of strata into
marine canyons. It is therefore important to meaningful tectonostratigraphic units should
integrate palaeocurrent data with data on sand- d e p e n d upon several criteria, such as contrasts
stone petrology before reliable provenance de- in sedimentary facies associations, sandstone
terminations can be made. petrology, structural style, palaeontology, and
We suggest that normal sedimentary pro- general physical properties.
cesses, rather than large-scale tectonic events,
may be responsible in some cases for lithologic-
al and compositional changes recorded within ACKNOWLEDGMENTS: We would like to thank the
following people for valuable contributions to this
many subduction complexes and accretionary
paper: M. Hampton, D. Howell, D. Karig, G. F.
margins. Clearly, the possible effects of pene- Moore, J. C. Moore, W. Normark, and W. Schwel-
c o n t e m p o r a n e o u s or subsequent tectonism ler. The manuscript was greatly improved by critical
must be considered, but the possibility of relat- reviews from C. Ando, J. Leggett, G. F. Moore, W.
ing observed changes or differences to a logical Schweller, and the Italian and English sedimentolo-
system of sediment transport and deposition gists who served as anonymous referees.

References
AALTO, K. R. 1976. Sedimentology of a melange: (eds). Sedimentation in Submarine Canyons,
Franciscan of Trinidad, California. J. sediment. Fans, and Trenches, 340-61. Dowden,
Petrol. 46, 913-29. Hutchinson & Ross, Stroudsburg, Pennsylvania.
ARTHUR, M., VON HtJENE, R. et al. 1981. Sedimenta- Coulbourn, W. T. & Moberly, R. 1977. Structural
tion in the Japan Trench. Initial Rep. Deep Sea evidence of the evolution of forearc basins off
drill. Proj. 57, U.S. Govt Printing Office, South America. Can. J. Earth Sci. 14, 102-16.
Washington, in press. CURRAY, J. R. & MOORE, D. G. 1974. Sedimentary
BACHMAN,S. B. 1978. A Cretaceous and early Terti- and tectonic processes in the Bengal deep-sea fan
ary subduction complex, Mendocino coast, and geosyncline. In: BURK, C. A. & DRAKE, C.
northern California. In: HOWELL, D. G. & L. (eds). The Geology of Continental Margins,
McDOUGALL, K. (eds). Mesozoic Paleogeogra- 617-28. Springer-Verlag, New York.
phy of the Western United States. Pacif. Sec. Soc. DAMUTH, J. E. 1979. Migrating sediment waves
econ. Paleontol. Mineral. Pacif. Coast Paleo- created by turbidity currents in the northern
geogr. Symp. 2, 419-30. South China Basin. Geology, 7, 520-3.
BACHMAN, S. B. 1981. The Coastal Belt of the DICKINSON, W. R. 1971. Clastic sedimentary se-
Franciscan: youngest phase of northern Califor- quences deposited in shelf, slope, and trough
nian subduction (this volume). settings between magmatic arcs and associated
BARNARD, W. O. 1978. The Washington continental trenches. Pacif. Geol. 3, 15-30.
slope: Quaternary tectonics and sedimentation. 1974. Sedimentation within and beside ancient
Mar. Geol. 27, 79-114. and modern magmatic arcs. In: DoTr, R. H. &
BENNETrS, K. R. & PILKEY, O. H. 1976. Characteris- SHAVER, R. H. (eds). Modern and Ancient
tics of three turbidites, Hispaniola-Caicos Basin. Geosynclinal Sedimentation. Spec. Publ. Soc.
Bull. geol. Soc. Am. 87, 1291-300. econ. Paleontol. Mineral. Tulsa, 19, 230-9.
BOUMA,A. H. & NILSEN, T. H. 1978. Turbidite facies DICKINSON, W. R. & SEELY, D. R. 1979. Structure
and deep-sea fans--with examples from Kodiak and stratigraphy of forearc regions. Bull. Am.
Island, Alaska. Proc. Offshore Tech. Conf. 1, Assoc. Petrol. Geol. 63, 2-31.
559-70. ELMORE, R. D., PILKEY, O. H., CLEARY, W. J. t~
CARLSON, P. R. & NELSON, C. H. 1969. Sediments CURRAN, H. A. 1979. Black Shell turbidite,
and sedimentary structures of the Astoria sub- Hatteras Abyssal Plain, western Atlantic Ocean.
marine canyon-fan system, northeast Pacific. J. Bull. geol. Soc. Am. 90, 1165-76.
sediment. Petrol. 39, 1269-82. FIELD, M. E. • CLARKE, S. H. JR 1979. Small-scale
CARSON, B. 1980. Sediment dewatering associated slumps and slides and their significance for basin
with subduction: experimental results. Abstr. slope processes, southern California Borderland.
Progr. geol. Soc. Am. 12, 399. In; DOYLE,L. J. & PILKEY, O. H. (eds). Geology
- - , YUAN, J., MYERS, P. B. t~ BARNARD,W. D. of Continental Slopes. Spec. Publ. Soc. econ.
1974. Initial deep-sea sediment deformation at Paleontol. Mineral. Tulsa, 27, 223-30.
the base of the Washington continental slope--a FISHER, R. L. 1961. Middle America Trench:
response to subduction. Geology, 2, 561-4. topography and structure. Bull. geol. Soc. Am.
CARTER, R. M., HICKS, M. D., NORRIS, R. J. & 72, 703-20.
TURNBULL,I. M. 1978. Sedimentation patterns in HAMPTON, M. A., BOUMA, A. H., CARLSON, P. R.,
an ancient arc-trench-ocean basin complex: Car- MOLNIA, B. F. t~ CLUKEY,E. C. 1978. Quantita-
boniferous to Jurassic Rangitata orogen, New tive study of slope instability in the Gulf of
Zealand. In: STANLEY, D. J. & KELLING, G. Alaska. Proc. Offshore Tech. Conf. 4, 2307-18.
548 M. B. U n d e r w o o d & S. B. B a c h m a n

HAYES, D. E. 1974. Continental margin of western KELLING, G. & STANLEY, D. J. 1976. Sedimentation
South America. In" BURK, C. A. & DRAKE, C. L. in canyon, slope, and base-of-slope environ-
(eds). The Geology of Continental Margins, 581- ments. In: STANLEY,D. J. • SWIFT, D. J. P. (eds).
90. Springer-Verlag, New York. Marine Sediment Transport and Environment
HOWELL, D. G., VEDDER, J. G., MCLEAN, H., JOYCE, Management, 379-435. Wiley Interscience, New
J. M., CLARKE, S. H. JR & SMITH, G. 1977. York.
Review of Cretaceous geology, Salinian and KRISSEK, L. A., SCHE1DEGGER, K. F. (~ KULM, L. D.
Nacimiento blocks, coast ranges of central Cali- 1980. Surface sediments of the Peru-Chile con-
fornia. In: HOWELL, D. G., VEDDER, J. G. & tinental margin and Nazca plate. Bull. geol. Soc.
McDOUGALL, D. (eds). Cretaceous Geology of Am. 91,321-31.
the California Coast Ranges, West of the San KULM, L. D. & FOWLER, G. A. 1974. Oregon con-
Andreas Fault. Pacif. Sec., Soc. econ. Paleontol. tinental margin structure and stratigraphy: a test
Mineral. Tulsa, Pacif. Coast Paleogeogr. Field of the imbricate thrust model, In: BURr, C. A. &
Guide, 2, 1-46. DRAKE, C. L. (eds). The Geology of Continental
& VON HUENE, R. 1978. Trench-slope basins off Margins, 261-83. Springer-Verlag, New York.
Kodiak Island, a modern analog for some Late -- (~ SCHEIDEGGER, K. F. 1979. Quaternary sedi-
Cretaceous rocks of central California. Abstr. mentation on the tectonicaly active Oregon con-
Progr. geol. Soc. Am. 10, 109-10. tinental slope. In: DOYLE, L. J. & PILKEY, O. H.
HsO, K. J., KELTS, K. (~ VALENTINE, J. W. 1980. (eds). Geology of Continental Slopes. Spec.
Resedimented facies in Ventura Basin, Califor- Publ. Soc. econ. Paleontol. Mineral. Tulsa, 29,
nia, and model of longitudinal transport of tur- 247-64.
bidity currents. Bull. Am. Assoc. Petrol. Geol. --, SCHWELLER, W. J. & MASIAS, A. 1977. A
64, 1034-51. preliminary analysis of the subduction processes
leRAnlM, A. K., LATHAM, G. V. & LADD, J. 1979. along the Andean continental margin, 6° to 45°S,
Seismic refraction and reflection measurements In: TALWANI, M. & PrrMAN, W. C., III (eds).
in the Middle America Trench offshore Guate- Island Arcs, Deep-sea Trenches, and Back-arc
mala. J. geophys. Res. 84, 5643-9. Basins. Am. Geophys. Union, M. Ewing Ser. 1,
INGERSOLL, R. V. 1978a. Submarine fan facies of the 285-301.
Upper Cretaceous Great Valley sequence, -- , VON HUENE, R. et al. 1973. Initial Rep. Deep
northern and central California. Sediment. Sea drill. Proj. 18, U.S. Govt Printing Office,
Geol. 21,205-30. Washington, 1077 pp.
- - 1978b. Paleogeography and paleotectonics of KULM, L. D., RESIG, J. M., THORNBURG, T. M. &
the late Mesozoic forearc basin of northern and SCHRADER, H.-J. 1981. Cenozoic structure, stra-
central California. In: HOWELL, D. G. & tigraphy and tectonics of the central Peru forearc
McDOUGALL, K. (eds). Mesozoic Paleogeogra- (this volume).
phy of the Western United States. Pacif. Sec., Soc. LADD, J. W., IBRAHIM, A. K., MCMILLEN, K. J.,
econ. Paleontol. Mineral. Pacif. Coast Paleo- LATHAM, G. V., VON HUENE, R., WATKINS, J. S.,
geogr. Syrup. 2, 471-82. MOORE, J. C. & WORZELL,J. L. 1978. Tectonics of
KARIG, D. E., CARDWELL, R. K., MOORE, G. F. & the Middle America Trench, offshore Guatamala.
MOORE, D. G. 1978. Late Cretaceous subduction Int. Symp. Guatamala 4 February Earthquake and
and continental margin truncation along the Reconstruction Process, 1, Guatamala City, May
northern Middle America Trench. Bull. geol. 1978.
Soc. Am. 89, 265-76. LANGSETH, M., OKADA, H. et al. 1978. Near the Japan
, INGLE, J. C. et al. 1975. Initial rep. Deep Sea Trench--transects begins. Geotimes, 23, 22-6.
drill. Proj. 32, 927 pp. U.S. Govt Printing Office, LEGGETr, J. K. 1980. The sedimentological evolution
Washington. of a Lower Palaeozoic accretionary fore-arc in the
• LAWRENCE,M. B., MOORE, G. F. & CURRAY,J. Southern Uplands of Scotland. Sedimentology,
R. 1980. Structural framework of the forearc 27, 401-17.
basin, NW Sumatra. J. geol. Soc. London, 137, -- , MCKERROW, W. S. & EALES, M. H. 1979. The
77-91. Southern Uplands of Scotland: a Lower
--, MOORE, G. F., CURRAY, J. R. & LAWRENCE, M. Palaeozoic accretionary prism. J. geol. Soc. Lon-
B. 1981. Morphology and shallow structure of don, 136, 755-70.
the trench slope off Nias Island, Sunda arc, in LEWIS, K. B. 1981. Quaternary sedimentation of the
special SEA TAR volume. Am. geophys. Union, Kikurangi oblique subduction and transform
in press. margin. In: BALLANCE, P. F. & READING, H. G.
& SHARMAN, G. F., III 1975. Subduction and (eds). Sedimentation in Oblique-slip Moblile
accretion in trenches. Bull. geol. Soc. Am. 86, Zones. Spec. Publ. int. Assoc. Sediment. 4,
377-89. 171-89. Blackwell Scientific Publications, Ox-
- - , SUPARKA, S., MOORE, G. F. & HEHANUSSA, P. ford.
E. 1979. Structure and Cenozoic evolution of the LONSDALE, P. 1978. Ecuadorian subduction system.
Sunda Arc in the central Sumatra region. In: Bull. Am. Assoc. Petrol. Geol. 62, 2454-77.
WATKINS, J. S., MONTADERT,L. & DICKERSON, P. MARLOW, M. S., SCHOLL, D. W., BUFFINGTON, E. C.
W. (eds). Geological and Geophysical Investiga- & ALPHA, T. R. 1973. Tectonic history of the
tions of Continental Margins. Mem. Am. Assoc. central Aleutian arc. Bull. geol. Soc. Am. 84,
Petrol. Geol. 29, 223-37. 1555-74.
 S e d i m e n t a r y facies associations within s u b d u c t i o n c o m p l e x e s 549

