ttBs§

tü l$$eBh$'\\$
$18\t

Quaternary Tectonic and Recent Seismic Crustal Movements

in the Arauco Peninsula and its Environs,
Central Chile

sohei I(arzuKA*. Tokihiko MArsuDA* *,

lvtichio Nocaui*, and Nobuyuki YoNEKURA***

. ABSTRACT

An outer part of the outer. arc in the arc-trénch system is a zone where great earthquakes
have occurred periodicatly .ánd remarkable crustal movements have been recorded both at
the great earthquake and at the inter-earthquake times. For comparing these short'term
historical crustal movements with long-term Quaternary ones, the Arauco Peninsula in the
Pacific coast of central Chile is one of the most suitable land on the earth to be studied. This
is the reason why we investigated the Quaternary crustal movements in and near the Arauco
Peninsula by means of geomorphology and Quaternary stratigraphy.
The Arauco Peninsula, which lies in the continental shelf zone bordering the Coastal
Range of central Chile, consists of five gemorphic surfaces: Las Nochas, Buena Esperanza,
Cañete, Lower teraces, and Holocene lowland surfaces in a descending order. The forrner
three are marine terraces, and the latter two are partly marine and partly alluvial in origin.
The topographic features and surface-making deposits of these su¡faces show that each of
the surfaces has been built during va¡ious periods of marine transgression, probably caused
by an interglacial or an i¡terstadial and a postglacial uplift of sea level. Among these
surfaces, the widest Cañete Surface, most probably formed in the last great interglacial or
Sangamon Interglacial, shows composite features of two modes of cnxtal movement§: a
landward tilting and an upwarping with a NW-SE trending axis. NW-SE trending upwarping
of the latter is also indicated by heights of elevated shorelines of the Buena Esperanza and
Las Nochas Sgrfaces, and even by heights of erosional surface¡ in the Nahuelbuta'Mountains,
a part of the Coast Range, lying east of the Arauco Peninsula. The e¡osional surfaces in the
mountains, which were probably built in the late Tertiary, imply that a N§ trending
upwarping has taken place during the Quaternary, superposing the above mentioned NW-SE
trending one. Two islets, Santa MarÍa and Mocha in the continental shelf zone, off the
northwestern and southwestern coasts of the Arauco Peninsula, consist of marine terraces.
These marine terraces show a landward tilting during the late Quatetnary. Radiocarbon
datings for Holocene terrace deposits in Mocha Island indicate that the mean rate of uplift
has been 5 5 cm per 1 00 years for the last several thousands years '
The mode of landward tilting during the late Quatemary in the shelf zone is just the same
as that of co-earthquake crustal movements in the historical age, and the amount of late

+ Departmentof Geography, Tokyo Metropolitan University

**Earthquake Research Institute, the University of Tokyo
*** Department of Geography, the University of Tokyo

-1-
Quaternary tilting is most likely to have been formed by the accumulation of repeated
co+arthquake tiltings; while the mode of the N-S trending Quaternary upwarping in the
Nahuelbuta Mountains is rather reverse to that of co+arthquake movements, and thus the
amount of the Quaternary uprnarping of the mountains would have been formed by the
accumnlation of inter-earthguake movements. Based on this, the terms, a zone of
co-etrthquake deformation and a zone of inter-earthquake deformation are proposed to the
shelf zone and the Coast Range respectively.

RESUMEN

La parte exterior del arco externo en el sistema arco-fosa es una zona donde han ocurrido
los grandes terremotos periodicamente y se has registrado importantes movimentos de la
corteza tanto en los terremotos mismos como en los períodos entre terremotos. La península
de Arauco en Chile central es u¡ra de las regiones más adecuadas del mundo para compárar
los movimientos de la corteza, sean históricos o cuaternarios. Este es el motivo porque
hernos estudiado los movimientos cuaternarios de la corteza en la región de la Península de
Arauco, con métodos geomorfológicos y estratigráficos.
La Península de Ar¿uco se halla en la z¡na de la plataforma continental que linda con la
Cordillera de la Costa, ye consiste en cinco superficies geomorfológicas por orden
descendiente: Las Nochas, Buena Esperanza, Cañete, Las Terr¿zas Bajas y el aluvio del
Holoceno. Estas últimas son marinas y aluviales en parte, mientras que las tres primeras son
terrazas marinas. Según las características de topografía y sedimentación de cada superficie,
consideramos que se han formado en un ambiente de transgresión marina quizás causada por
levantamientos, interglaciales, interestadiales o postglaciales del nivel del mar. La superficie
Cañete, la más extensa de las cinco, se ha formado seguramente en la riltima época
interglacial (Sangamon). Esta superficie exhibe rasgos compuestos de dos modos de
movimientos de la corteza: inclinación h¿cia el continente y combadura con eje NW-§E. Una
combadura con la misma di¡ección se observa también en el desplazamiento de la cost¿ de
emersión de Buena Esperanza y Las Nochas, además en el desplazamiento de las superficies
de erosión en la Cordillera Nahuelbuta, una parte de la Cordillera de la Costa ubicada al este
de la Península de Arauco. La superficie de erosión de esta Corüllera, que se ha formado tal
vez en el terciario superior, nos indica que ocwrió en el cuaternario la combadura con rumbo
NW-SE y además otra con rumbo N§. Las terrazas marinasque sehallanen dosislas, Santa
Mana y Mocha en la plataforma continental al noroeste y suroeste de la Penínsulá de {r¿spo,
fueron inclinadas hacia el continente durante el cuaternario superior. El levantamiento
medio de la Isla Mocha ha sido de 55 cm por siglo según las edades C-14 de los depósitos de
las terrazas.
La incünación hacia el continente en la zona de la plataforma continental durante el
cuaternario superior concuerda con 1os desplazamientos en los grandes terremotos históricos.
La inclinación cuaternario en esta zona se puede explicar por el efecto acumulado de los
terremotos. La combadura cuaternaria con eje N§ en la Corüllera de Nahuelbuta es opuesta
al desplazamiento de los terremotos, por lo tanto se debió formar por el movimiento de la
coÍteza entre terremotos. Por siguiente, proponemos aphcar los términos zona de
deformacíón m-sísmica y zona de deformación inter-sísmíca para la plataforma y la
Cordillera de la Costa, respectivamente.

1
 CONTENTS

Abst¡act I
Rcsumen 2
I. Int¡oduction 3
II. Regional setting 4
1. Andean Arc and Peru{hile Trench system 5
2. Physiographical and geological setting of the region sttrdied 6
III. Geomorphblogy. Quaternary geology and Quaternary crustal movements in the
Arauco Peninstüa and its environs 8
l. Previous studies 8
2. The Nahuclbuta Mountains 9
3. Pleistocene terraces and Holocene lowland in the Arauco Peninsula l0
A. Higher terraces l0
ts. Middle terrace (Cañete terrace) l3
C. Lower ter¡aces l8
D. Holocene lowland 19
4. Marine terraces of Santa Maria Island 20
5. Marine ten'aces of Mocha Island . 2t
6. Change of sea level and ages of the marine terraces 24
7. Summary of Pleistocene and Holocene vertical crrutal movements 26
IV. Relationshlp between recent deformations associated with. great earthquakes and
Quaternary crustal deformations . . . 28
1. Tectonic deformations associated with the 1960 Chilean great earthquake . . . . . 28
2. Previous evidence of tectonic deformations associated with great earthquakes in
historical time . . 30
3. Accumulation of seismic deformations in the late Quaternary . . . 32
V. Summary 35
Acknowledgments . 36
References JI
Photographs 39

I. INTRODUCTTON

The arc-trench system is the most active region on the surface of the earth. Around the
Pacific, there are many arc-trendr systems, whole of which makes the circum-Pacific
orogenic, seismic and volcanic belt.
Although arc-trench systems differ from one another in detail, every arc-trench system
has a common feature in the m¿cro-morphology, seismicity, crustal and subcrustal
structures. In general, the system consists of three topographical units; an oceanic trench, a
non-volcanic outer arc and a volcanic inner arc. The boundary between the non.volcanic
outer arc and volcanic inner arc is the front of volcanic belt or volcanic front (Sugimura,
1960), which runs parallel with the trench at the distance of 100-300 km from the axis of
the trench.
The outer arc is distinguished by its seismic activity; it is especially true near the trench
where great earthquakes, together with associated regional crustal deformations, have been

-3-
recorded. It has been noticed in the seaward parts of the outer arc of southwestern Japan
that the mode of Pleistocene and Holocene crustal deformations has a characteristic in
conunon feature, that is, an accumulation of landward tilting of the crust. This mode of
deformation is similar to that of crustal displacements associated with great earthquakes
occurred in the same regions (Sugimura and Naruse, 1954; Yoshikawa et al-, 1964 a, b,
1968; Yonekura, 1968). Nearly the same mode of crustal displacements associated with
great earthquakes has been reported from the Pacific coasts of Alaska and Chile, and the
mode of displacement has been explained by an underthrust faulting between oceanic and
continental lithospheres (Plafker, 1969; Plafker and Savage, l97A). The thrust-fault mech-
anism was also applied to explain the mode of crustal diqplacements associated with the
7923 and 1946 great earthquakes occurred in the Pacific coasts of southwestern Japan
(Ando, l97l;Fitch and Scholz, l97l).
Thus it is becoming clear that the seaward part of the outer arc has a similar mode of
deformation in the Quaternary and in the present; and also that the crustal deformation is
caused by underthrust ofthe oceanic lithosphere beneath the outer arc at the trench.
Our knowledge of crustal deformation in the seaward part of the outer arc, however, is
not enough to understand the whole features concerned, because: (l) the seaward part of
the outer arc is under the sea, and (2) a systematic study of the Quaternary and present
crustal deformations for the outer part of the outer arc started recently. Therefore, it is
disired to be done more studies for the seaward part of the outer arc.
The Arauco Peninsula in central Chile is one of the most suitable region for that study on
land, as well as the regions of the Mejillones Peninsula and the Tongoy area in northern coast
of Chile. Because all these regions lie at the distance of 100 km from the Peru-Chile Trench
axis. Moreover, the Arauco Peninsula has been attacked frequently, nearly once in every
century, by great earthquakes; and the documents of crustal displacements associated with
them are fairly abwrdant (Darwin, 1851;Plafker and Savage; l97O;Lamnitz,l970). These
are the reasons why we investigated this region under the financial support of the Japanese
Ministry of Education.
Our first seismo-geological mission to Chile and Peru (Mision sismo-geología del Japón al
Chile y Perú) spent about 40 days in the Arauco area from December 1970 to lanuary 1.971.
This is a primary report on the Arauco Peninsula and its environs. The main purpose of
this article is to describe the topography and Quatemary geology, especially the evidence of
crustal deformations during the Quaternary and the historical age in the region coacerned. A
discussion on the relationship between Quaternary crustal deformations and historical
seismic dislocations will be added briefly.
The more general discussion on the neotectonics of the outer arc in the Andean Arc
system will be stated in the near future, when the second seismo-geological mission in 1972
will finish the studies on crustal deformations in the Temuco-Valdivia region, southeast of
the Arauco Peninsula.

II. REGIONAL SETTING

The Arauco Peninsula is located on the Pacific coast of central Chile between 37" and
38"S. The peninsula is a part of the Andean Arc and the Peru-Chile Trench system (Figs. I
and 2).

