The Sauropods Evolution and Paleobiology - Kristina Curry Rogers Amp Amp Jeffrey Wilson PDF

THE SAUROPODS
THE SAUROPODS
Evolution and Paleobiology
Kristina A. Curry Rogers and Jeffrey A. Wilson
UNIVERSITY OF CALIFORNIA PRESS

Berkeley Los Angeles London
University of California Press, one of the most distinguished university
presses in the United States, enriches lives around the world by ad-
vancing scholarship in the humanities, social sciences, and natural sci-
ences. Its activities are supported by the UC Press Foundation and by
philanthropic contributions from individuals and institutions. For
more information, visit www.ucpress.edu.
University of California Press

Berkeley and Los Angeles, California
University of California Press, Ltd.

London, England
© 2005 by the Regents of the University of California
Library of Congress Cataloging-in-Publication Data
The Sauropods: Evolution and Paleobiology / [edited by] Kristina A. Curry

Rogers, Jeffrey A. Wilson.
p. cm.
Includes bibliographical references and index.
ISBN 0-520-24623-3 (cloth : alk. paper)
1. Saurischia. 2. Saurischia—Evolution. 3. Saurischia—
Morphology. 4. Saurischia—Anatomy. 5. Paleobiology. I. Curry
Rogers, Kristina A., 1974- II. Wilson, Jeffrey A.
QE862.S3S28 2005
567.913—dc22
2005010624
10 09 08 07 06 05
10 9 8 7 6 5 4 3 2 1
The paper used in this publication meets the minimum requirements

of ANSI/NISO Z39.48-1992 (R 1997) (Permanence of Paper). ∞
Cover: Titanosaurid Nesting Grounds, painting by Mark Hallett.

dedication
This volume is dedicated to our friend John S. Kohl, M. F. and J. S. McIntosh. 1997. Discovering
“Jack” McIntosh. Perhaps more than any other Dinosaurs in the Old West. The Field Journals of
Arthur Lakes. Smithsonian Institution Press,
single person, Jack has influenced the course of
Washington and London, 198 pp.
sauropod studies since World War II. In addi- McIntosh, J. S. 1989. The sauropod dinosaurs: a
tion to being one of the foremost students of brief survey. In: Padian, K., and Chure, D. J.
sauropods (McIntosh and Berman, 1975; (eds.). The Age of Dinosaurs. Short Courses in
Berman and McIntosh, 1978; McIntosh, 1989, Paleontology No. 2. University of Tennessee,
1990a, b), Jack has devoted himself to archiving Knoxville. Pp. 85–99.
———. 1990a. Sauropoda. In: Weishampel, D. B.,
and interpreting original map and quarry data
Dodson, P., and Osmólska, H. (eds.). The
from the early history of North American Dinosauria. University of California Press,
dinosaur paleontology (Ostrom and McIntosh, Berkeley. Pp. 345–401.
1966; Kohl and McIntosh, 1997). We close The ———. 1990b. Species determination in sauropod
Sauropods: Evolution and Paleobiology with an dinosaurs with tentative suggestions for their
classification. In: Carpenter, K., and Currie, P. J.
interview with Jack that records his thoughts on
(eds.) Dinosaur Systematics: Approaches and
the past, present, and future of sauropod stud- Perspectives. Cambridge University Press,
ies. Perhaps the greatest attribute of sauropod Cambridge.
dinosaurs is that they attract students like Jack, McIntosh, J. S. and D. S. Berman. 1975. Description
whose devotion and contagious enthusiam con- of the palate and lower jaw of Diplodocus
tinue after half a century. (Reptilia: Saurischia) with remarks on the
nature of the skull of Apatosaurus. Journal of
Berman, D. S. and J. S. McIntosh. 1978. Skull and Paleontology 49: 187–199.
relationships of the Upper Jurassic sauropod Ostrom, J. H. and J. S. McIntosh. 1966. Marsh’s
Apatosaurus (Reptilia, Saurischia). Bulletin of Dinosaurs. Yale University Press, New Haven,
the Carnegie Museum of Natural History 388 pp.
8: 1–35.
contents
Acknowledgments / ix 7 • POSTCRANIAL PNEUMATICITY IN SAUROPODS

AND ITS IMPLICATIONS FOR MASS
Introduction: Monoliths of the ESTIMATES / 201
Mesozoic / 1 Mathew J. Wedel
Jeffrey A. Wilson and Kristina Curry Rogers
8 • THE EVOLUTION OF SAUROPOD LOCOMOTION:
MORPHOLOGICAL DIVERSITY OF A
SECONDARILY QUADRUPEDAL RADIATION / 229
1 • OVERVIEW OF SAUROPOD PHYLOGENY AND
EVOLUTION / 15
Matthew T. Carrano
Jeffrey A. Wilson 9 • STEPS IN UNDERSTANDING SAUROPOD
BIOLOGY: THE IMPORTANCE OF SAUROPOD
2 • TITANOSAURIA: A PHYLOGENETIC
TRACKS / 252
OVERVIEW / 50
Joanna L. Wright
Kristina Curry Rogers
10 • NESTING TITANOSAURS FROM AUCA MAHUEVO
3 • PHYLOGENETIC AND TAXIC PERSPECTIVES ON AND ADJACENT SITES: UNDERSTANDING
SAUROPOD DIVERSITY / 104 SAUROPOD REPRODUCTIVE BEHAVIOR AND
Paul Upchurch and Paul M. Barrett EMBRYONIC DEVELOPMENT / 285
4 • SAUROPODOMORPH DIVERSITY THROUGH
Luis M. Chiappe, Frankie Jackson, Rodolfo
TIME: MACROEVOLUTIONARY AND
A. Coria, and Lowell Dingus
PALEOECOLOGICAL IMPLICATONS / 125 11 • SAUROPOD HISTOLOGY: MICROSCOPIC VIEWS
Paul M. Barrett and Paul Upchurch ON THE LIVES OF GIANTS / 303
Kristina Curry Rogers and Gregory M.
5 • STRUCTURE AND EVOLUTION OF A SAUROPOD
Erickson
TOOTH BATTERY / 157
Paul C. Sereno and Jeffrey A. Wilson A Conversation with Jack McIntosh / 327
6 • DIGITAL RECONSTRUCTIONS OF SAUROPOD
DINOSAURS AND IMPLICATIONS FOR List of Contributors / 335
FEEDING / 178
Index / 337
Kent A. Stevens and J. Michael Parrish
vii
acknowledgments
This volume was developed from a symposium their excellent papers and patience throughout
dedicated to Jack McIntosh presented at the the process of bringing this book to publication.
2001 Society of Vertebrate Paleontology annual We especially thank Blake Edgar (UC Press) for
meeting, which we convered with Dan Chure. initially supporting this project and his contin-
We thank Jack McIntosh for allowing us to ual guidance throughout.
organize a symposium in his honor. We also Most of all, we thank our partners, Ray
thank the participants of that symposium, most Rogers and Monica Wilson, for their love and
of whom authored chapters in this volume, for support.
ix
INTRODUCTION
Monoliths of the Mesozoic
eorge gaylord Simpson (1987:71) elongate neck, which in turn is balanced by a long
G expressed his impressions of the well-
known North American sauropod Diplodocus in
tail that tapers tipward. Numerous synapomor-
phies reflecting this general body plan diagnose
the form of a poem to his mother, written while the basalmost sauropod nodes, and small and
he was studying Mesozoic mammals at Oxford large differences in all regions of the skeleton
University: allow recognition of 121 sauropod species
(Upchurch et al. 2004) that were globally distrib-
Oh! Thou imbecile reptile Diplodocus! uted during most of the Mesozoic Era.
Whoever created so odd a cuss?
Sauropods are paradoxical animals because
With a tail like a neck,
And a neck like a tail—
they are built on an obvious and memorable body
I wonder, by heck, plan but are nonetheless one of the most taxonom-
If you ever do fail ically diverse dinosaur groups. This volume is ded-
To remember your ends, icated to exploring sauropod systematics and pale-
And when danger impends obiology by presenting an up-to-date summary of
Do stand still, which is bad,
our knowledge and remaining questions in these
or still more, run tail first,
Or indeed run both ways, which is rather worst! areas. While acknowledging and embracing the
paradoxical nature of sauropods, we hope to
Simpson adorned his poem with a caricature of explain how their body plan was constructed,
Diplodocus longus (fig. I.1), which—it must be explore its variations, and dispel the myth that it led
said—looks very sauropod-like. That the anatomy to evolutionary stagnation and eventual replace-
of a sauropod can be adequately conveyed in a ment by more “advanced” herbivorous dinosaurs.
humorous sketch attests to the relatively simple
and recognizable body plan that characterizes the
SAUROPODS AS MONOLITHS
group. Sauropods have deep, barrel-shaped chests
supported by four pillarlike legs. They have a rela- Sauropods are the largest animals known to
tively small skull that is perched at the end of an have walked the earth, as recorded in numerous
1
FIGURE I.1. Caricature of Diplodocus longus by George Gaylord Simpson (from Simpson 1987).
footprints found around the globe (e.g., Bird chapter and given approximately half a page
1941, 1944; Hunt et al. 1994). “Monolithic” seems per species suggests that they are at least per-
an apt descriptor of both their size and their ceived as monolithic by dinosaur specialists. Is
tight adherence to the body plan described there any justification or explanation for this
above, of which there are few reversals observed characterization?
during the nearly 150 million years of sauropod
evolution. SAUROPOD FOSSIL RECORD
The recently published second edition of The sauropod fossil record itself may be responsi-
The Dinosauria (Weishampel et al. 2004) con- ble for the monolithic perception of sauropods.
tains 22 chapters focused on “dinosaur system- Sauropods first appear in the fossil record during
atics.” The partitioning of chapters and number the Late Triassic, during which there are currently
of pages devoted to each offer insight into spe- several candidate earliest-appearing sauropods.
cialists’ perception of different dinosaur groups Together, these body fossils and ichnofossils sug-
as well as the attention each has been given his- gest a late Carnian or Norian origin for the group
torically (table I.1). Not surprisingly, theropods (summarized in Wilson 2005). Possible Carnian
have garnered the most attention, with their sauropods include Blikanasaurus (Yates 2003,
diversity (282 species) partitioned among nine 2004; Yates and Kitching 2003; Upchurch et al.
chapters and 185 pages. Thyreophorans (68 2004) and the Portezuelo Formation trackmaker
species, 58 pages), ornithopods (107 species, 71 (Marsicano and Barredo 2004); probable Norian
pages), and marginocephalians (56 species, 53 sauropods include Antenonitrus (Yates and
pages) are partitioned into three chapters each. Kitching 2003) and the Tetrasauropus trackmaker
Prosauropods (23 species, 27 pages) and (Lockley et al. 2001).
sauropods (121 species, 64 pages) are the only The notion of Triassic sauropods is new to this
major dinosaur groups represented by single century but was expected based on the first
chapters in The Dinosauria 2. This allotment appearance of other saurischians (e.g., Upchurch,
may be justified for Prosauropoda, which is the 1995; Wilson and Sereno 1998). Prior to this, a
smallest of the groups mentioned yet the only lengthy ghost lineage implied by these relation-
to receive more pages than its species count. In ships preceded the exclusively post-Triassic
contrast, Sauropoda is the second most diverse sauropod record (fig. I.2), which began with the
dinosaur group, representing 18%, or nearly fragmentary remains of the Lower Jurassic
one-fifth, of the 661 recognized dinosaur sauropods Vulcanodon (Raath 1972) and
species. Together, sauropods and theropods Barapasaurus (Jain et al. 1975) and the complete
encompass 60% of dinosaur species diver- remains of the Middle Jurassic Shunosaurus
sity. That sauropods are lumped into a single (Zhang 1988; Chatterjee and Zheng 2003).
2 MONOLITHS OF THE MESOZOIC

TABLE I.1
Species Counts and Pages Devoted to Systematics in The Dinosauria, 2nd Edition
CHAPTER PAGES SPECIES PAGES/SPECIES
1 Basal Saurischia 22 11 2.00
Theropoda
2 Ceratosauria 24 33 0.73
3 Basal Tetanurae 40 74 0.54
4 Tyrannosauroidea 26 19 1.37
5 Ornithomimosauria 14 12 1.17
6 Therizinosauroidea 14 12 1.17
7 Oviraptorosauria 19 18 1.06
8 Troodontidae 12 9 1.33
9 Dromaeosauridae 14 20 0.70
10 Basal Avialae 22 74 0.30
11 Prosauropoda 27 23 1.17
12 Sauropoda 64 121 0.53
13 Basal Ornithischia 10 4 2.50
Thyreopoda
14 Basal Thyreophora 8 5 1.60
15 Stegosauria 20 17 1.18
16 Ankylosauria 30 46 0.65
Ornithopoda
17 Basal Ornithopoda 20 24 0.83
18 Basal Iguanodontia 25 38 0.66
19 Hadrosauridae 26 45 0.58
Marginocephalia
20 Pachycephalosauria 14 17 0.82
21 Basal Ceratopsia 16 21 0.76
22 Ceratopsidae 23 18 1.28
Total Dinosauria 490 661 —
Sauropoda contribution 13.06% 18.31% —
NOTE: Introductory chapters that contained no species descriptions (i.e., “Saurischia,” “Ornithischia”) are not included in the page
tally.
Together, these taxa indicate that the sauropod dinosaurs. Upchurch (1998) recorded 60
body plan was constructed early in their evolu- synapomorphies diagnosing nodes basal to
tionary history and that the earliest sauropods Eusauropoda, and Wilson (2002) identified 74
resemble later sauropods more than they do synapomorphies arising at Sauropoda and
sauropod outgroups (such as prosauropods). Eusauropoda. These two analyses independently
That is, until recently, few fossils have been recognize that synapomorphies appearing at
available to document the transition between basal sauropod nodes represent 26% of all
sauropods and their hypothesized sister-taxa. synapomorphies identified. Few, if any, dinosaur
This fact of the fossil record was borne out in groups have such a base-heavy distribution of
lower-level phylogenetic analyses of sauropod synapomorphies. Thus the sauropod fossil
MONOLITHS OF THE MESOZOIC 3

FIGURE I.2. Temporal distribution and relationships of major lineages of dinosaurs during the Triassic and Jurassic. The vertical bar
representing sauropod lineage duration is divided into a black section (pre-2000) and a gray section with asterisk (post-2000). Icons
from Wilson and Sereno (1998) and Sereno (1999); timescale based on Harland et al. (1990). (Modified from Wilson 2002:fig. 2.)
record has in part led the monolithic depiction of bivores to quadrupedal, long-necked, herbivo-
sauropods. rous monoliths.
As new Triassic sauropods are discovered,
many of the synapomorphies identified by SAUROPOD SYSTEMATIC RECORD
Upchurch (1998) and Wilson (2002) will likely Romer (1968:137–138) bookended his discus-
diffuse stemward and articulate the transition sion of sauropods in his Notes and Comments on
from bipedal, short-necked, nonspecialized her- Vertebrate Paleontology with the following

laments: “A proper classification of the great and Kitching 2003; Allain et al. 2004) and India
amphibious sauropods has been the despair of (e.g., Jain et al. 1975; Yadagiri 2001); Middle
every one working on the group,” and “It will be Jurassic sauropods from China (Zhang 1988; He
a long time, if ever, before we obtain a valid, et al. 1988, 1998; Tang et al. 2001; Ouyang and Ye
comprehensive picture of sauropod classifica- 2002) and Argentina (Bonaparte 1986b); and well
tion and phylogeny.” The perception that sauro- preserved sauropod skeletons from the Cretaceous
pod interrelationships were intractable has con- of Asia (Borsuk-Bialynicka 1977; Suteethorn et al.
tributed to their monolithic status, perhaps 1995), South America (Salgado and Bonaparte
even more than do the circumstances of their 1991; Calvo and Salgado 1995; González Riga
fossil record. How can it be otherwise if sub- 2003; Martínez et al. 2004), India (Jain and
groups are not identifiable? Bandyopadhyay 1997), Africa (Jacobs et al. 1993;
Although a “valid” and “comprehensive” Sereno et al. 1999), and Madagascar (Curry
sauropod phylogeny was not available during Rogers and Forster 2001).
Romer’s time, steps had already been initiated The history of discovery of the 121 sauropod
to resolve the relationships of constituent taxa. species recognized as valid by Upchurch et al.
Sauropod interrelationships were resolved in (2004) is summarized in figure I.3. Prior to
stages, beginning with early classifications by the “Dinosaur Renaissance,” inaugurated with
Marsh (1895, 1898), Janensch (1929), and Ostrom’s (1969) description of Deinonychus,
Huene (1932), followed much later by sauropod discoveries rarely topped more than
Bonaparte’s (1986a, 1986b) recognition of a handful every five years. A notable outlier,
“eo” and “neo” sauropods, McIntosh’s (1989, however, is the burst of sauropod discoveries
1990a, 1990b) delineation of numerous and descriptions in the late 1870s that coin-
sauropod families, and the relatively recent cides with peak Cope–Marsh activity.
use of cladistic techniques (Russell and Zheng Following 1969, however, sauropod discover-
1993; Calvo and Salgado 1995; Upchurch 1995, ies always exceed five per five years, steadily
1998; Wilson and Sereno 1998; Wilson increasing through the late 1990s, when 25
2002; Upchurch et al. 2004). Together, these sauropods were named. The subsequent drop
and other studies have gained substantial in sauropod discoveries in figure 3 is an arti-
consensus on the interrelationships of fact of only counting up to the year 2002.
sauropods, although several contentious Projecting to 2005 based on these numbers
areas remain (see chapters by Wilson and suggests an excess of 10 sauropods for this
Curry Rogers). interval. This relatively sudden surge in sauro-
pod discoveries is more striking when we con-
THE PACE OF DISCOVERY sider that 50 sauropod species were named
Romer (1968:137–138) suggested that the frag- since the first edition of The Dinosauria, and
mentary nature of many sauropod taxa con- that one-third of all sauropod species have
tributed to their unresolved interrelationships: been named since the first cladistic analyses of
“The reasons for our difficulties are apparent. Sauropoda (Calvo and Salgado 1995;
Few complete skeletons exist; feet and skulls are Upchurch 1995).
rare; many of the numerous described forms are Both the steady improvement of the sauro-
based on fragmentary material.” This was cer- pod fossil record and the intensified interest
tainly the case, and the improvement in our in and resolution of sauropod phylogeny are
understanding of sauropod phylogeny is the beginning to differentiate the group into
result of an improved sauropod fossil record. more manageable pieces, such as basal
Recent key discoveries, in stratigraphic order, sauropods, macronarians, and diplodocoids.
include the discovery of primitive sauropods in Both of these advances are relatively recent—
Africa (e.g., Charig et al. 1965; Raath 1972; Yates cladistic analyses and escalation in sauropod

FIGURE I.3. Sauropod species named since the description of Cetiosaurus (Owen, 1841) and recognized as valid by
Upchurch et al. (2004).
discoveries postdate the first edition of The longed growth strategy common in other reptiles
Dinosauria (1990) but predate The Dinosauria 2 (e.g., Enlow and Brown 1956, 1957; Case 1978a;
(2004). Francillon-Vieillot et al. 1990). Interestingly,
although other vertebrates attained comparable
SAUROPODS LIVING LARGE body sizes, sauropods may be the only giants
Body size is the most recognizable characteristic whose young hatch from eggs. Even the largest of
of sauropods, and can be expected to have influ- adult sauropods began as hatchlings measuring
enced all aspects of their biology (Peters 1983; only one meter long and weighing less than 10 kg
LaBarbera 1989). The largest sauropods are esti- (Chiappe et al. 1998, 2001), a range of ontoge-
mated to have reached adult body masses of 40 to netic size exceeding that for any other dinosaur
70 metric tons or more (Peczkis 1994:appendix), lineage. In contrast, other large-bodied verte-
an upper bound reached independently within brates bore live young, including chon-
multiple sauropod lineages (Diplodocoidea, drichthyans (sharks [Dulvy and Reynolds 1997]),
“Seismosaurus”; Macronaria, Brachiosaurus; mammals (whales [Clapham et al. 1999]), and
Titanosauria, Argentinosaurus). The smallest marine reptiles such as plesiosaurs (Cheng et al.
sauropods (e.g., Magyarosaurus, Saltasaurus) may 2004), ichthyosaurs (Böttcher 1990; Maxwell
have weighed between 1.5 and 3 metric tons and Caldwell 2003), and mosasaurs (Caldwell
(Erickson et al. 2001) and represent one of the and Lee 2001).
few phylogenetic decreases in body size among The enormity of sauropod dinosaurs has cast
dinosaurs. a shadow over studies of their paleobiology.
Sauropods appear to have attained large adult Struggling to find adequate descriptors for sauro-
body size by rapid post-hatching growth (e.g., pod size—“ponderous,” “behemoth,” “enor-
Rimblot-Baly et al. 1995; Curry 1999; Sander mous,” “stupendous,” and “massive” being a
2000; Erickson et al. 2001; Sander and few—many early paleontologists assumed that
Tückmantel 2003) rather than by the slow, prosauropods could not support their body weight on

FIGURE I.4. Oliver P. Hay’s 1910 reconstruction of “The Form and Attitudes of Diplodocus.” Hay’s interpretation of
sauropod paleobiology—lazing on beaches of ancient rivers, lizardlike in their stance and habit—was common until quite
recently.
land (e.g., Owen 1875; Osborn 1899; Hatcher ing lineages (Gilmore 1946; Lucas and Hunt
1901). Although their anatomy was trumpeted as 1989; Dodson 1991).
a “marvel of construction . . . a mechanical tri- Ironically, some of these interpretations
umph for great size, lightness, and strength” themselves can be looked on as holdovers from a
(Osborn 1899:213), early life reconstructions previous era. Studies by Walter Coombs (1975,
depicted sauropods as large, lumbering, and near 1978) and Robert Bakker (1968, 1971a, 1971b,
or up to their necks in ancient swamps (fig. I.4). 1986) reinvented sauropods as dynamic, terres-
Even the first-described sauropod footprints were trial vertebrates that might have been agile
initially interpreted as having been made in an enough to feed tripodally, use their tails as
aquatic environment (Bird 1941, 1944). This per- weapons, and generate their own body heat via
ception of sauropods as unwieldy, archaic herbi- high food consumption capabilities (fig. I.5).
vores relegated to evolutionary backwaters was These interpretations challenged the prevailing
prominent as recently as the early 1990s: perception of sauropods as archaic dinosaurs
destined for extinction and helped underscore
Their large sizes, small heads, simple teeth,
and tiny brains served them well for millions
how little we know about sauropod paleobiology.
of years. But in the Cretaceous, more progres- The recent surge in sauropod discoveries around
sive, large-headed, larger-brained dinosaurs the world, combined with the taxonomic revision
appeared (the ornithopods and margin- of fragmentary genera and the first testable
ocephalians) and vegetation changed. . . . The hypotheses of relationship, provide the requisite
old giants retreated to southern continents,
framework for delving deeper into these ques-
where the newcomers did not flourish.
tions of sauropod paleobiology. In this volume,
(Dodson 1991:34).
we attempt to gain new understanding of
Consequently, Cretaceous Gondwanan sauropods “nature’s grandest extravagances” (Dodson
were viewed as Jurassic relics rather than thriv- 1991:34).