MCMILLEN, K. J., ENKEBOLL, R. H., MOORE, J. C., - - t~ NILSEN, T. H. 1974. Depositional trends of
SHIPLEY, T. A. & LADD, J. W. 1981. Sedimenta- modern and ancient deep-sea fans. In: D o f f , R.
tion in different tectonic environments of the H. • SHAVER, R. H. (eds). Modern and Ancient
Middle America Trench, Southern Mexico and Geosynclinal Sedimentation. Spec. Publ. Soc.
Guatemala (this volume). econ. Paleontol. Mineral., Tulsa, 19, 69-91.
-- ~¢ HAINES, T. R. 1981. Late Quaternary sedi- NILSEN, T. H. 1980. Modern and ancient submarine
ments of the southern Mexico margin. In: MOORE, fans: discussion of papers by R. G. Walker and
J. C., WATKINS,J. S. et al. (eds). Initial Rep. Deep W. R. Normark. Bull. Am. Assoc. Petrol. Geol.
Sea drill. Proj., 66, U.S. Govt Printing Office, 64, 1094-101.
Washington. - - & BOUMA, A. H. 1977. Turbidite sedimentology
MOBERLEY, R., SHEPHERD, G. L. & COULBOURN, W. and depositional framework of the Upper Cre-
T. 1981. Forearc and other basins, continental taceous Kodiak formation and related strati-
margin of northern and southern Peru and adja- graphic units, southern Alaska. Abstr. Progr.
cent Ecuador and Chile (this volume). geol. Soc. Am. 9, 1115.
MOORE, D. G., CURRAY,J. R. & EMMEE, F. J. 1976. -- & Moore, G. W. 1979. Reconnaissance study of
Large submarine slide (olistostrome) associated Upper Cretaceous to Miocene stratigraphic units
with Sunda arc subduction zone, northeast and sedimentary facies, Kodiak and adjacent
Indian Ocean. Mar. Geol. 21,211-26. islands, Alaska. Prof. Pap. U.S. geol. Surv.
MOORE, G. F. 1979. Petrology of subduction zone 1093, 1-34.
sandstones from Nias Island, Indonesia. J. sedi- NORMARK, W. R. 1970. Growth patterns of deep-sea
ment. Petrol. 49, 71-84. fans. Bull. Am. Assoc. Petrol. Geol. 54, 2170-95.
--, BILLMAN, H. G., HEHANUSSA, P. E. & KARIG, - - 1978. Fan valleys, channels, and depositional
D. E. 1980. Sedimentology and paleobathymetry lobes on modern submarine fans: Characters for
of Neogene trench-slope deposits, Nias Island, recognition of sandy turbidite environments.
Indonesia. J. Geol. 88, 161-80. Bull. Am. Assoc. Petrol. Geol. 62, 912-31.
& KARIG, D. E. 1976. Development of PESCADORE, T. 1978. The Irpinids: A model of tecto-
sedimentary basins on the lower trench slope. nically controlled fan and base-of-slope sedi-
Geology, 4, 501-4. mentation in southern Italy. In: STANLEY,D. J.
- - & -- 1980. Structural geology of Nias Is- & KELI.ING, G. (eds). Sedimentation in Sub-
land, Indonesia: implications for subduction marine Canyons, Fans, and Trenches, 325-39.
zone tectonics. Am. J. Sci. 280, 193-223. Dowden, Hutchinson & Ross, Stroudsburg,
MOORE, J. C. & KARIG, D. E. 1976. Sedimentology, Pennsylvania.
structural geology, and tectonics of the Shikoku PILKEY, O. H., LOCKER, S. D. & CLEARY, W. J. 1980.
subduction zone. Bull. geol. Soc. Am. 87, 1259- Comparison of sand layer geometry on the flat
68. floors of ten modern depositional basins. Bull.
--, WATKINS,J. S. etal. 1979. Progressive accretion Am. Assoc. Petrol. Geol. 64, 841-56.
in the Middle America Trench, southern Mex- PIPER, D. J. W., VON HUENE, R. & DUNCAN, J. R.
ico. Nature. London, 281,638--42. 1973. Late Quaternary sedimentation in the ac-
MOORE, J. C. et al. 1981. Facies belts of the Middle tive eastern Aleutian Trench. Geology, 1, 19-22.
America Trench ~and forearc region, southern PRINCE, R. A., SCHWEELER, W. J., COUEBOURN, W.
Mexico: results from Leg 66 DSDP (this T., SHEPARD, G. L., NESS, G. E. & MASIAS, A.
volume). 1980. Bathymetry of the Peru-Chile Continental
MUTrl, E., NILSEN, T. H. & RICcI-LUccHI, F. 1978. Margin and Trench. Geol. Soc. Am. Map Set.
Outer fan depositional lobes of the Laga Forma- PUDSEY, C. J. & READING, H. E. 1981. Sediment-
tion (upper Miocene and lower Pliocene) east- ology and structure of the Scotland Group,
central Italy. In: STANLEY, D. J. & KEELING, G. Barbados (this volume).
(eds). Sedimentation in Submarine Canyons, REIMNITZ, E. & GUITERREZ-ESTRADA,M. 1970. Rapid
Fans and Trenches, 210-23. Dowden, Hutch- changes in the head of the Rio Balsas submarine
inson & Ross, Stroudsburg, Pennsylvania. canyon system, Mexico. Mar. Geol. 8, 245-58.
-- & Rlco-LvccHl, F. 1975. Turbidite facies and Rlco-LuccHI, F. 1975. Depositional cycles in two
facies associations. In: Examples of turbidite and turbidite formations of northern Apennines
facies associations from selected formations of the (Italy). J. sediment. Petrol. 45, 3-43.
northern Apennines, 21-36. 9th Int. Congr. Ross, D. A. 1971. Sediments of the northern Middle
Sedimentology, Nice, Guidebook. America Trench. Bull. geol. Soc. Am. 82, 303-
NARDIN, T. R., EDWARDS, B. D. & GORSEINE, D. S. 22.
1979. Santa Cruz Basin, California Borderland: - - SHOR, G. G. Jr 1965. Reflection profiles
&

dominance of slope processes in basin sedi- across the Middle America Trench. J. geophys.
mentation. In: DOYLE, L. J. & PIEKEY, O. H. Res. 70, 5551-71.
(eds). Geology of Continental Slopes. Spec. SAGRI,M. 1979. Upper Cretaceous carbonate turbi-
Publ. Soc. econ. Paleontol. Mineral. Tulsa, 27, dites of the Alps and Apennines deposited below
209-22. the calcite compensation level. J. sediment.
NELSON, C. H. & KULM, L. D. 1973. Submarine fans Petrol. 49, 23-8.
and deep-sea channels. In: Turbidites and Deep SCHOLL,D. W. 1974. Sedimentary sequences in the
Water Sedimentation. Pacif. Sec., Short Course, north Pacific trenches. In: BURK, C. A. &
Soc. econ. Paleontol. Mineral. 39-78. DRAKE, C. L. (eds). The Geology of Continental
550 M. B. U n d e r w o o d & S. B. B a c h m a n

Margins, 493-504. Springer-Verlag, New York. trench-slope basin. In: HOWELL, D. G., VEDDER,
, CHRISTENSEN, M. N., YON HUENE, R. & J. G. & McDOUGALL, K. (eds). Cretaceous Geol-
MARLOW, M. S. 1970. Peru-Chile trench sedi- ogy of the California Coast Ranges, West of the
ments and sea-floor spreading. Bull. geol. Soc. San Andreas Fault. Pacif. Sec. Soc. econ.
Am. 81, 1339-60. Paleontol. Mineral., Pacif. Coast Paleogeogr.
--, MARLOW, M. S. & COOPER, A. K. 1977. Sedi- Field Guide, 2, 57-69.
ment subduction and offscrapping at Pacific mar- , BACHMAN, S. B. & SCHWELLER, W. J. 1980.
gins. In: TALWANI, M. & PITMAN, W. C., III Sedimentary processes and facies associations
(eds). Island Arcs, Deep Sea Trenches, and within trench and trench-slope settings. In:
Back-Arc Basins. Am. geophys. Union, M. FIELD, M. E., BOUMA, A. H. & COLBURN, I.
Ewing Ser. 1, 199-210. (eds). Quaternary Depositional Environments on
SCHWELLER, W. J. • KULM, L. D. 1978. Depositional the Pacific Continental Margin, 211-29. Pacif.
patterns and channelized sedimentation in active Sec. Soc. econ. Paleontol. Mineral., Tulsa.
eastern Pacific trenches. In: STANLEY, D. J. & VON HUENE, R. 1972. Structure of the continental
KEELING, G. (eds). Sedimentation in Submarine margin and tectonism at the eastern Aleutian
Canyons, Fans, and Trenches, 311-24. Dowden, Trench. Bull. geol. Soc. Am. 83, 3613-26.
Hutchinson & Ross, Stroudsburg, Pennsylvania. 1974. Modern trench sediments. In: BURK, C.
SEELY, D. R. 1979. The evolution of structural highs A. & DRAKE, C. L. (eds). The Geology of
bordering major forearc basins, In: WATKINS, Continental Margins, 207-11. Springer-Verlag,
J. S., MONTADERT,L. & DICKERSON, P. W. (eds). New York.
Geological and Geophysical Investigations of 1979. Structure of the outer convergent margin
Continental Margins. Mem. Am. Assoc. Petrol. off Kodiak Island, Alaska, from multichannel
Geol. 29, 245-60. seismic records. In: WATKINS, J. S., MONTADERT,
-- , VAIL, P. R. & WALTON, G. G. 1974. Trench L. & DICKERSON, P. W. (eds). Geological and
slope model. In: BURK, C. A. & DRAKE, C. L. Geophysical Investigations of Continental Mar-
(eds). The Geology of Continental Margins, 249- gins. Mem. Am. Assoc. Petrol. Geol. 29, 261-
60. Springer-Verlag, New York. 71.
SHEPARD, F. P., MARSHALL, N. F., McLOUGHLIN, P. -- , AUaOUIN, et al. 1980. Leg 67: The Deep Sea
A. & SULLIVAN, G. G. 1979. Currents in sub- Drilling Project Middle America Trench transect
marine canyons and other sea valleys. Stud. off Guatemala. Bull. geol. Soc. Am. 91,421-32.
Geol. Am. Assoc. Petrol. Geol. 8, 1-173. WALKER, R. G. 1978. Deep-water sandstone facies
SHIPLEY, T. H., MCMILLEN, K. J., WATKINS, J. S., and ancient submarine fans: Models for explora-
MOORE, J. C., SANDOVAL-OCHOA, J. H. & tion for stratigraphic traps. Bull. Am. Assoc.
WORZEL, J. L. 1980. Continental margin of Petrol. Geol. 62, 932-66.
Guerrero and Oaxaca, Mexico. Mar. Geol. 35, & Murrl, E. 1973. Turbidite facies and facies
65-82. associations, ln: Turbidites and Deep Water Sedi-
SILVER, E. A. 1971. Transitional tectonics and late mentation. Pacif. Sec., Short Course, Soc. econ.
Cenozoic structure of the continental margin off Paleontol. Mineral. 119-58.
northernmost California. Bull. geol. Soc. Am. WHITAKER, J. H. McD. 1974. Ancient submarine
82, 1-22. canyons and fan valleys. In: D o ~ , R. H. &
SMITH, G. W., HOWELL, D. G. & INGERSOLL, R. V. SHAVER, R. H. (eds). Modern and Ancient
1979. Late Cretaceous trench-slope basins of Geosynclinal Sedimentation. Spec. Publ. Soc.
central California. Geology, 7, 303-6. econ. Paleontol. Mineral., Tulsa, 19, 106-25.
SPEED, R. C. 1981. New views on the geology of WHITE, R. S. ~7. KLITGORD, K. 1976. Sediment de-
Barbados. Second Latin Am. Geol. Congr. formation and plate tectonics in the Gulf of
UNDERWOOD, M. B. 1977. The Pfeiffer Beach slab Oman. Earth planet. Sci. Lett. 32, 199-209.
deposits, Monterey County, California: possible

MICHAEL B. UNDERWOOD t~ STEVEN B. BACHMAN, Department of Geological

Sciences, Kimball Hall, Cornell University, Ithaca, New York 14853, U.S.A.
 Composition of modern deep-sea sands from arc-related basins

J. Barry Maynard, Renzo Valloni & Ho-Shing Yu

SUMMARY: Petrographic and chemical examination of modern deep-sea sands shows a
clear distinction between active and passive plate-tectonic settings. Among active settings,
sands from forearc basins of island arcs and basins from strike-slip continental margins can
be distinguished, but those from other arc-related settings overlap considerably in
composition. In particular, sands from continental margin subduction zones and those
from the backarc of island arcs appear to be indistinguishable.
Chemically, arc-related sands are very close in composition to ancient greywacke
sandstones. For instance, sodium almost always exceeds potassium. Thus it is not
necessary for sodium to be added to sands diagenetically to make greywackes, as has been
suggested.

The influence of tectonics on the composition their 'collision coasts' and 'marginal sea cate-
of sandstones has long been a central feature of gories'. The petrography of sands from the
sandstone petrology. The geosynclinal model of passive settings and the collision-related set-
global tectonics stimulated much thought on tings has been described in some detail by
this subject. Probably the best-known ideas are Dickinson & Valloni (1980). In this paper,
those of Krynine (e.g. 1942). Subsequently such therefore, we concentrate more on the arc-
research declined, but the plate model for related settings, and introduce the chemical
global tectonics has produced a new wave of properties of the sediments.
interest (e.g. Crook 1974; Schwab 1975; Potter Deep-sea sands are found in most of the
1978; Dickinson & Suczek 1979). Much of this environments listed in Table 1, but are most
work has relied on data from ancient sediments, common in strike-slip and subduction settings
where diagenetic changes and difficulty in de- and the later stages of intercontinental rifting.
ciphering the tectonic setting limit the useful- Of the subduction settings, the continent side of
ness of the conclusions. Much more work is a subducting continental margin (the backarc of
needed on recent sediments to establish a base- continental margin arcs) does not contain deep-
line against which ancient sediments can be sea sands. Instead, many of the world's big
compared. rivers derive their sediments from these areas
We have previously reported (Valloni & (e.g. the Amazon), so that the sands described
Maynard 1981) the results of our work on the by Potter (1978) may largely fill this category.
petrography of a set of recent deep-sea sands Thus the important environments from the
from piston cores. In this paper we shall com- standpoint of this paper are basins of island arcs
pare this petrographic information with bulk and basins lying seaward of continental margin
chemistry, and attempt to relate these to tecto- volcanic arcs or strike-slip plate boundaries.
nics in arc-related settings.

Petrography
Plate tectonic models of depositional
basins About 80 samples of deep-sea sands, mostly
from L a m o n t - D o h e r t y piston cores supple-
First, it is necessary to define the terms used. mented by some D S D P material, were studied
The classification of sedimentary environments petrographically (Fig. 1). The details of the
using plate tectonics has been extensively dis- methods used and the depository for all of the
cussed by Dickinson (1974), Dickinson & Seely original data are given in Valloni & Maynard
(1979) and Reading (1978, chapter 14). These (1981). For petrography, the sands were sieved
ideas are summarized in Table 1. The most to isolate the fine sand fraction, because of the
important subdivision is into passive and active strong effect of grain size on feldspar content
settings, as originally pointed out by Inman & (e.g. Field & Pilkey 1969, fig. 3). The results
Nordstrom (1971). In the classification used are summarized in Table 2 and Fig. 2.
here, passive settings are synonymous with The samples separate well into passive (trail-
their 'trailing-edge coast'. We have subdivided ing edge) and active settings on the basis of the

551
552 J. B. Maynard, R. Valloni & H.-S. Yu
TAaLE 1. Plate tectonic classification o f sedimentary environments (after Reading 1979, chapter 14;
Dickinson & Valloni 1980, fig. 2)

I Spreading-related or passive settings

(A) Intracratonic rifts (East Africa). Mostly filled by alluvial fans and lakes
(B) Failed rifts or aulacogens (Benue Trough). Thick sequence of deposits ranging from deep-sea fan to
fluvial
(C) Intercontinental rifts
1 Early (Red Sea). Evaporites and some clastics
2 Late (Atlantic). Early stage sediments along margins overlain by continental shelf or deltaic
deposits, passing seaward into oceanic crust overlain by turbidites and finally pelagics. Often termed
trailing-edge
II Active settings
(A) Continental collision-related
1 Remnant ocean basins (Bay of Bengal). Thick turbidite fan deposits eventually piled into imbricate
thrust sheets
2 Late orogenic basins (sub-Himalayas). Variety of terrestrial and shallow marine deposits, but
dominantly fluvial. Termed peripheral foreland basins if on the subducting plate
(B) Strike-slip fault-related settings (California). Thick sequences of deep-marine to fluvial sediments in
small basins, mostly derived from adjacent uplifts
(C) Subduction-related settings
1 Continental margin magmatic arcs (Andes)
(a) Forearc. Thin to thick deposits ranging from turbidites to fluvial. Termed leading-edge in some
classifications
(b) Backarc. Very thick, mostly terrestrial, accumulations. Forms much of the molasse facies of older
models. Sometimes termed ensialic backarc or retroarc
2 Intra-oceanic arcs (Japan, Aleutians)
(a) Forearc. Thin turbidite deposits plus pelagics
(b) Backarc. Thin to thick accumulations of clastics near the arc, passing into pelagics and possibly
into terrigeneous clastics again at the continental margin

I/"
S

/
/

I I I
A
I , I / x
I I • / ~,
I
/ • %

Fro. 1. Sample location map.