-4-
 1. The Andean Arc and the Peru-Chile Trench system
This giant arc-trench system has a typical arc-trench feature. A trcncli, a non-volc:anic
outer arc and a volcanic inner arc are arranged in this order from tire oceanic side to the
continent.

llo' loo' go" go'

SOUTH
AMERICA

re-
')/s+¡
Frgos
=-i928
.t193
9

Fig. 1 Index map showing the Andean Arc-Trench systerr and the
southeast Pacific.
1: Volcanic belt with volcanic front.
2: Region of great earthquake.
The Peru-Chile Trench extends from 4"N to 40"S, the deepest part of which reaching
about 8000 m below sea level offthe coast ofAntofagasta at 23"S (Fisher and Raitt, 1962).
The non-volcanic outer arc is defined as the region between the inner wall of trench and
the volcanic front, Generally, the outer arc is composed of a continental slope, a continental
shelf, an outer arc rise and a mid-arc trough. In Chile, the outer arc rise is expressed as the
Coast Range; the mid-arc trough, the Longitudinal (Central) Valley. While in Peru, both the
outer arc rise and the mid-arc trough are generally expressed as a wide continental shelf.
There are two exceptional areas; the northern coast of Peru and the landing place of the
Nazca Ridge in southern Peru, where the outer arc rises are seen on land as coastal ranges.
Along the Chilean Pacific coast, the continental shelf is narTow. It is especially true off
the coasts of the Mejillones and Arauco peninsulas, the distances from the trench axes to the

-5-
coasts being 70 km and 100 km respectively. Such a short distance from the trench axis to
the coast is rather unique in the world.
The Andean Mountains or the volcanic inner arc is 4000 to 6000 m high above sea level in
its central part, decreasing its height to the south gradually until about 2000 m at around
400s.
The outer arc is characterized by inteme shallow seismic activities both of smaller
earthquakes and of great ones (Fig 1). In the central and northem Andes intermediate and
deep earthquakes occur, and the deepest hypocenters are as great as 650 km deep, whereas
in the southern Andes, south of 29"S the deepest ones occur about 150-200km(Santo,
1969),and south of 46'5 no seismic activity is observed along the conünental margin.
The thickness of the crust increases progressively from the Pacific basin floor to the
Andes (Fisher and Raitt, 1962;l-omnitz,1962), with an excepüon of the crustal thinning
beneath the offshore flank of the trench (Hayes, 1966). The thickness reaches from about
10 km beneath the trench axis to 55-70 km beneath the Andes.

2. Physiographical and geological setting of the region studied

From south of Santiago to Puerto Montt (34"-42'5), there is a striking zonal
arrangement of macro-morphology from the oceanic side landward a trench, a
continental slope, a conünental shelf, the Coast Range, the Central Valley, -and the Andean
Mountains capped by volcanoes (Fig. 2). Generally, the width of the continental shelf is
several tens kilometers, the Coast Range about 100 kilometers, and the Central Valley
several tens kilometers.
The only one peninsula protrudíng to the Pacific along the coast south of Santiago is the
Arauco Peninsrfa. Off the west coast of this peninsula, the width of the shelf is
exceptionally narrow. According to the study of wismic reflection profiles (Sholl et a1.,
1970: Fig. 3), the continental shelf north of the Arauco Peninsula consists mainlyof thick
Tertiary strata, unconformably overlying upper Cretaceous rocks. The rocks of the Arauco
Peninsula are also mainly composed of Tertiary strata overlying Cretaceous rocks. Hence the
peninsula seems to be a raised part of the shelf zone.
The Tertiary strata in the Arauco Peninsula are of the Eocene, Miocene and Pliocene
(Tavera, 1942; Muñoz Cristi, 1946,1956; Brüggen, 1950; Zeil, 1964;Garcia. 1968). The
Eocene, consisting of ma¡ine and intercalated continental sediments with coal seams dips
westward at about 20", and is affected by block fatfting. The strata of the marine lower and
upper Miocene are named the Navidad and Ranquil Formation respectively. The Pliocene
marine Tubul Formation is distributed in the northern part of the peninsula. The Pleistocene
marine sediments, which will be described later in detail, truncate the above-mentioned
"basement rocks", and are distributed over nearly the whole peninsula.
The Coast Range in the latitude of 34o 42oS fs generally low-relief mountains as a whole
lower than 1000 m above sea level. However, the Nahuelbuta Mountains (Cordillera de
Nahuelbuta), which lies east of the Arauco Peninsula, attain an exceptional height of more
than 1000 m, the highest point reaching 1530 m. No river traverses the mountains. On the
northeast the mountains are bordered by the River Bío Bio, and on the southby the River
Imperial, both of which start from the Andes and empty into the Pacific Ocean.
The Coast Range consists mainly of Precambrian or Paleozoic metamorphic rocks (Schist
and phyllite) and granitic plutonic rocks. The Nahuelbuta Mountains also consist of these
two kinds of rocks. Along the western slope of the Nahuelbuta, marine strata of the upper
Cretaceous overlie the basement complex, striking N-S with a dip of l0o to 30o west as are

*6-
 I

,)
(- ',
)',
j'.,
)r §
la)

J
E
F
z
l¡J I

76" 74" 720 700 6go

Fig. 2 Generalized topography of the Andean arc-trench system in

certral Chile.
Contour interval: 100m. Soüd circle:volcano. C: Concep-
ción. M:Mendosa- P:Puerto Montt. S:Santiago. Va:Val-
divia- Vo:Valparaiso.

the Eocene strata in the Arauco Peninsula.

The present climate of the Arauco Peninsula is as follows. The annual precipitation is

-7 -
 ,l
2l
Sholl et ol. 1970
3

0
I

2
3

o
I

2
3
4
5 Verl Exog = ¡9
6

Fig. 3 Cross sections from the trench to the coastal area in and
near the Arauco Peninsula.
Submarine sections are drawn from Sholl et al., 1968, 1970
and from a chad, No. 3075 published by Hydrographer of
the NaYy, u.K,

about 1500 mm (mainly in winter). The mean temperature is around 8-9'C in winter,
16-17"C in summer, and the annual mean is about 13"C.

III. GEOMORPHOLOGY, QUATERNARY GEOLOGY AND QUATERNARY

CRUSTAL MOVEMENTS IN THE ARAUCO PENINSULA AND ITS
ENVIRONS

l. Previous studies
It has been known that the Tertiary formations in the Arauco Peninsula are covered
unconformably with marine Pleistocene deposits. The deposits were named the Arauco
Formation by Muñoz Cristi (1968), and subdivided into older and younger ones by Stiefel
(1968). The studies ofthe Pleistocene deposits, however, have scarecely been done so far.
A terrace surface widely distributed in the peninsula was called as"meseta pliocenica"
(Pliocene tableJand) by Brüggen (1950). According to Süefel (1968), the terrace surface is
partly an erosional surface cutting the underlying Tertiary formaüons, and partly a
depositíonal one of the Pleistocene deposits. This terrace surface is designated as middle
terrace surface (or Cañete Surface) in this paper.
It has been described by Briiggen (1950) and Stiefel (1968) that the terrace surface is the
highest in the east of Los Alamos, from where the altitude of the.surface descends both to
the north and south. This feature of the surface, which must be ascribed to crustal
deformation, is concerned with a main subject of this paper.

-8-
 tsruggen (1950) also pointed out thatbetween tlüs terrace and the NahuelbutaMountains
there exists a higher marine terrace with thin deposits, and that the highest part of the
terrace lies to the east of Los Alarnos. Tlús terrace is designated as highcr terracos in this
paper and their dclormations are also ati iurportant subject of this papcr.
As to the topography of the Nahuelbuta Mountains. it has been rriited that there are
renlnants of peneplanated surfaces, and their altitudes imply an upheaval of the mountains
in the Pleistocene (Brüggen, 1950;N'luñoz Cristi, 1956).

2. The Nahuelbuta Mountains

Fig. 4 is a Gipfelflur showing topographic features of the Nahuelbuta N{ountains and the
Arauco Peninsula. A projected profile of the mountains is shown in Fig. 13. The Gipfelflttr
c:ontour lines are drawn based on the highest points in 2x2 km quadrangles on 1: 50,000
topographic maps. The projected profile is drawn based on the liighest points in east-west
belts of 2 km in width.

74"
nil
Erosion
surfoce

o ,
'cloN / /t Slare- ne ol
Losl'lo,riós
S!, lcae

Shorel t ol

'r Si MARIA
o
I

(\
)

I
I

o 40km

Fig. 4 Gipfelflur of the Nahuelbuta Mountains and the Arauco

Peninsula.
Contour interval: 100 m
As shown in these figures and Photo I , the altitudes of the mountains decrease from the
trighest peak of 1530 m graclually to the north and abruptly to the south. On the west, the

-9 -
mountains are bordered by fairly steep slope running straight nearly north-south- While on
the east, the bordering slope is rather gentle and runs irregularty. The wid.th of the
mountains is larger in the south than in the north.
Referring to the 1000 m contour line, it is noted that the central high of the mountains
extends in the NW-SE direction- Interesting is that the mountains in the northwestern part
of the Arauco Peninsula lie just on the northwestern extension of this high, and also as seen
in the later part of this paper (Fig. 12), an axis of uplift of the Cañete Surface lies nearly
along this extension.
As shown in Fig. 4, low-relief erosion surfaces are distributed with various heights in
alrnost whole area of the Nahuelbuta. The distribution is drawn based on our interpretation
from aerial photographs (about 1:40,000 in scale) and topographic maps (1:50,000).
The origin of these erosion surfaces seerns to be due not to marine abrasion, but to
subaerial denudation. In the field, we examined the erosion surfaces at several places:
northeast of Coronel, northeast of Carampangue and southeast of Curanilahue. But we could
not find any marine deposits overlying the basement rocks.
Judging from the altitudes of the erosion surfaces, it canbe said that most of them were
made once as a peneplain under the same base-level of erosion. The age of the peneplanation
is not evident, but it must be younger than the Eocene strata,'and older than the Pleistocene
marine terraces. Probably the age falls in the late Tertiary. Thus the present altitudes of the
erosion surfaces seem to indicate an areal variation of crustal deformation since the late
Tertiary.
The Nahuelbuta Mountains are dissected by steep valleys of the youth stage of erosion.
This fact as well as the deformation of terraces in the eastern part of the Arauco Peninsula
implies that the deformation and upheaval of the mount¿ins took place mainly durir-rg the
Quaternary, as stated by Briiggen (1950) and Muñoz Cristi (1968).

3. Pleistocene terraces and Holocene lowland in the Arauco Peninsula

The landform of the Arauco Peninsula is composed of marine terraces, which are
subdivided into higher, middle and lower terraces. and Holocene lowland. These divisions are
shown in the geomorphological map (Fig. 5) based on airphoto interpretations and field
observations.
Altitudes of raised shorelines of the marine terraces were read on altimeters in the field or
on topographical maps, and are shown by numerals in Fig. 5. Altimeters of the American
Paulin System Terra MT-2 were used and correction for barometric changes was made by
using a Barograph of the same System.

A. Higher terraces
The higher terraces are distributed along the western foot of the Nahuelbuta (Photos I
and 2), and in the northwestern mountains of the Arauco Peninsula which were called as
"cerros de Yane-Lavapie" by Brüggen (1950).
The higher terraces along the weste(n foot of the Nahuelbuta are composed of two
terrace surfaces, the upper of which is designated here as the Las Nochas Surface, and the
lower as the Buena Esperanza Surface. Both surfaces are fair§ dissected by valleys, but
original terrace surfaces are partly preserved.

Las Nochas Surface

The deformed shoreline of the L¿s Nochas Surface reaches up to 500 m in altitude at

-10-
 PI LAVAPIE

ARAUCO BAY
RBIOB O
'§7
F
É 'ARAMPA\6UE

"(./,
,t € tr¡:ffiffi,
q (/
/
,/ Á,H'.i
e 0 /f.,
).,
/,4f (

/a
/fl
¿
Y ,.P*.;r#
§.Óñ 4 ,. "orJ
EF
-)
/_ o-;oa ;r, 'cD
)
eCl ,rzi6 u I
UJ

tlj. f
h--- íÍrff-,I\§lff Y,&,4 I
\,r '',oulrfS' z
,' ^"oÍl ÉtL:'i4Ál'.'i,\

W
< .ORONEL
z. ' @

t! U¡i:*lr:e ?r: o
(,
U iq!/¿¡;,+-l f
o ITii EDf ul
j
É.