FIGURE I.5. John Gurche’s dra-
matically posed Barosaurus pro-
tecting its young from an attack-
ing Allosaurus. This depiction
casts sauropods as active ani-
mals capable of rearing up on
their hind legs.
SAUROPOD EVOLUTION groups. Wilson highlights the difficulties in trac-

AND PALEOBIOLOGY ing two major parts of the sauropod phylogeny—
even the earliest known sauropods have the fea-
The Sauropods: Evolution and Paleobiology opens
tures of the axial and appendicular skeletons that
with three chapters aimed at delimiting the evo-
characterize all sauropods, implying a 10 mil-
lutionary history and diversity of Sauropoda.
lion- to 15 million-year ghost lineage. Similarly,
These chapters provide a phylogenetic context
the preponderance of partial material and rela-
for the subsequent chapters on sauropod paleo-
tive conservatism of the postcranial skeleton has
biology. Chapter 1 highlights our current view of
obfuscated the other end of sauropod evolution,
sauropod phylogeny and concludes that
that of Rebbachisauridae and Titanosauria.
sauropods were successful in terms of their geo-
Wilson highlights the importance of these taxa
graphic and temporal distributions, biomass,
for biogeography during the Late Mesozoic as
morphological complexity, and diversity at both
the continents attained their current positions.
higher and lower levels. Our understanding of
Similarly, rebacchisaurids and titanosaurs bear
sauropod phylogeny is particularly lucid with
dental specializations and were coeval with the
regard to the Diplodocoidea and Macronaria and
burgeoning ornithischian populations in both
the cranial specializations in each of these major
the northern and the southern hemisphere. The

connections between Late Cretaceous survivor- teries of Nigersaurus to those of euornithopods
ship and narrow-crowned dentition provide an and neoceratopsians, and conclude that diver-
interesting opportunity for future work. gent functions and uncorrelated progression sig-
In chapter 2, Curry Rogers expands on the nal independent causes for the evolution of den-
work of Wilson and examines the interrelation- tal batteries. This contrasts with the view that a
ships among Titanosauria, a significant yet single environmental cue (i.e., the evolution of
poorly understood sauropod clade. She outlines angiosperms) prompted the evolution of com-
the current consensus with regard to titanosaur plex dental batteries, and that sauropods and
phylogeny and highlights the skeleton of ornithischians were in direct competition for
Rapetosaurus krausei, the first titanosaur with resources during the last stages of the Mesozoic.
associated cranial and postcranial remains, as a In chapter 6, Stevens and Parrish present and
keystone taxon in current phylogenetic analy- implement their method for three-dimensional
ses. Her chapter concludes with one of the first reconstructions of sauropod skeletons, focusing
detailed analyses of lower-level titanosaur rela- on the pose of the neck and its implications for
tionships, which helps set the stage for a more sauropod herbivory. Their reconstructed feeding
detailed analysis of the paleobiology of this envelopes challenge the view that all sauropods
unique, derived group of macronarians. were high browsers. In contrast to the giraffelike
Building on recent phylogenetic analyses, reconstructions of Brachiosaurus, or the tripodal
Upchurch and Barrett investigate sauropod “suc- stance that is so often depicted in the popular
cess rate” in chapter 3 by taking a closer look at media, Stevens and Parrish identify most
taxic and phylogenetic estimates of sauropod sauropods as medium to low browsers. In their
diversity. Both estimates compare favorably with view, sauropod feeding is constrained not only by
other estimates of sauropod diversity, indicating the dentition, but also by the axial flexibility and
that important radiations occurred in the Middle forage availability and abundance.
and Late Jurassic, as well as at the end of the Although sauropod vertebrae have long been
Cretaceous. Such results contradict the long- touted as being specialized for reducing weight
standing view that sauropod diversity reached its while providing strength, the relationship has
zenith in the Jurassic and its nadir in the not been systematically analyzed and the impli-
Cretaceous, and highlights the high diversity of cations of potential weight-reducing methods
particular sauropod clades late in the history of qualified. In chapter 7, Wedel investigates the
sauropod evolution. Barrett and Upchurch axial skeleton from the perspective of pneu-
expand on this analysis of sauropod diversity in maticity and “efficiency of design.” Concluding
chapter 4 by exploring the effects of macroevolu- that most eusauropod vertebrae were pneu-
tionary mechanisms potentially responsible for matic, Wedel estimates that these vertebrae are
these patterns. The authors determine that pro- in most cases between 50% and 60% air. The
posed causes and their effects do not always fit effects of pneumaticity might reduce sauropod
the observed sauropod diversity pattern (i.e., mass estimates by as much as 10%, and provide
competition between rebacchisaurids and some interesting new possibilities for interpret-
titanosaurs; competition between sauropods and ing sauropod physiology.
ornithopods). In addition, the authors highlight Chapters 8 and 9 focus on sauropod limbs
the potential for coevolution among titanosaurs and their signficance for interpreting locomotor
and angiosperms. capabilities in sauropods. In chapter 8, Carrano
The theme of complex dentitions and her- utilizes a phylogenetic framework to address
bivory during the last radiation of sauropods is the morphological and functional diversity
apparent in the description of the unique dental observed in the secondarily quadrupedal
battery present in Nigersaurus presented in chap- sauropods. Without question, the appendicular
ter 5. Sereno and Wilson compare the dental bat- morphology of sauropods is constrained by

being big in a terrestrial environment. Carrano age 5, an African elephant is only ~1 metric ton; at
concludes that sauropods exhibit considerable the same age, Apatosaurus might have been closer
morphological diversity associated with varying to 20 metric tons.
locomotor patterns. For example, titanosaurs
can be characterized in part by their wide-gauge
FUTURE DIRECTIONS
gaits and the coeval development of reduced
body size in some clades, potential tripodality, New sauropods will continue to be discovered
and so forth. As outlined by Wright in chapter 9, in Mesozoic strata from around the world,
sauropod trackways have provided widely diver- despite the improbability of the preservation of
gent views of sauropod habits during the his- their enormous bodies and fragile skulls and
tory of discovery, with early tracks thought to axial bones. These future discoveries will surely
demonstrate that sauropods spent more time remedy some of the temporal and geographic
wading than walking in more terrestrial set- gaps in our current record of Sauropoda and
tings (Bird 1939, 1941, 1944). Wright outlines reveal anatomical surprises that will revise our
the history of discovery of sauropod tracks, estimates of their genealogy. The Sauropods:
summarizes our understanding of trackmak- Evolution and Paleobiology not only documents
ers, and critiques the ichnological evidence that what is currently known about sauropod evolu-
sauropods traveled in organized herds. tionary history, but also highlights deficiencies
Sauropod size and reproduction are dis- in our understanding. In that sense, we hope
cussed in chapter 10. Chiappe et al. document this volume sparks interest, provokes ques-
the first discoveries of identifiable sauropod tions, and provides fresh ground for continued
embryos and nests—the fantastic accumulation research on sauropod dinosaurs.
of fossils from Auca Mahuevo, Argentina. This Several gaps in our understanding of sauro-
site provides definitive evidence that sauropods pod paleobiology are not addressed in this vol-
were egg-laying reptiles. Chiappe et al. analyze ume. Many of these are directly related to the
the taphonomy and morphology of this amazing sauropod fossil record, which documents as
assemblage of sauropod eggs, providing us with few as 20 genera known from better than 90%
a clearer understanding of how sauropods repro- of the skeleton (Upchurch et al. 2004). One of
duced and what hatchlings looked like, an exam- the most conspicuous deficiencies is the
ple of nesting structure, and the first unspecula- absence of transitional Triassic forms bridging
tive interpretations of sauropod nesting behavior. the morphological gap between sauropods and
In chapter 11, Curry Rogers and Erickson sur- their closest relatives, prosauropods, which
vey sauropod growth rates from the perspectives obfuscates the origin of the sauropod body
of bone histology, long bone growth, and develop- plan. Other poorly sampled horizons include
mental mass. Instead of the century-long ontoge- the Middle Jurassic, during which nearly all the
nies predicted by early workers extrapolating from main neosauropod lineages are hypothesized to
reptilian growth rates for the large-bodied have evolved, and the Early Cretaceous, when
sauropods (e.g., Case 1978a, 1978b), Curry Rogers rebbachisaurid and titanosaur sauropods
and Erickson present a dramatically different view underwent interesting changes in herbivory
of the ontogeny for Apatosaurus. Growth rates on and locomotion that were manifest in later
par with those of some of the largest living verte- Cretaceous forms. Geographical biases in the
brates were normal for sauropods, and develop- sauropod fossil record also persist. Australia
ment of a characteristic trend for dinosaur growth and Antarctica stand out as southern continen-
allows predictions of growth rates for the largest tal landmasses that have produced only a few,
sauropods ever known. They suggest that fragmentary sauropod taxa. Similarly, with the
sauropods reached their enormous adult sizes exception of a few well-preserved taxa from
quickly—likely within 15 years. To compare—at North Africa, the continent has only provided a

glimpse of later-stage sauropod evolution. Bird, R. T. 1939. Thunder in his footsteps. Nat. Hist.
North America, despite its rich Late Jurassic 43: 254–261.
———. 1941. A dinosaur walks into the museum.
sauropod record, hints of an as-yet-undiscov-
Nat. Hist. 47: 74–81.
ered Cretaceous sauropod fauna. These large- ———. 1944. Did Brontosaurus ever walk on land?
scale evolutionary questions are not the only Nat. Hist. 53: 61–67.
ones impeded by an imperfect fossil record; Bonaparte, J. F. 1986a. The early radiation and phylo-
straightforward anatomical questions also per- genetic relationships of sauropod dinosaurs,
sist. Examples include the arrangement and based on vertebral anatomy. In: K. The Beginning
of the Age of Dinosaurs. Cambridge University
diversity of body armor (osteoderms), the con-
Press, Cambridge. Pp. 247–258.
figuration and embryological identity of carpal ———. 1986b. Les dinosaures (Carnosaures,
bones, and the nature and arrangement of gas- Allosauridés, Sauropodes, Cétiosauridés) du
tralia and clavicles. Jurassique moyen de Cerro Cóndor (Chubut,
Other questions are not necessarily depend- Argentina). Ann. Paléontol. 72: 325–386.
Borsuk-Bialynicka, M. 1977. A new camarasaurid
ent on discoveries of new and better fossils, but
sauropod Opisthocoelicaudia skarzynskii, gen. n.,
on our ability to interpret those already collected. sp. n. from the Upper Cretaceous of Mongolia.
Although we have made inroads, summarized in Palaeontol. Polonica 37: 1–64.
several chapters in this volume, sauropod life Böttcher, R. 1990. Neue Erkenntnisse über die
history stands out as a complex issue knotting Fortpflanzungsbiologie der Ichthyosaurier
physiology, development, ecology, and behavior. (Reptilia). Stuttgarter Beiträge Naturkunde (Ser
B) 164: 13–51.
Basic questions remain, including the trajectory
Caldwell, M. W., and Lee, M. S. Y. 2001. Live birth in
of body size change within the phylogenetic, tem- Cretaceous marine lizards (mosasaurids). Proc.
poral, and geographic distribution of sauropods; Roy. Soc. London B 268: 2397–2401.
the sequence of fusion of skeletal elements (par- Calvo, J. O., and Salgado, L. 1995. Rebbachisaurus
ticularly in the vertebral column) throughout tessonei, sp. nov. a new Sauropoda from the
Albian-Cenomanian of Argentina; New evi-
ontogeny; the masticatory forces producing wear
dence on the origin of Diplodocidae. GAIA 11:
facets on teeth; the composition of sauropod 13–33.
diets; and the relative roles of oral and gastric Case, T. J. 1978a. On the evolution and adaptive sig-
maceration of plant material. More challenging nificance of postnatal growth rates in the terres-
still are questions of sauropod thermal biology trial vertebrates. Q. Rev. Biol. 53: 243–282.
and energetics, reproductive behavior, and social ———. 1978b. Speculations on the growth rate and
reproduction of some dinosaurs. Paleobiology 4:
interactions. These and other questions will pro-
320–328.
pel future sauropod research. Charig, A. J., Attridge, J., and Crompton, A. W. 1965.
On the origin of the sauropods and the classifica-
tion of the Saurischia. Proc. Linn. Soc. London
LITERATURE CITED 176: 197–221.
Chatterjee, S., and Zheng, Z. 2003. Cranial anatomy
Allain, R., Aquesbi, N., Dejax, J., Meyer, C., of Shunosaurus, a basal sauropod dinosaur from
Monbaron, M., Montenat, C., Richir, P., Rochdy, M., the Middle Jurassic of China. Zool. J. Linn. Soc.
Russell, D., and Taquet, P. 2004. A basal sauro- 136: 145–169.
pod dinosaur from the Early Cretaceous of Cheng, Y.-N., Wu, X.-C., and Ji, Q. 2004. Triassic
Morocco. C.R. Palovol. 3: 199–208. marine reptiles gave birth to live young. Nature
Bakker, R.T. 1968. The superiority of dinosaurs. 432: 383–386.
Discovery 3: 11–22. Chiappe, L. M., Coria, R. A., Dingus, L., Jackson, F.,
———. 1971a. The ecology of brontosaurs.Nature Chinsamy, A., and Fox, M. 1998.Sauropod
229: 172–174. dinosaur embryos from the Late Cretaceous of
———. 1971b. Dinosaur physiology and the origin of Patagonia.Nature 396: 258–261.
mammals. Evolution 25: 636–658. Chiappe, L. M., Salgado, L., and Coria, R. A. 2001.
———. 1986. Dinosaur Heresies.William Morrow, Embryonic skulls of titanosaur sauropod
New York. dinosaurs. Science 293: 2444–2446.

Clapham, P.J., Wetmore, S.E., Smith, T.D., and Omeisaurus tianfuensis]. Sichuan Scientific and
Mead, J.G. 1999. Length at birth and at inde- Technological Publishing House, Chengdu. (In
pendence in humpback whales. J. Cetacean Res. Chinese with English abstract.)
Manage. 1: 141–146. He, X., Wang, C., Liu, S., Zhou, F., Liu, T., Cai, and
Coombs, W.P., Jr. 1975. Sauropod habits and habi- Dai, B. 1998. [A new sauropod dinosaur from the
tats. Palaeogeogr. Palaeoclimatol. Palaeoecol. 17: Early Jurassic in Gongxian County, South
1–33. Sichuan]. Acta Geologica Sichuan 18: 1–6. (In
———. 1978. Theoretical aspects of cursorial adapta- Chinese with English abstract.)
tions in dinosaurs. Q. Rev. Biol. 53: 393–418. Huene, F. von. 1932. Die fossil Reptil-Ordnung
Curry, K. A. 1999. Ontogenetic histology of Saurischia, ihre Entwicklung und Geschichte.
Apatosaurus (Dinosauria: Sauropoda): New insights Monograph. Geol. Palaeontol. 4: 1–361.
on growth rates and longevity. J. Vertebr. Paleontol. Hunt, A. P., Lockley, M. G., Lucas, S. G., and Meyer, C.
19: 654–665. A. 1994. The global sauropod fossil record. GAIA
Curry Rogers K.A., and Forster, C. 2001. The last of 10: 261–279.
the dinosaur titans: a new sauropod from Madagascar. Jacobs, L. L., Winkler, D. A., Downs, W. R., and
Nature 412: 530–534. Gomani, E. M. 1993. New material of an Early
Dodson, P. 1991. Lifestyles of the huge and famous. Cretaceous titanosaurid sauropod dinosaur from
Nat. Hist. 100: 30–34. Malawi. Palaeontology 36: 523–534.
Dulvy, N.K., and Reynolds, J.D. 1997. Evolutionary Jain, S. L., and Bandyopadhyay, S. 1997. New
transitions among egg-laying, live-bearing and titanosaurid (Dinosauria: Sauropoda) from the
maternal inputs in sharks and rays. Proc. Roy. Late Cretaceous of central India. J. Vertebr.
Soc. Lond. B 264: 1309–1315. Paleontol. 17: 114–136.
Enlow, D.H., and Brown, S.O. 1956. A comparative Jain, S. L., Kutty, T. S., Roy-Chowdhury, T., and
histological study of fossil and recent bone tissue Chatterjee, S. 1975. The sauropod dinosaur from
(Part 1). Tex. J. Sci. 8: 405–443. the Lower Jurassic Kota Formation of India. Proc.
———. 1957. A comparative histological study of fos- Roy. Soc. London B 188: 221–228.
sil and recent bone tissue (Part 2).Tex. J. Sci. 9: Janensch, W. 1929. Material und Formengehalt der
186–214. Sauropoden in der Ausbeute der Tendaguru-
Erickson, G., Curry Rogers, K., and Yerby, S. 2001. Expedition. Palaeontographica 2 (Suppl. 7): 1–34.
Dinosaurian growth patterns and rapid avian LaBarbera, M. 1989. Analyzing body size as a factor
growth rates. Nature 412: 429–433. in ecology and evolution. Annu. Rev. Ecol.
Francillon-Vieillot, H., de Buffrénil, V., Castanet, J., System. 20: 97–117.
Géraudie, J., Meunier, F.J., Sire, J.Y., Zylberberg, Lockley, M. G., Wright, J. L., Hunt, A. G., and Lucas,
L., and de Ricqlès, A. 1990. Microstructure and S. G. 2001. The Late Triassic sauropod track
mineralization of vertebrate skeletal tissues. In: record comes into focus: old legacies and new
Skeletal Biomineralization: Patterns, Processes, paradigms. New Mexico Geological Society
and Evolutionary Trends, Vol. 1. Van Nostrand Guidebook, 52nd Field Conference, Geology of
Reinhold, New York. Pp. 473–530. the Llano Estacado. Pp. 181–190.
Gilmore, C.W. 1946. Reptilian fauna of the North Lucas, S. G., and Hunt, A. P. 1989. Alamosaurus and
Horn Formation of central Utah. U.S. Geol. Surv. the sauropod hiatus in the Cretaceous of the
Prof. Paper 210C: 1–52. North American Western Interior. In: Farlow,
González Riga, B. 2003. A new titanosaur (Dinosauria, J. O., (ed.). Paleobiology of the Dinosaurs. Geolog.
Sauropoda) from the Upper Cretaceous of Soc. Am. Special Paper 238: 75–85.
Mendoza Province, Argentina. Ameghiniana 40: Marsh, O.C. 1895. On the affinities and classification
155–172. of the Dinosaurian reptiles. Am. J. Sci. 21: 417–423.
Harland, W. B., Armstrong, R. L., Cox, A. V., Craig, L. E., ———. 1898. On the families of sauropodous
Smith, A. G., and Smith, D. G. 1990. A Geologic dinosaurs. Am. J. Sci. 6(4): 487–488.
Time Scale 1989. Cambridge University Press, Marsicano, C. A., and Barredo, S. P. 2004. A Triassic
Cambridge, 263 pp. tetrapod footprint assemblage from southern
Hatcher, J.B. 1901. Diplodocus (Marsh): Its osteology, South America: palaeobiogeographical and evolu-
taxonomy, and probable habits, with a restoration tionary implications. Palaeogeogr. Palaeoclimatol.
of the skeleton. Mem. Carnegie Mus. 1: 1–63. Palaeoecol. 203: 313–335.
He, X.-L., Li, K., and Cai, K.-J. 1988. [The Middle Martínez, R. D., Giménez, O., Rodríguez, J., Luna,
Jurassic dinosaur fauna from Dashanpu, Zigong, M., and Lamanna, M. C. 2004. An articulated
Sichuan. Vol. IV. Sauropod Dinosaurs (2) specimen of the basal titanosaurian (Dinosauria:

Sauropoda) Epachthosaurus sciuttoi from the early Russell, D. A., and Zheng, Z. 1993. A large mamen-
Late Cretaceous Bajo Barreal Formation of chisaurid from the Junggar Basin, Xinjiang,
Chubut Province, Argentina. J. Vertebr. Paleontol. People’s Republic of China. Can. J. Earth Sci. 30:
24: 107–120. 2082–2095.
Maxwell, E.E., and Caldwell, M. W. 2003. First record Salgado, L. and Bonaparte, J. F. 1991. Un nuevo
of live birth in Cretaceous ichthyosaurs: closing sauropodo dicraeosauridae, Amargasaurus cazaui,
an 80 million year gap. Proc. Roy. Soc. London B gen. et. sp. nov., de la Formacion La Amarga,
270: 104–107. Neocomiano de la Provincia del Neuquen,
McIntosh, J.S. 1989. The sauropod dinosaurs: a brief Argentina, Ameghiniana 28: 333–346.
survey. Pp. In: Padian, K., and Chure, D.J. (eds.). Sander, P. M. 2000. Longbone histology of the
The Age of Dinosaurs. Short Courses in Tendaguru sauropods: implications for growth
Paleontology No. 2. University of Tennessee, and biology. Paleobiology 26: 466–488.
Knoxville. Pp. 85–99. Sander, P. M., and Tückmantel, C. 2003. Bone lamina
———. 1990a. Sauropoda. In: Weishampel, D.B., thickness, bone apposition rates, and age esti-
Dodson, P., and Osmólska, H. (eds.). The mates in sauropod humeri and femora. Paläontol.
Dinosauria. University of California Press, Z. 77(1): 139–150.
Berkeley. Pp. 345–401. Sereno, P. C. 1999. The evolution of dinosaurs.
———. 1990b. Species determination in sauropod Science 284: 2137–2147.
dinosaurs with tentative suggestions for their classi- Sereno P C., Beck, A.L., Dutheil, D.B., Larsson,
fication. In: Carpenter, K., and Currie, P.J. (eds.). H.C.E., Lyon, G.H., Moussa, B., Sadleir, R.W.,
Dinosaur Systematics: Approaches and Perspectives. Sidor, C.A., Varricchio, D.J., Wilson, G.P., and
Cambridge University Press, Cambridge. Wilson, J.A. 1999. Cretaceous sauropods from the
Osborn, H.F. 1899. A skeleton of Diplodocus. Mem. Sahara and the uneven rate of skeletal evolution
Am. Mus. Nat. Hist. 1: 191–214. among dinosaurs. Science 286: 1342–1347.
Ostrom J.H. 1969. Osteology of Deinonychus antir- Simpson, G. G. 1987. Simple Curiosity: Letters from
rhopus, an unusual theropod from the Lower George Gaylord Simpson to His Family,
Cretaceous of Montana. Bull. Peabody Mus. Nat. 1921–1970. Laporte, L. F. (ed.). University of
Hist. 30: 1–165. California Press, Berkeley. 340 pp.
Ouyang, H., and Ye, Y. 2002. [The first mamen- Suteethorn, V., Martin, V., Buffetaut, E.,
chisaurian skeleton with complete skull, Triamwichanon, S., and Chaimanee. 1995. A new
Mamenchisaurus youngi]. Sichuan Science and dinosaur locality in the Lower Cretaceous of
Technology Publishing House, Chengdu. (In northeastern Thailand. C. R. Acad. Sci. Paris (Ser.
Chinese with extended English abstract.) Iia) 321: 1041–1047.
Owen, R. 1875. Monographs of the fossil Reptilia of Tang, F., Jin, X., Kang, X., and Zhang, G. 2001.
the Mesozoic formations (Pt. II) (genera Omeisaurus maoianus, a complete Sauropoda
Bothriospondylus, Cetiosaurus, Omosaurus). Palaeon- from Jingyan, Sichuan. Research works of
togr. Soc. Monogr. 29: 15–93. Natural Museum of Zhejiang, China Ocean
Peczkis, J. 1994. Implications of body-mass estimates Press, Beijing. (In Chinese with extended English
for dinosaurs. Journal of Vertebrate Paleontology abstract).
14: 520–533. Upchurch, P. 1995. The evolutionary history of sauro-
Peters, R.H. 1983. The ecological implications of pod dinosaurs. Philos. Trans. Roy. Soc. London B
body size. Cambridge University Press, Cambridge/ 349: 365–390.
shill New York. ———. 1998. The phylogenetic relationships of sauro-
Raath, M. 1972. Fossil vertebrate studies in Rhodesia: a pod dinosaurs. Zool. J. Linn. Soc. 124: 43–103.
new dinosaur (Reptilia, Saurischia) from the near Upchurch, P., Barrett, P., (and) Dodson, P. 2004.
the Trias-Jurassic boundary. Arnoldia 30: 1–37. Sauropoda. In: Weishampel, D. B., P. Dodson,
Rimblot-Baly, F., de Ricqlès, A., and Zylberberg, L. and Osmólska, H. (eds.). The Dinosauria, 2nd ed.
1995. Analyse paléohistologique d’une série de University of California Press, Berkeley. pp.
croissance partielle chez Lapparentosaurus mada- 259–324.
gascariensis (Jurassique Moyen): Essai sur la Weishampel, D. B., Dodson, P., and Osmólska, H.
dynamique de croissance d’un dinosaure (eds.). 2004. The Dinosauria, 2nd ed. University
sauropode. Ann. Paléontolo. 81: 49–86. of California Press, Berkeley. 861 pp.
Romer, A. S. 1968. Notes and Comments on Wilson, J. A. 2002. Sauropod dinosaur phylogeny:
Vertebrate Paleontology. University of Chicago critique and cladistic analysis. Zool. J. Linn. Soc.
Press, Chicago. 304 pp. 136: 217–276.

Wilson, J. A. 2005. Integrating ichnofossil and body systematics of early dinosaurs. J. Syst. Palaeontol.
fossil records to estimate locomotor posture and 1: 1–42.
spatiotemporal distribution of early sauropod ———. 2004. Anchisaurus polyzelus Hitchcock: the
dinosaurs: a stratocladistic approach. Paleobiology smallest known sauropod dinosaur and the evo-
31: 400–427. lution of gigantism amongst sauropodomorph
Wilson, J.A., and Sereno, P.C. 1998. Early evolution dinosaurs. Postilla 230: 1–58.
and higher-level phylogeny of sauropod Yates, A. M., and Kitching, J. W. 2003. The earliest
dinosaurs. Soc. Vertebr. Paleontol. Mem. 5: 1–68 known sauropod dinosaur and the first steps
(supplement to J. Vertebr. Paleontol. 18). towards sauropod evolution. Proc. Roy. Soc.
Yadagiri, P. 2001. The osteology of Kotasaurus London B 270: 1753–1758.
yamanpalliensis, a sauropod dinosaur from the Zhang, Y. 1988. [The Middle Jurassic dinosaur fauna
Early Jurassic Kota Formation of India. J. Vertebr. from Dashanpu, Zigong. China. Volume III.
Paleontol. 21: 242–252. Sauropod dinosaur (I). Shunosaurus]. Sichuan
Yates, A. M. 2003. A new species of the primi- Publishing House of Science and Technology,
tive dinosaur, Thecodontosaurus (Saurischia: Chengdu. (In Chinese with extended English
Sauropodomorpha) and its implications for the summary.)

One
Overview of Sauropod Phylogeny

and Evolution
Jeffrey A. Wilson
SAUROPOD STUDIES FROM OWEN TO long bones” and “the toes being terminated by
THE PRESENT strong claws” (Owen 1842:102), but this assess-
ment was based on limited anatomical evidence
This year marks the one hundred sixty-fourth (Owen 1875:27). Key data emerged with the dis-
anniversary of Richard Owen’s (1841) description covery of abundant Cetiosaurus bones in
of the first sauropod—Cetiosaurus, the “whale Oxfordshire by John Phillips. Thomas Huxley
lizard”—on the basis of vertebrae and limb ele- examined this “splendid series of remains”
ments from localities across England. Although before the publication of Phillips’ (1871) mono-
these remains “had been examined by Cuvier graph and was the first to place Cetiosaurus within
and pronounced to be cetaceous” (Buckland Dinosauria (Iguanodontidae [Huxley, 1869:35]).
1841:96), Owen (1841:458–459) demonstrated Phillips (1871) interpreted Cetiosaurus as a plant-
the saurian affinities of Cetiosaurus on the basis eating dinosaur and hypothesized that its limb
of several features, including the absence of epi- bones were “suited for walking.” He could not
physes (growth plates) on caudal vertebrae (fig. rule out the possibility that it was amphibious,
1.1). He differentiated Cetiosaurus from other however, concluding that it was a “marsh-loving
extinct saurians on the basis of its large size and or riverside animal.” Owen (1875:27) later acqui-
characteristics of its vertebrae (see Upchurch esced, referring Cetiosaurus to the Dinosauria
and Martin 2003:215). Owen (1841:462) con- because of its four sacral vertebrae. He admitted
cluded his initial description with this assess- that it may have had some terrestrial capabilities
ment: “The vertebræ, as well as the bones of the but concluded that Cetiosaurus was an estuarine
extremities, prove its marine habits . . . the sur- or marine animal based on its “organ of swim-
passing bulk and strength of the Cetiosaurus ming,” the tail (Owen 1875:41).
were probably assigned to it with carnivorous These early interpretations, based on some-
habits, that it might keep in check the what limited samples, were followed by the
Crocodilians and Plesiosauri.” He regarded discovery of abundant sauropod skeletons in
Cetiosaurus as a crocodilian by the “form of the western North America and eastern Africa during
15
extended across the Atlantic surrounded the
posture of sauropods. American scientists
favored an upright, columnar posture, whereas
their German colleagues deemed a lacertilian
pose more appropriate (Holland 1910;
Desmond 1975). A second question, less con-
troversial but farther-reaching, emerged from
the study of these two large collections of sauro-
pod material—How should sauropod diversity
be classified?
TRADITIONAL CLASSIFICATION
When Marsh (1878) coined the suborder
Sauropoda, it included only a single family,
FIGURE 1.1. Sagittally sectioned posterior caudal vertebra
Atlantosauridae. Several of the features Marsh
of Cetiosaurus oxoniensis (OUM-J13697) with label in
Owen’s hand. This sectioned vertebra was used to demon- (1878:412) listed in that initial diagnosis of
strate the lack of epiphyses at either end of the caudal cen- Sauropoda are now well-corroborated synapo-
trum. Scale equals 5 cm. morphies for the group or for more exclusive
sauropod subgroups that were not identified
the late nineteenth and early twentieth centuries. at the time of Marsh’s writing. Marsh invented
O.C. Marsh and E.D. Cope described numerous new families to accommodate the increasing
new and well represented sauropod genera sauropod diversity revealed by new discoveries
from the Morrison Formation of the western worldwide (e.g., Atlantosauridae, Morosauridae,
United States, including the first complete Diplodocidae, Pleurocoelidae, Titanosauridae).
sauropod skull (Diplodocus [Marsh 1884]), The formal familial diagnoses for these groups
reconstructions of the skeletons of Brontosaurus (Marsh 1884, 1895) also recognized features
by Marsh (1883; fig. 1.2) and Camarasaurus by currently considered synapomorphies for sauro-
Cope (Osborn and Mook, 1921:pl. 82; fig. 1.2), pod subclades. These diagnoses, however, did
and the first mount of a complete sauropod not resolve how these groups were interrelated;
skeleton (Diplodocus [Anonymous 1905]). These Marsh’s ranked classifications did not function
discoveries provided the first examples of onto- as hypotheses of evolutionary descent.
genetic variation and phylogenetic diversity in On the basis of his burgeoning Tendaguru
sauropods. Later, German expeditions to East collection, Janensch (1929a) produced a very dif-
Africa (present-day Tanzania) produced sauro- ferent classification of Sauropoda that employed
pod material rivaling that from North America. higher level groupings. He recognized two prin-
Janensch and others led field crews at cipal sauropod subgroups, one with broad, later-
Tendaguru, where they collected more than ally facing nares and spatulate tooth crowns and
235,000 kg of fossils (Maier 2003:105) that rep- the other with elevated, dorsally facing nares and
resented many new genera described over the narrow tooth crowns. Janensch named these two
course of 50 years (e.g., Janensch, 1914, 1929a, families Bothrosauropodidae and Homalosauro-
1935–36, 1950, 1961). The abundance and podidae, and recognized three and four subfam-
diversity of sauropod remains unearthed in ilies within each, respectively. Huene (1956) fol-
North America and Africa not only answered lowed this dichotomous scheme, raising
many of the queries posed by early sauropod Janensch’s subfamilies to familial rank and
researchers (e.g., dinosaurian affinities and ter- Janensch’s families to “family-group” rank. In
restrial habits of sauropods) but also posed new contrast to that of Marsh, Janensch’s classifica-
ones. One of the major controversies that tion could be interpreted as an evolutionary
16 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

FIGURE 1.2. First reconstruction of sauropod dinosaurs. Top: “Brontosaurus” (=Apatosaurus) excelsus from Marsh (1883). From a lithograph later published in Ostrom and McIntosh (1966).
Bottom: Camarasaurus supremus, as drawn by Ryder in 1877 under the direction of Cope. Later published in Osborn and Mook (1921:pl. 82).
hypothesis that involved divergence between two derived feature characterizing Diplodocus-like
lineages differing in tooth morphology. taxa (i.e., Diplodocoidea) and titanosaurs.
A dichotomous scheme for higher-level clas- Salgado et al. (1997) were the first to depart
sification of sauropods based on tooth form and from this traditional dichotomy by providing
narial position became widely accepted, despite character evidence linking narrow-crowned
nomenclatural differences (Brachiosauridae titanosaurs to the broad-crowned Brachiosaurus,
versus Titanosauridae [Romer 1956, 1966]; rather than to the other narrow-crowned group
Camarasauridae versus Atlantosauridae [Steel (Diplodocoidea). This result was corroborated
1970]). Other traditional classifications of by Wilson and Sereno (1998). In a subsequent
sauropods, however, follow Marsh in recogniz- analysis, Upchurch (1998) produced a topol-
ing taxa of equivalent rank (usually families) ogy that agreed in many ways with those of
with no higher-level hierarchical information Salgado et al. (1997) and Wilson and Sereno
(e.g., McIntosh 1990). Bonaparte (1986a) also (1998) but also explored the relationships of
utilized serially ranked families, but he genera not treated by either. These three analy-
regarded Late Jurassic and younger sauropod ses agree on several topological points, includ-
families (“Neosauropoda”) as advanced relative ing the separation of early-appearing genera
to older forms (“Eosauropoda”). (e.g., Vulcanodon, Shunosaurus, Barapasaurus,
Numerical methods for assessing phyloge- Omeisaurus) from a derived clade called
netic relationships in sauropod dinosaurs were Neosauropoda (Bonaparte 1986a), the identifi-
first introduced by Gauthier (1986) in his analy- cation of the two constituent neosauropod lin-
sis of saurischian dinosaurs. His character eages Diplodocoidea (e.g., Apatosaurus) and
choice reflected those cited by previous authors Macronaria (e.g., Camarasaurus), and the posi-
(e.g., Romer 1956; Steel 1970) and his topology tioning of the titanosaur lineage within
consequently conformed to the traditional Macronaria (fig. 1.3).
dichotomy. Since then, more than a dozen Despite points of agreement, other topologi-
cladistic analyses focusing on Sauropoda or its cal differences persist. The most significant of
subgroups have appeared (Russell and Zheng these centers on the phylogenetic affinities of
1993; Calvo and Salgado 1995; Upchurch 1995, two groups of Asian sauropods: the Chinese
1998; Salgado et al. 1997; Wilson and Sereno “euhelopodids” (Shunosaurus,” Omeisaurus,
1998; Sanz et al. 1999; Curry 2001; Curry Mamenchisaurus, Euhelopus) and the Mongolian
Rogers and Forster, 2001; Wilson 2002; Calvo nemegtosaurids (Nemegtosaurus, Quaesitosaurus).
and González Riga 2003; González Riga 2003; Upchurch (1995) proposed “Euhelopodidae” as
Upchurch et al. 2004). Together these analyses a clade that evolved while China was geograph-
have scored 1,964 characters in 229 sauropod ically isolated from Europe from Middle
taxa, resulting in a variety of phylogenetic Jurassic until Early Cretaceous times (Russell
hypotheses that are discussed briefly below. 1993; Z. Luo 1999; Barrett et al. 2002;
Upchurch et al. 2002; Zhou et al. 2003). It
CLADISTIC HYPOTHESES evolved independently of its sister-taxon
The main topological disagreement among Neosauropoda but was eventually replaced by it
early cladistic analyses of Sauropoda centered during the Cretaceous (Upchurch 1995, 1998).
on the relationships of broad- and narrow- In contrast, Wilson and Sereno (1998) sug-
crowned sauropods. Upchurch (1995) pre- gested that Chinese sauropods are paraphyletic,
sented the first large-scale cladistic analysis of with Omeisaurus occupying the sister-taxon to
sauropods, in which he proposed a slightly Neosauropoda (as in Upchurch 1995, 1998),
modified version of the traditional dichotomy but Shunosaurus positioned basally and
that resolved broad tooth crowns as a primitive Euhelopus positioned apically. This result was
feature and narrow tooth crowns as a uniquely corroborated by Wilson (2002), whose analysis

Eu hio uru rus
Br ara CO rus
Br ara hos A
TI hios uru EA
E
ac sa au
u
IA
IA
am n ID
ac sa ID
am DO osa
he sa s
TA a s
TA u s
SA s
D isa rus
O osa rus
R
C oca CO
O pas us
TI lop uru
Eu isa us
N uru
U
U
H OD s
H lop s
Ba os n
Sh pas n
LO th
L ru
he uru
ra aur
e u
SA
e ur
un au
N s
s
un do
ra do
ap O
t
m a
ap u
IP n
IP u
D oca
Sh ano
Ba ano
O
l
l
lc
lc
m
Vu
Vu
C
NEOSAUROPODA
NEOSAUROPODA EUHELOPODIDAE
Wilson & Sereno Upchurch EUSAUROPODA
(1998) EUSAUROPODA (1998)
FIGURE 1.3. Hypotheses of the relationships of sauropod dinosaurs based on (left) Wilson and Sereno (1998) and (right)
Upchurch (1998).
resolved some of Upchurch’s (1998) “euhelopo- Rogers and Forster 2001; Wilson 2002, 2005a).
did” characters as supporting the monophyly of Although the weight of the evidence is in favor
Omeisaurus and Mamenchisaurus (Omeisauridae). of titanosaur affinities for Nemegtosaurus and
A Templeton test (e.g., Larson 1994) showed Quaesitosaurus, convergences with diplodocoids
that “euhelopodid” paraphyly could not be sta- are noteworthy (Upchurch 1999; Curry Rogers
tistically rejected by the matrix of Upchurch and Forster 2001; see below).
(1998), but the “euhelopodid” monophyly could In addition to areas of disagreement, there
be rejected by the matrix of Wilson (2002). are unresolved areas resulting from lack of
Thus far, no other analysis has specifically information. Two such areas involve the origin
investigated the relationships of these Chinese of sauropods and the diversification of their lat-
sauropods, but Upchurch’s most recent analysis est surviving lineage, Titanosauria. Sauropods
supported paraphyly of some “euhelopodid” have long been absent from Triassic rocks, but
genera (see Upchurch et al. 2004; Barrett and their two saurischian sister-taxa (Prosauropoda,
Upchurch, chapter 4). Theropoda) are found in lowermost Upper
A second area of disagreement involves the Triassic horizons. Recent discoveries of Triassic
relationships of the isolated skulls of the sauropod body fossils and ichnofossils (see
sauropods Nemegtosaurus and Quaesitosaurus below) have provided the first opportunity to
from the Late Cretaceous of Mongolia. These resolve sauropod origins, but additional field
slender-crowned taxa were originally described and museum research is needed. Renewed
as Dicraeosaurus-like (Nowinski 1971), a desig- interest in titanosaurs, whose interrelation-
nation consistent with the presumed diplodocid ships remain resolved, have been fueled by
affinities of the Late Jurassic Chinese descriptions of many new discoveries in the
Mamenchisaurus (McIntosh 1990), as well as field (Curry Rogers, chapter 2). These include
the conventional division of sauropods into nar- the first titanosaur with associated cranial and
row-crowned and broad-crowned groups. More cranial remains (Rapetosaurus Curry Rogers
recently, cladistic analyses have produced new and Forster 2001, 2004), the first embryonic
hypotheses of relationships for Nemegtosaurus titanosaur remains (Chiappe et al. 1998, 2001;
and Quaesitosaurus, including the mono- Salgado et al. 2005), and nearly complete asso-
phyletic sister-taxon of diplodocoids (Yu 1993; ciated or articulated postcranial skeletons from
Upchurch 1998, 1999; Upchurch et al. 2002), South America (Mendozasaurus González Riga
basal members of a clade including diplodocoids 2003; Epachthosaurus Martínez et al. 2004;
and titanosaurs (Upchurch 1995), and, most Gondwanatitan Kellner and Azevedo 1999),
recently, titanosaurs (Salgado et al. 1997; Curry Asia (Phuwiangosaurus Martin et al. 1994;
OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 19

us di
a
us ur s s
s us ur sa ru s rus us s s us s uru au s us
s u s a o u r r u r u s u l i c ur s
on uru aur aur uru his n th rus isa uru au au uru rus us r
au sau us aur au osa s coe aur nsa uru
u
d a s c a u h a a s s a u c s p is s t u o s e
no os pa gos isa en ria c
oc sa a ys g eo s a do
a
ar hio lo w to eg ur th o u sa
lca un ra ta me am ba pl ger bb a ar cra ato ros plo m ac he ala pe m isa pis am uq lta
Vu Sh Ba Pa O M Jo Ha Ni Re Lim Am Di Ap Ba Di Ca Br Eu M Ra Ne Is O Al Ne Sa
SALTASAURINAE
DIPLODOCIDAE OPISTHOCOEL’INAE
DICRAEOSAURIDAE SALTASAURIDAE
REBBACHISAURIDAE
DIPLODOCOIDEA TITANOSAURIA
SOMPHOSPONDYLI
TITANOSAURIFORMES
MACRONARIA
WILSON 2002
length = 430
mpt = 3
NEOSAUROPODA CI = 0.66
RI = 0.80
EUSAUROPODA
SAUROPODA
FIGURE 1.4. Phylogenetic relationships of sauropod genera based on Wilson (2002). Dashed lines indicate nodes that are
lost in trees two steps longer than the most parsimonious tree. Taxonomy has been updated for Rayososaurus (=Limaysaurus
[Salgado 2004]) and “Titanosaurus” colberti (=Isisaurus [Wilson and Upchurch 2003]).
Tangvayosaurus Allain et al. 1999), India is presented that has been identified in various
(Isisaurus Jain and Bandyopadhyay 1997), analyses (table 1.1). Appendix 1.1 lists each char-
Europe (Lirainosaurus Sanz et al. 1999; acter and its states.
Ampelosaurus Le Loeuff 1995, 2003), and Africa
(Malawisaurus Jacobs et al. 1993; Paralatitan
MAJOR EVOLUTIONARY EVENTS IN
Smith et al. 2001). Several analyses have investi-
SAUROPODA AND ITS SUBGROUPS
gated titanosaur phylogeny (most notably Curry
[2001] and Curry Rogers and Forster [2001]), and Sauropoda is a monophyletic group whose body
there are several points of agreement among plan (fig. 1.2) is supported by more than 40
them (Wilson and Upchurch 2003). These pre- synapomorphies, many of which were not lost
liminary analyses are the first step toward estab- within the 150 million-year history of the group
lishing a framework for titanosaur evolutionary (McIntosh 1990; Upchurch 1995, 1998; Wilson
history, but at least a dozen valid titanosaur gen- and Sereno 1998). Modification of this basic
era have yet to be accommodated by a phyloge- architecture, as it pertains to the evolution of her-
netic analysis, in addition to the many unde- bivory, neck elongation, and locomotion within
scribed specimens uncovered in recent years. five clades (Sauropoda, Eusauropoda, Neosau-
The topology of Wilson’s (2002) analysis of ropoda, Diplodocoidea, Macronaria) is explored
Sauropoda, based on 27 taxa scored for 234 here. Important to this discussion is the pre-
characters, is assumed in this paper (fig. 1.4). sumed ancestry of Sauropoda, which is not yet
Outgroup choice, character descriptions, char- agreed on. Whereas most researchers favor a
acter coding assumptions, character–taxon monophyletic Prosauropoda (Sereno 1989;
matrix, and tree statistics are given by Wilson Galton 1990; Wilson and Sereno 1998; Galton
(2002). Below, the evolutionary events diagnos- and Upchurch 2000, 2004; Benton et al. 2000),
ing several major sauropod clades are dis- recent analyses of sauropodomorph relationships
cussed. For each event, a set of synapomorphies (Yates 2001 2003, 2004; Yates and Kitching