 Composition of modern deep-sea sands 553

primary petrographic variables quartz (Q), Supplementary variables such as the type of
feldspar (F), and fine-grained rock fragments feldspar or rock fragments have often been
(L). Samples with more than 40% quartz are used as further discriminants. Fig. 4 shows that
almost exclusively from basins adjacent to pas-
sive continental margins (Fig. 2). In addition, 1.0
Quartz

0 Trailing- edge /o ~ I- CA
•
&
Strlke-slip
Back. . . .
/o
/ O00Zb
\
T
iss
• Fare - arc
on mar in arc
II

I, I I I
0 Lv/L
Feldspar Lithics

FIG. 2. Framework constituents of modern FIG. 4. Petrography of deep-sea sands from

deep-sea sands from various tectonic set- tectonic settings that overlap on Fig. 3.
tings. Note the very close similarity of samples
from the backarc of island arcs (BA) to
those from the forearc of continental mar-
samples from intraoceanic forearc basins are
gin arcs (CA). P/F is the ratio of plagioc-
easily distinguished: they are composed almost lase to total feldspar; Lv/L the ratio of
entirely of volcanic rock fragments. Samples volcanics to total lithics.
from the other three settings have extensive
areas of overlap on triangular plots (Fig. 3).
all three of these settings have about the same
Q proportion of plagioclase in the feldspar, but
that the strike-slip related settings have a signi-
ficantly lower proportion of volcanic grains in
the rock fragments. In ancient rocks, this differ-
ence may translate to less matrix. Thus, of the
five settings considered, three have a reason-
ably distinct petrography, but basins lying on
the ocean side of magmatic arcs on continental
margins seem to contain sands almost identical
in composition to those found in the backarc
basins of island-arcs.
All of the arc-related sands share the charac-
teristics of low quartz content and a high pro-
portion of volcanic lithic fragments. Valloni &
Maynard (1981) suggested several subdivisions
of this petrographic type. They identified:
F L
Volcanic-vitric sands (Vv). These sands have
virtually no quartz, and are made up almost
FIG. 3. Range of variation in framework entirely of volcanic grains having a vitric texture
petrography of deep-sea sands. The poly- or glass shards. They are almost entirely con-
gons show one standard deviation about
the mean. (See Ingersoll 1978 for a discus- fined to forearc basins of island arcs (Fig. 5).
sion of this representation.) TE = trailing- Volcanic-lithic sands (Vp and Vk). These sands
edge; SS = strike-slip; CA = continental contain about equal amounts of feldspars and
margin arc; BA = backarc of island arc; lithic fragments, with volcanic grains again
FA = forearc of island arc. making up almost all of the lithic fragments.
554 J. B. Maynard, R. Valloni & H.-S. YU
lOO ity: the forearc and backarc basins contain very
Q different sands for a given arc-type. To what
extent this difference between backarc and
80 F
forearc sands also applies to continental margin
arcs we cannot say, having no backarc sands of
6O Lk k this type. Studies of shelf sands of Indonesia
would provide the appropriate contrast. This
result suggests that in general more elaborate
a. 40 tectonic subdivisions will not be reflected in
petrography, although within a given basin such
20 variations may be detectable (e.g. Moore
1979).
A n o t h e r limitation is the surprising similarity
in composition of sands from the backarc of
TE SS .BA CA FA island arcs and the forearc of continental mar-
gin arcs. These two settings are likely to be
Fro. 5. Distribution of petrographic types preponderant in ancient arc-related terranes,
by tectonic setting. See text for abbrevia- and their close similarity is discouraging for the
tions. use of petrography to distinguish them. Why
this pattern occurs is not clear to us.

TABLE 2. Petrology of modern deep-sea sands

Tectonic setting No. of samples Q F L C/Q P/F Lv/L

Trailing-edge (TE) 29 61 26 13 0.10 (I. 13 0.13
Strike-slip (SS) 7 31 36 33 0.31 (I.58 0.45
Continent margin-arc (CA) 8 16 53 31 - (I.72 0.98
Oceanic arc
(1) Backarc (BA) 27 16 34 50 0.19 (I.64 0.89
(2) Forearc (FA) 9 3 16 81 - 0.90 0.99

Heavy minerals are particularly abundant (3- How do these results compare with ancient
6%). Vp refers to a sub-variety with almost all sediments? Dickinson & Suzcek (1979) found
plagioclase feldspar, Vk to one with appreci- similar compositions in ancient sandstones.
able ( > 2 0 % ) K-feldspar. These sands are char- Their arc-related samples fall into the same part
acteristic of both the backarc basins of island of the QFL triangle as ours (Fig. 6), except for
arcs and the forearc basins of continental mar- our samples from intra-oceanic forearc basins.
gin arcs. Probably such sands are rarely preserved in
Lithofeldspathic sands (Lp and Lk). These sandstones on the continents. Their 'continen-
sands are significantly richer in quartz than the tal block provenance' corresponds to our pas-
preceding varieties. Feldspar and lithic frag- sive settings. We have few samples from the
ments are again present in subequal amounts, area they designate 'recycled orogen prove-
but volcanic grains no longer predominate. Two nances' but sands of this composition make up
sub-types are distinguished based on the pre- the bulk of the samples reported from modern
dominance of plagioclase (Lp) or K-feldspar big rivers by Potter (1978). Crook (1974) has
(Lk). Among the quartz grains, polycrystalline presented data from analysis of over 300
varieties are abundant (20-30%). These sands ancient turbidites that differ in some respects
are most common in strike-slip settings, from our modern sands (Fig. 7). He divided
although they are also found off passive con- these rocks into quartz-rich, quartz-
tinental margins. Two quartz-rich types (Q and intermediate, and quartz-poor varieties, as
F) can also be distinguished, but they are shown. He suggested that these divisions cor-
confined to passive settings (Fig. 5). respond to trailing-edge continental margins,
The distribution of sand types is somewhat leading-edge continental margins (both strike-
curious, because there is apparently no rela- slip and subduction), and island arcs. Our mod-
tionship to the type of island arc. That is ern sands show both the passive margin sands
'mature' island arcs like Japan and 'primitive' and the arc-related sands crossing these divi-
arcs like the Marianas, seem to be producing sions. Instead, a division at about 40% quartz
the same sands. The big difference is in polar- separates the passive and active settings, with
 Composition of modern deep-sea sands 555
Q trast between recent and ancient sands needs
further study, especially considering the simi-
larity of Potter's (1978) river sands to Crook's
CONTINENTAL~RECYCLED
BLOCK/~ "..~OROGEN pattern.

Bulk chemistry
Most of these samples were also analysed for
bulk chemistry. The object was to make a
comparison between the composition of beds of
turbidite sand and the immediately overlying
mud. These can then be compared with ancient
turbidite sand-shale pairs to see whether there
is exchange of components between the sand
L and shale beds during diagenesis. Also we
wished to see if the tectonic subdivisions pre-
FIG. 6. Petrography of sands as related to viously described are associated with distinctive
tectonic setting of the source area, largely chemical compositions. During diagenesis, and
from ancient samples, from Dickinson & low-grade metamorphism, much of the type of
Suczek (1979, fig. 9). The distribution petrographic information described earlier be-
matches closely that of Fig. 2, except com- comes obscured. This is particularly true of the
positions assigned to recycled orogen types of rock fragments, many of which are
provenances are rare in our samples. converted to matrix. Therefore it would be
particularly helpful if chemical analyses could
be used, because these processes seem to pro-
duce little change in bulk chemistry, except for
water and CO2. In addition to the modern sands
and associated muds, 129 samples of ancient
turbidite sandstones and shales were analysed.
Age, location, and chemical analyses for these
samples are presented in the appendix. Deter-
minations were made using atomic absorption
following fusion of the sample with LiBO2 and
dissolution in HNO 3 (Yu 1979).
The set of ancient turbidite sands and shales
that we used is unfortunately rather limited.
The sampling is biased towards Cambro-
Ordovician rocks of the Appalachian Basin.
Also, virtually all of the samples we collected
turned out to be from active tectonic settings,
based on our interpretation of the literature,
Fro. 7. Petrography of ancient turbidite probably either forearc to a continental margin
sandstones, after Crook (1974). The QFR (e.g. Southern Uplands of Scotland) or from
triangle differs from the QFL triangle used backarc basins of island arcs (Cambrian of
in the other figures in shifting the values Appalachian Basin?). Usually, this kind of
somewhat towards the R corner (all poly- assignment to subvarieties of active settings is
mineralic grains counted as rock frag-
very tentative, so we have not attempted it. On
ments), as discussed in Valloni & Maynard
(1981). the other hand, the average sandstone composi-
tion from these samples is close to that reported
by Pettijohn (1963, table 12) for a similar
no possible further subdivision of the active sampling of greywackes, so we feel that our
settings based on quartz. Some of the difference samples are representative. There is no com-
between our results and his are probably attri- parable data set for the shales: ours appears to
butable to our use of only fine sands, which be the first compilation of chemical data on
might give higher feldspar. It seems unlikely shales of this type.
that all of the difference can be explained by As with petrography, the results show that
such analytical variables, however. This con- passive settings are quite distinct (Table 3, Figs
556 J. B. Maynard, R. Valloni & H.-S. Y u

TABLE 3. Bulk chemistry of modern deep-sea sands and associated muds

Tectonic setting SiO2 A1203 Fe203 MgO CaO K20 Na20

I Sands
Trailing-edge 77.9 9.8 2.9 1.3 4.1 2.0 1.9
Strike-slip 67.8 15.6 3.7 2.3 3.6 2.9 3.9
Continental 69.5 14.1 3.9 1.9 4.4 2.6 3.6
margin arc
Backarc 68.8 14.4 4.5 2.4 4.4 2.0 3.6
Forearc 61.5 15.2 7.7 3.8 6.7 1.4 3.8

2 Muds
Trailing-edge 65.9 13.7 5.3 2.8 8.2 2.6 1.5
Strike-slip 65.8 14.4 6.8 3.4 4.9 2.0 2.7
Continental 66.1 16.9 6.4 3.2 3.0 2.5 2.4
margin arc
Backarc 68.0 14.9 6.5 3.1 2.8 2.3 2.5
Forearc 68.9 12.1 7.2 3.0 4.9 1.5 2.6

Modern Sands
o Trailing-edge Fe203÷MgO there is a large amount of overlap. Thus it
0 Cont.marginarc // seems unlikely that bulk chemistry can be used
+ Strike-slip // to distinguish among types of arc-related sand-
Back-arc /
O Fore- arc / . ~ Lithic stones in ancient rocks.
Ancient Greywackes / • Greywackes, the typical sandstone type of
• Crook(1974) / []
• Pettijohn (1967 II turbidites, are exceptionally rich in NazO com-
pared with other sandstone types (e.g. Petti-
john et al. 1972, p. 211). It is not known,
however, whether this high N%O reflects Na-
rich source rocks or is a diagenetic feature. For
instance, both primary and secondary sources
of Na20 have been proposed for the albite-rich
Arkoses "~ Charny greywackes of Quebec (Middleton
1972; Lajoie 1973; Lajoie et al. 1974). Diagen-
N o 2 0 ~ K20
etic addition of Na20 could be either a local
FIG. 8. Comparison of the chemical com- process, caused by exchange of K + in sands for
position of modern deep-sea sands with Na + in interbedded shales, or could involve
ancient greywackes. Sands from trailing- introduction of Na + by migrating pore fluids. A
edge continental margins fall outside the comparison of the chemistry of modern and
compositional range for greywackes, but ancient turbidite sands should show whether
the arc-related samples all have composi- the high Na20 can reasonably be interpreted as
tions typical of greywackes. Fields from
a primary feature or whether diagenetic addi-
Blatt et al. (1972, fig. 9-2).
tions seem to be called for.
8 & 9). The sands are particularly rich in SiO 2 Fig. 9 shows that sands from the four active
and the K20/Na20 ratio is almost always settings are similar in chemical composition to
greater than one. The muds are even more ancient greywacke sandstones, particularly in
enriched in K20 (Fig. 10), but have SiO 2 values the K20/Na20 ratio. The shales, however,
similar to the other muds. Sands and muds from show a pronounced enrichment in K20 com-
forearc basins of island-arcs are also distinctive, pared with their modern counterparts (Table
having extremely low KzO/Na20 ratios and very 4). Does this reflect diagenetic changes? Fig. 11
high Fe203 + MgO (Table 3). The other three shows a comparison between the compositions
settings, however, have sands and muds of very of the sands and their associated muds or
similar composition. Samples from strike-slip shales. This diagram shows that the increase in
basins and continental margin-arc basins are KzO/Na20 in the ancient shales is not matched
chemically identical, and intra-oceanic backarc by a corresponding decrease in the sands. Thus
basins are distinguished only by slightly lower the sands could not have acted as a source of K +
KzO/Na20 and slightly higher Fe203 + MgO during diagenesis. In fact, the close similarity in
(Table 3). Figs 9 and 10 show, however, that composition of the recent and ancient turbidites
 Composition o f modern deep-sea sands 557

12 , , , , , , , ,

Sands

TE
I 0 I

8
o~
.<

o~

'1 I

FA ] ' ~ SS~. ~'nc'ent sands t

4 .8 1.2 1.6 2.0

K20/Na20

FIG. 9. Chemistry of deep-sea sands. Note the extensive overlap of the active tectonic settings. The
coincidence of arc-related sands and ancient greywackes suggests that diagenetic additions are not
necessary to explain the composition of these sandstones. Error bars are one standard deviation each
side of the mean.