U J
LL l\,ó« \o#t $".k1- a
- ) "as t"\9í ^+á E
o
U
tr

73"t5

PII PILCO
5 cpe ono s(o rp

6:.! Eiir::l ru.

ffi us r\ÚL¡'u5 )u'I r s
Nochos Surfoce
[j,:.J"
canauÁvron Vja 3uet,o Esperonzo Surfoce ] surfoces

j,.l
ó40 Coñete Surfoce
-vliddle lerroce S

E Lo*u, terroce g:rfoces

ffi !cto.en" low cnd surfoce

tl Flxed sond dune of lhe Pleisiocen¿
i-l-..1 fxeo sond oune cf the Recenl

Ll-il Active sond dune

i-:l Beoch rioqes

il> Loxe
B&- Shore lire of Los Nlochos S

fu ]fx:¡re-'¡necf Bueno Esperonzc S

I5 Shore-|rre,:f Coñcle S

e S¡ore l¡ne of lowerterroce s¡rfoces

.¿rc lleghl of shore , re reod by oltlrneler (m)
or5o Herghtof shore-line reodon rnop (rn)
x O Locolily of qeo cq c seclron
5 rO 15 ¿OKñ

Fig- 5 Geomorphological map of the Arauco Peninsula.

37o35'S, from where altitudes of the shoreline decline both to the north and to the south.
The width of the surface is in a few kfometers and is widest around the highest shoreline.
The terrace surface is, broadly speaking, an abrasion platform cutting the basement
complex of metamorphic and Eocene rocks. The terrace deposits are generally thin, and
composed of sand, pebbles and partly boulders. In places, however, thick marine terrace
deposits are found; for example, at Descabezado,5 km southeast of Curanilahue, sand and
gravel beds of more than 20 m thick are present.
The terrace deposits have been weathered to such an extent that they are reddish brown
to reddish orange in color (7 .5R 516 - 10R 6/8 in Munsell Color Chart).
The elevated shoreline of the Las Nochas Surface is not straight, but indented as shown in
Fig. 5. This feature of the shoreline indicates that there were mariy small embayments
curved in the Nahuelbuta Mountains when the marine terrace was built.
Considering such an irregular shoreline feature and the existance of thick terrace deposits,
it is obvious that the terrace was built under an environment of ria coast, in other words,
under the enüronment of marine transgression-

Bue na E s pe ran za S urfac e

This surface extends along the western foot of the Nahuelbuta Mountains between Las
Nochas and Cañete Surfaces. The highest shoreline is about 340 m high above sea level at the
south of Curanilahue, where the terrace surface reaches the greatest width. The shoreline
angle of this terrace is not so obvious as that of the Las Nochas Surface. To the north of
Lota, the shoreline and terrace surface are not preserved.
This terrace surface is mainly an abrasion platform, which truncates the basement of
schist, Cretaceous and Eocene strata. In places, however, we saw marine beds consisting of
sand, silt and pebbles, for instance at 3.5 km southeast of Curanilahue. The Buena Esperanza
shoreline is less indented than that of the I¿s Nochas Surface.

Higher terrace on tlze northwestern mountaíns in the Arauco Peninsula

The northwestern mountains, being less than 400 m high, consist of Eocene strata.
Valleys in the mountains are prevailingly in the NE-SW direction. They may have been built
under a structural control of the Eocene rocks, since the strata strike to the same NE-SW
direction.
On summits of the mountains there present flat or low-relief surfaces of 325-390 m in
height. Although no research works have been done as to the surface-making deposits,
judging from their topographic features they must have been built as a marine terrace
originally.
Altitudes of the surface are kept nearly constant in the N-S direction from the northern
end, near Punta Lavapie at 37"12'S to the south at 37o19'S, and they also vary little in the
E-lV direction. Hence it may be concluded that significant differential vertical displacements
have not been taken place in this mountains since the surface was built.
We tentatively correlate this higher terrace surface with the Buena Esperanza Surface in
age based on their altitudes.

B. Middle terrace (Cañete terrace)

Topography of the C.añete terrace and its drainage system
The middle terrace occupies most of the area of the Arauco Peninsula. The terrace surface
is widely preserved except in the western part of the peninsula where it has mostly been lost

-13-
 by fluvial erosion (Fig. 5). This surface is designated as the Cañete Surface after the name of
the largest town located on the surface,
As shown in Figs. 5 and 12, the altitude of the surface is highest (225-25Om) in the
western part of the peninsula, from where it declines northeastward and southeastward until
it comes to less than 100 m. The former shoreline angle of this terrace is well preserved along
the western border of the Nahuelbuta from near Concepción (36'45'5) to Cañete
(37"50'S). Near Cañete the raised shoreline enters valleys in the Nahuelbuta and makes an
irregular feature. Therefore, the shoreline must have been formed along a ria coast.
The highest altitude of the shoreline along the western border of the Nahuelbuta is 225 m
above sea level at the northeast of Los Alamos, from there decreasing to the north and south
(Figs. 5 and 12). Around the northwestern mountains of the peninsula, the shoreline angle
of the Cañete terrace is not well preserved, although the mountains must have been an island
surrounded by the sea shore at the time of the Cañete Surface. The altitude of the r¿ised
shoreline is 225-250 m above sea level in the southern part of the mountains.
Rivers which dissect the Cañete Surface in the northeastern and southeastern parts of the
peninsula flow consequenfly to he surface features that are shown by the restored contour
lines in Figs. 5 and 12, while in other parts of the peninsula rivers run inconsequently to the
surface features. For example, the largest river in the peninsula, the River Lebu (Rio Lebu),
flows to the west cutting through the highest part of the Cañete Surface (Photo 5). It is mmt
likely that the River Lebu is an antecedent river.
Of interest to mention is the fact that valleys running NNE-SSW are dominant in the
southem and also, to some extent, in the northern parts of the Cañete Surface. This peculiar
drainage system seems to be developed under topographic control of longitudinal dunes of
the NNE-SSW direction, which will be described in the succeeding section.

Middle tefface deposits: the C,oñete Formation and superficial deposits overlying it
The Pleistocene deposits forming the middle terrace are designated as the Cañete
Formation. The main part of the formation is marine, and the uppermost part is eolian in
origln. In some places fluvial deposits are found above the Cañete Formation, but eolian
loamy deposits are common in most places above it. A geological map of the Cañgte
Formation is shown in Fig. 6 and.columnar sections in Fig. 7.

Distribution, thickness and sequence of the &ñete Formstion. As shown in Fig. 6, the
thickness of the formation is large in the northeast and soutlreast parts of the terrace
reaching 60 m, whle small in the central part less 20 m, and especially small or almost
lacking in the central east and northwest parts, where the Eocene strata and Precambrian
schist are exposed as the basement rocks. The terrace surfaces in these central east and
northwest parts are located near the ancient shorelines, thus the surfaces are inferred as
abrasion platforms in origin. Since the detail basal topography of the Cañete Formation is
not clear, the thickness variation shown with contour lines in Fig. 6 indicates the general
tendency.
The Cañete Formation can be divided into the lower marine beds consisting mostly of
thick sand and the upper eolian beds of thin sand. The boundary between them is not
necessarily sharp.

Marine beds of the Cañete Formttion. The marine beds consist mainly of sand, but in some
horizons especially in the lower, of pebbles, cobbles and boulders, §ilt layers are also

-14-
Fig. 6 Geological map of the Arauco Peninsula showing geology of
the Cañete Formation.

15 -
 o @ @ @@
751- coñete
Surfoce

ro

20 Coñete
Formotion

30 C

40 35
Plio?
r58
50 Mio.

60
El sili, cloy, mudslone
Plio.,Pl iocene
70 E sond, fine-coorse
Mio. , Miocene
Fl§61 pebble, boulder
lF-7 pumice, pumiceous
Eo. ' Eocene
80 c ,cross-lominoied
% soil
w 'wind-blown
IFI shell
t-l burrow of lorvo Cicindelidoe

Fig.7 Columnar sections of the Cañete Formation. Localities are

shown in Fig.5 and 6. Numbers represent height above sea
level in meters.
intercaleted in some places (Fig. 7), and marine molluscan fossils are seen at the basal parts
of the formation (Column 7 in Fig. 7).
Gravels are round or subround in shape. Sand is generally well-sorted. In the middle part
sand layers show cross lamination in general. Some sand beds contain many grains of mica
and some others do many grains of quartz, whose origin is probably schist and granites in the
Nahuelbuta. It is noted that sand beds in the upper and middle horizons contain a lot of
pumice grains. The source of the pumice grains may be primarily due to the Andean
volcanoes southeast of this region. These pumice grains might be originated as pumice flows
and then transported by fluvial processes along ancestral courses of the present River
Imperial or Toltén from the Andes to the coast and further transported north to the Arauco
Peninsula by longshore currents.
This interpretation depends on the following facts that the rocfts of the Nahuelbuta could
not be sources of pumice grains and that the prevailing wind and longshore current in this
region would have been from south to north in the Pleistocene Í¡s well as in the present. Thus
the origin must be anywhere to the south of the peninsula. The secondary source area is
reasonably considered as the coastal area between the Imperial and Toltén rivers to which
pumiceous volcanic material would easily be supplied from the volcanoes of the Andes.

Eolian bed of the Cafiete Formation. The uppermost part of the Cañete Formation consists
of widely-distributed eolian sand beds. These beds are composed of well-sorted fine to
medium sand including whitish small pumice grains. The sand beds are firmly consolidated,
hence they can be said eolianite @hoto 3). The thickness of the beds are 1-3 m in general,

-16-
but where the sand beds make dune topography (see Photo l) the thickness reaches ntore
than several meters. Sometimes the sand bcds are separated into two or three horizons by
buried soil layers.
The sand dunes made of these sand beds extend widely in thc southern part <¡f tlie Car'iete
Surface. Some of them having distinct dune form are strorvn in Fig. 5 under the nomination
of "flxed sand dune of the Pleistooene". Generally they are longituürral dunes with
NNE-SSW elongation. To the east and southeast of Cañete town. sand dunes of the same
horizon mask the south-facing ancient shoreline angles of the Cañete Surface, so that the
altitudes of shoreline were unable to be measured there.
The eolian beds are generally ovcrlain by a bright brown to reddish browu loamy soil of
l-3 m thick (Photo 3), rvhich will be rnentioned later. Ileneath this soil and just in the
r.tppernrost part of the eolianite. there are seen fossil burrows in many places. Most of thenr
are found in the eolian beds as described bclow, but some are found in boulders of eolianite
or not-eolian sand beds on the abrasion platfonn of the Cañete Formation.

Fossil burrows in eolian sa¡td as an indicator of sedimentary environnlent. The burrorvs are
ordinarily dug from the surface of eolianite perpendicularly as deep as about 10 cm with a
diameter of 1 -3 cm (Photo 3 ). They are found not onl¡' in the eolianite of the trpper Cañete
Formation, but also on the surface of dune sand overlying the lower terrace surf'ace. and
even in recent dune sand on the lowland.
Fortunetly, rve could find a dned remain of an insect frr¡m a burrow in recent dune sand
on the lorvland of southeastern coast of tlie peninsula. This insect is identified as a larva of
tiger beetle (.Cictudelidae) by zoologrst Yuzo Kitazawa and entomologist Kazuyoshi Kurosa.
They suggest that the f<-rssil burrow features in the eolianite bed of the Caitete Formation
were made by the sar¡re kind of larva as what we found, and that the environntent, wherein
some kinds of tiger beetle larva like to live, is cotrsistent with the envirorunetrt that we
estimated from the eolianite, that is, a barren sand field.
From this line of evidence. it is apparent tl-rat the Cañete Surface was once widely masked
by rvind-blown sand when the surlace had enrerged from the sea. The direction of
sand-carrying wind was most probably south or southwest as indicated by the axial
directions of elongated sand dunes.

Superfical deposits overlyittg the Catúte Fomlation. On rare occasions there present tlún
beds 2-5 m thick, consisting of gravels, sand and silt. which overlie uncomformably the
Cañete Fo¡nation (Photo 4). These beds seem to be flunal deposits of streams, wlüch flew
temporarily on the Cañete Surface when the surtace emerged from the sea. These beds are
also mantled with loamy soil (Photo 4).
The loamy soil, which mantles the whole surface of the Cañete Formation and the
superficial fluvial deposits, is silty loam of 1-3 m in thickness and is bright brown to reildish
brown in color. This loamy deposit is a kind of air-laid nraterials, perhaps loess andior
volcanic ash in origin, judging from its mode of deposition and grain size. The samekind of
loamy deposit masks the lower terrace surface too, though the thick¡iess is less than that on
the middle terrace surface.