TABLE 1.1.
Synapomorphies for the Five Sauropod Clades Discussed
CHARACTER NUMBER, BY CLADISTIC ANALYSIS
SALGADO ET AL. UPCHURCH WILSON & SERENO WILSON

(1997) (1998) (1998) (2002)
Columnar, quadrupedal posture (Sauropoda)

elongate forelimbs — 158 1 172
elongate metatarsal V — — 15 225
straight limb elements 4 186 1 149
reduction of olecranon — 161 4 167
femur with eccentric cross-section — 191 10 198
unossified limb articular surfaces — — — —
unossified distal carpals — 164 — 173
unossified distal tarsals (3 & 4) — 197 13 216
Herbivorous specializations (Eusauropoda)
tooth rows shortened — 73 67 66
precise occlusion — — 35 67
tooth rows arched — 59 31 65
teeth overlap — — 34 69
enamel wrinkling — — 33 71
broad crowns — 71 32 70
dentary deepens anteriorly — 57 30 55
Neck elongation (Eusauropoda)
number of neck vertebrae 5 76 37 80
number of dorsal vertebrae — 95 70 91
Hindfoot posture (Eusauropoda)
pes shortened relative to tibia — — 50 223
spreading metatarsus — — 52 217
metatarsal I broader than II–V — — 51 221
pedal phalangeal count reduced — 200 57 233
metatarsal II broader than III–IV* — — 73 224
pedal unguals directed laterally* — — 64 228
Reduced ossification of wrist & ankle (Neosauropoda)
reduction to two carpals — 163 79 173
astragalus reduced — 195 85 210
Forefoot posture (Jobaria Neosauropoda)
bound metacarpus — 169 80 175
tightly arched metacarpus — 169 81 176
Herbivorous specializations (Diplodocoidea)
tooth row restricted anteriorly — 74 — 66
mandible squared in dorsal view — 59 — 65
jaw articulation shifted forward — ?27 — 46, 53
pterygoid flange and adductor — — — 37
fossa shifted forward
loss of crown overlap — — — 69
cylindrical tooth crowns — 70 — 70
enhanced tooth replacement rate — — — 74
TABLE 1.1. (continued)
CHARACTER NUMBER, BY CLADISTIC ANALYSIS
SALGADO ET AL. UPCHURCH WILSON & SERENO WILSON

(1997) (1998) (1998) (2002)
Presacral specializations (Flagellicaudata)

forked neural spines — 92 106 85, 89
elongate neural spines* — — — 93
number of cervical vertebrae* — 77 37 80
number of dorsal vertebrae* — 95 70 91
Tail specializations (Diplodocoidea)
elongate caudal centra — 134 — 137
biconvex caudal centra — 134 — 136
30 or more archless caudal centra* — 128 — 138
Wide-gauge limb posture (Saltasauridae)
femur distal condyles beveled — — — 201
eccentric femoral midshaft — — — 198
coracoid quadrangular† 29 153 — 156
scapular blade deflected dorsally† — — — 151
crescentic sternal plates† 26 154 — 158
humeral distal condyles exposed anteriorly — — — 163
humeral distal condyles divided — — — 164
humeral deltopectoral crest expanded — — 3 161
prominent olecranon† — 161 4 167
distal radius expanded transversely† — — — 170
distal tibia expanded transversely† 7 — — 205
iliac blade directed laterally† 28 172 — 187
femur deflected medially† 19 187 100 199
carpus unossified* — 165 79 173
manual phalanges absent* 27 — 43 181
NOTE: Character numbers are those employed in four major cladistic analyses of sauropod relationships. Asterisks (*) denote synapo-
morphies that apply at slightly less inclusive nodes; daggers (†) denote synapomorphies that apply at slightly more inclusive nodes (see
text for details).
2003) resolve taxa considered “prosauropods” to should be called Prosauropoda, the derived
be paraphyletic. Although the earliest of these sauropod-like forms should be included in
analyses supports a fully pectinate arrangement Sauropoda, and taxa resolved as outgroups to
of “prosauropods” (Yates 2001, 2003:fig. 22), the those clades are non-sauropodomorph saurischi-
most recent analyses resolve a monophyletic core ans. The phylogenetic definitions for this
of prosauropods flanked basally by primitive node–stem triplet are as follows (Sereno
forms and apically by sauropod-like forms (Yates 1998:table 4) (boldface type indicates node-based
and Kitching 2003:fig. 4; Yates 2004:fig. 13). definitions; regular type indicates stem-based
Sereno (1998) specified phylogenetic definitions definitions):
that designate Prosauropoda and Sauropoda
reflexive stem-based clades that comprise the Sauropodomorpha Huene 1932—
node-based Sauropodomorpha. Applying this Plateosaurus engelhardti, Saltasaurus lorica-
phylogenetic definition to the Yates and Kitching tus, their most recent common ancestor
(2003:fig. 13) topology, the monophyletic core and all descendants.

Prosauropoda Huene 1920—All columnar, graviportal posture, which involved
sauropodomorphs closer to Plateosaurus independent changes in limb proportions, pos-
engelhardti than to Saltasaurus loricatus. ture, and ossification.
Sauropoda Marsh 1878—All
sauropodomorphs closer to Saltasaurus COLUMNAR, QUADRUPEDAL POSTURE
loricatus than to Plateosaurus engelhardti. Outgroups to Sauropoda are primitively bipedal
and characterized by relatively short forelimbs
Below, I summarize the major specializa- that generally represent less than half the length
tions relating to herbivory, neck elongation, and of the hindlimb. In these forms, the proximal
locomotion for each of five major sauropod hindlimb is shorter than the distal hindlimb,
clades. The synapomorphies discussed are nearly half of whose length is provided by the
listed in table 1.1, alongside their usage in vari- metatarsus. Sauropoda is characterized by modi-
ous cladistic analyses of sauropod relation- fications of proportions both within and between
ships. Appendix 1.1 gives a full character list the fore- and the hindlimbs, a modification
with primitive and derived states. related to quadrupedalism. Sauropods have elon-
gate forelimbs that are at least 70% of the
SAUROPODA hindlimb length, nearly twice that of their out-
Probable sauropod body fossils and ichnofossils groups (table 1.2, fig. 1.5). This change was
are present in Upper Triassic (Carnian) sedi- accommodated by an overall lengthening of the
ments, but their referrals require confirmation forelimb, especially the distal elements (fig. 1.6),
(summarized in Wilson 2005b). The partial and an overall shortening of distal hindlimb ele-
hindlimb of Blikanasaurus is proportioned simi- ments relative to the proximal element (fig. 1.7).
lary to those of later sauropods (Yates 2003, Reduction of the distal hindlimb did not include
2004; Yates and Kitching 2003; Upchurch et al. metatarsal V, which attains at least 70% of the
2004), but correlation with body size cannot yet length of metatarsal IV in all sauropods, effecting
be ruled out. Likewise, trackways from the Upper a more symmetrical pes with five weight-bearing
Triassic (Carnian) Portezuelo Formation of digits. Although lengthening of the forelimb and
West–Central Argentina resemble those of later relative shortening of the distal hindlimb charac-
sauropods, but their identification remains tenta- terize all sauropods, future discoveries may sug-
tive (Marsicano and Barredo 2004). The oldest gest that these features are not correlated. Both
definitive sauropod fossils are the Tetrasauropus early sauropod body fossils and ichnofossils sug-
trackways preserved in the Chinle Group of west- gest that quadrupedalism evolved in sauropods
ern North America, which are Norian–Rhaetian sometime prior to the Late Triassic (Wilson
in age (ca. 210 mya [Lockley et al. 2001; see 2005b). Adoption of a quadrupedal pose within
Wright, chapter 9]). Slightly younger or coeval Sauropoda represents one of four such acquisi-
?Rhaetian strata in Thailand preserve the frag- tions within Dinosauria, each of which is associ-
mentary remains of Isanosaurus (Buffetaut et al. ated with body size increase (Carrano 2000,
2000). Isanosaurus may be more derived than the 2005; see Carrano, chapter 8).
slightly younger Vulcanodon (Raath 1972), which Associated with the proportional changes
is generally considered the most primitive sauro- that facilitate a quadrupedal pose are specializa-
pod (Wilson 2002: fig.13, table 13). Because the tions that allow a columnar, rather than flexed,
basalmost sauropods Vulcanodon, Isanosaurus, limb posture. In basal dinosaurs, the disposition
and Gongxianosaurus (fig. 1.5) lack complete cra- of limb articular surfaces and shaft curvature
nial remains and much of the vertebral column, suggest a slightly flexed resting pose for the hip,
the majority of the features diagnosing knee, shoulder, and elbow joints. In sauropod
Sauropoda are appendicular synapomorphies. Of outgroups, for example, the anteroposterior cur-
these, many are related to the adoption of a vature of the femur offsets proximal and distal

1m
FIGURE 1.5. Silhouette skeletal reconstruction of Gongxianosaurus shibeiensis in left lateral view. Reconstruction based on
unnumbered specimens pertaining to three individuals described by He et al. (1998). The majority of the skeleton pertains
to a possibly subadult individual represented by an articulated pectoral girdle and forelimb and an articulated hindlimb that
were discovered in association (He et al. 1998:1). The two series of articulated caudal vertebrae likely pertain to a distinct,
adult individual, as does the premaxilla. Both the caudal series and the premaxilla have been scaled to the size of the appen-
dicular elements. The relative size of missing elements (i.e., skull, neck, trunk, manus) was based on the basal sauropods
Vulcanodon and Shunosaurus. Additional elements attributed to Gongxianosaurus (Luo and Wang 2000) are not included in
this reconstruction because they have not yet been figured or described in detail.
condyles of the femur approximately 20° from dorsal surface of the ulna and tibia, respectively,
horizontal (fig. 1.8A, B). Likewise, bony extensor in Vulcanodon, Gongxianosaurus (fig. 1.5), and
processes on the ulna (olecranon) and tibia most other sauropods (fig. 1.7). Reduction of
(cnemial crest) are prominent in immediate these processes suggests a more columnar
sauropod outgroups but do not project above the alignment of the elbow and knee joints. In addi-
TABLE 1.2
Limb Proportions in Selected Saurischian Genera
FORE:HIND MT III:TIBIA REFERENCE(S)
Theropoda
Eoraptor 0.43 0.43 Sereno (pers. comm.)
Herrerasaurus 0.47 0.52 Sereno (1993), Novas (1993)
Prosauropoda
Jingshanosaurus 0.42 0.57 Zhang & Yang (1994)
Lufengosaurus 0.50 0.57 Young (1941)
Plateosaurus 0.52 0.48 Huene (1926)
?Sauropoda
Blikanasaurus — 0.36 Galton & van Heerden (1985)
Antenonitrus 0.81 0.38 Yates & Kitching (2003)
Sauropoda
Vulcanodon 0.78 0.37, 0.32 Raath (1972), Cooper (1984)
Gongxianosaurus 0.62 0.38 He et al. (1988)
Shunosaurus 0.67 0.27 Zhang (1988)
Omeisaurus 0.90 0.28 He et al. (1988)
Jobaria 0.88 0.28 Sereno et al. (1999)
Apatosaurus 0.72 0.21 Gilmore (1936)
Camarasaurus* 0.83 0.24 Gilmore (1925)
Camarasaurus (0.85) 0.24 McIntosh & al. (1996)
Opisthocoelicaudia 0.79 0.25 Borsuk-Bialynicka (1977)
NOTE: Forelimb length equals the sum of the lengths of the humerus, radius, and longest metacarpal; hindlimb length equals the sum
of the lengths of the femur, tibia, and longest metatarsal. Asterisk(*) indicates measurement of a juvenile individual; parentheses indi-
cate an estimated value. Abbreviation: mt, metatarsal.

FIGURE 1.6. Forelimb proportions in the prosauropods Lufengosaurus and Plateosaurus and the basal sauropods Vulcanodon,
Shunosaurus, and Omeisaurus. Forelimbs have been scaled to the same humeral length. Based on Young (1947), Huene
(1926), Raath (1972), Zhang (1988), and He et al. (1998), respectively.
FIGURE 1.7. Hindlimb proportions in the prosauropods Lufengosaurus and Plateosaurus and the basal sauropods
Vulcanodon, Shunosaurus, and Omeisaurus. Hindlimbs have been scaled to the same femoral length. Based on Young (1947),
Huene (1926), Raath (1972), Zhang (1988), and He et al. (1998), respectively.
FIGURE 1.8. Femoral curvature
in the saurischian dinosaurs
Herrerasaurus (A),
Massospondylus (B), Vulcanodon
(C), and Isanosaurus (D).
Femora are figured in right
medial view and have been scaled
to the same length to facilitate
comparison. The left femur of
Isanosaurus has been reversed.
Based on Novas (1993), Cooper
(1981), Cooper (1984), and
Buffetaut et al. (2000),
respectively. A B C D
FIGURE 1.9. Left humerus of Camarasaurus grandis (YPM 1901) in anterior (left), lateral (middle), posterior (right), proximal
(top), and distal (bottom) views (from Ostrom and McIntosh 1966:pl. 49). Scale bar equals 30 cm.
tion, the longest weight-bearing elements in the ular ends have a rugose, irregular surface,
skeleton (humerus, femur) have eccentric mid- whereas their shafts are smooth. Owen
shaft cross sections that are broader mediolater- (1841:461) recognized this feature in Cetiosaurus,
ally than anteroposteriorly. Distal limb elements noting that “the articular surfaces which are pre-
(radius/ulna, tibia/fibula) do not share this served are covered with large tubercles for the
cross-sectional geometry, but they bear weight attachment of thick cartilage.” Similarly, Marsh
in tandem and are together broader mediolater- (1878:413) described the humerus of Cama-
ally than anteroposteriorly. rasaurus as “rough, and well covered with carti-
Reduced ossification of limb elements repre- lage” (fig. 1.9). The thickness of this cartilage cap
sents the third major appendicular specialization has not yet been estimated but is implied in
characterizing Sauropoda. A conspicuous fea- articulated skeletons by the difference in vol-
ture of sauropod limb elements is that their artic- umes of the acetabulum and femoral head. The

proximal carpal and tarsal elements have few to HERBIVOROUS SPECIALIZATIONS
no nonarticular surfaces and are completely The Middle Jurassic Shunosaurus is the earliest-
made up of rough, rugose bone. Consequently, appearing sauropod known from well-preserved
the configuration of sauropod wrist and ankle cranial remains. All cranial synapomorphies of
elements relative to adjacent elements is difficult Eusauropoda are ambiguous and may later be
to determine because little of the articular shown to characterize more inclusive groups.
surfaces remains. Distal carpals have not been Shunosaurus possessed a sophisticated dental
identified in any sauropod skeleton, and distal apparatus that is highly modified relative to that
tarsals have only been recovered for of prosauropods, indicating that eusauropods
Gongxianosaurus, in which discoidal ossifica- modified the shape of the crowns as well as their
tions are preserved atop metatarsal III and arrangement along the tooth row. Principal
between metatarsal IV and metatarsal V (He et among these changes is the acquisition of pre-
al. 1998:fig. 4C; fig. 1.5). Retention of ossified cisely occluding dentition, a feature that is
distal tarsals may suggest that Gongxianosaurus unknown elsewhere in Saurischia.
is the most primitive sauropod, but their absence Prosauropods and theropods primitively
in other basal sauropods (e.g., Vulcanodon) has have lower tooth rows that extend the length of
not yet been confirmed by articulated material. the dentary but upper tooth rows that extend
farther posteriorly to midorbit. With different
EUSAUROPODA lengths and numbers of teeth, upper and lower
Eusauropoda is the node-based group including teeth have mismatched occlusion that gener-
Shunosaurus lii, Saltasaurus loricatus, their most ates no regular wear pattern. Additionally,
recent common ancestor, and all descendants prosauropods and theropods have tooth rows
(fig. 1.4). This definition specifies all named that are relatively straight in dorsal or ventral
sauropods except Vulcanodon, Gongxianosaurus, view. Right and left sides meet at an acute
and Rhoetosaurus, as well as the possible early angle, and none of the teeth are oriented trans-
sauropods Blikanasaurus and Antenonitrus. The versely (fig. 1.10A, B). Sauropods differ in all of
oldest well-preserved eusauropod is the Middle these respects. Nearly all sauropods known by
Jurassic Shunosaurus, which is known from cranial remains have tooth rows that are of even
several complete skeletons (Zhang 1988). length and contain similar numbers of teeth.
Consequently, many of the synapomorphies The upper tooth row terminates at or in front of
diagnosing Eusauropoda are ambiguous and the antorbital fenestra, and the dentary always
may obtain a broader distribution once basal has an edentulous region posterior to the last
forms are known more completely. Eusauropod tooth. In dorsal view, the tooth rows are curved
synapomorphies greatly outnumber those of rather than straight, and at least two teeth are
any other node within Sauropoda—Wilson oriented transversely (fig. 1.10C, D). Together,
(2002) reported 53, fewer than half of which these changes signal precise occlusion in
could be scored in more basal taxa. Thus, the sauropods, as evidenced by crown wear facets
ambiguous (i.e., cranial and axial) synapomor- generated by tooth-to-tooth wear (Calvo 1994).
phies may have evolved as early as the diver- In dorsal view, the tooth rows are outwardly
gence of Sauropoda from Prosauropoda in ear- arched rather than straight, and at least two
liest Late Triassic (Carnian, 220 mya [Flynn et teeth are oriented transversely (fig. 1.10C, D).
al. 1999]). The unambiguous (i.e., hindlimb) The entire tooth row is transversely oriented in
synapomorphies, on the other hand, signal some sauropods (see “Diplodocoidea,” below).
more recent modifications since the divergence Most sauropods develop an imbricate arrange-
of Eusauropoda from Sauropoda in the Late ment of teeth in which the mesial edge of each
Triassic (Rhaetian, 210 mya [Buffetaut et al. tooth is overlapped by the distal edge of the pre-
2000]). ceding tooth.

A B C D
FIGURE 1.10. Snouts of the theropod Herrerasaurus (A), the prosauropod Plateosaurus (B), and the sauropods Brachiosaurus
(C) and Diplodocus (D) in dorsal view. Based on reconstructions from Sereno (1993) and Wilson and Sereno (1998)
Eusauropod tooth crowns also have distinc- vertebrae to achieve the primitive eusauropod
tive shape and texture. All teeth have a charac- precaudal count of 13-13-4.
teristically wrinkled enamel texture whose Later, Patagosaurus, Omeisaurus, and more
function is unknown. Coarseness of enamel derived sauropods acquire a fifth sacral without
wrinkling varies to some extent within changing the precaudal count (13-12-5), which
sauropods, with narrow-crowned teeth usually most likely represents the incorporation of a
exhibiting much finer wrinkling than broad dorsal vertebra into the sacrum (rather than the
tooth crowns. Sauropod tooth crowns are prim- addition of a cervical and loss of a dorsal). Nearly
itively spatulate, with a D-shaped cross section. all of the dozen subsequent neck-lengthening
Precise tooth-to-tooth occlusion is not lost events characterize individual neosauropod gen-
within Sauropoda, but many of the other her- era and are not synapomorphies of larger clades.
bivorous innovations are modified in later line- The exception is Diplodocidae (15-10-5), which
ages, principally Diplodocoidea (see Sereno and incorporated two dorsal vertebrae into the cervi-
Wilson, chapter 5). Because all known sauropod cal series. Thus, there is no progressive increase
skulls share these features, their sequence of in neck length within Sauropoda; rather, indi-
acquisition is not yet known. vidual genera were specialized for their neck
length. All three means of neck lengthening
NECK ELONGATION (incorporation, duplication, elongation) were
The primitive saurischian precaudal vertebral employed within Sauropoda.
count is 27, although the relative number of cer-
vical, dorsal, and sacral vertebrae vary in HINDFOOT POSTURE
Theropoda (9-15-3,respectively) and Prosauropoda Theropods and prosauropods are interpreted as
(10-14-3). Vertebral counts are not known for non- having a digitigrade pes, a posture in which the
eusauropods, but Vulcanodon has a sacrum with heel and proximal metatarsals were held off the
four coosified vertebrae (Raath 1972). The fourth ground, and the distal metatarsals and pha-
sacral vertebra in sauropods is a caudosacral, langes contacted the substrate (Carrano 1997).
based on osteological and developmental evi- Eusauropods are characterized by several
dence (Wilson and Sereno 1998). The eusauro- changes that together result in a unique hind-
pod Shunosaurus (13-13-4) is the basalmost sauro- foot posture that is easily recognized in foot-
pod genus for which the vertebral count is prints (fig. 1.11). These include the independent
known. Compared to outgroups, eusauropods modification of the length, arrangement, and
are characterized by two neck elongation events: robustness of the metatarsus, as well as the
(1) incorporation of one dorsal vertebra into the reduction in the number and size of the pedal
cervical series and (2) duplication of two cervical phalanges.