TABLE 4. Comparison between modern deep-sea sands and ancient turbidites

SiO 2 AI203 Fe203 MgO CaO K20 Na20 KzO/Na20

1 Sands
Modern* 68.8 14.5 4.4 2.2 4.3 2.2 3.7 0.60
Ancient] 70.4 14.2 5.4 2.2 2.6 2.1 3.1 0.68
Ancients 66.9 14.4 4.6 2.1 7.7 2.0 2.5 0.78

2 Shales
Modern* 67.3 15.3 6.5 3.2 3.1 2.3 2.5 0.92
Ancients 61.3 18.3 6.5 3.1 5.1 4.0 1.4 2.9

* Average of SS, BA and CA samples.

t Data of Pettijohn (1963).
:~ One hundred and twenty-nine samples of sand and shale from ancient turbidites (Appendix).

suggests that no diagenetic additions are neces- sandstones. Note in Fig. 11 that the Lower
sary to explain the chemistry of these sand- Palaeozoic sandstones have a higher K 2 0 / N a 2 0
stones. ratio than younger ones. This trend again
How then can the chemistry of the shales be argues against diagenetic addition of Na20, but
explained? Note the much higher K20 content it also suggests that Lower Palaeozoic clastics
of the older shales. This pattern is not confined have an anomalously large K20 content. There
to this type of shale, but seems to be a world- are many reports (Bowie et al. 1966; O d o m
wide secular trend in shale composition (Weav- 1975; Stablein & Dapples 1977; Swett & Smit
er 1967; Ronov & Migdisov 1971; Van Moort 1972) of highly feldspathic sands of this age
1972). Some preliminary work in our laboratory from environments not normally associated
suggests that this high K20 in Lower Palaeozoic with arkoses (e.g. beaches and tidal flats).
shales is due to detrital K-feldspar, and so is a Interpretations of their origin vary, but the
primary rather than a diagenetic effect. A simi- abundance of this type of occurrence suggests
lar conclusion probably also applies to the older some non-uniform event in the Earth's history.
558 J. B. Maynard, R. Valloni & H.-S. Yu
, , l , i l , l

Muds

o~
<
8

,-,o 6
I
FA[ • I

TEl
O-
s,_~_s ~ BA I
T.-I-: = ~ 1 - L~x I

Ancient® Shales1

A ~ A , A , 1'.6 J i
.4 .8 1.2 2.0
K20/No20

FIG. 10. Chemistry of muds interbedded with deep-sea sands. As with the sands, the arc-related
settings show extensive overlap. Note the strong enrichment of ancient shales in K20.

.!
1.4 1:1

C)TE
1.2

Lower
,~ 1.0 ® Palaeozoic

>
c .8 Mesozoic -
< #SS ® Cenozoic
o
~.6 ABA
o
::x? .4
• FA

.~ .~ 1'.2 ;.6 2'.0~ 4'.4 4'.s s'.2

K2OJNa20 in shales

FIG. 11. Comparison of the compositions of turbidite sands and associated muds or shales shows that
older sands do not become depIeted in KzO relative to the shales. The very high K20/Na20 in
Palaeozoic shales is due to detrital K-feldspar, and so is a depositional rather than a diagenetic effect.

Conclusions
Tectonics was found to exert a strong influence For example, active and passive settings are
on the composition of m o d e r n deep-sea sands, quite distinct in both petrography and bulk
but only broad categories are distinguishable. chemistry. Quartz contents higher than 40%
 Composition of modern deep-sea sands 559

and K20/Na20 greater than 1.0 indicate sands clusion. High K20 in Lower Palaeozoic shales
from passive (trailing-edge) continental mar- o f this t y p e a p p e a r s to be a p r i m a r y r a t h e r t h a n
gins. Subdivisions of active settings are much a d i a g e n e t i c effect.
less distinct, with the forearc of continental
margin arcs (leading-edge margins) having
almost identical sands to those from the backarc ACKNOWLEDGMENTS: We thank our many colleagues
of island arcs. who have discussed these ideas with us, particularly
The bulk chemistry of sands from active Paul Potter, Bill Dickinson, and Chris Suczek.
Maynard wishes to express grateful appreciation to
settings is quite similar to the reported composi-
the staff of Grant Institute of Geology, University of
tion of average greywacke. From this we infer Edinburgh, for help with preparation of the manu-
that it is not necessary to have post-depositional script. Research supported in part by NSF Grant
changes in composition to make greywackes. EAR79--03919. The Lamont-Doherty Core Collec-
Comparison of some ancient turbidite shales tion is maintained through NSF Grants OCE76-
with modern turbidite muds supports this con- 18049 and O N R contract CN14-75-C-0210.

References
BLATT, H., M1DDLETON, G. V. & MURRAY, R. 1972. sandstones and conglomerates between
Origin of Sedimentary Rocks, Prentice-Hall, B e a u m o n t and Bic, Quebec. Can. J. Earth Sci.
New Jersey, 634 pp. 11,951-63.
BowiE, S. H. U., DAWSON, J., GALLAGHER, M. J. & MIDDLETON, G. V. 1972. Albite of secondary origin in
OSTLE, D. 1966. Potassium-rich sediments in the Charny Sandstones, Quebec. J. sediment. Petrol.
Cambrian of northwest Scotland. Trans. lnstn 42, 341-9.
Ming Metall. 78, B125-45; 79, B60-9. MOORE, G. F. 1979. Petrography of subduction zone
CROOK, K. A. W. 1974. Lithogenesis and tectonics: sandstones from Nias Island, Indonesia. J. sedi-
the significance of compositional variation in ment. Petrol. 49, 71-84.
flysch arenites (greywackes). Spec. Publ. Soc. ODOM, I. E., 1975. Feldspar grain-size relations in
econ. Paleontol. Mineral. Tulsa, 19, 304-10. Cambrian arenites, upper Mississippi Valley. J.
DICKINSON, W. R. 1974. Plate tectonics and sedi- sediment. Petrol. 45, 636-51.
mentation. Spec. Publ. Soc. econ. Paleontol. PE]q'IJOHN, F. J. 1963. Chemical composition of sand-
Mineral. Tulsa, 22, 1-27. stones. Prof. Pap. U.S. geol. Surv. 440--S, 21 pp.
DICKINSON, W. R. & SEELY, D. R. 1979. Structure -- POTTER, P. E. & SIEVER, R. 1972. Sand and
and stratigraphy of forearc regions. Bull. Am. Sandstone. Springer-Verlag, New York, 628 pp.
Assoc. Petrol. Geol. 63, 2-31. POTrER, P. E. 1978. Petrology and composition of
- - & SUCZEK, C. A. 1979. Plate tectonics and modern big river sands. J. Geol. 86, 423-49.
sandstone composition. Bull. Am. Assoc. Petrol. READING, H. G. (ed). 1978. Sedimentary Environments
Geol. 63, 2164-72. and Facies. Blackweli Scientific Publications, Ox-
- - & VALLONI, R. 1980. Plate settings and prove- ford, 576 pp.
nance of sands in modern ocean basins. Geology, RoNov, A. B. & MIGDISOV, A. A. 1971. Geochemical
8, 82-6. history of the crystalline basement and the
FIELD, M. E. & P1LKEV, O. H. 1969. Feldspars in sedimentary cover of the Russian and North
Atlantic continental margin sand off the south- American platforms. Sedimentology, 16, 137-87.
eastern U.S. Bull. geol. Soc. Am. 80, 2097-102. SCHWAB, F. L. 1975. Framework mineralogy and
INGERSOLL, R. V. 1978. Petro-facies and petrologic chemical composition of continental margin-type
evolution of the late Cretaceous forearc basin, sandstone. Geology, 3, 487-90.
northern and central California. J. Geol. 86, STABI.EIN,N. K. & DAPPLES,E. C. 1977. Feldspars of
335-52. the Tunnel City Group (Cambrian), western
INMAN, D. L. & NORDSTROM, C. E. 1971. On the Wisconsin. J. sediment. Petrol. 47, 1512-38.
tectonic and morphologic classification of coasts. SWETr, K. & SMIT, D. E. 1972. Paleogeography and
J. Geol. 79, 1-21. depositional environments of the Cambro-
KRYNINE, P. D. 1942. Differential sedimentation and Ordovician shallow marine facies of the North
its products during one complete geosynclinal Atlantic. Bull. geol. Soc. Am. 83, 3223-48.
cycle. Ann. Primero congresso Panamericano de VALLONI, R. & MAYNARD, J. B. 1981. Detrital modes
Ingeniera de Minas y Geologia (Santiago, Chile) of modern deep-sea sands as related to plate
Geologia 1st Pt. 2, 537-61. tectonics: a first approximation. Sedimentology,
LAJOIE, J. 1973. Albite of secondary origin in Charny 28, 75-83.
Sandstones, Quebec: discussion. J. sediment. VAN MOORT, J. C. 1972. The K20, CaO, MgO and
Petrol. 43, 575-6. CO 2 contents of shales and related rocks and
-- HEROUX, Y. & MATHEY, B. 1974. The Precam- their implications for sedimentary evolution
brian shield and Lower Paleozoic shelf: the un- since the Proterozoic. 24th Int. geol. Congr.,
stable provenance of the Lower Paleozoic flysch Sect. 10, 427-39.
560 J. B. Maynard, R. Valloni & H.-S. Yu
WEAVER,C. E. 1967. P o t a s s i u m , illite a n d the o c e a n . compaction and cementation of quartz arenites
Geochim. cosmochim. Acta, 31, 2182-96. and chemical changes in graywackes. D i s s e r t a -
Y u , H . 1979. Three aspects of sandstone diagenesis: tion, P h D , U n i v . C i n c i n n a t i ( u n p u b l . ) .

J. BARRY MAYNARD, D e p a r t m e n t o f G e o l o g y , U n i v e r s i t y o f Cincinnati, Cincinnati,

O h i o 45221, U . S . A .
RENZO VALLONI,I s t i t u t o P e t r o g r a f i a , U n i v e r s i t a di P a r m a , 43100 P a r m a , Italy.
H o - S m N G Y u * , D e p a r t m e n t of G e o l o g y , U n i v e r s i t y o f Cincinnati, Cincinnati,
O h i o 45221, U . S . A .
* Present address: P e n n z o i l I n t e r n a t i o n a l , P . O . B o x 2967, H o u s t o n , T e x a s 77001,
U.S.A.

Appendix
Bulk chemistry of some ancient turbidite sands (sd) and Shales (sh)
# of Inferred
samples SiO 2 AI203 Fe203 MgO CaO K20 Na20 SiO2/Al203 K20/Na20 tectonic
setting

Pico, Repetto Fmns. Sd 3 66.6 15.0 2.62 0.88 8.84 3.14 2.94 4.44 1.07 Active
Pliocene
Ventura Basin, Cal, Sh 4 67.5 15.9 5.47 2.51 3,11 2.92 2.54 4.25 1.15
Marnoso Arenacea Sd 1 60.5 13.9 2.61 2.60 16.5 1.89 2.07 4.35 0.91 Active
Lower Miocene
Muraglione Pass, Sh 3 49.2 16,7 5.27 5.85 18.9 2.88 1.14 2.95 2.53
Toscana, Italy
Macigno Sd 2 66.8 14.5 3.48 2.32 8.20 2,21 2.56 4.61 {).86 Active
Oligocene
Pieve pelago, Italy Sh 4 66.4 15.0 5.33 3.28 4.37 3.11 1.97 4.43 1.58
Loreto Fmn Sd 1 58.4 16.6 4.76 2,43 15.4 0.88 2.19 3.52 0.40 Active
Oligocene
Magallenes Prov.,
Chile
Prealpine Flysch Sd 4 62.4 13.7 3.21 2.92 14.8 1.32 1.67 4.55 0.79 Active
Oiigocene
Val d'Illiez Sh 7 58.7 16.4 5.61 3.34 12.5 2.46 !.09 3.59 2.26
Switzerland
Cretaceous sands of Sd 13 66.8 16.4 6.60 2.72 2.26 1.81) 3.49 4.07 0.52 Active
Tierra del Fuego
Magallenes Prov.,
Chile
Lewes River Group Sd 4 63.9 15.5 5.75 2.67 6.28 2.00 3.79 4.14 0.53 Active
Laberge Grp.
Triassic-Jurassic
Whitehorse Yukon
Escuminac Fmn. Sd 1 68.3 10.5 4.13 2.00 12.3 1.98 0.77 6.50 2.57 Passive
Devonian
Miguasha, Quebec Sh 2 66.7 16.0 6.72 3.00 3.64 3.00 0.94 4.18 3.19
Gala Greywacke Sd 3 67.3 15.0 5.35 2.18 1.67 3.61 4.91 4.49 0.73 Active
Silurian
Innerliethen, Scot. Sh 2 67.7 15.7 5.43 3.23 1.20 3.97 2.65 4.31 1,50
Aberystwyth Grit Sd 3 71.8 15.0 6.70 1.80 2.03 1.35 1.41 4.78 0.96 Active
Silurian
Aberystwyth, Wales Sh 7 66.7 17.4 7.22 2.02 1.25 4.19 1.29 3.83 3.25
Frenchville Fmn Sd 1 57.1 11.9 2.70 1.15 19.3 3.57 4.39 4.82 0.81 Active
Silurian
Stockholm, Maine Sh 3 65.7 12.5 5.80 3.12 8.01 3.07 1.77 5.26 1.73
Martinsburg Fmn Sd 16 83.0 1t.4 3.93 0.63 0.29 0.48 0.30 7.25 1.60 Active
Ordovician
Lehigh Gap, Penn. Sh 17 58.7 25.9 6.26 1.66 0.36 6.51 0.53 2.27 12.3
Martinsburg Fmn Sd 4 73.5 11.3 5.80 1.71 4.11 2.06 1.51 6.50 1.36 Active
Ordovician
 Composition of modern deep-sea sands 561
Williamsport, Md. Sh 4 61.9 21.8 7.53 2.07 0.66 5.22 0.82 2.84 6.36
Utica Fmn. Sd 4 69.1 15.0 4.51 2.15 5.72 1.84 1.70 4.62 1.08 Active
Ordovician
St Antoine de Tilley, Sh 3 63.3 16.8 6.41 4.37 3.31 4.72 1.06 3.77 4.45
Quebec
Normanskill Fmn Sd
Ordovician
Hudson Valley, N.Y. Sh 4 56.9 19.6 8.00 3.66 7.14 4.09 0.71 2.90 5.76 Active
Sillery Grp. Sd 5 69.1 15.2 6.38 2.48 2,84 1.43 2.63 4.55 0.54 Active
Cambrian
St Lawrence Valley, Sh 4 54.6 25.8 9.74 2.44 0.64 5.42 1.37 2.12 3.95
Quebec
Average sand 15 66.9 14.4 4.6 2.1 7.7 2.0 2.5 4.65 0.78
Average shale 13 61.3 18.3 6.5 3.1 5.1 4.0 1.4 3.35 2.86
Pettijohn's average
greywacke
(volatile-free basis) 61 70.4 14.2 5.4 2.2 2.6 2.1 3.1 4.96 0.68
 Initiation of subduction zones: implications for arc evolution
and ophiolite development

D. E. Karig

SUMMARY: The initial location of trenches with respect to continental margins is a

critical factor in the nature of the forearc region and its subsequent evolution, as well as
for the interpretation of ophiolites. Although recent studies of ancient margins and highly
evolved arcs suggest that wide strips of oceanic crust are commonly trapped between the
trench and volcanic arc, review of young active arcs does not support this conclusion.
Neogene trenches have formed predominantly very close to the crustal interface between
oceanic crust and continental or older island arc crust. A few arcs have developed along
transform zones, but none can be identified as having formed by breakage within an
oceanic plate. Apparently the high strength of normal oceanic lithosphere causes the
initial fracture to utilize pre-existing zones of weakness or pre-stressing, such as the
downbowed and fractured crustal interfaces along passive margins. There seems to be no
physical basis for the initial rupture to form smooth, large radius arcs, thereby trapping
oceanic crust in re-entrants. Moreover, oceanic crust or ophiolites originating adjacent to
a continental or arc margin ought to betray that heritage by a high terrigenous or
volcaniclastic component in their overlying sediment. Large ophiolite sheets that can be
interpreted as forearc slabs may develop in areas with relatively unusual plate geometries,
as where oceanic spreading zones cut diagonally across the forearc. Disrupted ophiolites
could, in many cases, represent slices of oceanic crust and upper mantle accreted to and
responsible for the growth of oceanic arcs. Evidence for accreted slices in the Mariana arc
and the range of forearc geometries among oceanic arcs allow construction of an
evolutionary sequence in which an initial steep and narrow trench shape is differentiated
into upper and lower sections by intermittent accretion of oceanic crustal slices. As a
result, the slope apron on the upper section rotates arcward and can rise to form a basin as
is displayed in the southern Middle America arc.