Tertiary forntations underlyittg the Canerc Formation

Tertiary formations in the Arauco Peninsula are, as stated already, composed of the
Pliocene, Miocene and Eocene sediments. The Pliocene Tubul Formation. which is

-17 -
distributed in the northern half of the peninsula (Fig. 6). consists of grey massive siltstorre
including marinc lossil shells in some places. Foraminifera yielded in it has been studie<l by
Martínez (1968 a).
The Miocene and Eocene formations are erposed in the eastem part and central
(Curanilahue-Lebu) area o[ the peninsula. These formations dip to the west in general. The
Miocene in the central area is mainly of rnassive rnudstone scarcely including fossil shells.
The Eocene fbrrnaLion occurring the eastern part consists mainly of sandstone beds with
coal seams. The sandstone beds have been weathered to yellow to brown fur color generally.
Someti¡nes it is difficult to distinguish the weathcred Eocene rocks tiom the Cañete
Fonnation.

Physiographic devebpntent oJ- the middle terrace

The Cañete Formation is a sedimentary unit formed during a marine transgression. Before
the transgression, there was a dissected land in the area of the present Arauco Peninsula on
the western foot of the Nahuelbuta. As the rnarine transgression progressed, the land sub-
merged to form ria coasts in this region alorrg the western border of the Nahuelbuta
Mountains,
During the maxi¡lum stage of the transgression sequence. a rvide abrasion platform was
built and a depositional surface of the Cañete Formation as well. The duration years of
stillstand of sea lcvel at the maximum stage can be roughly estimated frorn the width of the
abrasion platform, based upon our experience on postglacial abrasion platforms in .Iapan, as
10,000 years or so.
During the following regression stage, the Cañete Surface emerged above the sea levelas a
newly born coastal lowland, which was successively covered by' sand and sand dunes under
southerly wind. The eolían sand was accumulated much more ir the soutirern coast than in
the northern coast of the Arauco Peninsula. Such a dit'lerence between the nodhern and
southern coasts in the Pleistocene time is cc¡nsistent with the difference in the prescnt state
of sand accurnulation betweeu the both coasts (see Fig. 5).
On the way of landing of the Cañete Surface, many rivers extended fronl the surrounding
mountains on t<¡ the Cañete Surface, and also many newly born strcams having their
catchment areas in the Cañete Surface appeared. These rivers and streams cut down the
surface. T'he large extended rivers took their course consequently to the general submarine
topography of the Cañete Surface, while small stream courses were determined by the smaLll
relief on the Cairete Surface such as the NNE-SSW longitudinal dunes.
Physiographic development of the Cañete Surlace is much affected by the deformatiorr of
this region, which will be mentit¡Iled in later sections.

C. Lower terraces
Lower terraces are both marine and fluvial in origin. Marine lower tcrraces are distributed
Iocally along the coast of the peninsula, and fluvial terraces along rivers in the pcninsula (Fig.
5). Relatively wide lower terraces are described Ín the followings.

Atong the River Cararnphatryue. northeast of the peninsula

Near I¿ Meseta, east of Carampangue town, there are low table-lands which are 16-17 m
high above sea level. These terraces consist of fluvial sand and silt.
The 16 m high lower terrace at Ramadillas, 6 km southwest <¡f Carampangue town, is
composed of fluvial sand and subangular gravels.

-18-
 Alang the Ritter Tubul, northem coast ofthe pcninsulu
Around the alluvial lowland of the River Tubul, there are at least three levels olterrace
surfaccs. 20-30 m, 50 (r0 m, and about 80 m lúgh above sea level, but we havc no
knowledge abor,rt their terrace deposits.

Near Lebu, wastenT coast oJ the pcn¡nsulu

There are several small steps of terraces just near the town of Lebu, as follows The
highest terrace (a) has the shoreline at about 150m in height, the second (b) at about
125 m. the third (c) at about 115 m, the fourth (d) at about 100 m; and further lower
fragmental terraces (e) exist. The terraces (a), (b), (c). and (d) are inferred as marine terraces.
and each surface seems to be an abrasion platform. At the (c) terrace we observed deposits
of 3-5 m thick pumiceous sand and gravels.
Along the River Lebu there are fluvial terraces. 150-200 m high above sea level.
Around l0 km south of Lebu, there is a marine lower terrace having a 116 m-high
shoreline. This may be correlated to the (b) or (c) terrace near Lebu. The surface of this
terrace is masked with eolianite. From this te¡race to the south no marine lower terraces can
be seen, notwithstanding there present a higlr colrtinuous sea cliff betweel the middle
terrace and lowland (Fig. 5 and Photo 6). This can be crplained in such a way that near
Lebu the basal rocks of the Eocene is hard enough to protect tlie lower terraces trom marine
erosion, but to the soutl¡ basal rocks consist rnainly of the unconsolidated Cañete Formation
irstead of the Eocene, so lower terraccs rvas eroded completell'to fonn the lúgh sea cliff
behind i.t b1'rnarine erosion.

Along the River Paicati, solttheast o.f the peninsula

Along the River Paicaví, south of Cañete town. occur fluvial terraces of 25 30 m in
height, the surtaces of which are several meters lower tha¡r tire Cañete Surface. On thcse
terraces dune sand (eolianite) is distributed.
As stated in the preceeding paragraphs, there are many' levels belonging to the lower
terraces in separate localities, but it is difficult to correlate with each other. It is also
difficult to mention whether marine transgression occurred or not during the age of the
lower terraces, However, it is convincing that magnitude of the marine transgression, even if
it occurred. was much smaller than that of marine transgressions at the time of ttre Las
Nochas, Cañete and Holocene lowland surfaces, the last one of which will be descríbed in the
succeeding section.

D. Hr¡locene lowland
Holocene coastal lowlands are distributed afong the coasts. and Holocene alluvial
lowlands along the downstrcams of large rivers (see Photo 5).
The topographic l-eatures of alluvial lowlands in l.he downstrealrs suggest that there rvere
drowned valleys before the deposition of alluvial sedimets. This view has been supportcd by
borc-holc data in the lowland of the mouth of the River Lebu (García. 1968), which show
that there lie 60 70 m thick "Quaternary" deposits upon the basal rocks. A similar view was
so far expressed by Martinez (1968 b), who studied the Holocene deposits near Concepción.
He concluded that the Holocene or Frandrian transgression occurred therc.
As is shown in Fig. 5, the coastal lowland along Arauco Bay is characterized by many
lines of beach ridges, whfe the lowland along the southern coast of the peninsula, by sand
fields and a large number of sand dunes. Some of these sand dunes are fixed by plants, and

-19-
others are barren and active at the present, as represented in Fig.5 and Photo 6.

4. Marine terraces of Santa María Islarrd

Santa María Island (Isla Santa María) lies north of the northwestern cape (Punta Lavapie)
of the Arauco Peninsula. This island is 12km in north-south length and consists of marine
terraces and a lowland morphologically. The altitude of this island is less than 80 m above
sea level (Fig. 8). On this island we spent two days, studying the marine terraces.

ffi ills

rE
H

lst terroce
2nd

H 3rd
4th

E Low Lond
-37 Height of
shore-l i ne
ond
surf oce

O 3km
Lower Unit

Fig.8 Geomorphological map of Santa Maria Island.

General geology
As far as our observation is concerned, the main part of the island is composed of two
sedimentary units: the lower one consists of massive mudstone and the upper, of stratified
tuffaceous sandstone and siltstone. According to the geological map preparcd by Muños
Cristi (192t6), the lower unit is Eocene in age, but after R. Martlnez (oral comunication) it
belongs to Miocenefüocene in age. The upper unit may be correlated with the Cañete
Formation.
A conspicuous unconformity between the upper and the lower units are exposed widely
on sea cliffs of the west coast and partially on sea cliffs of the east coast, Along the westem

-20-
sea cliffs, the bor-rndary is situated 20-40 m ligh above sea level in the northern part. below
sea level around Cape Cadenas in the central part, and up to the top of the 60 m high sea
cliff at 2km south of Cape Cadenas in the southern part. Along the eastern sea cliffs.
altitudes of the boundary are lower than 5 m above sea level Hence, east-rvest cross section
of the isla¡rd, as is shown in thc lower section of Fig.8, suggests a general tendency of
eastward dip of the unconformity.

Marine tenac'es
This island consists of a series of terraces and a fairly wide lorvland which occurs in thc
eastern half of'the island. The terraces carr be divided into fcrur levels, termed the First,
Second, Third, and Fourth terraces in descending order (Fig.8). Above the First terracc
surf'ace there exist scattered small hills (I'hoto 7). Judgrng from their topograplúcal leatures,
most terraces seem to be marinc in origin though our observation of terrace deposits was
extrerncly limited. The highest First terrace surface is likel.v- to be a depositional sr,trface of
the upper sedirnentary unit. and tlie scattered hills are most probabl¡ composed of the lowcr
unit.
In Fig. 8 altitudes of shorelines are indicated by nun.rerals at our measured locations as
well as altitudes of some flat terrace surfaces after topographic maps. In the ficld wc notlced
that eacli terrace surface inctnes to the east-northeast (see Phc¡t<¡ 7). The same trend of
inclination is also found from the distribution of altitudes of each shoreline. 1'his tendency is
obvior»ly seen at projected altitudes of the Third terrace shoreline as shown in the lower
section of Fig. 8. The altitudes of this shoreline range from 60 m to 34 m in height above sea
level i¡r a distance of 2 kni, so the degree of tilting since the age of the Third terrace is about
l" (I.3 per cent).
Although we have no rcliable information regarding agcs of thc terraces, the First tcrrace
is possibly correlative with the Cañete Surface in the Arauco Penilsula accorcling to their
altitudes and morphology.

5. Marine terraces of Mocha Island

Mocha Island, a spindle-shaped island of l3 km in length (Fig. 9),lies near the outer edge
of the continental shelf southwest off thc Arauco Peninsula. The axis of thc Ctrile Trench
lies about 90km off the island. Morphologically the island is composed of two units:
mountains capped by lúgher and middle terrace surfaces, and a lower terrace surface
surrounding the mountains (Fig. 10 and Photo 9). We investigated the lower terrace for two
days on thc island.

General geology
According to Tavera and Veyl (1958), this island consists of marine Tertiary rocks. of
which the Mioccne Navidad and Ranquil Formations strike NE-SW to NW-SE with westward
dip of 5-40o. These Miocene formations occupy the maln part of the island. The Plioccne
formation is also exposed along the southeastern coast of the island ilipping gently
westward.

Higher and Middle teruaces

The Higher and Middle terraces of 200 400 m in height are distinctly separated from the
l¡wer terrace lower than 33 m by a steep ancient sea cliff, which was formed
contemporaneously with the Lower terrace (Photo 9). The Higher terrace of 330-390 m in

-21-
 Pta. Pi edra Los Lobos

r9 :aa

_18
:9
Caleta
La Hacienda

Pta.
Anegadiza

4\
Pta de I

1
/ Caleta
Islas /I Derrumbe
/ §-4J ¡ N
V,l I\
uVn \
.t/
U Pta. Chales _t_-
I
l

:a /
0 4KM

Fig. 9 Topographical map of Mocha Island. Contour interval: 50 m.

height is fairly dissected, and is partly underlain by uriconsolidated 5 m thick deposits

consisting of round gravels and fine sand (Tavera and Veyl, 1958). The deposits and the
geomorphological features indicate that the Higher terrace is a marine terrace. This terrace
sudace tilts eastward at a gradient of 2-4 per cent. Most rivers dissecting the mountains run
eastward consequently to the inclination of the surface.
The Middle terrace is distributed separately on the mountains. An altitude of the
shoreline angle is about 250 m above sea level at the southern part of the island.

aa
 1
F-.l
Ptá.Piedra Los Lobos :-=
0
- 2
t- l_] 3

ñl
f,\'--_ |
1

5
Caleiá a'\,--.]
La HacLenda
@_r b
l'r r 7
t -.='1
íra >tt
l\-:1
I
i-.'- I

negadiza

4
Pta. de Caieta
Las Islas Derrumbe
N
'ofrí^
"l/ \
U
-f
I

4KM

,/\

Fig. l0 Geomorphologial map of Mocha Island.