A
FIGURE 1.11. Right pes of Apatosaurus in proxi-

mal (A) and dorsal (B) views; C, right pes print of
?Brontopodus oriented relative to the trackway mid-
line (arrow). Apatosaurus modified from Gilmore
(1936:figs. 25, 27, 28); ?Brontopodus modified from
Thulborn (1990:fig. 6.16f). Abbreviations: I-V, dig-
C its I-V.
Sauropod outgroups have long distal ends are not in mutual contact, as they are in
hindlimbs, in which the metatarsus accounts for sauropod outgroups. In dorsal view, for example,
40% to 50 of the tibial length (table 1.2). In con- the metatarsal shafts are separated by interven-
trast, the eusauropod metatarsus is markedly ing spaces (fig. 1.11B). These changes effect a
abbreviated and comprises less than 25% of the more spreading hindfoot posture in which the
tibial length. The proportions of the basal metatarsus was held in a subhorizontal, rather
sauropods Vulcanodon and Gongxianosaurus, as than subvertical, orientation. Wilson and Sereno
well as those of Blikanasaurus and Antenonitrus, (1998:41) recognized this as a “semi-digitigrade”
are intermediate between sauropod outgroups foot posture. Carrano (1997:fig. 1B) termed the
and eusauropods such as Shunosaurus (figs. 1.5, inferred foot posture in sauropods “sub-
1.7, table 1.2). In addition to these proportional unguligrade,” referring to the specialized foot
changes, the eusauropod metatarsus attains a posture of hippopotamids, rhinoceratids, and
spreading configuration in which the proximal proboscideans, in which the metatarsus is held

vertically, a fleshy pad supports the foot, and metatarsus (table 1.2), implying that body weight
the penultimate and ungual phalanges contact was borne subequally by its constituent elements.
the substrate. Although the hypothesized sauro- The eusauropod pes, in contrast, displays marked
pod hindfoot posture is similar to “sub- asymmetry of metatarsal shaft diameters in
unguligrady” (viz. the fleshy heel pad), the which metatarsal I is broader at than all others.
metatarsus is thought to have been held in a The disparity among metatarsals II–V becomes
nearly horizontal rather then a vertical orienta- more pronounced in more derived sauropods.
tion, and the nonungual phalanges are hypothe- Omeisaurus, Mamenchisaurus, and Neosauropoda
sized to have contacted the substrate. The term are diagnosed by a metatarsus in which the min-
semi-digitigrady is used here to refer to the foot imum shaft diameters decrease laterally such that
posture hypothesized for Eusauropoda. the diameters of metatarsals III and IV are 50%
Although the earliest sauropods are inter- to 60% that of metatarsal II (table 1.3). Hatcher
preted as having a digitigrade posture, there are (1901:51) noted this pattern and suggested that
no footprints attributed to sauropods that indi- “the weight of the body was borne by the inner
cate such a foot posture. Rather, the earliest side of the foot.” This feature is manifest in well-
sauropod trackways bear elongate pes prints that preserved sauropod footprints, in which the inner
indicate a semidigitigrade hindfoot posture (e.g., margin is more deeply impressed than the outer
Portezuelo trackways [Marsicano and Barredo margin (e.g., Pittman and Gillette 1989:322).
2003], Tetrasauropus [Lockley et al. 2001]). This The acquisition of a semidigitigrade hind-
22 million- to 44 million-year discrepancy foot posture is accompanied by reduction of the
between the first appearance of semi-digitigrade phalangeal portion of the pes. Prosauropods
pedal posture in the body fossil and that in the and basal theropods retain a full complement of
ichnofossil records may indicate early appear- pedal phalanges on digits I–IV that invariably
ance of eusauropods, homoplasy, or that hindfoot number 2-3-4-5, each digit bearing an ungual
posture has been erroneously interpreted in early phalanx (table 1.4). The possible basal sauropod
sauropods. Based on a stratocladistic analysis of Blikanasaurus retains the same count. Although
ichnological and body fossil data, Wilson (2005b) the pedal phalangeal formula is not known in
suggested that semi-digitigrady evolved in the Vulcanodon, its penultimate phalanges resemble
Late Triassic and was either reversed or misinter- those of prosauropods and are not drastically
preted in the early sauropods Vulcanodon and shortened (Cooper 1984:figs. 34, 35). The articu-
Gongxianosaurus. A preliminary study of skeletal lated hindfoot of Gongxianosaurus confirms that
remains referred to Plateosaurus has inferred a basal sauropods maintained a high number of
less digitigrade posture than traditionally posited phalanges (2-3-4-5), which themselves were
for prosauropods (Sullivan et al. 2003), which longer than broad (fig. 1.5). The pes of
underscores difficulties in determining locomo- eusauropods is reduced in both the number and
tor posture from osteology. the size of phalangeal elements. The penulti-
In addition to revising the temporal origin mate phalanx in digits II–IV is reduced to a
of semi-digitigrade hindfoot posture in early plate-shaped disc or lost in Shunosaurus,
sauropods, ichnofossils indicate that the sub- Omeisaurus, and various neosauropods. The
horizontal foot was supported by a fleshy heel greatest number of phalanges retained in
(fig. 1.11C). eusauropod digit IV is three (e.g., Shunosaurus,
Eusauropods are also characterized by a Omeisaurus, Camarasaurus), two fewer than the
departure from the within-pes proportions that outgroup condition of five. Despite the loss of two
characterize other saurischians. Body weight in phalanges, an ungual is maintained on digit IV.
theropods and prosauropods is accommodated Eusauropods clearly demonstrate non-terminal
by three and four pedal digits, respectively. In phalangeal reduction, which maintains the size
these taxa, shaft breadth varies little across the and functionality of the unguals amid substantial

TABLE 1.3
Pedal Proportions in Select Saurischian Genera
MT I:II:III:IV MT V:IV
(MINIMUM BREADTH) (LENGTH)
Theropoda
Eoraptor — 0.56
Herrerasaurus 0.60:0.90:0.95:1 0.59
Prosauropoda
Jingshanosaurus 1.0:1.0:1.0:1.0 0.55
Lufengosaurus 0.84:0.84:0.84:1 0.56
Plateosaurus 1.0:1.0:1.0:1.0 0.61
?Sauropoda
Blikanasaurus 1:1:0.89:0.65 0.53
Antenonitrus 0.62:0.86:1? —
Sauropoda
Gongxianosaurus — 0.64
Vulcanodon 0.73:0.64:0.82:1 0.75
Shunosaurus 1:0.92:0.85:0.85 0.70
Omeisaurus 1:0.87:0.56:0.62 0.90
Apatosaurus 1:0.75:0.50:0.55 —
Camarasaurus* — 0.75
Camarasaurus 1:0.69:0.44:0.50 0.81
Opisthocoelicaudia 1:0.78:0.67:0.44 0.78
“Barosaurus” 1:0.68:0.41:0.36 0.95
NOTE: References as in Table 1.2; “Barosaurus” data from Janensch (1961). Asterisk (*) indicates measurement of a juvenile individual.
digital shortening. Non-terminal phalangeal record in the Lower Jurassic Barapasaurus, they
reduction may have also produced a mammal- appear 13 million to 35 million years earlier in the
like phalangeal count in cynodont-grade synap- ichnofossil record. Upper Triassic Tetrasauropus
sids (Hopson 1995). trackways (fig. 1.11) clearly preserve impressions
Further modification of the pedal configura- of unguals deflected laterally relative to the axis of
tion described above diagnoses sauropods more the pes (Lockley et al. 2001), indicating that this
derived than Shunosaurus. In Barapasaurus, feature evolved earlier than implied by body fos-
Omeisaurus, and all neosauropods, the four pedal sils alone (Wilson 2005b).
unguals are directed laterally with respect to the
digit axis. This reorientation of the pedal unguals NEOSAUROPODA
is accomplished by a beveled proximal articular Neosauropoda is the node-based group includ-
surface and twisting of the axis of the ungual. ing Diplodocus longus, Saltasaurus loricatus, and
Wilson and Sereno (1998) scored the basal all descendants of their most recent common
sauropods Vulcanodon and Shunosaurus with the ancestor (Wilson and Sereno 1998; fig. 1.4).
primitive condition (i.e., anteriorly directed Within this node-based group, the two reflexive
unguals), and the primitive condition appears to stem-groups (Diplodocoidea, Macronaria)
characterize Blikanasaurus, Antenonitrus, and form a stable node–stem triplet (boldface type
Gongxianosaurus. Although laterally directed indicates node-based definitions; regular type
pedal unguals first appear in the body fossil indicates stem-based definitions).

TABLE 1.4
Manual and Pedal Phalangeal Counts in Select Saurischian Genera
MANUS PES
Theropoda
Eoraptor 2*-3*-4*-1-0 2*-3*-4*-5*-1
Herrerasaurus 2*-3*-4*-1-0 2*-3*-4*-?-1
Prosauropoda
Jingshanosaurus 2*-3*-4*-3*-1 2*-3*-4*-5*-1
Lufengosaurus 2*-3*-4*-3-1 2*-3*-4*-5*-1
Plateosaurus 2*-3*-4*-3-2 2*-3*-4*-5*-2
?Sauropoda
Blikanasaurus — 2*-3*-4*-?5*-?
Antenonitrus — —
Sauropoda
Vulcanodon — —
Gongxianosaurus — 2*-3*-4*-5*-?
Shunosaurus 2*-2-2-2-2 2*-3*-3*-3*-2
Omeisaurus 2*-2-?-?-1 2*-3*-3*-3*-2
Diplodocus — 2*-3*-3?-2-0
Camarasaurus 2*-1-1-1-1 2*-3*-4*-2*-?
Brachiosaurus 2*-1-1-1-1 —
Opisthocoelicaudia 0-0-0-0-0 2*-2*-2*-1?-?
NOTE: Asterisk (*) indicates a clawed digit. References as in Table 1.2; Diplodocus data from Hatcher (1901).
Neosauropoda Bonaparte 1986b— Cretaceous neosauropods include rebbachisaurid

Diplodocus longus, Saltasaurus loricatus, diplodocoids and Titanosauria (left out of
their most recent common ancestor, and Bonaparte’s definition), and the near-simultane-
all descendants. ous appearance of the principal neosauropod lin-
Diplodocoidea Upchurch 1995—All eages in the Late Jurassic implies that one or
neosauropods more closely related to more of them were present in the Middle
Diplodocus longus than to Saltasaurus lori- Jurassic. Neosauropoda accommodates the
catus. majority of sauropod genera and encompasses
most of its morphological diversity.
Macronaria Wilson and Sereno 1998—All
The recently described Jobaria has been
neosauropods more closely related to
resolved as the outgroup of Neosauropoda
Saltasaurus loricatus than to Diplodocus
(Sereno et al. 1999; Wilson 2002; fig. 1.4) on
longus.
the basis of a number of advanced features they
Bonaparte (1986b:369) originally referred to share. Jobaria has been alternatively resolved
neosauropods as the “end-Jurassic” sauropods— within Neosauropoda as a basal macronarian
members of Dicraeosauridae, Diplodocidae, (Upchurch et al. 2004), but the evidence sup-
Camarasauridae, and Brachiosauridae. Although porting this hypothesis is presently outweighed
no definitive skeletal remains referable to this by the retention of several primitive characters.
group have been recorded prior to the “end The relevant synapomorphies of Jobaria
Jurassic,” the phylogenetic definition of Neosauropoda is discussed alongside those dis-
Neosauropoda is not temporally bounded. tinguishing Neosauropoda. Jobaria and

Neosauropoda can be distinguished by a novel metacarpals suggests that the primary axis of the
forefoot posture in which the manus is wrist joint was positioned more proximally,
arranged into a tight semicircle and held verti- between these carpals and the bones of the fore-
cally. Neosauropoda, in turn, is distinguished arm. Other neosauropods have a single carpal ele-
by marked reduction in the number and ossifi- ment positioned above metacarpals II and III
cation of carpal and tarsal elements. (e.g., Apatosaurus [Hatcher 1902; Gilmore 1936]).
One individual of Apatosaurus, however, preserves
REDUCED OSSIFICATION OF CARPALS a carpal element hypothesized to articulate with
AND TARSALS metacarpals IV and V (Filla and Redman 1994).
The evolutionary history of sauropod dinosaurs As discussed later, some sauropods lack ossified
documents reduced ossification of carpal and carpus altogether (see “Macronaria” below).
tarsal elements. This tendency may be related to In prosauropods and basal theropods, the
the reduced ossification that characterizes all body of the astragalus (i.e., the portion below
sauropod weight-bearing elements (fig. 1.9; see the ascending process) is trapezoidal. In the
Sauropoda, above). In the carpus of Herrerasaurus prosauropod Massospondylus and the basal thero-
(Sereno 1993:fig. 15) and Eoraptor (P. Sereno, pod Herrerasaurus, the proximodistal and antero-
pers. comm.), a large radiale, ulnare, and series of posterior depth of the medial side of the astra-
four distal carpals are present. In prosauropods, galus equals or exceeds that of the lateral side
the proximal carpals are very reduced or absent (Cooper 1981:fig. 71f; Novas 1989:figs. 2.5–10).
(unossified), but the medial three distal carpals The basal sauropod Shunosaurus appears to retain
(dc 1–3) are present and articulate with the proxi- the primitive condition, based on the only avail-
mal ends of metacarpals I–III, respectively (e.g., able view (anterior) of the astragalus (Zhang
Massospondylus [Cooper 1981:figs. 35, 36], 1988:fig. 54). In contrast, the astragalus in Jobaria
Lufengosaurus, [Young 1941:fig.15]). The earliest and Neosauropoda appears wedge-shaped in both
sauropod for which a carpus is known preserves proximal and anterior views. In Jobaria, for exam-
only three block-shaped carpals that decrease in ple, the proximodistal and anteroposterior depth
size laterally, based on their presumed position of the astragalus diminishes markedly toward its
(Shunosaurus [Zhang 1988:figs. 2, 48]). Because medial side. In addition, in proximal and distal
they were found closely associated with views, the primitive posteromedial corner of the
metacarpals I–III, rather than with the radius astragalus is absent, and the astragalus has a sub-
and ulna, Wilson and Sereno (1998:47) regarded triangular, rather than subrectangular, shape.
them as distal carpals. Other nonneosauropods
show a similar pattern: Omeisaurus has three FOREFOOT POSTURE
carpals of decreasing size (He et al. 1988), as Prosauropods and basal ornithischians retain the
does Jobaria (Sereno et al. 1999). In Jobaria, one primitive dinosaur condition in which the proxi-
surface of the largest carpal has two triangular mal ends of the metacarpals are not closely
facets that match the proximal surfaces of appressed and are only slightly arched in articu-
metacarpals I and II, suggesting that it is a distal lation. The metacarpals are subrectangular in
carpal; the other surface bears no discernible proximal view, and their intermetacarpal articular
articular surface. Thus prosauropods and basal surfaces do not extend down the shaft. For exam-
sauropods retain only three ossified distal ple, the articulated manus of Massospondylus is
carpals and lack ossified proximal carpals. cupped approximately 90 between metacarpal I
Neosauropods further reduce the number of ossi- and metacarpal V, and most of this arch occurs in
fied carpals to two or fewer. In Camarasaurus, two metacarpals I and III, whose lateral articular
block-shaped carpals are present and fitted to the surfaces form an acute angle with the anterior
metacarpals. As Osborn (1904:182) noted, the fit- surface (Cooper 1981:fig. 37). A similar condition
ted articulation between the carpals and the is present in basal ornithischians (e.g.,

Lesothosaurus [Sereno 1991]). The condition in document sauropod trackmakers with a digiti-
theropods, however, differs from that of grade manus (fig. 1.12). These trackways record
prosauropods and basal ornithischians. Although the appearance of a digitigrade forefoot posture
the manus is bound proximally in theropods, the 22 million to 44 million years earlier than pre-
metacarpus retains the same 90 proximal curva- dicted by Upchurch (1998) and 57 million years
ture present in prosauropods and ornithischians earlier than predicted by Wilson and Sereno (57
(Herrerasaurus [Sereno 1993,fig. 15], Deinonychus, my). This discrepancy can be interpreted as the
[Ostrom 1969:fig.62]). The configuration and early appearance of Neosauropoda or the early
pose of the metacarpus of basal sauropods such appearance of digitigrade foot posture in non-
as Shunosaurus and Omeisaurus are not agreed neosauropods. Assessment of forefoot posture in
on. Whereas Wilson and Sereno (1998) sug- non-neosauropods was based solely on the pub-
gested that their forefoot posture resembles that lished illustrations of the only basal taxa preserv-
of prosauropods and basal ornithischians, in ing manual remains, Shunosaurus (Zhang
which the manus is spreading and only slightly 1988:fig. 49, pl. 14) and Omeisaurus (He et al.
arched ventrally, both Upchurch (1998) and 1988:figs. 47, 48; pl. 14, figs. 4–6). On the basis
Bonnan (2003) interpreted them as having the of these illustrations, Wilson and Sereno
derived, digitigrade forefoot posture that charac- (1998:48) argued that Shunosaurus and
terizes Jobaria and neosauropods. Omeisaurus lacked a digitigrade forefoot pos-
The metacarpus of Jobaria and neosauropods ture because their metacarpals have poorly
is arranged into a tightly bound, digitigrade defined intermetacarpal articular surfaces. Other
structure that is hypothesized to have contacted dinosaurs with a vertically oriented, digitigrade
the substrate at the metacarpal–phalangeal foot posture have metapodials that are tightly
joints. The metacarpals arranged into a vertical appressed (bound) proximally and have well-
cylinder in which all are subequal in length and marked intermetapodial facets that extend down
have well-developed intermetacarpal articular their shafts (e.g., Herrerasaurus pes, Iguanodon
surfaces that extend distally to midshaft (e.g., manus). Additionally, the metacarpals of
Camarasaurus [Ostrom and McIntosh 1966:figs. Shunosaurus and Omeisaurus are subrectangular
55–59]). Proximally, the metacarpal heads are proximally, implying that they were only slightly
wedge-shaped and articulate in a tight arc of arched proximally (~90) in articulation, unlike
approximately 270. This tubular arrangement Jobaria and neosauropods (Wilson and Sereno
of the metacarpals is due to their medial and lat- 1998:fig. 40). Upchurch (1998:68), however,
eral articular surfaces meeting the external argued that despite these considerations,
(anterior) aspect of the metacarpal at an acute Shunosaurus and Omeisaurus had forefeet that
angle. A tightly curled, digitigrade manus, were both digitigrade and U-shaped proximally,
defined by these osteological features, is pres- features he regarded as a single character (table
ent without exception in Neosauropoda. 1.1). Trackways from Italy, Poland, and Morocco
Although cladistic studies have regarded digi- preserve a digitigrade manus that is not tightly
tigrade forefoot posture as diagnostic of arched (fig. 1.12), suggesting that the bound
Eusauropoda (Upchurch 1998), Neosauropoda metatarsus and its tightly arched configuration
(Upchurch 1995; Wilson and Sereno 1998), or are independent characters. The trackways fur-
Jobaria Neosauropoda (Wilson 2002), ichno- ther suggest that the bound metacarpus was
fossils suggest a much earlier origin. Upper acquired earlier in sauropod history than was the
Triassic trackways from North America (Lockley tubular metacarpus.
et al. 2001) and South America (Marsicano and
Barredo 2004), as well as Lower Jurassic track- DIPLODOCOIDEA
ways from Italy (Dalla Vecchia 1994), Poland One of two reflexive neosauropod stem-groups,
(Gierlinski 1997), and Morocco (Ishigaki 1988), Diplodocoidea includes all neosauropods more

FIGURE 1.12. Early sauropod footprints. (A), Tetrasauropus trackway from the Upper Triassic (Chinle Group) of Cub Creek
(based on Lockley et al. (2001); (B), trackway from the Lower Jurassic (Pliensbachian) of the Atlas Mountains, Morocco
(based on Farlow 1992:fig. 2a, b); (C), trackway (ROLM 28) from the Lower Jurassic (Hettangian–Pliensbachian) of Lavini di
Marco, Italy (based on Dalla Vecchia 1994:fig. 2); (D) manus–pes pair of ?Parabrontopodopus, from the Hettangian of the
Holy Cross Mountains, Poland (based on Gierlinski 1997:fig. 1b). (A–C) are oriented relative to trackway midline; trackway
midline cannot be determined for (D). Scale equals 50 cm for (A), 10 cm for (B–D).
closely related to Diplodocus longus than to Flagellicaudata Harris and Dodson 2004—
Saltasaurus loricatus (Wilson and Sereno 1998). Diplodocus longus, Dicraeosaurus hanse-
By this definition, Diplodocoidea unites manni, their most recent common ances-
Haplocanthosaurus, Rebbachisauridae, Dicraeo- tor, and all descendants.
sauridae, and Diplodocidae. The position of Dicraeosauridae Janensch 1929b—All
Haplocanthosaurus as the basalmost diplodocoid diplodocoids more closely related to
is weakly supported and awaits further confir- Dicraeosaurus hansemanni than to
mation by additional material. Apart from the Diplodocus longus.
position of Haplocanthosaurus, the relationships
Diplodocidae Marsh 1884—All diplodocoids
within and between diplodocoid families are
more closely related to Diplodocus longus
stable. The three families have stem-based defi-
than to Dicraeosaurus hansemanni.
nitions specifying all taxa more closely related to
their namesake genus than to either of the other This arrangement of taxon names affords a
two namesake genera. Sereno (1998) formally node–stem triplet within Diplodocoidea that
defined Dicraeosauridae and Diplodocidae; unites two well known stem-based groups
Rebbachisauridae is phylogenetically defined (Diplodocidae, Dicraeosauridae) whose sister-
for the first time here. A revised phylogenetic taxon relationship has been long recognized.
nomenclature for Diplodocoidea and its sub- Like all neosauropod lineages, earliest
groups is proposed below (boldface type indi- diplodocoids are found in Upper Jurassic
cates node-based definitions; regular type indi- rocks. Diplodocidae is currently restricted to
cates stem-based definitions): the Late Jurassic of North America (Diplodocus,
Apatosaurus, Barosaurus, Seismosaurus) and
Diplodocoidea Upchurch 1995—All Africa (“Barosaurus” africanus). Dicraeosauridae
neosauropods more closely related to Diplo- is also known from the Late Jurassic of Africa
docus longus than to Saltasaurus loricatus. (Dicraeosaurus) but survives into the Early
Rebbachisauridae Bonaparte 1997—All Cretaceous of South America (Amargasaurus).
diplodocoids more closely related to Rebbachisauridae is the latest surviving
Rebbachisaurus garasbae than to diplodocoid clade and is restricted to the Creta-
Diplodocus longus. ceous of Africa (Nigersaurus, Rebbachisaurus),