Evolution of the Earth's crust beneath trench- accretion proceeds but could even migrate land-
arc systems remains one of the more conten- ward if there were compressional deformation
tious topics in the geological sciences. The within the forearc basin (Dickinson & Seely
problem bears strongly not only on the original 1979; Seely 1979). These authors imply that
geometry of arc systems and their subsequent broad initial forearc regions, often related to
evolution, but also on the d e v e l o p m e n t of e m b a y m e n t s in the continental margin, are
ophiolites. Models of convergent plate margins typical or at least very c o m m o n .
commonly assume that much of the arc, in Studies of the morphotectonic geometries of
particular the forearc, is underlain by oceanic contemporary arc systems have led to some-
crust (Dewey & Bird 1970; Mitchell & Reading what different conclusions concerning the in-
1971). A major reason for this assumption is the itiation and evolution of arc systems (Karig
widespread belief that the initial crustal fracture 1974, 1977; Karig et al. 1980). The systematic
leading to trench formation occurs within increase in the width of the accretionary prism
oceanic crust. Support for such a model is also or inner trench slope with increasing maturity
derived from the occurrence of ophiolites that of the arc system suggests an arc evolution
cap structurally imbricated stacks of accreted involving a significant upgrowth and outgrowth
material (Dewey 1976) and sometimes form the of the trench slope and slope break from an
seaward flanks of ancient forearc basins (Dick- initial position generally along interfaces be-
inson & Seely 1979). tween oceanic crust and either continental or
A recent corollary to this m o d e l identified older arc crust. This had led me (Karig 1977) to
the trench slope break as the seaward edge of consider trapping of oceanic crust beneath fore-
the trapped oceanic crustal slab, which thus arcs as a relatively unusual situation.
serves as a mechanical buttress against which an In this paper I approach the problem of
accretionary prism is built (e.g. Hamilton initiation and early evolution of arcs by examin-
1977). As a result of this process, the trench ing examples of young arc systems (less than
slope break would not migrate seaward as 10 Myr old), where the initial characteristics

563
564 D.E. Karig

can still be observed. I also use data from older Mammerickx 1972), possibly by the migration
active arc systems in which the early evolution of the pole of relative plate motion (e.g. Dewey
is relatively well constrained by geological and 1975). As a modification to this scheme, Falvey
geophysical data. (1978) suggests that the change in boundary
type occurs at a triple junction, from which a
E x a m p l e s of arc initiation spreading ridge heads northward into the Fiji
Basin.
Most identifiable young arc systems occur The inner trench slope of the entire New
along pre-existing interfaces between oceanic Hebrides arc is steep and narrow in comparison
and continental crust, or between oceanic and to those of older arc systems (Karig & Sharman
island arc crust, at least in a regional sense. 1975). Although there are local variations in
Commonly these interfaces represent rifted, slope morphology (Ravenne et al. 1977; Daniel
passive margins (e.g. New Hebrides). Although 1978), there is no clear tectonic division into
only back arc rift margins have been utilized lower and upper slope sections as recognized
during the past 10 Myr, passive continental in the larger Tonga of Mariana arcs. Local
margins bordering main ocean basins must also sedimentary aprons on the upper slope sur-
be assumed as likely sites. Often crustal inter- round the larger active volcanoes but even
faces represent rejuvenated convergent margins these are lacking in the Hunter-Matthew sec-
(e.g. northern Middle America). Less common- tion of the arc, where the inner slope is narrow-
ly, arc initiation appears to have utilized pre- er still, and displays irregular, unsedimented
existing transform boundaries (e.g. McQuarie). benches (Karig & Mammerickx 1972).
With the possible exception of the Mussau Recent refraction and detailed seismological
Trough (which might also have been a trans- studies in the central section of the arc (Ibrahim
form) I cannot identify any young arc that has et al. 1980; Isacks et al. 1981) indicate that the
originated within an oceanic lithospheric plate. arc is built of crust some 30 km thick, having a
Plate convergence within marginal basins has velocity structure intermediate between oceanic
often been assumed to differ from that in and continental in character, and similar to that
oceanic basins, primarily to explain the em- of other oceanic arcs (e.g. Murauchi et al. 1968;
placement of some ophiolites (e.g. Dewey Shor et al. 1971). This type of crust appears to
1976). Such assumptions are not supported by underlie the upper part of the trench slope
observations in active arc systems. Crustal con- (Fig. 2), whereas low-velocity material, sugges-
vergence occurs by subduction, initiated along tive of accreted material, underlies the lower-
a basin flank (e.g. Shikoku, Solomons, New most slope. In between lies an area of crust,
Hebrides), even when the basin crust is quite 20-30 km wide, where the velocity structure is
young. ambiguous or unconstrained. Beneath this area
These regional observations, however, are could lie marginal basin crust trapped behind
not adequate to show what might occur on a the trench, thin arc crust representing the nor-
more local scale; one which is appropriate to mal-faulted, transitional zone between arc and
the understanding of ophiolites and subduction marginal basin, or accreted slices of oceanic
complexes. Description of several better known crustal rocks.
examples of initial or early arc geometries (Fig. The sum of acquired geophysical data defines
1) serves that purpose and as a basis for discus- a narrow, steep trench slope constructed pri-
sion of the possible constraints on arc initiation. marily of igneous material in an unknown
structural geometry. This lack of a wide forearc
New Hebrides arc basin or terrace, is consistent with the very
narrow flattened section in the shallow section
The present New Hebrides trench has de- of the seismic zone (e.g. Isacks & Barazangi
veloped along the rear (western and southern) 1977). Ophiolitic slices, or even narrow
flank of an older arc, apparently as a result of ophiolite sheets, could lie beneath part of the
an arc polarity reversal that occurred about trench slope, but any such ophiolite sheet ought
6 Myr BP (Faivey 1978; Carney & MacFarlane to carry a distinctive cover of volcaniclastic
1978). Regional tectonic reconstruction (Falvey strata, as it would have developed immediately
1978) indicates that the trench developed along west of the pre-reversal, Miocene volcanic arc.
the rifted flank of an early Tertiary back arc
basin. The section of the arc south of 21°S, and Philippine archipelago
swinging eastward into the Hunter-Matthew
fracture zone appears more primitive and may The Philippine archipelago has been an area
have developed from that transform (Karig & with a very complicated pattern of rapidly
 Initiation o f s u b d u c t i o n z o n e s 565
120 ° 180 ° 120 ° 60 °

40 ° _ 40 °

20 ° - 20 °

0° _ O °

20 ° 20 °

40 ° - 40 °

60° t I 60 °
I I I I 1 I I 1 .... I I I I I I 1 I ] I I I I
120 ° 180 ° 120 ° 60 °

Fl6. 1. Active convergent margins of the Indo-Pacific and Atlantic regions, noting arc systems either
initiated or re-activated within the past 10 Myr. Systems used as examples in the text are indicated
and identified by letter: (A) Vanuatu (New Hebrides); (B) Luzon; (C) Sunda; (D) Middle America.

changing plate consumption during the tends from the crest of the Sierra Madre Range
Neogene, including formation of several new to a very small prism of accreted sediments at
convergent margins (Fig. 1). Most is known the slope base (Karig 1977). Because this slope
about the margins bounding Luzon, where re- marks the eastern boundary of a Cretaceous-
peated arc polarity reversals are suspected. Palaeogene subduction complex (Murphy 1973;
Subduction has been occurring along the west- Karig 1973), the nascent trench would appear
ern margin (Manila Trench; Fig. 3) since some to have rejuvenated an older convergent mar-
time in the Miocene, but a new trench is now gin. If the descending oceanic lithosphere con-
developing along the eastern coast (Karig 1977; tinues beneath this slope with the geometry
Bowin et al. 1978). This eastern zone appears to implied by its inclination at the trench, there is
be the northward propagating extension of the no room for a slice of trapped oceanic crust.
Philippine Trench (Fig. 3), all of which is quite Nor does the non-magnetic character of the
young and which is extending southward as well slope suggest that any such oceanic crust
(Karig 1975; Hamilton 1979). Profiles across exists.
the Philippine Trench at widely scattered loca- Subduction has persisted along the western
tions show a steep inner trench slope, without a margin of Luzon since early or mid-Miocene,
significant accretionary prism (Karig 1975; producing a well-developed accretionary prism
Hamilton 1979). This trench seems to have of moderate size (e.g. Ludwig et al. 1967; Fig.
developed along or near the pre-existing margin 3). A discussion of this margin bears on the
of archipelagic crust, although in most areas problem of arc initiation and the origin of
the data are not capable of determining ophiolites because that section of the arc be-
whether or not a narrow strip of oceanic crust tween 15°N and 16°N exposes the Zambales
might be trapped west of the site of trench ophiolite, in a position east of the present
initiation. forearc basin (Fig. 3). This large ophiolite is
The nascent trench east of Luzon has clearly one of the youngest and, at least in its upper
developed at the base of the slope between 16 structural levels, one of the least deformed
and 17°N, where the most extensive surveying examples known. As important, it retains a
has been undertaken (Karig, unpublished re- recognizable tectonic position within an active
suits of SI0 Antipode I V cruise). In that region arc.
an extremely steep slope (averaging > 10°) ex- Tectonic studies of this body are only begin-
566 D. E. Karig

Io ~ - ~ ~ : . ;
Accreted Materiel

<-~ru-~_.~ . "~ <:<:~.':'::,:.i(':f):,>tA'rc ,>:Crusi'}')'k:<: ~~:<'.i<g:'

,, i:::,:..,:....... 0

Io
Km

401 0 iO 20 30 40 50Kin\ ~x
50 t I ~ i t I \ 50
no vertical exaggeration ~,
6O
\
6O
\
FnG. 2. Crustal and tectonic section of the central New Hebrides arc: an example of a relatively young
(6 Myr) arc system. This section is constrained by data from Isacks & Barazangi (1977); lbrahim et
al. (1980); Luyendyk et al. (1974); Ravenne et al. ( 1977); and unpublished seismological data of B. L.
lsacks, This section illustrates the relatively narrow and steep forearc area, and the limited area
beneath which oceanic crust might be trapped.

(Miocene Volcanic Arc )

Cordillera
Manila Trench Luzon Trough Central 5

-~, I-z° Km
I-3o
.~ L4o

(Miocene
(Old Foreorc Basin) Volcanic Arc)
(New Accretionary Prism ) Zombales Luzon Central Basin Cordillera
Manila Trench Luzon Trough Ophiolite t ~ Central I- 5
o ~ ~ -~-=~=---_- . . . . . ~ ,'o
o

• B " I- 20
1200 122° 124° ' \~'~ L
vm,,.~._, ~ I ' " ' "~ ~ 0 20 40 60 80F
A',[-~ ~ ) , ~ . . K E Y "' ~ I t I i I , i , I ~-40
18" : "" : '\ Km Km
~t,~" ~ ,"-'~'. ~ ~I Zamboles Ophiolite '~__ VE=I -
,-.'~/" t ~ ~'JA '~ I~'~'~o ~'~ " "
,,t[~-,- ) :-'~'~" .~'~/~# A ActiveVolcanic Arc ~ -60
•,a pl. . .," ~ 't
[I| P ~ . ~ ~" Z'_', Miocene Volcanic Arc '~
16"
- ~ ~ ~ t k-~ Deformed Sediment ~- 80

'f' z"~Z :"~' A l~ Metomorphics ~1~ I00

Intrusives

FIG. 3. Tectonic sketch map and two schematic tectonic sections across western Luzon, where the
Zambales ophiolite slab occurs. The southern section crosses the arc through the Zambales ophiolite
and the Luzon Central Basin (an old forearc basin) and shows its relationship to the present forearc
basin (Luzon Trough), The northern section lacks the older forearc and ophiolite but exposes a
deformed terrane including disrupted ophiolitic fragments. Also shown on the map is the northward
propagating end of the Philippine Trench, along the east coast of Luzon. Data are from Hayes &
Ludwig (1967); Ludwig (1970); Schweller & Karig (1979); Cardwell et al. (1980) and various
unpublished industrial and Philippine governmental sources.
 Initiation of subduction zones 567