1: Higher terrace. 2: Middle terrace. 3: Lower terrace.
4: Shoreline of the Lower terrace and its altitude (in meters
above sea level). 5: Sand dune and dune crest. 6: Rockfall
scarp and rockfall deposits. 7: Stream. 8: Generalized con-
tour line (interval: l00m) and altitude (in meters above sea
level). 9: Sampling locality for radiocarbon dating.

There is no strict evidence for the ages of the Higher and Middle terraces, but it is likely
that the higher one may be correlated to the Cañete Surface in the Arauco Peninsula.

-23-
 Lower terrace and its absolute age
The Lower terace fringes the island entirely along the coast. This terrace is subdivided
into several sublevels by small cliffs, whose bases are raised shorelines. The highest sublevel
develops on the east coast, and the highest raised shoreline reaches 33 m above sea level, near
Caleta La Hacienda (Fig. l0). Many elevated beach ridges are observed along the
northeastem coast (Photo 9). These raised shorelines and beach ridges indicate that the
island has been uplifted repeatedly.
The Lower terrace as a whole is a roc§ wave-cut platform underlain by thin beach
deposits of gravels, sand and shell fragments (Photo 10), and partly covered by sand dunes.
As shown in Fig. 10, the large elevated sea cliff has a number of rockfall sc¿rrps, some of
which appeared at the time of the 1960 great earthquake. These rockfaü deposits also cover
the Lower terrace surface at the foot of the sea cliff.
Radiocarbon dating for beach deposits (fossil shell fragments, Photo 10) from four
localities on the Lower terrace (Fig. 10) was carried on at the Laboratory of Dating,
Gakushuin University, Tokyo, and the results obtained are shown in Table 1.

Table l. Radiocarborl dates ol'shell lragments on

the Lorver terrace in Mocha lsland.

l-ocation Altitude Carbon-14 age

(shorvn in Fig. l0) (m above sea level (years B.P.) Code no.
)

I 25 3920190 GaK-3709
II 20 i9:l0tI 00 GaK-3710
III l1 259011 l0 GaK -3708
IV l0 I 990r80 GaK-37 I 1

The calculation is based on ttre l.ibby's l-Lali life of (l-14.5570 years. The sample
of present shellswasdated as "mode¡n", Dt-14=+1.75tl.0l';(caK l7 I2I

As is indicated in Fig. I l, the radiocarbon ages suggest that 1) the average rate of uplift is
about 55 cm/100 years since 4000 yean ago, and 2) üe age of the highest shoreline of the
Lower terrace (33 m higir above sea level) is about 6000 years ago if the average rate of
uplift is extraporated. Thus the obtained age of 6000 years is compatiblc with the
well-known stage of 6000 years ago, when the world-wide sea level rised up to nearly the
present sea level.
Therefore, it can be concluded that the general feature of the l,ower terrace had been
formed by marine abrasion up to 6000 years before under the rising sea level, and that since
then Mocha Island has been uplifted intermittently with average rate of 55 cm/l00 years
under nearly a stable sea level.

6. Change of sea level and ages of the marine terraces

From the foregoing description about the landforms of marine terraces and the height of
elevated shorelines, it is clear that the Arauco Peninsula and its surroundings have been
üsplaced by vertical crustal movements during the late Quaternary. Mocha Island, above all,
shows remarkable rate of uplift as revealed by the high altitude of the Holocene terrace.
Of interest to mention is a fact that Mocha Island has an evidence of the Frandrian
transgression reoorded by the dated Holocene terrace and the high sea cliff behind it,
although the island has been uplifted strongly. This means that tlrc rate of eustatic sea level

-24
 iddlelon ls.

ot2345678gto
thousond yeors B.P
Fig. l1 Rate ol upheaval in the Holocene in I,focha Island, Chile.
jlliddleton Island, Alaska and the Boso Peninsula, Japan.
Origirtal data 1br Middleton Island arc in Plafler and Rubin
(1967), for Mocha lsland are-iu this^paper. and for the
Boso Peninsuia are unpublished data of N. Yottekura.

rise is higtrcr than the rate of land uplift. Numerically, the forrner is known as abilut
I mi 100 years for 16,000 to 6,000 years ago, and the latte¡ is 0.55 mi 100 years.
Generally speaking, in an uplifting land, such a superiority of the rising rate of sea level
over the uplifting rate of land is necessary for coastal tcrrace building. Therefore, when wc
find a marine terrace ir uplifting lands, especially when we find thick manne terrace deposits
andior raised ria coast features. it is safely concluded that the terrace was built urlder a
circumstance of rising sea level, in other rvords, according to the theory of glacial eustasy,
under world-wide interglacial or interstadial circumstances
This line of thinking, which was born so far at the time of coastal terrace study in
Shikoku, southwestem Japan by Yoshikawa, Kaizuka and Ota (1964 a, b, 1968), may be
applicable to the Pleistocene terraces of the Las Nochas, Buena Esperanza and Cañete
tcrraces as well as the Holocene tcrrace surface in Mocha Island. Ilence these Pleistocene
surfaces seem to l-mve been formed in world-widc interglacials or interstadials. The Cañete
Surface, the youngcst and widest one among them may be of the last great interglacial or the
Sangamon interylacial age. about 100,000 years ago. This estimation has not so much
discrepancy with an estimatcd age of 60,000-80,000 yean ago for the Higher terrace of
Mocha Island, which seems to be correlated to the Cañete Surface as mentioned already. The
age estimation of the l{igher terrace in Mocha Island is based on the assumption that the

-25-
mean rate of uplift obtained from the Lower terrace,55 cm/100 years, wasresponsible for
elevating the Higher terrace to the present altitu<les, 330-390 m above sea level.
Accordingly the Buena Esperanza Surface and the Las Nochas Surface should have been
formed in sorne of interglacials or intcntadials preceding the last great ilterglacial, and thc
lower terraces of the Arauco Peni,nsula have been fbrmed itr some n-rinor intcrglacials
preceding the last glacial or some iltterstadials in thc last glacial age.

7. Summary of Pleistocene and Holocene vertical crustal moYements

It is concluded from our rcsults that the dominant mode of crustal movements i¡ the
Arauco Peninsula and its surroundings is up - and down-warping or tilting. Any active fault,
wlrich cut the Quaternary terrace surfaces, is not recognizedby us, although the presence of
a fault or faults have been presumed (e.g., Saint Amand, 1961).

A Bo
Coñete Surfoce,
ils shoreline ond ¡sobose
o t
Shoreline of Bueno
-a. -9.
Esperonzo Surfoce rO (-o
Shoreline of
Los = =
Nochos Surfoce (D 3
(D
ct
u @

f, o
o-
o
E g
o <
:
(O

E =
3
= o
o
3
o
E

m
a
-0
n
TTI

@Z
CI\
,0>
C)
zt
iló
m m
-l
m
s
zfoo o ln
Fig. 12 Isobase or restored contour lines of the Cañete Surface (A)
and longitudinal projection of shoreline heights (B).

-26-
 In the left half of Fig. 12. isobase contours of the Cañete Surface are shown. Strictly
speaking, the isobase contours do not represent an exact mode of crustal movemetlts, but
represent restored cor-rtour iines of thc Cañete Surface. Roughly speaking. however the
restored contour lines are regardecl as lines of equal uplitl, since the initial form of the
Cañete Surface. which was a cut-and-built submarinc platform, might have been as small as
less than 20 meten irt relief.
In the right half of Fig. 12, longitudinal projected profiles of the higher and middle
terracc shoreli.nes are shown. From this figure it can be said that the mode of tectonic
deformation of the peninsula is just like as a saddle of horse. In othcr words the mode is
contposed of two series of warping. one nearly N-S, the other nearly NW'SE itt trend.
As for N-S trending warping, axes of uplitt are situated along the western coast of the
penilsula and in tl-re Nahuelbuta, and an axis of subsidence is nearly in the centtal paft of
the pelinsula. This subsiding axis seeurs to extend north to Arauco Bay, and s<¡uth to the
shelf area between Mocha Island and the c:ontinetit. The eastward tilting of Santa Maria and
Mocha Islands also indicates that thc uplift axis is situated on the outcr rim of the
continental shelf zone, and the subsiüng axis on the coast side'
As for the NW-SE trending rvarping, an axis of uplift is given by' the line that cronnects the
highest parts of the higher and the ¡niddle terrace shorelines. The southeastward and
northwestward extensioni of this line pass the highest pañ of the Nahuelbuta, and nearly thc
northwestern mountains of the pe¡insula respeclively, as is previously noted'
It is worthy of note that thc mode of shoreli¡re defonlalions of the higher and middle
terraces borderilg the Nahuelbuta is closelt relatcd to the heiglrts of the Nahuelbrrta

of the highest poinf

Heiqht
' in Zx?km squore in Nohuelbuto lvlls

¡ Height of shore-line reod by oliimeter

o Heighl ol shore-line reod on mop

.--\\o Los Nochos Surfoce

Bueno Esperonzo S

Coñefe S.

- l3 Projected profile of the Nahuclbuta

Fig. Mountains and heights
of sltorelines to 'longitudinal plane.
11
Mountains, as shown in Fig. 13. The longitudinally projected crest line of the mountains
seems to be an amplified shape of the deformed shorelines of the higher and middle terraces
Accordingly, it is concluded that the NW-SE trending warping is dated back far older than
the time when those shorelines were formed. Probably they have continued since the erosion
surfaces had been formed in the Nahuelbuta.
The NW'SE trending axis of uplift must have created the Arauco Peninsula as an
exceptional protrusion along the Chilean coast. If there had been no NW-SE uplift in this
area, the land of the peninsula should have been a part of the submarine continental shelf
zone.
The above-mentioned saddle-shaped deformation in the Arauco Peninsula is consistent
with the distribution of the Plicane Tubul Formation, which is seen only in the downwarped
of the northern peninsula (Fig.6). Thus, the saddle-shaped deformation may have continued
since the Pliocene Epoch. This mode of deformation, however, is inconsistent with the
structure of the Miocene and Eocene forrnations, which dip westward in general in the
peninsula. Therefore it is expected that the change of tectonic mode happened in the
Püocene Epoch. The change of tectonic mode is also found in Mocha Island, where the
Miocene strata dip westward whfe the Pleistocene terraces dip eastward.

TV. RELATIONSHIP BETWEEN RECENT DEFORMATIONS ASSOCIATED WITH

GREAT EARTHQUAKES AND QUATERNARY CRUSTAL DEFORMATIONS

1. Tectonic deformations associated with the 1960 Chilean great earthquake

The 1960 great earthquake sequence started on May 21, 1960 with an intensive
foreshock, whose epicenter was located in the Arauco Peninsula. The main shock (Richter
magnitude of about 8.5) occurred on May 22, 1960. Its epicenter was lqcated in the
continental slope off Mocha Island (Fig. 14). The main strock was followed by after shock
series. The aftershock area w¿ls very long along the continental margin, from 36.5"S to 48oS,
and mainly between the trench and the volcanic front in east-west extension (Saint Amand,
r961).
The 1960 great earthquake was, according to Plafker and Savage (1970), accompanied by
up - and down-warping along the continental margin between 37"S and 46o5, and from the
trench to the volcanic front. The extent of the deformation coincided with the aftershock
area. The regional vertical displacement was characterizedby a broad asymmetric warping
elongated parallel to the trench without any significant faults on land. A marked uplift
occurred in the continental shelf zone and probably much or all of the continental slope as
indicated by tsunami generation. The vertical upward displacement at the tsunami source
area is estimated at 5.7 to 10 m on an average from the tsunami energy (Hatori, 1966). The
maximum observed uplift on land was 5.7 m on Guamblin Island (43.5"5) located close to
the outer edge of the continental shelf. On the other hand, a remarkable subsidence occurred
in the Coast Range. The maximum measured subsidence was2.7 m near Valdivia. Relatively
minor uplift less than I m was observed along the western margin of the Andes Mountains
This asymmetric warping is diüsible into two sectors in profile by an axis of the maximum
subsidence which lies nearly along the crest of the Coast Range. To the west of the axis, it is
characterized by a steep continentward tilting, whereas to the east it shows a gentle seaward
tilting. The width of a steep continentwarrrl tilting zone, namely the distance from the inner
wall of the trench to the axis of the maximum subsidence, is about 150 km.