TABLE 1.5
Data Support in the Two Neosauropod Lineages Macronaria and Diplodocoidea
MACRONARIA DIPLODOCOIDEA
number of taxa 11 9
cranial
% missing data 58 55
% character support 30 39.5
axial
% character support 37 45.5
appendicular
% character support 33 15
NOTE: The relative proportions of cranial, axial, and appendicular characters supporting the interrelationships of these clades are com-
pared below. Missing data scores were based on Wilson (2002:table 8). Total percentage missing data was higher in Diplodocoidea (48%)
than Macronaria (44%).
South America (Limaysaurus), and Europe anterior to the antorbital fenestra and arranged
(Histriasaurus, Salas rebbachisaurid). in jaws that are rectangular in dorsal view (fig.
The features supporting the relationships 1.10). In dicraeosaurids and diplodocoids, most
within Diplodocoidea and differentiating its teeth are positioned on the transverse portion
composite genera are supported by a predomi- of the jaw ramus. Rebbachisaurids further this
nance of cranial and axial synapomorphies trend by restricting all teeth to the transverse
(table 1.5). These include a major transforma- portion of the jaw, which extends lateral to the
tion in skull shape and a highly modified verte- ramus (see Sereno and Wilson, chapter 5).
bral column, discussed below. Transversely oriented tooth rows are unknown
elsewhere in Dinosauria.
HERBIVOROUS SPECIALIZATIONS Other modifications of the diplodocoid skull
As discussed above, the basic sauropod skull plan are novelties that have no precedent in sauropod
is quite distinct from those of basal saurischians evolution. One set of such features that charac-
(see “Eusauropoda” below). The set of features terizes Diplodocoidea is the reorientation of the
comprising this plan evolved sometime prior to braincase and part of the palate relative to the
the first appearance of Eusauropoda (Middle dermal skull. In sauropod outgroups and in
Jurassic) and, with few exceptions, was retained most sauropods, the jaw articulation lies at the
until their last appearance (latest Cretaceous). posterior extreme of the skull, behind the orbit.
The diplodocoid skull is perhaps the most unique Likewise, the basipterygoid processes are short
among Sauropoda, and may be thought of as the and point ventrally, and the adductor fossa is
result of exaggeration of several eusauropod fea- positioned on the posterior half of the lower jaw,
tures combined with novelties that evolved step- just below the orbit. The diplodocoid skull dif-
wise within Diplodocoidea. fers in each of these respects, due to a reorien-
The broadening of the snout and shortening tation of the dermal skull relative to the brain-
of the tooth row that characterizes Eusauropoda case. In diplodocoids the quadrate is oriented
is exaggerated in diplodocoids, which evolved anteriorly such that the jaw joint is positioned
upper and lower tooth rows that are restricted below the orbit in lateral view. The pterygoid and

its connection to the braincase via the basiptery- the anterior cervical neural spines to those of the
goid processes are shifted forward. These two mid-dorsal region, but they may extend to the
changes effectively shorten the lower jaw and anterior caudal neural spines in some taxa (e.g.,
shift anteriorly the adductor fossa. Although Diplodocus). The forked neural spines are longest
they diagnose the same node, these features are in Amargasaurus and Dicraeosaurus, in which they
not considered to be correlated. Some but not all are more than four times the centrum height, and
of these changes are manifest in other taxa, such shortest in Apatosaurus, in which they are shorter
as pachycephalosaurs, which have anteriorly ori- than the centrum height. A median tubercle may
ented basipterygoid processes with otherwise occasionally be present between the two rami of
typical quadrate articulation and lower jaw the neural spines of the pectoral region, but this
length. This change in the shape of the feature is not present in all diplodocoids. Forked
diplodocoid skull effects an inclined line of neural spines may have been a specialization that
action for the principal jaw closing musculature, allowed passage of elastic ligaments (Janensch
which may have resulted in more fore–aft 1929b; Alexander 1985), such as the ligamentum
motion than present in other sauropods nuchae and ligamentum elasticum interspinale
(Upchurch and Barrett 2000). (Tsuihiji 2004). The presence of forked neural
One aspect of diplodocoid skulls represents a spines and implied ligaments in some taxa but
reversal from the basic eusauropod condition. not in others remains unexplained.
Relative to their outgroups, eusauropods have As mentioned earlier, more than a dozen
broad crowns that overlap one another along the neck lengthening events appear within
tooth row. Diplodocoids, in contrast, reduce Neosauropoda, whose basic complement of cer-
crown size and lose the crown overlap diagnostic vical, dorsal, and sacral vertebrae is 13-12-5. All
of Eusauropoda. These features are correlated, are autapomorphies for genera, except for one
because crown overlap requires some expansion event that characterizes Diplodocidae. The
of the crown relative to the root. Reduction of diplodocids Apatosaurus and Diplodocus incorpo-
crown size may also be correlated with the rela- rated two dorsal vertebrae into the neck to
tive shortening of the tooth row, if tooth number obtain the precaudal count of 15-10-5. Although
remains constant. A significant consequence of precaudal counts are not known for other
crown reduction is that additional replacement diplodocids, dicraeosaurids and rebbachisaurids
teeth can pack the jaw ramus. Up to five replace- do not appear to share this feature.
ment teeth fill a given position in Diplodocus
(Holland 1924:fig. 3), whereas up to seven are TAIL SPECIALIZATIONS
present in Nigersaurus (Sereno et al. 1999:fig. The number of caudal vertebrae comprising the
2D). This specialization may allow enhanced tail is fairly similar in outgroups to Sauropoda.
tooth replacement rates (see Sereno and Wilson, Caudal counts are known for the prosauropods
chapter 5). Jingshanosaurus (44), Lufengosaurus (43), and
Plateosaurus (41), as well as the basal theropods
PRESACRAL SPECIALIZATIONS Eoraptor (45; Sereno, pers. comm.) and
Neural spines vary in length, shape, and orienta- Herrerasaurus (43–45). This number is retained in
tion throughout Dinosauria, but only within the basal sauropod Shunosaurus (43) and slightly
sauropods are they completely divided. Forked increased in more derived sauropods such as
neural spines appear several times in Sauropoda, Omeisaurus (50–55) and Camarasaurus (53). A
usually as an autapomorphies for genera (e.g., marked increase in the number of caudal verte-
Camarasaurus, Euhelopus, Opisthocoelicaudia). brae characterizes Diplodocidae, which nearly
Flagellicaudata (Diplodocidae Dicraeosauridae) doubles the primitive count (Diplodocus, 80 ;
is the only suprageneric group characterized by Apatosaurus, 82). Tail elongation in diplodocids is
forked neural spines. These usually extend from the result of supernumerary distal caudal verte-

brae, of which there are more than 30. These Brachiosaurus brancai than to Saltasaurus
archless distal caudal centra are not only numer- loricatus.
ous, but also distinctly biconvex and elongate. Somphospondyli Wilson and Sereno
Although biconvex distal caudal centra are known 1998—All titanosauriforms more closely
in other neosauropods, none are as elongate or related to Saltasaurus loricatus than to
numerous as in diplodocids (Wilson et al. 1999). Brachiosaurus brancai.
Together, this series of 30 or more elongate,
biconvex centra constitute a “whiplash” tail, which Titanosauria Bonaparte and Coria 1993—
has been interpreted as a defensive (e.g., Holland Andesaurus delgadoi, Saltasaurus loricatus,
1915) or noisemaking (Myhrvold and Currie their most recent common ancestor, and
1997) specialization. Although their caudal all descendants.
counts are unknown, the presence of elongate, Saltasauridae Powell 1992—
biconvex caudal centra in the rebbachisaurids Opisthocoelicaudia skarzynskii, Saltasaurus
Limaysaurus (Calvo and Salgado 1995; Salgado loricatus, their most recent common
2004) and Nigersaurus (personal observation), as ancestor, and all descendants.
well as the dicraeosaurid Dicraeosaurus (Janensch Opisthocoelicaudiinae McIntosh 1990—All
1929b), suggests that the whiplash tail may have saltasaurids more closely related to
been a general diplodocoid feature. However, this Opisthocoelicaudia skarzynskii than to
can only be confirmed by future discoveries of Saltasaurus loricatus.
articulated remains.
Saltasaurinae Powell 1992—All saltasaurids
MACRONARIA more closely related to Saltasaurus lorica-
tus than to Opisthocoelicaudia skarzynskii.
The second of the two reflexive neosauropod
stem-groups, Macronaria includes all neosauro- Macronaria is more taxonomically diverse
pods more closely related to Saltasaurus loricatus and widespread than its neosauropod counter-
than to Diplodocus longus (Wilson and Sereno part Diplodocoidea. Like other neosauropod lin-
1998). By this definition, Macronaria unites eages, macronarians first appear in the Late
Camarasaurus, Brachiosaurus, Euhelopus, and Jurassic. However, the simultaneous appearance
Titanosauria (fig. 1.4). Two node–stem triplets of Camarasaurus, Brachiosaurus, and the possible
are recognized within Macronaria, one for titanosaur Janenschia suggests an earlier origin
Titanosauriformes (Wilson and Sereno 1998) for the group. Furthermore, trackway evidence
and the other for Saltasauridae Sereno (1998), a may suggest a Middle Jurassic origin for
titanosaur subgroup. Phylogenetic definitions titanosaurs (Wilson and Carrano 1999; Day et al.
within Macronaria are as follows (boldface type 2002, 2004; see below) and thus all neosauropod
indicates node-based definitions; regular type lineages. Macronarians are the only sauropod
indicates stem-based definitions): subgroup to persist until the end of the
Cretaceous, represented as titanosaurs in the
Macronaria Wilson and Sereno 1998—All
Maastrichtian in North America (Alamosaurus
neosauropods more closely related to
[Gilmore 1946]), India (Isisaurus [ Jain and
Saltasaurus loricatus than to Diplodocus
Bandyopadhyay 1997]), Europe (Magyarosaurus
longus.
[Huene 1932], Ampelosaurus [LeLoeuff 1995]),
Titanosauriformes Salgado et al. 1997— Asia (Nemegtosaurus [Nowinski 1971],
Brachiosaurus brancai, Saltasaurus lorica- Opisthocoelicaudia [Borsuk-Bialynicka 1977]),
tus, their most recent common ancestor, Madagascar (Rapetosaurus [Curry Rogers and
and all descendants. Forster 2001]), Africa (cf. Titanosauria
Brachiosauridae Riggs 1904—All [Rauhut and Werner 1999]), and South
titanosauriforms more closely related to America (Gondwanatitan [Kellner and Azevedo

1999]). Titanosauria includes several
extremely large forms (e.g, Antarctosaurus
giganteus, Argyrosaurus, Argentinosaurus; fig.
1.13, left), but also genera diminuitive by
sauropod standards (e.g., Saltasaurus,
Neuquensaurus; fig. 1.13, right). The body size
range in Titanosauria exceeds that in other
sauropod subgroups and provides an opportu-
nity to evaluate temporal and morphological pat-
terns of body size change within the group, once
a genus-level phylogeny is established.
Appendicular synapomorphies comprise a
substantial proportion of character support
within Macronaria, particularly within its latest-
surviving clade, Titanosauria (table 1.5).
Although several appendicular synapomor-
phies apply at basal macronarian nodes, per-
haps the most striking changes occur within
Titanosauria and are related to the acquisition
of a wide-gauge limb posture.
WIDE-GAUGE LIMB POSTURE

Animals with parasagittal limb stance walk or
run on land with their limbs held close to the
body midline. In these forms, the supporting
elements swing anteroposteriorly and contact
the substrate near the body midline. As the ani-
mal reaches higher speeds, these contacts
approach and sometimes touch or cross the
midline. A parasagittal limb stance can be FIGURE 1.13. Left femora (posterior view) of the ti-
observed directly in living therian mammals tanosaurs Antarctosaurus giganteus (length, 231 cm) and
Neuquensaurus australis (length, 70 cm). To scale (from
and in birds (e.g., Muybridge 1957). Squamates Huene 1929:pl. 20, 36). Both femora correspond to adult
and crocodylians, in contrast, have a sprawling individuals.
gait in which the proximal limb elements are
oriented close to the horizontal plane and the defined as those in which manus and pes
limbs contact the substrate at some distance impressions are “close [to] or even intersect the
from the body midline (e.g., Blob 2001). trackway midline,” whereas “wide-gauge” track-
Fossilized trackways provide indirect evidence ways are “well away from the trackway midline”
for parasagittal locomotion in extinct dinosaurs (Farlow 1992:108, 109). Variation in gauge
(e.g., Thulborn 1982). As observed in living the- width has been inferred to be taxonomic, with
rians and birds, theropod and ornithopod fore- narrow-gauge stance presumed to be primitive
and hindfoot impressions are quite close to or and wide-gauge stance derived (Wilson and
overlap the trackway midline. Likewise, sauro- Carrano 1999). Further, the presence of certain
pod trackways evidence a parasagittal limb morphological characteristics of saltasaurid
stance, although the placement of the fore- and titanosaurs has suggested that they are the wide-
hindfeet relative to the midline varies within the gauge trackmakers. Wilson and Carrano (1999)
clade. “Narrow-gauge” sauropod tracks are recognized three hindlimb features that support

A B
FIGURE 1.14. Limb skeleton of the wide-gauge saltasaurid Opisthocoelicaudia. Pectoral girdle and forelimb (A) and pelvic gir-
dle and hindlimb (B) in anterior view. Forelimb reconstruction based on illustrations and photographs in Borsuk-Bialynicka
(1977:fig. 9B, pl. 7–9, 11); hindlimb reconstruction modified from Wilson and Carrano (1999).
the hypothesis that saltasaurids were wide- been mentioned that all sauropod femora (except
gauge trackmakers. In addition, they recognized some diplodocines) are broader mediolaterally
forelimb features that are related to wide-gauge that anteroposteriorly. This shape provides
locomotion. Still other features are merely asso- greater resistance to mediolateral bending.
ciated with wide-gauge limb posture but are not Saltasaurids, however, exaggerate this feature well
required by it. These are discussed below. beyond that of typical sauropods. This increased
Acquisition of wide-gauge limb posture femoral eccentricity may have offered greater
requires the manus and pes to contact the ground resistance to the increased bending moment
at some distance from the midline. This was imposed by a wide-gauge limb posture. The distal
achieved in saltasaurids by two modifications that tibia, whose distal end is diagnostically broader in
allowed the femur to angle outward from the titanosaurs than in other sauropods, may also be
body wall. First, the proximal third of the femur is specialized to counter mediolateral bending.
canted inward relative to the rest of the shaft (fig. A series of pectoral girdle and forelimb fea-
1.14B). A similar characteristic is present in forest tures is related to the acquisition of wide-gauge
bovids that walk with their femora more abducted limb posture in saltasaurids. The anterior thorax
than do their closest relatives (Kappelman 1988). and the shoulder girdle are broader in
Second, the distal condyles are not aligned saltasaurids than in other sauropods, owing to
orthogonal to the long axis of the femur, as in the combined effects of the elongate coracoids
other sauropods. Instead, the distal femoral and the enlarged, crescentic sternal plates (fig.
condyles of saltasaurids are beveled 10 dorsome- 1.14A). However, because the pectoral girdle
dially. As shown in figure 1.14, this conformation has no bony connection to the vertebral col-
orients the axis of the knee parallel to the ground umn, the absolute distance between the glenoid
and perpendicular to the ground reaction force and the midline cannot be determined. The
when the limb is angled away from the body. So forelimb is characterized by several reversals of
far, this feature is restricted to saltasaurids. A early sauropod synapomorphies associated with
third feature that may facilitate a wide-gauge limb the evolution of a columnar, graviportal pos-
posture in saltasaurids is the highly eccentric ture. The humerus in saltasaurids is unique in
femoral midshaft cross section. It has already that it bears a prominent deltopectoral crest,

and its distal condyles are both divided and among the articulated forelimb elements of
exposed anteriorly. These features are not pres- Alamosaurus and Opisthocoelicaudia (Gilmore
ent in other sauropods. Likewise, the ulna is 1946:pl. 4; Borsuk-Bialynicka 1977:29). In
characterized by a prominent olecranon process both cases, radius, ulna, and metacarpals were
that projects above the articular surface of the all preserved in articulation, but no intervening
ulna, as it does in sauropod outgroups but not in carpal elements were found. Manual phalanges
other sauropods. These reversals—particularly have been reported only rarely in association
extension of the ulnar articular surface onto the with titanosaur skeletons, and never has an
anterior surface of the humerus and the promi- ungual been reported. The manual phalanges
nent olecranon process—suggest a more habit- that have been reported are extremely reduced
ually flexed forelimb posture in saltasaurids. (e.g., Borsuk-Bialynicka 1977), and it is likely
Other features are consistent with this interpre- that the manus had no fleshy digits. Like the
tation, including the increased transverse diam- carpus and manus, the ankle is extremely
eter of the distal radius, which is a shape that reduced in the saltasaurids. The saltasaurid
better resists mediolateral bending moments. astragalus is distinct among dinosaurs in the
A third set of features is novelties of wide- extreme reduction of its mediolateral diame-
gauge limb trackmakers that do not signal a ter, which is subequal to the anteroposterior
modified limb posture but they may offer and proximodistal diameters. In articulation,
insight into the function of this novel locomo- the pyramidal astragalus of the saltasaurid
tory specialization. Saltasaurids are character- Opisthocoelicaudia contacts the fibula and the
ized by a short tail that consists of approxi- lateral aspect of the tibia but does not reach the
mately 35 stout caudals, many fewer than in medial extreme of the distal tibia (fig. 1.14B).
sauropods primitively and less than half the
number in diplodocids. The articular surfaces
CONCLUSIONS
of all titanosaur caudal centra are concavo-
convex; in all but Opisthocoelicaudia the ante- Sauropods were “successful” dinosaurs by
rior face of the centrum is concave (pro- virtue of their geographic distribution, tempo-
coelous). Borsuk-Bialynicka (1977) and Wilson ral survivorship, biomass, generic diversity,
and Carrano (1999) suggested that this short- higher-level diversity, and morphological com-
ened tail may have functioned as a third sup- plexity. Although historically studies of sauro-
port when saltasaurids reared during feeding or pod systematics have lagged behind those of
mating. A second saltasaurid synapomorphy other dinosaur subgroups, a burst of analyses
may also be related to occasional bipedal in the last decade has begun to elucidate the
stance or tripodal rearing. The preacetabular evolutionary history of the group.
processes or saltasaurid ilia are flared laterally The stratigraphic distribution of the first
such that they are oriented nearly perpendicu- representatives of Sauropoda and of their sister-
lar to the body axis. Wilson and Carrano (1999) taxon Prosauropoda implies a 10 million- to 15-
suggested that, among other effects, flared ilia million year missing lineage during which the
move the origination site of the femoral pro- score of features diagnosing sauropods evolved.
tractor muscles laterally, bringing them into Synapomorphies related to precisely occlusion,
anteroposterior alignment with the direction of neck elongation, and columnar posture evolved
travel. Finally, one feature peculiar to during the Late Triassic and Early Jurassic and
saltasaurids and their subgroups is a pro- characterize all sauropods. Although all main
nounced reduction in the ossification of the neosauropod lineages appear simultaneously
carpus, manus, and tarsus. Carpal elements around the globe in the Late Jurassic, it is prob-
have not been found associated with manual able that neosauropods were present in the
elements in any titanosaur and are not present Middle Jurassic and possible that they were

present in the Early Jurassic. Future sampling ments. Future discoveries and analyses are
of these poorly sampled intervals will better required to better understand the relationship,
illuminate early neosauropod evolution. The if any, between Late Cretaceous survivorship
two principal neosauropod subgroups, Macro- and narrow-crowned dentition.
naria and Diplodocoidea, are the predomin-
ant sauropods during the Late Jurassic and
ACKNOWLEDGMENTS
Cretaceous. The descent and diversification
within these two groups were shaped by I thank Jack McIntosh for the advice, encour-
changes in different regions of the skeleton. agement, criticism, and sauropod details he has
Diplodocoids are characterized by cranial and shared with me during the past decade but, also,
axial synapomorphies that led to the evolution for his single-handed championing of sauropods
of a dental battery in rebbachisaurids and a for so many years. I thank Paul Sereno for dis-
whiplash tail in flagellicaudatans. In contrast, a cussion of many of the ideas presented in this
series of appendicular changes led to the evolu- chapter. I also profited from correspondence
tion of a wide-gauge limb posture in the with Ricardo Martínez, Leo Salgado, and Paul
macronarian subgroup Titanosauria, which Upchurch. I thank Adam Yates for sharing
was the dominant sauropod lineage of the with me his in press manuscript on
Cretaceous, represented on nearly all continen- Anchisaurus. Chris Sidor provided useful
tal landmasses by more than 40 species. information on Blikanasaurus. I am grateful
Despite advances in understanding of the for detailed reviews by Cathy Forster and
group, substantial gaps in our knowledge of Thomas Holtz, Jr. Figures 1.5–1.8, 1.10, and
sauropod history still exist. Like the animals 1.14 were skillfully completed by B. Miljour
themselves, our understanding of sauropod from illustrations by J.A.W. This research was
history is deepest in the middle but somewhat supported by grants from The Dinosaur
thinner on both ends. The sequence of changes Society, the American Institute for Indian
leading to the sauropod body plan from the Studies, The Hinds Fund, and the Scott Turner
primitive saurischian condition is still poorly Fund. I obtained translations of several papers
understood. For instance, it is not known from the Polyglot Paleontologist (www.infor-
whether herbivorous specializations preceded matics.sunysb.edu/anatomicalsci/paleo/):
large body size and quadrupedality. New dis- Bonaparte (1986a, 1997), Bonaparte and Coria
coveries of basal sauropod taxa are needed to (1993), He et al. (1998), and Powell (1992).
address this question. At the other end of sauro- C. Yu translated excerpts from He et al. (1988)
pod history, phylogenetic understanding of and Zhang (1988). Janensch (1929b) was
the two latest-surviving sauropod groups, translated by S. Klutzny through a Jurassic
Rebbachisauridae and Titanosauria, are as yet Foundation grant (to J.A.W . and M. Carrano).
unknown, but new discoveries have already
begun to bring clarity to this problem. These LITERATURE CITED
two lineages are important biogeographically
Alexander, R. M. 1985. Mechanics of posture and gait
during the end of sauropod history and may
of some large dinosaurs. Zool. J. Linn. Soc. 83:1–25.
signal an interesting survivorship pattern. Allain, R., Taquet, P., Battail, B., Dejax, J., Richir, P.,
Despite the fact that both broad- and narrow- Véran, M., Limon-Duparcmeur, F., Vacant, R.,
crowned sauropod taxa were present on most Mateus, O., Sayarath, P., Khenthavong, B., and
continental landmasses, only narrow-crowned Phouyavong, S. 1999. Un nouveau genere de
dinosaure sauropode de la formation des Grés
taxa survived into the Late Cretaceous each
supérieurs (Aptien-Albien) du Laos. C.R. Acad.
independent case (see Barrett and Upchurch, Sci. Paris Sci. Terre Planétes 329: 609–616.
chapter 4). No broad-crowned sauropod teeth Anonymous. 1905. The presentation of a reproduc-
have been reported from Late Cretaceous sedi- tion of Diplodocus carnegiei to the trustees of the