ning, but preliminary mapping and analysis of S u n d a arc

available data define the ophiolite as a gently
east-dipping sheet that forms the western flank Although the present subduction geometry
of the Luzon Central Basin (Irving 1950; along the western margin of Sumatra appears to
Schweller & Karig 1979; de Boer et al. 1980). date from the late Oligocene (Karig et al. 1980),
The strata filling that basin and covering the the large amount of subsurface and onland data
ophiolite begin with tuffs and tuffaceous pelagic from the Nias transect (Kiechefer et al. 1980:
limestone of late Eocene to Oligocene age Moore & Karig 1980) lead to a number of
(Hashimoto et al. 1977; Garrison et al. 1979), conclusions pertinent to the initiation and
changing upward to lithic and volcaniclastic evolution of that subduction pulse. The Oli-
turbidites and related sediments. A shallowing gocene trench seems to have developed along
of the depth of deposition through time, from an irregular continental margin, disrupted by
several kilometres for the basal strata to sea- strike-slip faults (Karig et al. 1980). Basins,
level during the Pliocene, is reasonably certain trapped within the re-entrants, can be identified
(Ingle 1975; Garrison et al. 1979). One surpris- on multichannel reflection profiles, but most of
ing observation is the relative lack of deforma- these appear to overlie the continental slope.
tion in the upper levels of the ophiolite and in Refraction data from the area east of Nias
its sediment cover that might be associated with (Kieckhefer et al. 1980) indicate that that slope
trench initiation. At most we note a few widely is underlain by continental crust or more likely,
spaced folds and a submarine unconform- an older subduction complex (Karig et al.
ity across which there may be differential 1980).
tilting. Extensive palaeodepth and structural control
The Zambales ophiolite is most easily inter- from offshore wells and from the island of Nias
preted as a crustal sheet, formed in an oceanic demonstrate that the trench slope evolved from
(or back-arc) basin west of Palaeogene Luzon, a small, deep terrace stage in the early Miocene
and trapped between the initial position of the to one with a broad upper slope, near sea-level,
Manila Trench and a volcanic arc (Cordillera by the end of the mid-Miocene (Karig el al.
Central) sometimes early in the Miocene. This 1980). One observation pertinent to the evolu-
tectonic setting, together with subsurface stra- tion of arcs is that Miocene slope strata, depo-
tigraphic and magnetic data from unpublished sited on the upper trench slope with a signifi-
oil company reports would indicate that the cant initial dip, now lie on the trenchward flank
ophiolite and the Luzon Central Valley formed of the forearc basin, and have been rotated
an initially broad forearc basin, and evolved arcward into a nearly horizontal orientation.
much in the manner suggested by Dickinson & In the section of the Sunda arc including Java
Seely (1979). and Bali, refraction studies have been inter-
The Zambales region would appear to con- preted as favouring trapped oceanic crust be-
tradict the major premise of this study, but neath the forearc basin (Curray et al. 1977), but
several observations suggest that an unusual this interpretation is vulnerable on several
plate geometry was responsible for its creation. grounds. The thickness and velocity of seismic
First, the ophiolite neither occupied an embay- units beneath the forearc basin are quite vari-
ment in the mid-Tertiary margin of Luzon nor able and are significantly different from those
continues along strike to the north or south. To of typical ocean crust. They bear at least as
the north, the magmatic (frontal) arc is flanked much resemblance to arc crust (Shor el al. 1971)
instead by an assemblage of highly disrupted or to subduction complexes (Eaton el al. 1970).
ophiolitic and sedimentary fragments (e.g. Fan- In particular, the 4.5 to 5 km s -l material can
ning 1912). Secondly, magnetic anomalies be as easily or better interpreted as dewatered
obtained during oil company surveys over the on slightly metamorphosed sediments as basal-
eastern flank of the ophiolite and the Central tic second layer (e.g. Kieckhefer el al. 1980).
Luzon Valley, as well as preliminary studies of A Cretaceous-Palaeogene subduction com-
dykes near the top of the ophiolite ( e . g . J . F . plex, comprising m61ange and slices of oceanic
Violette, written comm. 1980) indicate that the crustal lithologies, is exposed in several areas
spreading ridge which generated the ophiolite near the south coast of Java (Ketner et al. 1976;
trended north-easterly. If, as suspected, the Hamilton 1979). Subsurface data from the area
ophiolite is autochthonous with respect to cen- offshore reveal a regional unconformity, cut at
tral Luzon, the rift zone along which the sea-level during the late Oligocene to early
ophiolite was generated somehow had to be Miocene (Bolliger 8: de Ruiter 1976).
resolved against another type of plate boundary Together, these data suggest that at least the
in that vicinity. inner (northern) side of the forearc basin is
568 D. E. Karig

underlaid by a subduction complex that was from other studies suggest that evolution of this
emergent during the initiation of the Neogene arc is open to alternative interpretations.
subduction event, rather than by oceanic crust The basement beneath the outer half of the
trapped behind the new trench. In short, the forearc basin emerges on the Nicoya, Osa and
data from the eastern Sunda arc do not support Burica peninsulas of Costa Rica and Panama
the presence of an oceanic crustal slab beneath where it has been mapped as consisting of slices
that forearc basin. of oceanic basalt with zones of disruption which
contain m61anges of oceanic crustal rocks and
sediments (e.g. Dengo 1962; Henninsen &
M i d d l e A m e r i c a arc
Weyl 1968; Galli-Olivier 1979). The term
Finite plate reconstructions imply that con- ophiolite has been applied to this complex but
sumption of oceanic crust has been a general the apparent degree of disruption seriously
feature of the west coast of Mexico and Ce.ntral detracts from interpretation of this basement as
America since well back into the Mesozoic. a trapped slab of oceanic crust.
Nevertheless, the response of this margin has Uplifted basement masses are delineated be-
varied greatly, both spatially and temporally, neath and seaward of the shelf edge by many
affording several glimpses of the initiation and seismic reflection and magnetic profiles, and
evolution of trench slopes. although interpreted by Seely (1979) as diapirs
Initiation of renewal of subduction during the rising from the oceanic slab, might as well be
late (?) Miocene along the Middle America Arc viewed as thrust slices (Ladd et al. 1978). This
north of the Tehuantepec Ridge followed re- interpretation would be more consistent with
moval of a pre-existing continental margin by structural nature of the exposed terrane.
an as yet poorly defined process (Karig et al. In the Nicoya complex, highly deformed
1978a). The narrow steep trench slope is clearly cherts and limestones of Jurassic to Late Cre-
divided into upper and lower slope sections. taceous age are structurally mixed with the
Single channel seismic reflection and magnetic oceanic igneous rocks (Galli-Olivier 1979;
profiles (Karig et al. 1979a) indicated approx- Schmidt-Effing 1979) and probably represent
imately 25-30 km of outgrowth by accretion of sediments of the ocean basin. These highly
sediments directly against the old continental deformed rocks are overlain by deep water
basement. Deep sea drilling (Moore, Watkins siliceous limestones and shales of late Cre-
et al. 1979, 1981) and associated multichannel taceous and Palaeocene age (Galli-Olivier
reflection profiles (Shipley et al. 1980, 1981) 1979: Harrison 1953), which appear from their
confirmed this geometry and more clearly out- position, degree of deformation, and
lined the existence of an upper slope apron or sedimentological character to be trench slope
incipient forearc basin over the contact between sediments. This unit is dominated by hemipela-
accreted sediments and continental crust. Slope gic sediments near the base of the section in
parallel strata in this slope apron provide seaward areas but form a transgressive elastic
another example of initially inclined basal units sequence futher east and upsection (Galli-
in the forearc setting. A well-defined trench Olivier 1979). This latter facies is several
slope break lies 10-15 km seaward of the edge kilometres thick and has reasonably been inter-
of the continental crust, well within the accre- preted as upper slope or forearc basin strata
tionary prism. In this rejuvenated arc there is (Galli-Olivier 1979). In places a conglomerate
again no trapped oceanic slab. Although accre- basal sequence is observed to unconformably
tion began against a buttress of continental overlie basalts of the subduction complex
basement, the trench slope break has already (Galli-Olivier 1979). The geological rela-
begun to migrate outboard and is rising. tionships in the Nicoya and related complexes
The character of the arc south of the Tehuan- indicate that subduction was operative by Con-
tepec Ridge contrasts strongly with the juvenile iacian time.
aspect to the north, but merits discussion be- Combination of palaeodepth information
cause it has served as the model of forearc from the borehole on the shelf edge off Guata-
development by Seely et al. (1974) and Seely mala (Seely 1979) with the sedimentary data
(1979). This arc has been cited as a well- from the emergent subduction complex along
constrained example where a wide slab of tectonic strike to the SE suggests that,
oceanic crust was initially trapped behind a rather than oceanic sediments, basal strata of
trench as the basement for a large forearc basin the forearc section are Coniacian slope apron
(Seely 1979). Although many of the well and sediments, rotated arcward from an original
seismic data used in that interpretation have not seaward dip. They overlie disrupted oceanic
been made public, the available information crustal slices and m61ange that were very likely
 Initiation of subduction zones 569

accreted to the trench slope. Deposition at edge of oceanic crust abuts densified and struc-
abyssal depths and subsequent shallowing turally thinned continental (transitional) crust,
would characterize lower slope sediments which thickens and acquires typical continental
equally as well or better than they would characteristics over a width of 50-100 km
oceanic sediments. Overlying and to the east of (Grow et al. 1979).
these strata are turbidites which could reflect During the subsequent evolution of passive
the increase in depositional area on a widening margins, the sediment load of the continental
and flattening upper slope or forearc basin. The rise prism commonly results in the downbowing
distribution of Late Cretaceous sediments sug- of the landward flank of the oceanic crust and
gests that the trench was initially in the vicinity lithosphere (Grow et al. 1979; Watts & Steckler
of the shelf-edge high and that the forearc 1979; Lehner & de Ruiter 1977). This broad
region was steep and narrow, similar to that of flexure involves dips of only 1° or less but it may
young oceanic arcs. With accretion of oceanic well be responsible for major faulting and even
crustal slices, together with oceanic sediment, decoupling of the crustal interface during the
the trench migrated westward, the trench slope early stages of the margin evolution (Watts &
flattened, and as the trench slope break rose, a Steckler 1979; Turcotte et al. 1977).
basin developed. The basal strata overlying the initially formed
oceanic crust along passive continental margins
show a significant terrigenous component for at
least several tens of kilometre seaward of the
Sites of t r e n c h initiation crustal interface. This terrigenous input may be
Physical contraints and definitive characteristics expressed as very rapidly deposited, landward
thickening clastic wedges (Tucholke & Moun-
As important as the documentation of the tain 1979; Klitgord & Grow 1980), at times
initial geometry of arc systems is the identifica- mixed with basalt (Moore 1973), or by more
tion of the characteristics of potential sites of subtle dispersion of the terrigenous component
subduction that constrain the position of the in hemipelagic deposits (Whitmarsh, Weser,
initial rupture zone. It is likewise necessary to Ross et al. 1974; Thiede 1974). There is adequ-
know what characteristics of the pre-existing ate evidence that the basal strata are not typi-
crust and sediment cover might be used to cally oceanic in character.
identify and differentiate loci of trench initia- A second kind of passive margin forms as a
tion. These characteristics are relevant not only result of back-arc spreading. Regardless as to
to the structure of arc systems but also to the whether or not this process of crustal formation
origin of ophiolite sheets. is identical with spreading within main ocean
The difficulty in spontaneously initiating sub- basins, and produces identical crust, there are
duction within oceanic plates, because of the significant differences in the sediment cover of
high stress levels that would be required to fail the two settings. Because back-arc spreading
the lithosphere, have been discussed by begins near the volcanic chain, the first sedi-
McKenzie (1977) and by England & Wortel ments deposited on initial basin crust are rapid-
(1980). Thus any pre-existing zone of weakness, ly deposited volcaniclastics (Karig & Moore
or of suitable pre-stressing in the lithosphere 1975; DSDP Staff, Leg 60, 1978). As the basin
would provide a preferential setting for subduc- widens, the arcward flank receives a large
tion. Review of young arcs indicates that these apron of volcaniclastic debris, which appears to
settings include both interfaces in crustal type depress and possibly fault the area of initial
and pre-existing lithospheric plate boundaries. basin crust (Karig 1972; Mrozowski & Hayes
Of the former, I have identified the passive 1979) in a manner similar to that of the loaded
margins, created by spreading in either main passive margins. The rear or continental flank
ocean or back-arc basins and previously active of these basins is rapidly separated from the
convergent margins. Of pre-existing plate volcanic arc and either collects a small apron of
boundaries, transforms appear to be utilized debris from the adjacent remnant arc or be-
most commonly. comes an apparently typical passive contin-
ental margin, depending on the configuration
of regional tectonic elements.
Passive margins
The point to be made is that the area of initial
Extensive geophysical studies of the Atlantic crust in marginal basins, like that of passive
and Red Sea margins (e.g. Grow et al. 1979; margins, tends to become down flexed and pos-
Lehner & de Ruiter 1977; Lowell & Genik sibly faulted during the evolution of those
1972; Fairhead 1973) indicate that the landward basins. Marginal basin flanks are also the sites
570 D.E. Karig

of quite distinctive sedimentation, which should et al. 1981). Larger offsets would probably be
easily differentiate that setting not only from preserved as trench-trench transforms, an ex-
normal ocean basins but from other parts of ample of which occurs near 16°N along the
marginal basins (Karig & Moore 1975). These Philippine Trench.
differences ought to be clearly expressed in the Secondly, it is not at all clear that most rifted
basal strata of ophiolite and in forearc basins margins are irregular on the scale inferred by
where marginal basins are involved. the 'trapping' model. For example, if, as is
The structural geometry of both types of generally assumed, the crustal interface of the
passive margins suggests that the interface be- North American passive margin is marked by
tween oceanic and transitional crust is the most the East Coast Magnetic Anomaly (e.g. Klit-
suitable zone for the initiation of subduction. gord & Behrendt 1979), then that interface
Not only is there a good possibility of litho- would have curvatures similar to those of some
spheric weakness along that interface, but the arcs.
basin crust is already downflexed into a suitable In short, it is not evident why large oceanic
geometric configuration. This argument is en- slabs would be trapped by initiation of subduc-
tirely qualitative, but accords with the common- tion along passive margins. Small sheets are
ly observed initiation of young arcs near that more logical products but in any case, trapped
interface. oceanic crust in this setting ought to betray its
This model is easily visualized in a two- origin in the terrigenous or volcaniclastic char-
dimensional setting, but it has become com- acter of the capping sediments and possibly in
monly accepted that most passive margins have anomalous crustal structure.
developed along highly irregular lines. This in
turn has led to an intuitive model which Inactive subduction margins
assumes that the initial rupture surface of a
subduction zone cannot follow the intricacies of Subduction can also be reinitiated along pre-
such a margin, but instead cuts across oceanic viously convergent, but deactivated continental
re-entrants in order to develop the smooth or arc margins. The location of renewed sub-
curvature that characterizes arc systems (Karig duction appears to be close to the seaward
1977; Dickinson & Seely 1979; Seely 1979). margin of the older subduction complex, prob-
These re-entrants would then be loci of forearc ably controlled by the geometry and mechanical
basins built on oceanic crust, and following a conditions during and following cessation of the
collision event, could become large ophiolites. previous subduction episode.
A more careful analysis suggests that this model After subduction ceases, the geometry of the
is incorrect and without a firm physical basis. convergent margin does not appear to change
First, there is no reason why an initial break significantly, because the shallow morphotecto-
could not closely follow an irregular crustal nic framework of that margin is primarily con-
interface. Motivation for the idea of a smooth trolled by the elastic loading of the oceanic
break was that, because mechanical flexures of lithosphere (Karig et al. 1976). If both plates
oceanic lithosphere have half-wavelengths on were completely elastic, cessation of conver-
the order of 100 km, the initial geometry of the gence should affect the geometry only by the
descending plate would not be able to reflect removal of deeper level stresses through ther-
small or sharp irregularities along the pre- mal resorption of the descending slab. In mar-
existing passive margin. However, the oceanic gins with well-developed accretionary prisms,
lithosphere is more correctly visualized as being this effect would be uplift, primarily in the
initially slipped beneath the upper plate at a upper sections of the trench slope. Continued
very low angle rather than being bent down deposition in the trench would decrease the
along that boundary. Slippage could readily mass imbalance, but would cause additional
occur along a tooth-like pattern of subduction crustal depression of the region. Departure
and transform segments, analogous to the pat- from perfectly elastic conditions results in elas-
tern of floating ice-fingers (see Weeks & tic-plastic bending at a hinge area that generally
Anderson 1958) used by Wilson (1966) to illus- forms near the trench (McAdoo et al. 1978), and
trate plate convergence. The flexure of oceanic should lead to local uplift seaward of that hinge.
lithosphere at the outer swell would simply To the extent that these conclusions concern-
reflect the average shape of the initial bound- ing preservation of subduction zone geometries
ary. Small irregularities in the subduction zone are correct, renewal of subduction would be
pattern would quickly be smoothed and masked expected to utilize the previous slip surface.
by preferential sedimentation and accretion in This has apparently occurred along eastern
indentations along the inner trench slope (Karig Luzon, and possibly along the Sunda and east-
 Initiation of subduction zones 571