-28
 .--- - 20ñ

_\ l./_
+ ,, l\
I

2 1t(
''' lr'
I
\
t, +¡ I
I
/ l\.
.//
ii.o/
L:/.
)",?''.,
tA./)

tlh
I'r .!)
Ir*
IL
t..
o 2'
@

\, ll
\\ tt ir
''
I::: .
-- :oo -- 10
I
j.:'
N,I UCO

.Í l.

Fig. l4 Seismic crustal movements accompanied wlth the 1835

Concepción earthquake and ihe 1960 earthquake.
1 :Epicenter of thc foreshock of 1960 earthquake. 2:Epicen-

iL,"":,'i:,ff :: j:l;i-":.'"Hr'"i,!i"l,iTn;;.'',tTo"T'J,l
quake (after Plaiker and Savage, 1970). 4:Leveling linc
along which vertical dispiacements accornpanied with the
1960 earthquake were deterrnined. 5:Isobase contour show-
ing approxi.mate amount of land-level change, in mete¡s.
Dotted where inf'erred (after Plafker and Savage, l97O).
6:Amount of vertical displacement acccmpanied with the
1835 earthquake (after Darwin, 1851 and Lomnitz, l97O).
7:Isobase contour showing approximate amount of land-
level change, in meters. 8:Volcano. 9:Quatenrary deposits of
the Central Valley. 10:Approximate outer edge of contlneu-
tal shelf, depth in Íleters.

As seen in Fig. 14, which prepared after Plafker and Savage (1970) as for the
displacement accompanied with the 1960 carthquake, not only Moctra Island but also the
Arauco Peninsula except for its northern part was upLifted. The amount of uplift was
relatively large in the southwestern part of the Arauco Peninsula and Mocha lsland whcre the
uplift was more than I meter. On the contrary, the Nahuelbuta Mountains especially its
southern part, probably subsided as shown in Fig. 14, though any evidence of displacements
has not been recorded for the Mountains. Ttre northern parts of boththe AraucoPeninsula

10
and the Nahuelbuta were not affected by the earthquake-related crustal movements in 1960.
As will be mentioned in the succeeding sections it has been noted in general that the
vertical co-earthquake uplift or subsidence is followed by the opposite recovery motion
during the aftershock sequence. However as for the 1960 earthquake sequence, any
evidence of the post-earthquake recovery motion opposite to the sudden co-earthquake
displacement has never been noticed as far as our knowledge is concerned.
Such terms as earthquake-related, pre-earthquake, co-earthquake post-earthquake, and
inter-earthquake, which are used hereinafter in this paper, are designated as illustrated in Fig.
15.

Fig. l5
Schematic diagram of earth-
quake-relatcd and intcr-earth-
quake displacements-
c: co-earthqnake displacement.
pr: pre-earthquake dis placement.
pj post-earthquake displacement.
pr + C * /, = e:earthquake-related
displacement.
i : intercarthq uake drsplacement.
7: recurrence interval.
t'r 7pr:duration of pre+arthquake
dis pl aceme nt.
I t ?"p:duration of post-eart hquake
Eorthquoke Eorthquoke recover)'.
R = pr + c - p - ilT:mean rate of
vertical displace ment.

2. Previous evidence of tectonic deformations associated with great earthquakes in

historical time
Lomnitz (1970) made a catalogue of great Chilean earthquakes greater than 7.5 in
magnitude for the pedod of 1562 1953. These historical data indicate that major shacks
seem to have occurred periodically in almost the same regions, such as the regions off
Valparaiso, off Concepción and off Valdivia. Many great earthquakes accompanying
tsunamis or vertical crustal movements were located off and near the coast Besides these
offshore earthquakes some great ones occurred on land between Santiago and Concepción.
Crustal movements associated with these on-land earthquakes were thought to be connected
with uplift of the Coast Range, but no surface fault has been obsewed (Lomnitz, 1970). To
the south of Concepción, on the other hand, great earthquakes occurred off coast
accompanying vertical uplift along the coast, instead of along the Coast Range. The area
between Santiago and Concepción was thus regarded by Lomnitz(1970) as a tectonicunit
different from the area between Concepción and the Chile Rise to the south.
In Fig. 16, time-space distribution of great offshore earthquakes (magnitude 8 or more)
are illustrated according to Lomnitz (1970) and Kelleher (1972). As seen in this figure, in
and just north of the region of this study (37"-38"5), the so-called Concepeión earth-
quakes occurred periodically in years 1570, 1657 , l'151 and 1835 with the mean
reccurence period of 92 years.In the studied region and its southern extension, the so-called
Valdivia earthquakes occurred in 1575, 1737 and 1837. The 1960 great earthquake was a
succeeding one of this sequence. Thus the mean reccurrence period for these four

*30-
 r960 (8.5)
t928(S.4)

t835 (8- I /4) t837 ( I +)

_____-Uiúdnzj
tTgO tEt/1t 173717r2-Al

t6E7'( I )

r 600
(Brá
Ezoie_s,2i s?5

30ó 34o 3go 4zo 46" S

Fig. l6 Tinte-space diagrartt -ol ('hilean major offshorc carthquakes

bet weclr 30' and 4Ér'S.
Tlte diagrarn is basccl on Kcllchcr (1971), ttre original data
oi lvhich by Lornnitz (1970). S,¡lid horizont¡l ltne seBrlelrts
reprcscnt the latcral extenf of relatively ivell-detennined
estilnates of rupture zoltes. Dashed lit.ies are less reliablc
ones. Solid circles represent evettts lor u'hich no rupture
zorre rvas estirriated. Nun.rbers in tlre parentheses are rnagni-
tudes of earthquakes by Lomtiitz (1970).

earthquakes in the Arauco-Valdivia region is 128 years. It is said that tlie rnagnitude and the
areal extension of the 1960 earthcluake was larger than those of the preceded two great
earthquakes, but was nearly as same as those of the 1575 earthqualie. It is also noticed that
there seems to be a tcndcncy of time-related occurrence between the Concepción and
Valdivia earthquakes, as the cases of 1570 1575 and lB35 1837. Judging from its great
n-ragnitude and wide areal extension, the 1960 great earlhquake seems to have occurrerl over
the both focal regions at a time.
The above-lnentioned historical data show that ttrc mcan recurrerrce interval of these
great of}'shore earthquakcs i¡ these regions is nearly a centurlr fbr recent 400 years, thougit
the interval of Valdivia earthquakcs is a little longer than that of Concepción earthquakes.
Vertical displacements associated with ttrese earthquakes are partly documented on
coasts since the 1751 carthquake (Lomnitz. 1970). At the time of the 1751 Concepción
earthcluake an uplift of the offshore islands was recorded, but thc amounts ¿rre ttnknown, As
for the earthquakes of 1835 and 1837, Darwin (1851) described significant vertical crustal
displacements on coasts.
The amounts of uplift associated with the 1835 Concepción earthquake along the coast
of the Arauco Peninsula and ofl'shore islands are givcn in Fig. l4 by numerals according to

3l -
Darwin (185 l) and Lomnitz (1970). Isobase contour lines for the upüft in Fig. 14 are drawn
based on those amounts. The maximum uplift was reported from the northern points of
Santa María Island as 10 feet. It is noteworthy that both Santa María tsland and the city of
Concepci'on lost a part of their uplift by the post-earthquake subsidence in the course of
some weeks after the earthquake (Darwin, 185 l). According to Darwin and others, Lomnitz
(1970) wrote that most of tfie uplift accompanied with the 1835 earthquake had
disappeared through a slow recovery within the period of the aftershock sequence.
Accompanying with the 1837 Valdivia earthquake 2.4m uplift was observed at Lemus
Island (44.5'5) on the ocean side of the Chonos Archipelago, but no other observations on
coastal changes appear to exist.

3. Accumulation of seismic deformations in the late Quaternary

If we sum up both the vertical displacements associated with the 1835 and 1960
earthquakes affected to the Arauco region (Fig. 14) and compare it with the late Quatemary
displacements of the same area (left half of Fig. 12), it is noticed that the eastward tilting in
the continental shelf zone including the Arauco Peninsula is the common characteristics to
both the short-term earthquake-related deformation and the long-term late Quatemary
deformation, whfe in the Coast Range the two deformations are not simlar to each other,
but rather reversed. Th¿t is, the earthquake-related deformation in the Coast Range is a
down-warping, but the late Quaternary deformation in the same area is, on the contrary, an
up-warping as stated previously. The same fact has been pointed out by Plafker (1972) trt
southem Chile and Alaska in connection with the 1960 Chilean and 1964 Alaskan
earthquake-related tectonic deformations, and also has been noticed by Yoshikawa (1968,
1970) in Shikoku, southwestern Japan in connection with the 1946 Nankai earthquake.
Herein let us examine a problem of acctmrulation of short-term displacements
quantitatively, at first for vertical displacements and next for tilting.

A ccumula tio nof v ert ícnl displnceme nt s

The amounts of vertical displacements used in the following discussion are given in Table
2, based on data obtained by previous literatures and by the present study.
The column (1) in Table 2 shows amounts of earthquake-rclated uplift for the latest three
great offshore earthquakes in thrce regions; Mocha Island, tfie western part of the Arauco
Peninsula and Santa María lsland. Here p signifies the amount of poat-earthquake recovery as
is given in Fig. 15. The amounts for 1960 earthquake \ñ'ere measr¡red in 1968 (Plafker and
Savage, 1970), so these are amounts not only for pre-and co-earthquake displacements, but
also contain post-earthquake displacements.
The column (2) shows mean rates of earthquake-related uplifts in historical age which
were obtained by summing up the three amounts of earthquake-related upüft in (1) and then
by dividing them by 200 years, nearly two recurrence periods of great earthquakes from the
1737 and lT5l earthquakes to the 1960 earthquake. However, if the displacements of the
775L, L73'1 and 1657 earthauqkaes were small enough, we should use four recurrence
periods, 400 years, instead of 200 years, because as mentioned above the 1575 earthquake
was as large as the 196O earthquake, and miglrt have been accompanied by significant
displacements (cf. Fig. 16). In Tables 2 and 3, the former case (200 years) is used for
calculation.
The column (3) shows altitudes of late Quaternary marine terraces in the three regions
The column (4) shows mean rates of uplifts during the late Quaternary, which were

-32-
 Table 2. (iornparison of meau rates of vertical displacements betwecn historical age an'J
late Quaternar).