British Museum. Annals of the Carnegie Cenomanian of Argentina; new evidence on the
Museum 3: 443–452. origin of Diplodocidae. GAIA 11: 13–33.
Barrett, P.M., Hasegawa, Y., Manabe, M., Isaji, S., Carrano, M. T. 1997. Morphological indicators of
and Matsuoka, H. 2002. Sauropod dinosaurs foot posture in mammals: a statistical and bio-
from the Lower Cretaceous of eastern Asia: taxo- mechanical analysis. Zool. Linn. Soc. 121: 77–
nomic and biogeographical implications. 104.
Palaeontology 45: 1197–1217. ——–. 2000. Homoplasy and the evolution of
Benton, M.J., Juul, L., Storrs, G. W., and Galton, dinosaur locomotion. Paleobiology 26: 489–512.
P. M. 2000. Anatomy and systematics of the ——–. 2005. Body-size evolution in the Dinosauria. In:
prosauropod dinosaur Thecodontosaurus antiquus Carrano, M.T., Gaudin, T.J., Blob, R.W., and Wible,
from the Upper Triassic of southwest England. J. J.R. (eds.). Amniote Paleobiology: Phylogenetic and
Vertebr. Paleontol. 20: 77–108. Functional Perspectives on the Evolution of
Blob, R. W. 2001. Evolution of hindlimb posture in Mammals, Birds and Reptiles. Chicago: University
nonmammalian therapsids: biomechanical tests of of Chicago Press (in press).
paleontological hypotheses. Paleobiology 27: 14–38. Chiappe, L. M., Coria, R. A., Dingus, L., Jackson, F.,
Bonaparte, J. F. 1986a. The early radiation and phylo- Chinsamy, A., and Fox, M. 1998. Sauropod
genetic relationships of sauropod dinosaurs, based dinosaur embryos from the Late Cretaceous of
on vertebral anatomy. In: Padian, K.(eds.). The Patagonia. Nature 396: 258–261.
Beginning of the Age of Dinosaurs. Cambridge Chiappe, L.M., Salgado, L., and Coria, R.A., 2001.
University Press, Cambridge. pp. 247–258. Embryonic skulls of titanosaur sauropod dinosaurs.
——–. 1986b. Les dinosaures (Carnosaures, Science 293: 2444–2446.
Allosauridés, Sauropodes, Cétiosaurides) du Cooper, M. R. 1981. The prosauropod dinosaur
Jurassique moyen de Cerro Cóndor (Chubut, Massospondylus carinatus Owen from Zimbabwe:
Argentine). Ann. Paléontol. 72: 325–386. its biology, mode of life and phylogenetic signifi-
——–. 1997. Rayososaurus agrioensis Bonaparte 1995. cance. Occas. Papers Natl. Mus. Rhodesia B Nat.
Ameghiniana 34: 116. Sci. 6: 689–840.
Bonaparte, J.F., and Coria, R.A. 1993. Un nuevo y ——–. 1984. Reassessment of Vulcanodon and the ori-
gigantesco sauropódo titanosaurio de la gin of the Sauropoda. Palaeontol. Africana 25:
Formación Río Limay (Albiano-Cenomanio) de la 203–231.
Provincia del Neuquén, Argentina. Ameghiniana Curry, K. A. 2001. The Evolutionary History of the
30: 271–282. Titanosauria. Ph.D. thesis, State University of
Bonnan, M.F. 2003. The evolution of manus shape New York, Stony Brook.
in sauropod dinosaurs: implications for func- Curry Rogers, K. A., and Forster, C. 2001. The last of
tional morphology, limb orientation, and phy- the dinosaur titans: a new sauropod from
logeny. J. Vertebr. Paleontol. 23: 595–613. Madagascar. Nature 412: 530–534.
Borsuk-Bialynicka, M. 1977. A new camarasaurid Curry Rogers, K. A., and Forster, C. 2004. The skull
sauropod Opisthocoelicaudia skarzynskii, gen. n., of Rapetosaurus krausei (Sauropoda: Titanosauria)
sp. n. from the Upper Cretaceous of Mongolia. from the Late Cretaceous of Madagascar. J.
Palaeontol. Polonica 37: 1–64. Verteb. Paleontol. 24:121–144.
Buckland, W. 1841. Geology and Minerology Dalla Vecchia, F. M. 1994. Jurassic and Cretaceous
Considered with Reference to Natural Theology, sauropod evidence in the Mesozoic carbonate
2nd ed. Lea & Blanchard, Philadelphia. 468 pp. platforms of the southern Alps and Dinards.
Buffetaut, E., Suteethorn, V., Cuny, G. , Tong, H., Le GAIA 10: 65–73.
Loeuff, J., Khansubha, S., and Jongautchariyakul., Day, J. J., Upchurch, P., Norman, D. B., Gale, A. S.,
S. 2000. The earliest known sauropod dinosaur. and Powell, H.P. 2002. Sauropod trackways, evo-
Nature 407: 72–74. lution, and behavior. Science 296: 1659.
Calvo, J. O. 1994. Jaw mechanics in sauropod Day, J.J., Norman, D.B., Gale, A.S., Upchurch, P., and
dinosaurs. GAIA 10: 183–193. Powell, H.P. 2004. A Middle Jurassic dinosaur
Calvo, J. O. and González Riga, B. J. 2003. trackway site from Oxfordshire, U. K. Palaeontology
Rinconsaurus caudamirus gen. et sp. nov., a new 47: 319–348.
titanosaurid (Dinosauria, Sauropoda) from the Desmond, A. J. 1975. The Hot-Blooded Dinosaurs.
Late Cretaceous of Patagonia, Argentina. Revista Blond and Briggs, London. 238 pp.
Geologica de Chile 30: 333–353. Farlow, J.O. 1992. Sauropod tracks and trackmakers:
Calvo, J.O., and Salgado, L. 1995. Rebbachisaurus tes- integrating the ichnological and skeletal records.
sonei, sp. nov. A new Sauropoda from the Albian- Zubía 10: 89–138.

Filla, B. J., and Redman, P.D. 1994. Apatosaurus yah- Harris, J.D., and Dodson, P. 2004. A new diplodocoid
napin: a preliminary description of a new species sauropod dinosaur from the Upper Jurassic
of diplodocid dinosaur from the Late Jurassic Morrison Formation of Montana, USA. Acta
Morrison Formation of southern Wyoming, the Palaeontol. Polonica 49: 197–210.
first sauropod found with a complete set of “belly Hatcher, J. B. 1901. Diplodocus (Marsh): its osteology,
ribs.” Forty-fourth Annual Field Conference— taxonomy, and probable habits, with a restoration
1994 Wyoming Geological Association Guidebook of the skeleton. Mem. Carnegie Mus. 1: 1–63.
44: 159–178. ———. 1902. Structure of the forelimb and manus
Flynn, J. J., Parrish, J. M., Rakotosamimanana, B., of Brontosaurus. Ann. Carnegie Mus. 1: 356–376.
Simpson, W. F., Whatley, R. L., and Wyss, A. R. He, X. -L., Li, K., and Cai, K.-J. 1988. [The Middle
1999. A Triassic fauna from Madagascar, Jurassic dinosaur fauna from Dashanpu, Zigong,
including early dinosaurs. Science 286: Sichuan. Vol. IV. Sauropod Dinosaurs (2)
763–765. Omeisaurus tianfuensis]. Sichuan Scientific and
Galton, P.M. 1990. Prosauropoda. In: Wieshampel, Technological Publishing House, Chengdu. 143 pp.
D.B., Dodson, P., and Osmólska, H. (eds.). The He, X. -L., Wang, C., Liu, S., Zhou, F., Liu, T., Cai, K.,
Dinosauria. University of California Press, and Dai., B. 1998. [A new sauropod dinosaur
Berkeley. Pp. 320–345. from the Early Jurassic in Gongxian County,
Galton, P.M. and Upchurch, P. 2000. Prosauropod South Sichuan]. Acta Geol. Sichuan 18: 1–6.
dinosaurs: homeotic transformations (“frame Holland, W.J. 1910. A review of some recent criticisms
shifts”) with third sacral as a caudosacral or a dor- of the restorations of sauropod dinosaurs existing
sosacral. J. Vertebr. Paleontol. 20: 43A. in the museums of the United States, with special
——–. 2004. Prosauropoda. In: Weishampel, D. B., reference to that of Diplodocus carnegiei in the
Dodson, P., and Osmólska, H. (eds.), The Carnegie Museum. Am. Nat. 44: 259–283.
Dinosauria, 2nd ed. University of California Press, ———. 1915. Heads and tails: a few notes relating to
Berkeley. Pp. 232–258. the structure of the sauropod dinosaurs. Annals of
Galton, P.M., and van Heerden, J. 1985. Partial the Carnegie Museum 9: 273–278.
hindlimb of Blikanasaurus cromptoni n. gen. and ——–. 1924. The skull of Diplodocus. Mem. Carnegie
n. sp., representing a new family of prosauropod Mus. 9: 379–403.
dinosaurs from the Upper Triassic of South Hopson, J. A. 1995. Patterns in evolution of the
Africa. Géobios 18: 509–516. manus and pes of non-mammalian synapsids. J.
Gauthier, J. A. 1986. Saurischian monophyly and the Vertebr. Paleontol. 15: 615–639.
origin of birds. In: Padian K., ed. The origin of Huene, F.V. 1920. Bemerkungen zur Systematik und
birds and the evolution of flight. Memoirs of the Stammesgeschicht einiger Reptilien. Z. Induktive
California Academy of Sciences 8: 1–55. Abstammungslehre Vererbungslehre 22: 209–212.
Gierlinski, G. 1997. Sauropod tracks in the Early ——–. 1926. Vollständige Osteologie eines
Jurassic of Poland. Acta Palaeontol. Polonica 42: Plateosauriden aus dem schwäbischen Keuper.
533–538. Geol. Palaeontol. Abhandlungen 15: 129–179.
Gierlinski, G., Pienkowski, G., and Niedzwiedzki, G. ——–. 1932. Die fossil Reptil-Ordnung Saurischia,
2004. Tetrapod track assemblage in the ihre Entwicklung und Geschichte. Monograph.
Hettangian of Soltyków, Poland, and its paleoen- Geol. Palaeontol. 4: 1–361.
vironmental background. Ichnos 11: 195–213. ——–. 1956. Paläontologie und Phylogenie der
Gilmore, C.W. 1925. A nearly complete articulated Niederen Tetrapoden. VEB Gustav Fischer Verlag,
skeleton of Camarasaurus, a saurischian dinosaur Jena. 716 pp.
from the Dinosaur National Monument, Utah. Huxley, T.H. 1869. On the classification of the
Mem. Carnegie Mus. 10: 347–384. Dinosauria, with observations on the Dinosauria
Gilmore, C.W. 1936. Osteology of Apatosaurus with of the Trias. Q. Rev. Geol. Soc. London 26: 32–51.
special reference to specimens in the Carnegie Ishigaki, S. 1988. Les empreintes de dinosaures du
Museum. Mem. Carnegie Mus. 11: 175–300. Jurassique inférieur du Haut Atlas central maro-
———. 1946. Reptilian fauna of the North Horn cain. Notes Serv. Géol. Maroc 44: 79–86.
Formation of central Utah. U. S. Geol. Surv. Prof. Jacobs, L. L., D. A., Winkler, Downs, W.R., and
Paper 210C: 1–52. Gomani, E.M. 1993. New material of an Early
González Riga, B. J. 2003. A new titanosaur Cretaceous titanosaurid sauropod dinosaur from
(Dinosauria, Sauropoda) from the Upper Malawi. Palaeontology 36: 523–534.
Cretaceous of Mendoza Province, Argentina. Jain, S. L., and Bandyopadhyay, S. 1997. New
Ameghiniana 40: 155–172. titanosaurid (Dinosauria: Sauropoda) from the

Late Cretaceous of central India. J. Vertebr. Maier, G. 2003. African Dinosaurs Unearthed.
Paleontol. 17: 114–136. Indiana University Press, Bloomington. 380 pp.
Janensch, W. 1914. Ubersicht uber die Wirbeltierfauna Marsh, O. C. 1878. Principal characters of American
der Tendaguru-Schichten, nebst einer kurzen Jurassic dinosaurs. Pt. I. Am. J. of Sci. (Ser. 3) 16:
Charakterisierung der neu aufgefuhrten Arten von 411–416.
Sauropoden. Arch. Biontol. 3: 81–110. ——–. 1883. Principal characters of American
——–. 1929a. Material und Formengehalt der Jurassic dinosaurs: Part VI. Restoration of
Sauropoden in der Ausbeute der Tendaguru Brontosaurus. Am. J. Sci. (Ser. 3) 26: 81–86.
Expedition. Palaeontographica (Suppl. 7) 2: ——–. 1884. Principal characters of American
1–34. Jurassic dinosaurs. Part VII. On the
——–. 1929b. Die Wirbelsäule der gattung Diplodocidae, a new family of the Sauropoda.
Dicraeosaurus. Palaeontographica (Suppl. 7) 7: Am. J. Sci. (Ser. 3) 27: 161–168.
39–133. ———. 1895. On the affinities and classification of
——–. 1935–1936. Die Schädel der Sauropoden the dinosaurian reptiles. Am. J. Sci. (Ser. 3) 50:
Brachiosaurus, Barosaurus, und Dicraeosaurus aus 413–423.
den Tendaguru-Schichten Deutsch-Ostafrikas. Marsicano, C. A., and Barredo, S.P. 2004. A Triassic
Palaeontographica (Suppl. 7) 3: 147–298. tetrapod footprint assemblage from southern
——–. 1950. Die wirbelsäule von Brachiosaurus bran- South America: palaeobiogeographical and evolu-
cai. Palaeontographica (Suppl. 7) 3: 27–93. tionary implications. Palaeogeogr. Palaeoclimatol.
——–. 1961. Die Gliedmaszen und Gliedmaszengürtel Palaeoecol. 203: 313–335.
der Sauropoden der Tendaguru-Schichten. Martin, V., Buffetaut, E., and Suteethorn, V. 1994. A
Palaeontographica (Suppl. 7) 3: 177–235. new genus of sauropod dinosaur from the Sao
Kappelman, J. 1988. Morphology and locomotor Khua Formation (Late Jurassic or Early
adaptations of the bovid femur in relation to habi- Cretaceous) of northeastern Thailand. C. R. Acad.
tat. J. Morphol. 198: 119–130. Sci. Paris (Ser. II) 319: 1085–1092.
Kellner, A.W.A., and Azevedo, S.A.K. de 1999. A new Martínez, R., Giménez, O., Rodríguez, J., Luna, M.,
sauropod dinosaur (Titanosauria) from the Late and Lamanna, M. C. 2004. An articulated speci-
Cretaceous of Brazil. Proc. Second Gondwanan men of the basal Titanosaurian (Dinosauria:
Dinosaur Symp. 15: 111–142. Sauropoda) Epachthosaurus sciuttoi from the early
Larson, A. 1994. The comparison of morphological Late Cretaceous Bajo Barreal Formation of
and molecular data in phylogenetic systematics. In: Chubut Province, Argentina. J. Vertebr. Paleontol.
Schierwater, B. S. B., Wagner, G. P., and DeSalle, R. 24: 107–120.
(eds.). Molecular Biology and Evolution: McIntosh, J.S. 1990. Sauropoda. In: Weishampel,
Approaches and Applications. Birkhauser Verlag, D. B., Dodson, P., and Osmólska, H. (eds.). The
Basel. Pp. 371–390. Dinosauria. University of California Press,
Le Loeuff, J. 1995. Ampelosaurus atacis (nov. gen., nov. Berkeley. Pp.345–401.
sp.), un nouveau Titanosauridae (Dinosauria, McIntosh, J.S., Miles, C.A., Cloward, K.C., and
Sauropoda) du Crétacé supérieur de la Haute Parker, J.R. 1996. A new nearly complete skele-
Vallée de l'Aude (France). C.R. Acad. Sci. Paris ton of Camarasaurus. Bull. Gunma Mus. Nat.
(Ser. II) 321: 693–699. Hist. 1: 1–87.
Le Loeuff, J. 2003. The first articulated titanosaurid Myhrvold, N. P., and Currie, P. J. 1997. Supersonic
skeleton in Europe. In: Meyer, C.A. (ed.). First sauropods? Tail dynamics in the diplodocids.
European Association of Vertebrate Paleontologists Paleobiology 23: 393–409.
Meeting. Natural History Museum, Basel. Muybridge, E. 1957. Animals in Motion. Brown,
Lockley, M.G., Wright, J.L., Hunt, A.G., and Lucas, L. S.(ed.). Dover, New York.
S. G. 2001. The Late Triassic sauropod track Novas, F. E. 1989. The tibia and tarsus in
record comes into focus: old legacies and new Herrerasauridae (Dinosauria, incertae sedis) and
paradigms. New Mexico Geological Society the origin and evolution of the dinosaurian tar-
Guidebook, 52nd Field Conference, Geology of sus. J. Paleontol. 63: 677–690.
the Llano Estacado. Pp. 181–190 ———. 1993. New information on the systematics
Luo, Y., and Wang, C. 2000. A new sauropod, and postcranial skeleton of Herrerasaurus
Gongxianosaurus, from the Lower Jurassic of ischigualastensis (Theropoda: Herrerasauridae)
Sichuan, China. Acta Geol. Sinica 74: 132–136. from the Ischigualasto Formation (Upper
Luo, Z. 1999. A refugium for relicts. Nature 400: Triassic) of Argentina. J. Vertebr. Paleontol. 13:
23–25. 400–423.

Nowinski, A. 1971. Nemegtosaurus mongoliensis n. gen., Russell, D. A. 1993. The role of central Asia in
n. sp., (Sauropoda) from the uppermost Cretaceous dinosaurian biogeography. Can. J. Earth Sci. 30:
of Mongolia. Palaeontol. Polonica 25:57–81. 2002–2012.
Osborn, H.F. 1904. The manus, sacrum and fore- Russell, D. A., and Zheng, Z. 1993. A large mamen-
limb of Sauropoda. Bull. Am. Mus. Nat. Hist. 20: chisaurid from the Junggar Basin, Xinjiang,
181–190. People's Republic of China. Can. J. Earth Sci. 30:
Osborn, H.F., and Mook, C.C. 1921. Camarasaurus, 2082–2095.
Amphicoelias, and other sauropods of Cope. Mem. Salgado, L. 2004. Lower Cretaceous rebbachisaurid
Am. Mus. Nat. Hist. 3: 247–387. sauropods from the Cerro Aguada del León
Ostrom, J.H. 1969. Osteology of Deinonychus antir- (Lohan Cura Formation), Neuquén Province,
rhopus, an unusual theropod from the Lower northwestern Patagonia, Argentina. J. Vertebr.
Cretaceous of Montana. Bull. Yale Peabody Mus. Paleontol. 24: 903–912.
Nat. Hist. 30: 1–165. Salgado, L., Coria, R. A., and Calvo, J. O. 1997.
Ostrom, J. H., and McIntosh, J. S. 1966. Marsh's Evolution of titanosaurid sauropods. I: Phylogenetic
Dinosaurs, the Collections from Como Bluff. analysis based on the postcranial evidence.
Yale University Press, New Haven, CT, London. Ameghiniana 34: 3–32.
388 pp. Salgado, L., Coria, R. A., Chiappe, L. M. 2005.
Owen, R 1841. A description of a portion of the skele- Osteology of the sauropod embryos from the
ton of Cetiosaurus, a gigantic extinct saurian occur- Upper Cretaceous of Patagonia. Acta Paleontol.
ring in the Oolitic Formation of different parts of Polonica 50: 79–92.
England. Proc. Geol. Soc. London 3: 457–462. Sanz, J. L., Powell, J. E., Le Loeuff, J., Martínez, R., and
——–. 1842. Report on British fossil reptiles, Part II. Pereda-Suberbiola, X. 1999. Sauropod remains
Reptiles. Rep. Br. Assoc. Adv. Sci. 1841: 60–204. from the Upper Cretaceous of Laño (northcentral
——–. 1875. Monographs on the British fossil Reptilia Spain). Titanosaur phylogenetic relationships.
of the Mesozoic formations. Part II. (Genera Estud. Mus. Ci. Nat. de Álava 14: 235–255.
Bothriospondylus, Cetiosaurus, Omosaurus). Sereno, P. C. 1989. Prosauropod monophyly and
Palaeontol. Soc. Monogr. 29: 15–93. basal sauropodomorph phylogeny. J. Vertebr.
Phillips, J. 1871. Geology of Oxford and the Valley of Paleontol. 9: 38A.
the Thames. Clarendon Press, Oxford. 529 pp. ——–. 1991. Lesothosaurus, "fabrosaurids,” and the
Pittman, J.G., and Gillette, D.D. 1989. The Briar early evolution of Ornithischia. J. Vertebr.
Site: a new sauropod dinosaur tracksite in Lower Paleontol. 11: 168–197.
Cretaceous beds of Arkansas. In: Gillette, D.D., ——–. 1993. The pectoral girdle and forelimb of the
and Lockley, M.G. (eds.). Dinosaur Tracks and basal theropod Herrerasaurus ischigualastensis.
Traces. Cambridge University Press, Cambridge, J. Vertebr. Paleontol. 13: 425–450.
New York. Pp. 313–332. ——–. 1998. A rationale for phylogenetic definitions,
Powell, J.E. 1992. Osteologia de Saltasaurus loricatus with application to the higher-level phylogeny of
(Sauropoda-Titanosauridae) del Cretácico Superior Dinosauria. Neues Jahrb. Geol. Paläontol.
del Noroeste argentino. In Sanz, J. L. and Abhandlungen 210: 41–83.
Buscalioni, A. D., (eds.). Los Dinosaurios y su Sereno, P. C., and Novas, F.E. 1993. The skull and
Entorno Biotico. Instituto “Juan de Valdes”, neck of the basal theropod Herrerasaurus
Cuenca. Pp. 165–230. ischigualastensis. J. Vertebr. Paleontol. 13: 451–476.
Raath, M. 1972. Fossil vertebrate studies in Rhodesia: a Sereno, P. C., Beck, A. L., Dutheil, D. B., Larsson,
new dinosaur (Reptilia, Saurischia) from the near H. C. E., Lyon, G. H., Moussa, B., Sadleir, R. W.,
the Trias-Jurassic boundary. Arnoldia 30: 1–37. Sidor, C. A., Varricchio, D. J., Wilson, G. P., and
Rauhut, O.W.M., and Werner, C. 1997. First record Wilson, J. A. 1999. Cretaceous sauropods from
of a Maastrichtian sauropod dinosaur from Egypt. the Sahara and the uneven rate of skeletal evo-
Palaeontol. Africana 34: 63–67. lution among dinosaurs. Science 286: 1342–
Riggs, E.S. 1904. Structure and relationships 1347.
of opisthocoelian dinosaurs. Part II: The Smith, J. B., Lamanna, M. C., Lacovara, K. J., Dodson,
Brachiosauridae. Field Columbian Museum P., Smith, J. R., Poole, J. C., Giengengack R., and
Geological Series 2: 229–248. Attia, Y. 2001. A giant sauropod dinosaur from an
Romer, A. S. 1956. Osteology of the Reptiles. Upper Cretaceous mangrove deposit in Egypt.
University of Chicago Press, Chicago, 772 pp. Science 292: 1704–1706.
——–. 1966. Vertebrate Paleontology. University of Steel, R. 1970. Saurischia. Handb. Paläoherpetol. 13:
Chicago Press, Chicago. 368 pp. 1–88.