ern Aleutian arcs (Karig et al. 1976). Where this demonstrate that the modal configuration of
mechanism is operative, no oceanic slab should arc initiation includes a steep inner trench
be stranded beneath the new forearc region. slope, resulting in a narrow forearc region and
Tectonic disruption of a quiescent convergent the lack of a forearc basin. Entrapment of strips
boundary might also occur, leading to a number of oceanic crust beneath the inner trench slope,
of possible configurations in which large ex- especially those wider than several tens of
panses of oceanic crust could become stranded kilometres, appears from that perspective to be
beneath the area later to be converted to a a relatively unusual event. A positive correla-
forearc region. One such pattern is a spreading tion in width and height of various accretionary
ridge that rifts the upper plate and older or prisms with the width of the upper flattened
pre-existing subduction complex. An excellent section of the seismic zone continues to provide
example is offered by the Gulf of California. A strong evidence for the upward and outward
slight migration of the Pacific-North American growth of the forearc region. However, it is
pole could lead to renewal of subduction along clear that the forearc does not merely expand with
Baja California, leaving the mouth of the Gulf a fixed geometry.
of California trapped as a very large oceanic The well-controlled history of the Sunda arc
sheet in the forearc area. illustrates a strong upward growth of the accre-
Strike-slip disruption of a quiescent subduc- tionary prism and outward migration of the
tion margin could also disrupt the simple con- trench but with changes in tectonic style since
vergent plate geometry. Whether such strike- the early Miocene (Karig et al. 1980). The lesser
slip faults are confined to the upper plate or not migration of the trench slope break requires
and how they affect the subduction interface is that the lower trench slope broadens and poss-
a problem that has not yet been addressed. ibly flattens as the arc evolves. This broadening
of the lower slope can be observed in the
evolutionary spectrum of arc systems, from a
Transform boundaries
20 km width in the northern Middle America
Although it is almost certain that some arc (Karig et al. 1978b; Shipley et al. 1981) to a
oceanic arc systems were initiated along trans- maximum observed width of 250 km in the
forms or fracture zones, it is less obvious that Makran (Farhoudi & Karig 1977).
this process traps a 60-100 km wide belt of In arcs dominated by accretion of sediments,
typical oceanic crust between the trench and lateral growth of the accretionary prism with
volcanic arc (e.g. Dickinson & Seely 1979). age certainly involves onlap and transgression
Long, active transform faults are more promis- of the old continental or arc slope, as the
ing loci of trench initiation than are fracture tectonic dam provided by the trench slope
zones because they are through-going lithos- break rises (e.g. Dickinson & Seely 1979).
pheric ruptures, along which pole migration However, use of the old slope as a reference
could produce suitable relative motions. (Karig 1977), as well as more detailed studies of
However, neither the longer transforms nor individual arcs, shows that there is also an
major fracture zones are simple fault bound- outward migration of the trench slope break.
aries, but rather consist of a structurally com- The trench slope break initially forms near the
plex zone that can exceed 50 km in width (e.g. base or seaward edge of the old slope (and
Bonatti et al. 1979). Within such zones, slices of probably marks that boundary) but it is much
oceanic crust are apparently interfaulted oversimplified to extrapolate that assumption
against comminuted mixtures of crustal and and to identify the trench slope break as a
mantle and oceanic sedimentary rocks, some of buttress in wider, more evolved arcs involving
which may qualify as m61anges (Choukroune et sediment accretion (e.g. Hamilton 1977).
al. 1978; Saleeby 1977). If subduction were Evolution of oceanic arc systems, where the
initiated along a long transform, the resultant sediment influx is low and oceanic crust is
forearc area might well be floored with oceanic principally involved is more difficult to deter-
material but most likely with very atypical mine, largely because internal structures cannot
structural characteristics. be observed on reflection profiles. The increase
in size of these forearc regions with increased
duration of subduction points toward some
C o n c l u s i o n s and i m p l i c a t i o n s form of cumulative accretion. Contrary to the
Implications for the structure and evolution of characteristics expected of such a model, recent
arc systems deep sea drilling results in this type of arc have
generally been interpreted as supporting subsi-
Profiles across very young arc systems dence and even tectonic erosion of the forearc
572 D. E. K a r i g

area (DSDP Staff, Leg 60 1978). These conclu- most data from the range of active oceanic arcs
sions, particularly in the Mariana arc, are open (Fig. 4). In this model, accretion of oceanic
to serious question. Counter-evidence for rela- slices to the base of a steep initial trench slope
tive uplift of the trench slope break in that arc would result in broadening of that slope and
can be cited from the nature of dredged and uplift of at least its trenchward section. This in
drilled sediments (Karig 1971; Fischer, Heezen turn would enhance the development of a slope
et al. 1971; DSDP Staff, Leg 60 1978), from apron by increasing the width over which de-
deformation along a surveyed submarine can- position could occur and by decreasing the
yon (Karig 1971), and from dredging of ser- gradient of the upper slope. The assumption
pentinized ultramafics from structural highs that the Tonga and Mariana arcs are more
along the trench slope break (Evans & Hawkins evolved examples in this evolution implies that
1979). South of the drilling transect, the trench a trench slope break develops, dividing the
slope appears to consist of long basement inner slope into upper and lower sections. The
ridges, convex toward the trench and uplifted upper slope section could continue to rotate
relative to sediments trapped behind (Karig & arcward until it develops into a forearc basin, as
Ranken, in preparation). These ridges are most displayed along the southern Middle America
easily interpreted as intermittently accreted arc. Basal strata in these forearc basins are
slices of oceanic crust and upper mantle, similar interpreted as arcward-rotated, originally
to those outlined in the Peru-Chile (Kulm et al. trenchward-dipping slope deposits, overlying
1973) and southern Middle America (Ladd et either pre-existing crust, accreted material, or
al. 1978) arcs. possibly even early arc volcanics.
Combining morphology and apparent struc- Emplacement of large, coherent ophiolite
ture of very young and immature arcs such as sheets has been attributed to many processes,
the New Hebrides with the concept that slices only a few of which are consistent with recent
of oceanic crust can be accreted to the inner tectonism observed in the oceans. The nature of
slope leads to a model that is compatible with the thrust sheets capped by most such ophiolites

O ] ... ~ 0

km
5 ,;._-,.,,':,g,?;,,,'...':&:': 5

J °
km

0 --__-_- . . . . . . . . . ~ 0
km

Advanced, e.g.S.Middle A m e r i c a - " ~ "'#;".';~.,~;.~-~.r;fX[;~:S,,'.'.~!"r.'::'~,;

I i I w I i 1 I I 1 I
0 20 40 60 80 I00 km

V.E.-- 3×

FIG. 4. Schematic evolution of oceanic arc systems based on the assumption that structural variations
in such arcs of different ages are equivalent to temporal variations. Growth is by intermittent
accretion of relatively large slices of oceanic crust and upper mantle, manifested in emergent terrane
as disrupted ophiolite. Uplift, rather than lateral growth, dominates, converting an originally steep
undifferentiated slope into an upper section (forearc basin) and a steep lower slope.
 Initiation of subduction zones 573

(e.g. Dewey 1976), coupled with a realistic lack of wide forearc regions in young arcs and
model of arc-continent collisions as is now their common initiation along crustal interfaces
occurring in the Banda Arc (e.g. v o n d e r Borch argues strongly against trapping of wide bands
1979), strongly imply that these ophiolites occu- of oceanic crust behind the site of trench initia-
pied a forearc setting during part of that evolu- tion, and beneath forearc basins. The apparent
tion. The Zambales ophiolite is one of these contradiction might be resolved if large
forearc sheets, uplifted before collision with ophiolite sheets did evolve in a forearc setting
and emplacement on to a continental margin. but in relatively unusual circumstances.
This sheet is autochthonous with respect to Anomalous plate boundary geometries, in
interior Luzon, and closely fits the models of particular those in which oceanic spreading has
Seely (1979), Dickinson & Seely (1979) and transected the forearc region (Fig. 5) do occur
Hamilton (1977). along convergent margins at present, and could
Quite a different picture emerges from the account for several perplexing characteristics
foregoing description of young arc systems. The commonly observed in ophiolites. U p p e r plate

~'4.k •,

~ .:~. aTg,
, ~,g. •. . . . . . . 7,~.~

~~ ~.(lrl(ln(l ,ype Tong(i-NeWB.HebridesType

/ ,~,,X / ....

/( / .:

t '7
y *y
/
/ .
..1

" ":~'C-I. C-2.

Gulf of Californio Type
FIG. 5. Several examples of plate geometries along convergent margins in which transverse upper
plate spreading zones could lead to creation of ophiolite sheets in a forearc setting. (A) In this case,
modelled after the southern end of the Mariana arc, forearc rifting results from the extension of a
back-arc spreading zone to the trench. Cessation of back-arc spreading could lead to the creation of a
volcanic chain across very young 'oceanic' crust. (B) In a related case, drawn from the southern New
Hebrides arc, upper plate rifting separates trench from transform segments of plate boundary.
Initiation of subduction and of a volcanic arc would follow the north-easterly migration of the triple
junction. (C) In a different set of circumstances, as could evolve in the Gulf of California, a main
ocean basin spreading zone could rupture a pre-existing arc system, creating a large area of oceanic
crust within the arc or continental crust (1). Migration of poles of relative motion might subsequently
reinitiate subduction along the margin, trapping oceanic crust in and behind the forearc area (2).
574 D. E. Karig
rifting could be c o n t e m p o r a n e o u s with subduc- ally disrupted, either before or during emplace-
tion ( s o u t h e r n M a r i a n a ) or c o u l d be followed ment on continental margins. Some of these,
by s u b d u c t i o n , as w o u l d be the situation if a especially those d i s r u p t e d before e m p l a c e m e n t ,
slight pole shift re-initiated s u b d u c t i o n along could r e p r e s e n t slices of far-travelled o c e a n i c
Baja California and across the G u l f of Califor- crust a c c r e t e d to i n n e r t r e n c h slopes. T h e great
nia. In either case, o c e a n i c crust w o u l d u n d e r l i e difference in tectonic significance b e t w e e n fore-
the entire forearc region as well as the superim- arc ophiolite sheets, initially a u t o c h t h o n o u s
p o s e d volcanic arc. N o t only could the age with respect to the u p p e r plate and disrupted
difference b e t w e e n the o c e a n i c (ophiolitic) oceanic slices, a l l o c t h o n o u s with respect to both
crust and the arc volcanics be very small, but plates supports the r e c o m m e n d a t i o n of the
the interval b e t w e e n creation of the crust and ophiolite P e n r o s e C o n f e r e n c e ( C o n f e r e n c e Par-
e m p l a c e m e n t on a c o n t i n e n t a l m a r g i n could be ticipants 1972) to restrict the t e r m ophiolite to
very short. F o r e a r c rifting could also account structurally integral sheets. It is also a blatant
for s p r e a d i n g oblique to the margin and for the plea for i n c r e a s e d study of the s e d i m e n t s associ-
c o m m o n lack of t e r r i g e n o u s basal strata. ated with ophiolites and of p o s t - m a g m a t i c
M a n y so-called ophiolites have b e e n structur- structures of these bodies.

References
BOI,LIGER, W. & DE RUITER, P. A. C. 1976. Geology orogene: eastward subduction, tectonic rota-
of the south-central Java offshore area. Proc. tions, and volcanism in the western Pacific (Phi-
Indonesian Petrol. Assoc., 4th Ann. Cony., 1975, lippines). Tectonophysics, 67, 251-82.
1, 67-81. DEN(;(), G. 1962. Tectonic-igneous sequence in Costa
BONATH, E., CHERMAK, A. & HONNORt-Z, J. 1979. Rica. Mem. geol. Soc. Am. 133-61.
Tectonic and igneous emplacement of crust in DEWEY, J. F. 1975. Finite plate ew~lution: implica-
oceanic transform zones. In: De~7~ Drilling Re- tions for the evolution of transforms, triple junc-
suhs in the Atlantic Ocean: ocean crust, Am. tions, and orogenic belts. Am. J. Sci. 275-A,
geophys. Union, M. Ewing Ser. 2, 239-48. .260-84.
BOWIN, C., EEl-:, R. S., LEE, C. S. & SCHOUTF~N,H. 1976. Ophiolite obduction. Tectonophysics, 31,
1978. Plate convergence and accretion in the 93-120.
Taiwan-Luzon region. Bull. Am. Assoc. Petrol. - - & BIRD, J. M. 1970. Mountain belts and the
Geol. 62, 1645-72. new global tectonics. Y. geophys. Res. 75, 2625-
CARDWELI,, R. K., ISACKS, B. L. & KARIG, D. E. 47.
1980. The spatial distribution of earthquakes, DICKINSON, W. R. & SI-EI.Y, D. R. 1979. Structure
focal mechanism solutions, and subducted lithos- and stratigraphy of forearc regions. Bull. Am.
phere in the Philippine and northeastern Indone- Assoc. Petrol. Geol. 63, 2-31.
sian islands. Geophys. Monogr. Am. geophys. EATON, J. P., O'NEILL, M. E. & MURDOCK, J. N.
Union, 23, 1-35. 1970. Aftershocks of the 1966 Parkfield-
CARNEY, J. N. & MACFARLANE, A. 1978. Lower to Cholame, California, earthquake: a detailed
middle Miocene sediments on Maewo, New study. Bull. seisrnol. Soc. Am. 60, 1151-98.
Hebrides, and their relevance to the develop- ENGLAND, P. C. & WORTEL, R. 1980. Some conse-
ment of the outer Melanesian arc system. Aust. quences of the subduction of young slabs. Tecto-
Soc. Explor. Geophys. 9, 123-30. nophysics, 47, 403-15.
CHOUKROUNE, P., FRANCHETEAU,J. & LEPICHON, X. EVANS, C. & HAWKINS, J. W. 1979. Mariana arc-
1978. In situ structural observations along trans- trench system: petrology of 'seamounts' on the
form fault A in the FAMOUS area, Mid-Atlantic trench-slope break. Trans. Am. geophys. Union,
Ridge. Bull. geol. Soc. Am. 89, 1013-29. 60, 968.
Conference Participants 1972. Ophiolites. Geotimes, FAIRHEAD,J. D. 1973. Crustal structure of the Gulf of
17, 24-5. Aden and the Red Sea. Tectonophysics, 20,261-7.
CURRAY, J. R., SHOR, G. G. JR., PAITI, R. W. & FALVEY, O. 1978. Analysis of palaeomagnetic data
HENRY, M. 1977. Seismic refraction and reflec- from Viti Levu, Fiji. Aust. Soc. Explor. Geo-
tion studies of crustal structure of the eastern phys. 9, 117-23.
Sunda and western Banda arcs. J. geophys. Res. FANNING, P. R. 1912. Geologic reconnaissance of
82, 2479-89. northwestern Pangasinan. Philippine J. Sci. 7A,
DANIEL, J. 1978. Morphology and structure of the 255-82.
southern part of the New Hebrides island arc FARHOUDI, G. & KARIG, D. E. 1977. Makran of Iran
system. J. Phys. Earth, 26, 5181-90. and Pakistan as an active arc system. Geology, 5,
DSDP Staff, Leg 60 1978. Leg 60 ends in Guam. 664-8.
Geotimes, 23, 19-22. FISCHER, A. G., HEEZEN, B. C. et al. 1971. lnit. Rep.
DE BOER, J., ODOM, L. A., RAGLAND, P. C., SNIDER, Deep Sea drill. Project, 6, U.S. Govt Printing
F. G. & TILFORD, H. R. 1980. The Bataan Office, Washington.
 Initiation of subduction zones 575