(l) l.\ (:l) (.1 )

Earthquake- lvlean rate of Altitude of late lvfean rate r¡f
related uplilt earthq uak c- Quaternarl, terrace uplilt duling
(m) relatetl uplilt (m) the late
in histcrrical age Quaternary
(ml I 00y') (mi100y)
1,0-1.8 (1960) iHighest part
/,
It'locha Island 0('¡) (1837) r0.8-1.2)-{ -1.r l ol the f ower i
0.5 5
06-p {lEi5) 2
i tefra r'e : i
(p(0.6) i6000t'. B P.
Western part of 0-r .3+ l 196U) r'Cañete Surface:i
the A rauccr o('r) (r 837) (o.6-t 6,J-: .;,, I
--' \ 0.2 5
Peninsula t-2-p ll 83s) ' r *ios' B.[,. I
{p<l-l)
0 (1q60) ¡F:rst terrace.,
!2
Santa Nfaria 0 C) 1837)
( (1 .2-t .» - S0 ', lxl0:v llP.l
Island 2.,1--1 .0-¿r t18,15)
Íp<2.4-3.0)

obtained by dividing the altitudes listed in (3) by each age of the terrace. Strictly speaking.
fbr getting true mean rates of uplift, the height of sea levei at the time of terrace formation
is necessary' to be given. We do not know the relative heiglrts of the former sea levels. but we
know that they are srnall relatively to the altitudes of the late Quaternary nrarine terraces in
this region. Thus the amounts in (3) can be used for the rates of uplifting in á rough
approximation. The more uncertailty lies on the age estimation of the Cañete Surface and
the Fint terrace of Santa Maria Island.
From (2) and (4) in Table 2, it is apparent that the mean rates of earlhquake-related
uplifts in the historical time are larger than those <¡f the late Quaternary uplift,if pissmall
enough. Namely, the mean rates ix (2) are much larger than those in (a); 1-2 times for
Mocha Island. 2-6 times for the western Arauco Peninsula. and 15 20 times for Santa
María lsland. These difl'erences between amounts in (2) and (4) become smaller when
amouts p arc taken into consideration, The amount of ¡r is especially of importance for the
case of Santa María Island, because it is known that the co-earth<1uake uplift (c in Fig. 15)
on the island mostly disappeared through postearthquake recovery (¡r) whithin the period of
the after shock sequence as cited before. For the cases of N{ocha Island a¡rd the western
Arauco Peninsula the amounts of p are unknown. Thus the difference between amounts in
(2) and (4) may be due partially to unceftainty in amounts in (l) and.(3), and partiallyto
the existance of inter-earthquake subsidence (amount i in I,ig. 15).
In the case of Muroto Promontry, southeastern Slúkoku, Japan, it has been known that
the radio (p +ilpr +c) of tfie total recovery (p + i) to the pre- and co-earthquake uplift
(pr + c; pr rnay be small) in one recurrence time is alS; ¡ otlier words, the ratio of the
residual displacement to the pre- and co-earthquak€ ones (.pr+c-p-ilpr +c) is l/5
(Yoshikawa et al., 1964 a, b,). Muroto Promontory has been thc only place in the world
where these ratios were known in success.
Then, if we assume for the cases of Mocha Island and the western part of the Arauco
Peninsula that:

JJ-
l) amounts p are negligibly small, i.e., amounts in (2) represent ntean amounts of pr + c,
2) amounts in (2) and (4) are reliable enough,
3) the mean rates of uplift for the short-term rvhich are shown in (2) are equal to the mean
rates for the long-tenn shown in (4), and
4) amounts pr are negligibly small,
then we can obtain the ratio of the residual displacement to the co-earthquake one
(pr+c-p'ilprtc-c -l/c) through dividing amounts in (a) by amounts in (2). The
results obtained are: 1-1/2 for Mocha Island and ll2-ll6 for the westenl Arauco Peninsula.

A c cu ntula t io n o f e ar t h c¡ uake-r e ln t ed ti ltitlg

The amounts for tiltings are given in Table 3, based on data from previous literatures and
our measurements and calculatio ns.
Tablc 3, ('omparison between mean rate of historical earthquake-relatcd tilting and that
of the late Quaternary tilting.

(r) I (2) (:r .¡

(.4)
Earthquake-related Mean rate r>f F.astward inclination N{can rate
eastu'ard tilting earthquake- of late Quaternary of eastward
related east- terrace surfaoe tilting dur-
ward tilting rng the late
rn historical Quaternary
age (per 100-v )
(per 100y)
4-5xt0-s (19ó0)
I\{ocha Island 0('r) (1837) I ^-2 ..5 x | 0'5 2-4x10-2 tcrrace; 3-6x 10'5
? (r 83s) ,6-8x10ay
Western part of 5-(r xI 0-s (1e60)
the Arauco 0 ('r) (r 837) -l-4x I 0 s 0.5- 1xl0's
Peninsula I -2x 10-s (183-5 )

0 ( l e60)
Santa Maria iThud \l
0 (?) (l 837) l-'1,5x10's 2xl0-2 I terrace: 4xl 0
Island
2-3x10's (183s) \s*loayi

The column (1) in Table 3 shows amounts of earthquake-related eastward tilting for the
three regions as ¡n the case of Table 2. As for the amounts of 1835 earthquake, some
corrections must be necessary in connection with amounts p in Table 2, but we have no data
for these corrections, hence we tentatively regard them as the approximate amounts of
earthquake-related tilting.
The column (2) shows amounts of mean rates of earthquake-related eastward tilting in
the historical age, which were obtained by the same procedure as in the case of Table 2.
The column (3) shows amounts of eastward inclinations of the late Quaternary terrace
surfaces which \i¡ere measured in this study.
The column (4) shows amounts of mean rates of eastward tilting during the late
Quatemary, which were calculated by dividing the inclinations in (3) by each age of terrace.
Herein the given age for the Third teÍrace in Santa María Island is very tentative.
Comparing (2) and (4) in Table 3, it can be said that the meafl rates of (2) and (4) for
mch region are same in a rough approximation. As to Mocha and Santa Maúa Islands
amounts of (2) are smaller than those of (4), contrarily to the cases in Table 2.

-34-
 A pntposal of two seismo-teck¡nit' zones
Through the above-mentioned comparisons of uplift and tilting between the short term
historical age and the long-term late Quaternary period, it is concluded that the displacement
or defbrmation of the earth's crust in the continental shelf zone has been accumulated for
the past 103-l0s yearsatthesarnerateinorderandrviththesamemodeof defornlationas
those in historical age. In other words, the continental shelf zone seems to have been
cleformed for at le¿rst 105 yean by the accumulation of co-earthquake deformations Thus,
the continental shelf zone in this region can be understood as a zone rvhcrc secular
defomration through recent geologic time has been co-earthquake deformation as revealed in
historical age. In this sence, we designate such tectonic zone as the continental shelf zone irt
tlús regron a zone of co-earthqttake defontwtio¡t. The accumulation of co-earthquake
deformation should be explained in connection rvith the elastic rebound theory for
earthquake, on which we will devotc a discussion irr a separate paper.
On the other hand, in the Nahuelbtrta Mountains the mode of acute co-earthqnake
defonnation is a downwarping. and it is reverse to the mocle of long-term Quarternary
deformation, whicl-r is an uprvarping as mentioned in the foregoing sections. Consequently, it
is quite likely that in the mountains the interearthqualte deformation lras a mo«le similar to
the Quatcrnary dcfcrrmation of upwarping. and anrounts of this itrter-earthquake upwarping
is superior to those of co-earthquake dou'nrvarping. Thi» the resultant upwarping
deformation cluring an earthquake recurrence time rvould have accuntulated tbr long titrte so
as to nrake the Nahuelbuta Mountains. If this 1i¡re of thinking is correct, the term a zone o,f
inter-earthqttake deformatioil can be applied to the nlountains. a part of the Coast Range of
central Chile.

V. SUMMARY

This work has been carried on as a study on the Quaternary crustal movement$ in the
outer part of the outer arc near a trenctr, where great earthquakes have frequently been
experienced. The main results obtained in and near the Arauco Peninsula, central Chile, are
as follows.
1) In the Nahuelbuta Morxrtains there are low relief surfaces formed during the late
Tertiary probably under subaerial denudation. The features of the surfaces indicate that an
upwarping has occurred in the mountains with the N-S and NW-SE trending axes of uplift
during the Quaternary.
2) The topography of the Aratrco Peninsula and the westem border of Nahuelbuta
Mountairts is subdivided into five geomorphic surfaces: the Las Nochas, Buena Esperanza,
Cañete, Lower terraces, and Holocene lowland surfaces, of which the former three are
rn¿rine terraces, and the later two are partly alluvial and partly marine in origin. Our field
studies of their elevated coastal features and terrace deposits lead us to á conclusion that the
formation of each surface was made under a marine transgression, probably caused by the
eustatic rise of sea level during either an inte¡glacial, or an interstadial and a postglacial age.
The widest Cañete Surface is presumably ascribed to the last great interglacial age.
3) Four Pleistocene marine terraces are distinguished on Santa María Island and two
terraces on Mocha Island- Thei¡ eastward tiltings in the late Qr.raternary are revealed by
altitudes ofelevated shorelines and also by inclinations ofthe terrace surfaces.
The radiocarbon datings of Holocene terrace deposits on Mocha Island give a mean rate of

-35-
uplift for the last several thousands years which seems to be one of
as 55 cm per 100 years
the most rapid rate of land uplift in the world except for the rates reported from the regions
which had been coversed by ice sheets in the Pleistocene.
4) The mode of Quatemary crustal movements in and near the Arauco Peninsula is a
tilting or warping without any faulting on land. The continental shelf zone, including the
Arauco Peninsula, Santa Maria and Mocha Islands, has been tilted to the east that is,
landwa¡d as a whole, and partially has been upwarped with a NW-SE trending axis in the
central part of the Arauco Peninsula. Our morphological studies disclose that this upwarping
became active at latest after the formation of erosional surfaces in the Nahuelbuta
Mountains and has continued to after the formation of the Cañete Surface.
The eastward tilting of the shelf zone in the Quaternary Period is markedly different from
the mode of crustal movements in the Paleogene and Neogene Periods in which a predomi
nantly westward tilting indicated by the structure of Eocene and Miocene strata. Inference is
that a change occurred in the Pliocene Epoch in the mode of crustal movement.
5) This region has been attacked nearly once in every century by great earthquakes of
magnitude 8 or more such as the so,called Valdivia and Concepción earthquakes. No
substantial difference is found between the mode of regional, acute co'earthquake
deformation and that of the long-term Quaternary eastward tilting, so far as the shelf zone is
concBrned. Besides in the shelf zone the mean rate of the co+arthquake tilting in the
historical age is of as the same order as that in the late Quaternary. Therefore, the term ¿
zone of co-earthquake deformation can be applied to the continental shelf zone in this
region.
In the Nahuelbuta Mountains, however, the mode of long-term Quatemary deformation is
rather reversed to the co-earthquake deformation. Thus it is very likely that the
inter-earthquake deformation, instead of the co-earthquake deformation, has been accumu-
lated through the Quatemary in the Nahuelbuta Mountains. So we give the term a zone of
inter<arthquake deformation to the Nahuelbuta Mountains
In the strelf zone, the ratios of the ¡esidual displacement to the co-earthquake
displacement in an earthquake recurrence time were estimated under some assufnptions at
lll*L12 for Mocha Island, and ll2-l16 for the western Arauco Peninsula. This implies that
the ratio varies from place to place even in not a so great distance'

ACKNOWLEDGMENTS

We wish to express our sincere thanks to the members of the Research Group for
Quaternary Tectonic Map of Japan, especially to Prof. Torao Yoshikawa, Department of
Geography, the University of To§o and Dr. Arata Sugimura, Department of Geology, the
University of To§o for their eocouragements and discussions regarding this research. We are
grateful to Prof. Rubén Martinez P., Escuela de Geologia, Universidad de Chile, Prof. Cinna
Lomnitz, Insütuto de Geofisica, Universidad Nacional Autonoma de Mexico, and Dr. Geoge
Plafker, U.S. Geological Suwey for their helpful discussions and suggestions We are also
indebted to Dr. Yuzo Kitazawa, Department of Biology, To§o Metropolitan University and
Dr. Kazuyoshi Kurosa for identifying the insect larva and suggesting its ecological behavior.
Thanks are also due to the Chilean people for their generous assistance and hospitality in
carrying out our field work.
This research has been financially supported by a Grant in Aid for Fundamental Scientific
Research provided by Japanese Ministry of Education, which is also gratfully acknowledged.

-36-
 REFERENCES

Ando, Ir{. (1971) A fault-origin model of thc great Kanto eafihqu?ke of 1923 as deduced
from geodetic data. Bull, E'arthq. Res. Inst. ,49 , 19 -32.
Brüggen. J. ( 1950) FundornentL¡.t de la gcología tle Chile. lnst. Geogr. trIiJitar. 374p.
Santiago.
Darwin. C. (185 !) Geologit'al t¡bservations t¡n cot'al reef's, t:olc{lrtic islands dtld ort SoLtth
ilrnerica, being the geolog.v oJ the t:o.1.,oge ctf the Beagle. under the cotntnand of Cdptdin
Fitz Rq,. R.N., during the ¡'ears 1832 ro 1836. 168p. London.
Fisher, R. and Raitt, W. (1962) Topography and structLlre of the Peru{hile trencir.
Deep-Seo Res., 9, 423 443.
Fitch, T. J. and Scholz, C. H. (191 | I A mec:hanism for underthnuting in solrthrvest Japan. a
model for convergent plate itrteractton..lour. Geopltl'.r. Res ..76. 260 i292. ,1

García, F. (1968) Estratigraña del tercia¡io de Chile central. in Cecioni. G. ed., E'l Tercioriu
de Cltile, zona central, Sociedad Geológica de Chile,25-51 .

Hatori. T. (1966) Vertical displacentent in a tsunami source area and the topography of the
sea bottonr. Bull. Earthq. Res. Inst.. 44. 1449 -1464.
I{ayes, D. E. (1966) A geophysical lnrcstigation on tl're Peru{-.liilc t¡¿rlch. ,\larine Gac¡1.,4,
309 351.
Kelleher. J. A.(,1912) Rupture zone of large South American earthquakes and sonie
predictions. Jottr. Qeoplt.l.,,r. R e.r.. 7 7 .2087
-2103 .

Lomnitz. C. (1962)On the Andean structurc.Jottr. Geopl¿,r'.i R¿s.,67,,35 1 363.

Lomnitz, C. (1970)Major earthquakes and tunamis in Chile during the period 1553 to 1955.
Geol. Rundschau, 59,9 38 960.
Martínez, R. (.l968a) Foraminlferos pliocénicos de Clhile central ll, Eclad y paleoecología clc
la formación'I'ubul. in Cecir.rni, G. ed.. El Tcrt'iario de Chile, zone central. Sociedad
Geológica de Clhile, I .5 5 I 65 .

It{a¡tínez. R- (1968b) Foraminife¡os y ev6lución de 1a Linea de cosla holocénica en la zona de

Concepción. in Cecioni. G. ed., L'l Tercíario de Cltile, zona centrdl. Sociedad Geológir:a de
Chile,2l l 2-58.
Muños Cristi, J. (1946) Estado actual de1 conocimiento sobre la geologra de la proüncia de
Arauco. Anal. Irac. Cien. h'is. )t frlatem., Linty. de Chile,n.3,30, 63.
Muños Cristi. J" (1956) Chile. in Jenks, W. F. ed,, Ilandbook oJ sottth American geolog¡,.
Geol. Soc. An.rerica, Memoir 65. 187 -)14,
Muños Cristi, J. (1968) Contribución al conocimiento geológico dc la región situacla al sur cle
Arauco y participación de material volsinico en los sedimeutos eocenos. in Cecionj , G.
ed..El Terciario de Chile. z<¡na central. Sociedad Geológica de Chile.63-93.
Plafker, G. (1969) Tcctonics of the March 27,1964 Alaska earthquake. Ll. S. Geot. Survey
ProJ'. Pape r, 54 3 -1, 1 7 4 p.
Plafker, G. (1972) Alaska¡ earthquake of 1964 and Chilean earthquakes of t960:
lmplications for arc teclonics. Jttur. Gt'<tph¡,s. Res., 77 ,gOl -g25.
Plafker, G. and Rubin, I!1. (1967) Vertical teclonic displacement in south-central Alaska
during and prior to the great 1964 earthquake. Jour. Geosci. Osaku City univ..l0. art.
t-7 .53'-66.
Plafker. G. and Savage. i. C. (1970) Mechanism of the Chilcan carthquakes of Nlay 2t and

-37 -
 22, 1960. Geol. Soc. Americu, Bull-,81, 1001 -103 1.
Saint Amand, P. (196i) Observaciones e interpretación delos terremotos Chi.lenos de 1960.
Cr¡mmunit'a. b'sc. Geol., Univ. de Chile, Año 1, n.2, 1-54.
Santo, T. (1969) Characteristics of seisn.ricity in SouthAmerica. Bull. Earthq. Res. Inst.,47,
635 -67 2 .
Scholl, D. W.. von Huene, R., and Ridlon, J. B. (1968) Spreading of the ocean floor;
undeformed sediments in the Peru-Chile trench. Science. 159 869-871 .

Scholl, D.W., Christensen, M. N., von Huene. R,, and N{arlow, M. S. (1970) Peru-Chile
trench; sediments and sea-floorspreading. Geol. Soc. Ameríca, Bull .,81 , 1339 1360.
Stiefel, J. (1968) Sedimentological reconnaissance of some Quaternary deposits of central
and southern Chile. in Morrison. R. B. and Wright, Jr., H. E. eds.,Means ol'correlation oJ'
Quaternary sttccessions, Univ. Utah Press.559 576.
Sugimura, A. (1960) Zonal arrangement of some geophysical and petrological features i¡
Japan and its environs. Jour. Fac. Sci., Utzir,. Tok-y"o, Sect.2, 12, 133 153.
Sugimura, A, and Naruse. Y. (1954) Change in sea level, seismic upheavals, and coastal
terraces in the southern Kanto region. Japan (l'l .Jap. Jour. Geol. Geogr.,74,l0l 113.
Tavera, J. J. (1942) Contribución al estudio de la estratigrafiay paleontologia del Terciario
de Arauco. Att. Primer Congr. Part. Ing. de ltíitt' ¡' Geol.,2, S80 632. Santiago-
Tavera, J. J. an¿ Veyl, C. O. (1958) Reconocirniento geológico de la Isla Mocha, 1955.
Publica. Dept. Geol., (Jnít,. de Chile ,n. 1l, 157 1'7'7 .
Yoneku¡a, N. (1968) Geomorphic development and mode of crustal movement on the south
coast of the Kü Peni¡suia, southwestern Japan. lour. Geogr., Tokyo, 77, 1-23 (in
Japa:rese with English abstract).
Yoshikawa, T. (1968) Seismic crustal deformation and its relation to Quaternary tectonic
movement on the Pacific coast of Southwest Japan. Daiyortki'kenkyu (The.Quaterndry
Research),7 ,157 - 170 (in Japanese with English abstract)'
Yoshikawa, T. (1970) On the relations between Quaternary tectonic movement and seisrnic
crustal deformation in Japan. Bull. Dept. Geogr. Univ. Tokyo,2, | 24.
Yoshikawa, T., Kaizuka, S., and Ota, Y. (1964a) Mode of crustal movement in
late
Quaternary on the southeast coast of Shikoku, southwestern Japan. Geogr. Rev' Japan,
37 ,627 -648 (in Japanese with English abstract).
Yoslúkawa, T., Kaizuka. S., and Ota, Y. (1964b) Crustal movement in the late Quaternary
revealed with coastal terraces on the southeast coast of Slúkoku, southwestern Japan.
Jour. Geodet. Soc. Japan, 10, i 16- 122.
Yoshikawa, T., Kaizuka, S., and Ota. Y. (1968) Coastal development of the Japanese Islands.
in Morrisorr, R. B. and Wright, Jr.. H. E. eds., Means of eorrelation of Quaternar¡'
succession. Univ. Utah Press, 457-465.
Zeil, W. ( 1964) Geologíe von Cltilc.233p. Berlin.

-38-
PhotograPhs
o;tE.!
! r: ¡ É-
¡- , J :9É
€! y d :co
\Ji)a¡d"
qFq¡Ec
!;H - Y
@¿.
o c.E..
t () c úY ¿Ul t)ñ
L ('.r
if,.9r
--!
o h.tr o ó +!o
É'-l< L.l il:a U
ñoéro -AÑ
=trño!
;-; o_rJ !
Z-É
Erv't¿
s ü,e iE c
i.aúü=
Yo=-; Em
!
§L
§!l!.
LIJ
< B oi4 ú
Jü

E O
- .- o
ó)-: Hri !-
--L
*¿9^¿
. <E =u)
z 2a.
4=C)+:
E - e-l
!¡i
a, ,zo
á.: *=tr
6E rr O
_ € ij oou
!/B\-
o ^ >Y >'::
ñ o?
4dJJt.J 36 c.,

=;r
;!d.,H
u'a
+o

+iÉY=5
E
*iá e
:lúb¡
!.r3
...!
úePL
9F.Etso:
d¡¡
-D"
t
- :p¿L
k:Y-v ! ,6*
O u rJ -. o UL:'.
É--!uO
*!l€t> z I 2L)
sE#> q ÉÉ a
>r!
! o
¡r 2 liu
o O " iÉq
-o0
!{ ^-E
eE!á¿) O o
(, sü
C¿ s ei '
=q b¡u =
!< d.6j: :1o,:¡
.¡
-É,
 .r{

:, l::iF
..:::t)
aa.,.:!:

r.:r'

N
!

Or
Photo 3 Whitish eolian sand (eoiianite, lower) and
brown loamy soil (upper) on the Cañete Surface.
Fossil burrows are secn in the lrppermost part of
the eolian sand bed, rvlúch consists of the upper
part of the Cañete Formation.

Photo 4 The middle horizon of this road cut is the

fluvial sand bed overlying the uppermost Cairete
Formation. The base of the fluvial bed shows small
three channels, which are filled with muddy depos-
its including pebbles. An eolian bed of about I
meter in thickness overlies the fluvial dcposits.
Photographed at 6 km north of Cañete town.

- 42-
Photo 3

E:'
:'*
-¿-
L' "1

r 'l'
:i
:i, '

Photo 4

-43-
 uq.i¿ c>
Q>:
OF
cio?
O6
-
,E>
LOE pO
.úo G3
^>,
4 9_.C

E!
^aF- 7'a
U",O
o39
\ .=
I *B o3
ú2r-
cH:
^r!
¿o
CIN
5 iltq.
(ioo
-
- ='=
É o.¡;
!=

Hp í or
+91,
33-s
i.ig=
-Oo:
ñ\J q
xo¿
-= o
5O->
!=^. ^
J o.9 ! r bo:
U-!
tuC-I o'"
> !-c co
ri¡!l
7á c
*u) _=
|-
'J*
€,
ú7É

tr!
"E- @.=
s.

ño@
!ooi'i -dE
Sobo
Da-
a .1 1 >E.o
Lt:>
' 3
^!
<x;
xx
rr) \o

Or
i/) \o
f
 o
+
I
Oo -J
(DP*5'
okJ
l;.!=r tr.--^i>
Ota C1*¿
o o
:Jry/.
-
r(Da:{
==
:+ -§P)-
4uñ9
4)J 1ñóó
J@r
-)'ü a
-, =
-N)=
=1óa
P<c
<3 H U ñU
,+
P J,v
É^O
5
É. É É.,
5;J0a*=lJ
o+á É+"c
'..:od+
-/El
^.m1, O^v!
hX¿4
ojdc¡
io=
^J^
rlOH §
-+@
;, *FD
6'o ¡ ÉÉ;<s
<o. o+-E
-B(!^1

3É¡i : r 5*
) 4C ¿
o*!' JdL^
ll-
E. I
3lJ =
4aaa
6¿+ ñ¿.;
§ o;bE
*^(Dc
{U4+
tg ñ,EP
:_H
E¡O
3 U:n -l
+{e +
B.=-c
€oo
;<ái
aG-+
4
(!o!.5- +ñh
C7 P3-
*. a)
>áñc
:¿ó
fo 4<a-'
sr*2
g ilJ-
of"
^V-^
 @ *¡

t
[ .,t',',.

iwi§r,5gi¡
,,, :,:,,,

f::'l:11::l:!:i!ii
rl:ri; ;¡; ;¡;iiiiiti :r,
i,ii,: t a i t;¡,i;;ii:il

¡,.]i;
Photo 9 An ai¡ view of the eastern half of Mocha Island.
The steep ancient sea cliff separates the mountains
and the flat Lower terrace, an uplifted abrasiotr
platform. On the Lower terrace, a pattern of
elevated beach ridges is seen, Note the flatness of
the crests of the mountains, whicir represents the
existance of the Higher and Middle teraces.

Photr¡ l0 Bea¡;h deposits on the Lower terrace of Nfocha

Island, consisting of mainly sand, pcbbles and shell
fragments. Such shell fragrnents were used for
radiocarbon dati¡rg. The surface soil is thin

-48-
Photo 9

Photo 10

-49