GALLI-OLIVIER. C. 1979. Ophiolite and island-arc 1973. Plate convergence between the Philip-
volcanism in Costa Rica. Bull. geol. Soc. Am. pines and the Ryukyu islands. Mar. Geol. 14,
90, 444-52. 153-68.
GARRISON, R. E . , ESPIRITU, E . , HORAN, L. J. & - - 1974. Evolution of arc systems in the Western
MACK, L. E. 1979. Petrology, sedimentology, Pacific. Ann. Rev. Earth planet. Sci. 2, 51-75.
and diagenesis of hemiplegic limestone and tuf- - - 1975. Basin genesis in the Philippine Sea. In:
faceous turbidites in the Aksitero Formation, KARIG, D. E., INGLE, J. C. JR et al. (eds). Init.
Central Luzon, Philippines. Prof. Pap. U.S. Rep. Deep Sea drill. Pro]. 31,857-79. U.S. Govt
geol. Surv. 1112, 16. Printing Office, Washington.
GROW, J. A., BOWIN, C. O. & HUTCHINSON, D. R. --1977. Growth patterns on the upper trench
1979. The gravity field of the U.S. Atlantic slope. In: TALWANI, M. & PITMAN, W. C. III
continental margin. Tectonophysics, 59, 27-52. (eds). Island Arcs, Deep Sea Trenches, and
HAMILTON, W. 1977. Subduction in the Indonesian Back-arc Basins. Am. geophys. Union, M. Ew-
region. In: TALWANI M. & PITMAN, W. C. (eds). ing Set. 1, 175-86.
Island Arcs, Deep Sea Trenches and Back-arc -- , CALDWELL, J. G. & PARMENTIER,E. M. 1976.
Basins Am. Geophys. Union, M. Ewing Ser. 1, Effects of accretion on the geometry of the
15-32. descending lithosphere. J. geophys. Res. 81,
-- 1979. Tectonics of the Indonesian region. Prof. 6281-91.
Pap. U.S. geol. Surv. 1978, 345 pp. --, CARDWELL, R. K., MOORE, G. F. & MOORE, D.
HARRISON, J. V. 1953. The geology of the Santa Elena G. 1978a. Late Cenozoic subduction and con-
Peninsula in Costa Rica, Central America. 7th tinental truncation along the northern Middle
Pacif. Sci. Congr, 102-14. America Trench. Bull. geol. Soc. Am. 89, 265-
HASHIMOTO, W., MATSUMARU, K., KURIHARA, K., 76.
DAVID, P. P & BALCE, G. R. I977. Larger -- , LAWRENCE, M. B., MOORE, G. F. & CURRAY,J.
foraminiferal assemblages useful for the correla- R. 1980. Structural framework of the forearc
tion of the Cenozoic marine sediments in the basin, northwest Sumatra. J. geol. Soc. London,
mobile belt of the Philippines. Geol. PaleD. 137, 77-91.
South-east Asia, 18, 103-24. - - & MAMMERICKX, J. 1972. Tectonic framework
HAYES, D. E. & LUDWIG, W. J. 1967. The Manila of the New Hebrides island arc system. Mar.
Trench and West Luzon Trough--II. Gravity Geol. 12, 187-205.
and magnetics measurements. Deep Sea Res. 14, -- & MOORE, G. F. 1975. Tectonically controlled
545-60. sedimentation in marginal basins. Earth planet.
HENNINSEN, D. & WEYL, R. 1968. Ozeanische kruste Sci. Lett. 26, 233-8.
in Nicoya-komplex von Costa Rica (Mittelameri- - - & RANKEN, B. in preparation. Tectonic ele-
ka). Geol. Rdsch. 57, 33-47. ments of the southern Mariana arc.
IBRAHIM, A. K., PONTOISE, B., LATHAM, G., LARUE, -- & SHARMAN, G. F. 1975. Subduction and accre-
M., CHEN, T., ISACKS, B., RECY J. & LOUAT, R. tion in trenches. Bull. geol. Soc. Am. 86,
1980. Structure of the New Hebrides arc-trench 377-89.
system. J. geophys. Res. 85, 253-66. KETNER, K. B., KASTOWO, MODJO, S., NAESER,
INGLE, J. C., JR 1975. Summary of late Paleogene- C. W., OBRADOVICH, J. D., ROBINSON, K.
Neogene insular stratigraphy, paleobathymetry, SUZrANDAR, T. & WIKARND, 1976. Pre-Eocene
and correlations, Philippine Sea and Sea of rocks of Java, Indonesia. J. Res. U.S. geol. Surv.
Japan region. In: KARIG, D. E., INGLE, J. C., JR 4, 605-14.
et al. (eds). Init. Rep. Deep Sea drill Proj. 31, KIECKHEFER, R. M., SHOR, G. G. JR • CURRAY,J. R.
317-856. U.S. Govt Printing Office, Washing- 1980. Seismic refraction studies of the Sunda
ton. Trench and forearc basin. J. geophys. Res. 85,
IRVING, E. M. 1950. Review of Philippine basement 863-89.
geology and its problems. Philippine J. Sci. 79, KLITGORD, K. D. & BEHRENDT, J. C. 1979. Basin
267-307. structure of the U.S. Atlantic margin. Mere. Am.
ISACKS, B. L. & BARAZANGI, M. 1977. Geometry of Assoc. Petrol. Geol. 29, 85-112.
Benioff zones: lateral segmentation and down- - - & GROW, J. A. 1980. Jurassic seismic stratigra-
wards bending of the subducted lithosphere. In: phy and basement structure of western Atlantic
TALWANI, M. & PITMAN, W. C. III (eds). Island margin quiet zone. Bull. Am. Assoc. Petrol.
Arcs, Deep Sea Trenches and Back-arc Basins, Geol. 64, 1658-80.
Am. geophys. Union, M. Ewing Set. 1, 99-114. KULM, L. D., SCHEIDEGGER, K. F., PRINCE, R. A.,
CARDWELL, R . K . , CHATELAIN, J . - L . , DYMOND, J., MOORE, T. C. & HUSSONG, D. M.
BARAZANGI, M., MARTHELOT, J. M., CHINN, D. 1973. Tholeiite basalt ridge in the Peru Trench.
& LOUAT, R. 1981. Seismicity and tectonics of Geology, 1, 11-14.
the central New Hebrides island arc. Am. geo- LADD, J. W., IBRAHIM, A. K., MCMILLEN, K. J.,
phys. Union, Ewing Series, 4. LATHAM, G. V., VON HUENE, R. E., WATKINS, J.
KARIG, U. E. 1971. Structural history of the Mariana S., MOORE, J. C. & WORZEL, J. L. 1978. Tecto-
arc system. Bull. geol. Soc. Am. 82, 323-44. nics of the Middle America Trench. In: Int
1972. Remnant arcs. Bull. geol. Soc. Am. 83, Symp. Guatamalan Earthquake and Reconstruc-
1057-68. tion Process, Feb. 4, 1976, 2.
576 D . E. K a r i g

LEHNER, P. & DE RUITER, P. A. C. 1977. Structural of the Kings-Kaweah ophiolite belt, southwest
history of Atlantic margin of Africa. Bull. Am. Sierra Nevada, California. In: COLEMAN, R. G.
Assoc. Petrol. Geol. 61, 961-81. & IRWIN, W. P. (eds). North American Ophiolite
LOWELL, J. D. & GENIK, G. J. 1972. Sea-floor Volume. Bull. Oregon, Dep. Geol Miner. Ind.
spreading and structural evolution of Southern 95, 141-60.
Red Sea. Bull. Am. Assoc. Petrol. Geol. 56, SCHMIDT-EFFING, VON R. 1979. Alter und Genese des
247-59. Nicoya-Komplexes, einer ozeanischen Palaok-
LUDWIG, W. J. 1970. The Manila Trench and West ruste (Oberjura bis Eozan) im sudlichen Zentral-
Luzon Trough--III: seismic refraction measure- amerika. Geol. Rdsch. 68, 457-94.
ment. Deep Sea Res. 17, 553-71. SCHWELLER,W. J. & KARIG,D. E. 1979. Constraints on
-- , HAYES, D. E. & EW1NG, J. I. 1967. The Manila the origin and emplacement of the Zambales
Trench and West Luzon Trough--I: bathymetry ophiolite, Luzon, Philippines. Abstr. Progr. geol.
and sediment distribution. Deep Sea Res. 14, Soc. Am. 11, 512-3.
533-44. SEELY, D. R.-1979. The evolution of structural highs
LUYENDYK,B. P., BRYAN,W. B. & JEZEK, P. A. 1974. bordering major forearc basins. Mere. Am.
Shallow structure of the New Hebrides island Assoc. Petrol. Geol. 29, 245-60.
arc. Bull. geol. Soc. Am., 85, 1287-300. -- , VAIL, P. R. & WALTON,G. G. 1974. Trench slope
McADoo, D. C., CALDWELL, J. G. & TURCOTrE, D. model. In: BURK,C. A. & DRAKE,C. L. (eds). The
L. 1978. On the elastic-perfectly plastic bending Geology of Continental Margins, 249-60. Springer-
of the lithosphere under generalized loading with Verlag, New York.
application to the Kuril trench. Geophys. J. R. SHIPLEY, T. H., MCMILLEN, K. J., WATKINS, J. S.,
astron. Soc. 54, 11-26. MOORE, J. C., SANDOVAL-OCHOA,J. H. & WORZEL,
MCKENZlE, D. P. 1977. The initiation of trenches: a J. L. 1980. Continental margin of Guerrero and
finite amplitude instability. In: TALWANI, M. & Oaxaca, Mexico. Mar. Geol. 35, 65-82.
PITMAN, W. C. III (eds). Island Arcs, Deep Sea - - , LADD, J. W., BUFFLER, R. T. & WATKINS,J. S.
Trenches and Back-arc basins. Am. geophys, 1981. Tectonic processes along the Middle
Union, M. Ewing Ser. 1, 57-61. America Trench inner slope (this volume).
MITCHELL, A. H. G. & READING, H. G. 1971. Evolu- SHOR, G. G. JR., KIRK, H. K. & MENARD, H. W.
tion of island arcs. J. Geol. 79, 253-84. 1971. Crustal structure of the Melanesian area. J.
MOORE, D. G. 1973. Plate edge deformation and geophys. Res. 76, 2562-86.
crustal growth, Gulf of California structural pro- THIEDE, J. 1974. Sediment coarse fractions from the
vince. Bull. geol. Soc Am. 84, 1883-906. western Indian Ocean and the Gulf of Aden,
MOORE, G. F. & KARIG, D. E. 1980. Structural Deep Sea Drilling Project, Leg 24. In: FISCHER,
geology of Nias Island, Indonesia: implications R. L., BUNCE, E. T. et al. (eds). Init. Rep. Deep
for subduction zone tectonics. Am. J. Sci. 280, Sea drill. Pro]. 24, 561-766. U.S. Govt Printing
193-223. Office, Washington.
MOORE, J. C., WATKINS,J. S. et al. 1979. Progressive TUCHOLKE, B. E. & MOUNTAIN, G. S. 1979. Seismic
accretion in the Middle America Trench, south- stratigraphy, lithostratigraphy and paleosedi-
ern Mexico. Nature. London, 281,638-42. mentation patterns in the North American basin.
- - et al. 1981. Facies belts of the Middle In: Deep Drilling Results in the Atlantic Ocean:
America Trench and forearc region, southern continental margins and paleoenvironment, Am.
Mexico: results from Leg 66 DSDP (this geophys. Union, M. Ewing Ser. 3, 58-86.
volume). TURCOTrE, D. L., AHERN, J. L. & BIRD, J. M. 1977.
MROZOWSKI, C. L. & HAYES, D. E. 1979. The evolu- The state of stress at continental margins. Tecto-
tion of the Parece Vela basin, Eastern Philippine nophysics, 42, 1-28.
Sea. Earth planet. Sci. Lett. 46, 49-67. VON DES BORCH, C. C. 1979. Continental island arc
MuRAucHI, S., DEN, N., ASANO, S., HOITA, H., collision in the Banda Arc. Tectonophysics, 54,
Yosnn, T., ASANUMA, T., HAGIWARA, K., 169-93.
ICHIKAWA, K., SATO, T., LUDWIG, W. J., EWING, WATTS, A. B. & STECKLER, M. S. 1979. Subsidence
J. I., EDGAR, N. T. & HOUTZ, R. E. 1968. Crustal and eustacy at the continental margin of eastern
structure of the Philippine Sea. J. geophys. Res. North America. In: Deep Drilling Results in the
73, 3143-71. Atlantic Ocean: continental margins and paleoen-
MURPSY, R. W. 1973. The Manila Trench-West vironment. Am. geophys. Union, M. Ewing Ser. 3,
Taiwan fold belt: a flipped subduction zone. 218-34.
Bull. geol. Soc. Malaysia, 6, 27-42. WEEKS, W. F. & ANDERSON,D. L. 1958. Sea ice thrust
RAVENNE, C., PASCAL, G., DUBOIS, J., DUGAS, F. & structures. J. Glaciol. London, 3, 173.
MONTADERT, L. 1977. Model of a young intra- WHITMARSH, R. B., WESER, O. E., Ross, D. A. et al.
oceanic arc: the New Hebrides island arc. Int. 1974. Initial Rep. Deep sea drill. Proj. 23, 11-31.
Syrup. Geodynamics in South-West Pacific, 63- U.S. Govt Printing Office, Washington.
78. WILSON, J. T. 1966. Are the structures of the Carib-
SALEEeY J. B. 1977. Fracture zone tectonics, con- bean and Scotia arc regions analogous to ice
tinental margin fragmentation and emplacement rafting? Earth planet. Sci. Lett. 1,335-8.
D. E. KARIG, Department of Geological Sciences, Cornell University, Ithaca, New
York 14853, U.S.A.