Sullivan, C., Jenkins, F.A., Gatesy, S.M., and Shubin, Wilson, J.A., and Carrano, M.T. 1999. Titanosaurs
N.H. 2003. A functional asssessment of hind foot and the origin of “wide-gauge” trackways: a bio-
posture in the prosauropod dinosaur Plateosaurus. mechanical and systematic perspective on sauro-
J. Vertebr. Paleontol. 23: 102A. pod locomotion. Paleobiology 25: 252–267.
Thulborn, R. A. 1982. Speeds and gaits of dinosaurs. Wilson, J.A., and Sereno, P.C. 1998. Early evolution
Palaeogeogr. Palaeoclimatol. Palaeoecol. 38: and higher-level phylogeny of sauropod dinosaurs.
227–256. Soc. Vertebr. Paleontol. Mem. 5: 1–68 (supplement
——–. 1990. Dinosaur Tracks. Chapman and Hall, to J. Vertebr. Paleontol. 18).
London. 410 pp. Wilson, J. A., Martinez, R. A., and Alcobar, O. A.
Tsuihiji, T. 2004. The ligament system in the neck of 1999. Distal tail of a titanosaurian sauropod from
Rhea Americana and its implication for the bifur- the Upper Cretaceous of Mendoza, Argentina.
cated neural spines of sauropod dinosaurs. J. Journal of Vertebrate Paleontology 19: 592–595.
Vertebr. Paleontol. 24: 165–172. Wilson, J. A., and Upchurch, P. 2003. A revision of
Upchurch, P. 1995. The evolutionary history of Titanosaurus Lydekker (Dinosauria–Sauropoda),
sauropod dinosaurs. Philos. Trans. Roy. Soc. the first dinosaur genus with a ‘Gondwanan’ dis-
London B 349: 365–390. tribution. Journal of Systematic Palaeontology 1:
———. 1998. The phylogenetic relationships of sauro- 125–160.
pod dinosaurs. Zool. J. Linn. Soc. 124: 43–103. Yates, A. M. 2001. A new look at Thecodontosaurus
———. 1999. The phylogenetic relationships and the origin of sauropod dinosaurs. J. Vertebr.
of the Nemegtosauridae (Saurischia, Sauropoda). Paleontol. 21: 116A.
Journal of Vertebrate Paleontology 19: 106–125. ——–. 2003. A new species of the primitive dinosaur,
Upchurch, P., and Barrett, P. M. 2000. The evolution Thecodontosaurus (Saurischia: Sauropodomorpha)
of sauropod feeding. In: Sues, H.-D. (ed.). Evolution and its implications for the systematics of early
of Feeding in Terrestrial Vertebrates. Cambridge dinosaurs. J. Syst. Palaeontol. 1: 1–42.
University Press, Cambridge. pp.79–122. ——–. 2004. Anchisaurus polyzelus Hitchcock: the
Upchurch, P., and Martin, J. 2003. The anatomy and smallest known sauropod dinosaur and the evolu-
taxonomy of Cetiosaurus (Saurischia, Sauropoda) tion of gigantism amongst sauropodomorph
from the Middle Jurassic of England. J. Vertebr. dinosaurs. Postilla. Pp 1–58.
Paleontol. 23: 218–231. Yates, A. M., and Kitching, J. W. 2003. The earliest
Upchurch, P., Hunn, C.A., and Norman, D.B. 2002. known sauropod dinosaur and the first steps
An analysis of dinosaurian biogeography: evi- towards sauropod evolution. Proc. Roy. Soc. London
dence for the existence of vicariance and disper- B 270: 1753–1758.
sal patterns caused by geological events. Proc. Young, C.C. 1941. A complete osteology of
Roy. Soc. London B 269: 613–621. Lufengosaurus huenei Young (gen. et sp. nov.).
Upchurch, P., Barrett, P.M., and Dodson, P. 2004. Palaeontol. Sinica (C) 7: 1–53.
Sauropoda. In: Weishampel, D.B., Dodson, ——–. 1947. On Lufengosaurus magnus (sp. nov.) and
P., and Osmólska, H. (eds.). The Dinosauria, 2nd additional finds of Lufengosaurus huenei Young.
ed. University of California Press, Berkeley. Pp. Palaeontol. Sinica (C) 12: 1–53.
259–322. Yu, C. 1993. The skull of Diplodocus and the phy-
Wilson, J.A. 2002. Sauropod dinosaur phylogeny: logeny of the Diplodocidae. PhD Thesis:
critique and cladistic analysis. Zool. J. Linn. Soc. University of Chicago.
136: 217–276. Zhang, Y. 1988. The Middle Jurassic dinosaur fauna
——–. 2005a. Redescription of the Mongolian sauro- from Dashanpu, Zigong, Sichuan. J. Chengdu Coll.
pod Nemegtosaurus mongoliensis Nowinski Geol. 3: 1–87.
(Dinosauria: Saurischia) and comments on Late Zhang, Y., and Yang, Z. 1994. [A New Complete
Cretaceous sauropod diversity. J. Syst. Palaeontol. Osteology of Prosauropoda in Lufeng Basin,
3: 283–318. Yunnan, China; Jingshanosaurus]. Yunnan
——–. 2005b. Integrating ichnofossil and body fos- Publishing House of Science and Technology,
sil records to estimate locomotor posture and Kunming. 100 pp.
spatiotemporal distribution of early sauropod Zhou, Z., Barrett, P. M., and Hilton, J. 2003. An
dinosaurs: a stratocladistic approach. Paleo- exceptionally preserved Lower Cretaceous ecosys-
biology 31: 400–423. tem. Nature 421: 807–814. re

APPENDIX 1.1. CHARACTER DESCRIPTIONS 93. Dorsal neural spines, length: approximately
twice (0) or approximately four times (1) cen-
Note: Complete character descriptions for synapomor- trum length.
phies discussed in text and listed in Table 1.1 (from
136. Distalmost caudal centra, articular face shape:
Wilson 2002). Primitive state is indicated as 0;1–5 repre-
platycoelous (0); biconvex (1).
sent derived states.
137. Distalmost biconvex caudal centra, length-to-
37. Pterygoid, transverse flange (i.e., ectoptery- height ratio: 4(0); 5(1).
goid process) position: posterior of orbit (0); 138. Distalmost biconvex caudal centra, number:
between orbit and antorbital fenestra (1); 10 or fewer (0); more than 30 (1).
anterior to antorbital fenestra (2). 149. Posture: bipedal (0); columnar, obligately
46. Basipterygoid processes, length: short, quadrupedal posture (1).
approximately twice (0); or elongate, at least 151. Scapular blade, orientation: perpendicular to
four times (1) basal diameter. (0) or forming a 45 angle with (1) coracoid
53. Basipterygoid processes, orientation: perpendi- articulation.
cular to (0) or angled approximately 45 to (1) 156. Caracoid, anteroventral margin shape:
skull roof. rounded (o); rectangular (1).
55. Dentary, depth of anterior end of ramus: 158. Sternal plate, shape: oval (0); crescentic (1).
slightly less than that of dentary at midlength
(0); 150% minimum depth (1). 161. Humeral deltopectoral crest, shape: relatively
narrow throughout length (0); markedly
65. Tooth rows, shape of anterior portions: nar- expanded distally (1).
rowly arched, anterior portion of tooth rows
V-shaped (0); broadly arched, anterior portion 163. Humeral distal condyles, articular surface
of tooth rows U-shaped (1); rectangular, tooth- shape: restricted to distal portion of humerus
bearing portion of jaw perpendicular to jaw (0), exposed on anterior portion of humeral
rami (2). shaft (1).
66. Tooth rows, length: extending to orbit (0); 164. Humeral distal condyle, shape: divided (0);
restricted anterior to orbit (1); restricted ante- flat (1).
rior to subnarial foramen (2). 167. Ulnar olecranon process, development:
67. Crown-to-crown occlusion: absent (0); pres- prominent, projecting above proximal articu-
ent (1). lation (0); rudimentary, level with proximal
articulation (1).
69. Tooth crowns, orientation; aligned along jaw
axis, crowns do not overlap (0); aligned 170. Radius, distal breadth: slightly larger than
slightly anterolingually, tooth crowns overlap (0) or approximately twice (1) midshaft
(1). breadth.
70. Tooth crowns, cross-sectional shape at mid- 172. Humerus-to-femur ratio: 0.60 (0); 0.60 (1).
crown: elliptical (0); D-shaped (1); cylindrical (2). 173. Carpal bones, number: three or more (0); two
71. Enamel surface texture: smooth (0); wrinkled or fewer (1).
(1). 175. Metacarpus, shape: spreading (0); bound,
74. Replacement teeth per alveolus, number: two with subparallel shafts and articular surfaces
or fewer (0); more than four (1). that extend half their length (1).
80. Cervical vertebrae, number: 9 or fewer (0); 10 176. Metacarpals, shape of proximal surface in
(1); 12 (2); 13 (3); 15 or greater (4). articulation: gently curving, forming a 90
arc (0); U-shaped, subtending a 270 arc (1).
85. Anterior cervical neural spines, shape: single
(0); bifid (1). 181. Manual digits II and III, phalangeal number:
2-3-4-3-2 or more (0); reduced, 2-2-2-2-2 or
89. Posterior cervical and anterior dorsal neural less (1); absent or unossified (2).
spines, shape: single (0); bifid (1).
187. Iliac preacetabular process, orientation:
91. Dorsal vertebrae, number: 15 (0); 14 (1); 13 (2); anterolateral to (0) or perpendicular to (1)
12 (3); 11 (4); 10 (5). body axis.

198. Femoral midshaft, transverse diameter: sube- 217. Metatarsus, posture: bound (0); spreading (1).
qual to (0), 125%–150%, or (1) at least 185% 221. Metatarsal l, minimum shaft width: less than
(2) anteroposterior diameter. (0) or greater than (1) that of metatarsals II–IV.
199. Femoral shaft, lateral margin shape: straight 223. Metatarsal III length: more than 30% (0) or
(0); proximal one-third deflected medially (1). less than 25% (1) that of tibia.
201. Femoral distal condyles, orientation: perpen- 224. Metatarsals III and IV, minimum transverse
dicular or slightly beveled dorsolaterally (0) or shaft diameters: subequal to (0) or less than
beveled dorsomedially approximately 10 (1) 65% (1) that of metatarsals I or II (1).
relative to femoral shaft.
225. Metatarsal V, length: shorter than (0) or at
205. Tibia, distal breadth: approximately 125% (0) least 70% (1) length of metatarsal IV.
or more than twice (1) midshaft breadth.
228. Pedal unguals, orientation: aligned with (0)
210. Astragalus, shape: rectangular (0); wedge- or deflected lateral to (1) digit axis.
shaped, with reduced anteromedial corner (1). 233. Pedal digit IV ungual, development: subequal
216. Distal tarsals 3 and 4: present (0); absent or in size to unguals of pedal digits II and III
unossified (1). (0); rudimentary or absent (1).

two
Titanosauria
A PHYLOGENETIC OVERVIEW
Kristina Curry Rogers
The discovery of preserved titanosaur eggs and

T itanosaur body fossils have
been recovered from every landmass ex-
cept Antarctica and are present in Upper Juras-
embryos at Auca Maheuvo in Argentina illumi-
nates aspects of their ontogeny and reproduc-
sic to Upper Cretaceous strata. Their unique, tive biology (Chiappe et al. 2001; see Chiappe
wide-gauge trackways extend their record back et al., chapter 10). New bone histological research
still farther, to the Middle Jurassic (Santos et al. promises to provide important new insight
1994; Wilson and Carrano 1999; Day et al., into the growth strategies of various titanosaur
2002; Wilson and Upchurch 2003). In spite of taxa (e.g., Erickson et al. 2001; see Curry
their extensive temporal and geographic ranges, Rogers and Erickson, chapter 11), including the
their fossil record remains relatively poor, with largest terrestrial vertebrates of all time (e.g.,
most genera based only on fragmentary and/or Argentinosaurus). There is even a hint of
disassociated postcranial remains. Presently, Ti- dwarfism among some titanosaur taxa (e.g.,
tanosauria includes over 30 currently accepted Magyarosaurus [Jianu and Weishampel 1999],
genera representing more than one-third of Saltasaurus and Neuquensaurus [Coiki 1999;
sauropod diversity, and new titanosaur genera Powell 2003).
are being erected more prolifically than for any In spite of these advances, Titanosauria has
other group of sauropods (see Wilson and Curry remained a phylogenetic puzzle. Although recent
Rogers, Introduction). consensus regarding the macronarian origin of
New titanosaur finds from around the globe titanosaurs has been attained, included taxa and
paint a picture of a dramatically diverse group, their interrelationships are still debated (e.g.,
which includes bizarre armored forms (e.g., Salgado et al. 1997; Wilson and Sereno 1998;
Agustinia, [Bonaparte 1998], Magyarosaurus Upchurch 1998; Sanz et al. 1999; Smith et al.
[Csiki 1999], Neuquensaurus [Powell 2003], 2001; Curry Rogers and Forster 2001, Wilson
Saltasaurus [Bonaparte and Powell 1980]) and 2002; Wilson and Upchurch 2003, Martínez et al.
imposing giants (Argentinosaurus [Bonaparte 2004; Apesteguía 2004). Even the type genus
and Coria 1993], Paralititan [Smith et al. 2001]). and species, ‘Titanosaurus indicus,’ has been
50
invalidated (Wilson and Upchurch 2003). The Coria 1993) and placed diplodocoids within
only means of improving phylogenetic resolu- Titanosauridae. Similarly, Steel (1970) included
tion for Titanosauria is through more detailed Titanosauria sensu stricto and diplodocoids
analyses and addition of well-preserved speci- within a new group, Atlantosauridae. More
mens. Taxa that are known from both cranial recent workers following traditional methods
and postcranial data are essential if we are to for higher-level sauropod classification either
bridge the gaps in our current understanding of restored Titanosauridae to family level (e.g.,
titanosaur anatomy. This “whole-skeleton” view Berman and McIntosh 1978; Powell 2003;
of Titanosauria allows controversial points in McIntosh 1989, 1990a, 1990b) or ignored
existing phylogenetic hypotheses to be tested titanosaurs altogether (e.g., Bonaparte 1986).
with a more robust data set. Titanosauria was established by Bonaparte
In this chapter I consider some of the con- and Coria (1993), who recognized the need for
tentious aspects of titanosaur phylogeny, with a a higher taxon to subsume two distinct
focus on within-group relationships. The dis- titanosaur groups: (1) Andesauridae, whose
cussion includes an historical primer and brief members are characterized by the presence of
description of Rapetosaurus krausei, a nearly hyposphene-hypantra and amphiplatyan cau-
complete titanosaur from the Upper Cretaceous dals, and (2) Titanosauridae, whose members
Maevarano Formation of Madagascar. The chap- lack hyposphene-hypantra and exhibit pro-
ter concludes with an analysis of titanosaur phy- coelous caudals. Andesauridae, also coined by
logeny that utilizes cranial and postcranial data. Bonaparte and Coria (1993), is based only on
This phylogenetic treatment is intended to serve primitive characters that specify a paraphyletic
as a framework and catalyst for continued analy- group and must be considered an informal
ses of titanosaur biology, evolution, and paleo- name until taxa are identified that share
biogeography. synapomorphies with Andesaurus (Wilson and
Upchurch 2003).
In a recent revision of 14 Titanosaurus
BACKGROUND ON TITANOSAUR species, including ‘T. indicus’ (the type species)
SYSTEMATICS Wilson and Upchurch (2003) recognized only
five valid species and deemed the type species
TAXONOMIC BACKGROUND ‘T. indicus’ a nomen dubium due to its basis on
The first titanosaur specimens were discovered obsolete characters. Consequently, they pro-
in the late 1800s in India (‘Titanosaurus indicus’ posed the abandonment of all coordinated rank-
Falconer 1868; Lydekker 1877; ‘T.’ blanfordi taxa, including Titanosaurinae, Titanosauridae,
Lydekker 1879), Europe (Macruosaurus Seeley and Titanosauroidea, and proposed and defined
1869, 1876), Argentina (Argyrosaurus Lydekker a standardized node and stem-based nomencla-
1893), and Madagascar (‘T.’ madagascariensis ture for Titanosauria (table 2.1). This nomen-
Depéret 1896a, 1896b). The definition and posi- clatural system is particularly significant given
tion of Titanosauridae have remained ambigu- the inconsistent usage of names and taxa
ous since its original usage in 1893, and the employed in the cladistic analyses outlined
taxon has long been used as a receptacle for enig- below (table 2.2, fig. 2.1).
matic and fragmentary Cretaceous sauropods,
with little agreement on included group mem- PHYLOGENETIC BACKGROUND
bers (Gilmore 1946; Romer 1956, 1966, 1968; The phylogenetic relationships of Titanosauria
Steel 1970; McIntosh 1990a, 1990b). For exam- have received relatively little attention, in part
ple, Romer (1956, 1966, 1968) united the because of the frustratingly fragmentary nature
Cretaceous taxa with procoelous caudal verte- of most genera. In the past decade, only a hand-
brae (Titanosauria sensu stricto Bonaparte and ful of phylogenetic analyses have examined
T I TA N O S AU R I A 51
TABLE 2.1
Phylogenetic Definitions for Titanosauria and Its Subclades
Titanosauria (Bonaparte and Coria 1993) Andesaurus delgadoi (Calvo and Bonaparte 1991),
Saltasaurus loricatus (Bonaparte and Powell 1980),
their most recent common ancestor, and all
descendants.
Lithostrotia (Upchurch et al. 2004) Malawisaurus dixeyi (Haughton 1928), Saltasaurus
loricatus (Bonaparte and Powell 1980), their most
recent common ancestor, and all descendants.
Saltasauridae (Bonaparte and Powell 1980) Opisthocoelicaudia skarzynskii (Borsuk-Bialynicka
1977), Saltasaurus loricatus (Bonaparte and Powell
1980), their most recent common ancestor, and all
descendants.
Saltasaurinae (Bonaparte and Powell 1980) All saltasaurids more closely related to Saltasaurus
loricatus (Bonaparte and Powell 1980) than to
Opisthocoelicaudia skarzynskii (Borsuk-Bialynicka
1977).
Opisthocoelicaudiinae (McIntosh 1990) All saltasaurids more closely related to
Opisthocoelicaudia skarzynskii (Borsuk-Bialynicka
1977) than to Saltasaurus loricatus (Bonaparte and
Powell 1980).
NOTE: As outlined by Wilson and Upchurch (2003) and as employed in phylogenetic analysis, node-based definitions are in boldface
type; stem-based definitions are in regular type.
Titanosauria in detail (Salgado et al. 1997; Curry appendicular) in 10 titanosaurs and six basal
2001; Curry Rogers and Forster 2001; Curry sauropod taxa. Two most parsimonious trees
Rogers 2001; Wilson 2002; Upchurch et al. (length, 54 steps; CI 0.81) were generated, dif-
2004). Each of these analyses employed a differ- fering only in the placement of Malawisaurus and
ent array of taxa, though a core group of relatively Epacthosaurus. Their consensus hypothesis was
well-known titanosaurs was common to all (table well resolved at the base of the tree, particularly
2.2). Most other cladistic studies have focused on with regard to all nontitanosaurs (fig. 2.1A).
higher-level sauropod phylogeny (Gauthier 1986; Salgado et al. (1997) were first to formally recog-
Yu 1993; Wilson and Sereno 1994, 1998; Calvo nize the close relationship between Brachiosaurus
and Salgado 1995; Upchurch 1995, 1998, 1999; and Titanosauria within Titanosauriformes and
Wilson 1999; Sanz et al., 1999). Several of these diagnosed Titanosauria with three unambiguous
studies have included one or more titanosaurs or postcranial synapomorphies. Though character
have used the higher-level grouping Titanosauria support for nodes within Titanosauria was fairly
as a terminal taxon. However, relationships poor, Salgado et al. (1997) refuted Bonaparte
within Titanosauria in these analyses tend to be and Coria’s (1993) hypothesis of two distinct
poorly supported (e.g., Salgado et al. 1997; families within Titanosauria (Andesauridae and
Upchurch 1998; Sanz et al. 1999; Curry Rogers Titanosauridae) and, instead, only supported
and Forster 2001) or utilize only a small fraction Titanosauridae monophyly. They included all
of available characters. Several of these previous titanosaurs except Andesaurus within this group
analyses are summarized below (table 2.2, figs. based on the presence of proximal caudals with
2.1, 2.2). strongly procoelous “ball and socket” articula-
Salgado et al. (1997) provided the first treat- tions. Saltasaurinae constituted the only other
ment of titanosaur ingroup relationships. They named taxon in the phylogenetic hypothesis of
analyzed 38 postcranial characters (20 axial, 18 Salgado et al. (1997). It was supported by two
52 T I TA N O S AU R I A
TABLE 2.2
Genera Employed for Seven Analyses of Titanosaur Relationships
TITANOSAUR GENERA TITANOSAUR GENERA
Salgado et al. 1997 Aeolosaurus Curry 2001 Augustinia

Alamosaurus Epacthosaurus
Andesaurus ‘Jainosaurus’
Argentinosaurus Janenschia
Epacthosaurus Lirainosaurus
Malawisaurus Magyarosaurus
Neuquensaurus Malawisaurus
Opisthocoelicaudia Malagasy Taxon B
Saltasaurus Nemegtosaurus
‘Titanosaurinae indet.’ Neuquensaurus
Opisthocoelicaudia
Upchurch 1998 Alamosaurus
Paralititan
Andesaurus
Phuwiangosaurus
Malawisaurus
Quaesitosaurus
Opisthocoelicaudia
Rapetosaurus
Phuwiangosaurus
Rocasaurus
Saltasaurus
Saltasaurus
Sanz et al. 1999 Andesaurus Santa Rosa indet.
Argyrosaurus Isisaurus
Epacthosaurus (Titanosaurus)
Lirainosaurus colberti
Opisthocoelicaudia
Wilson 2002 Alamosaurus
Peirópolis form
Malawisaurus
Saltasaurus
Nemegtosaurus
Curry Rogers Neuquensaurus
and Forster 2001 Alamosaurus Opisthocoelicaudia
Antarctosaurus Rapetosaurus
Malawisaurus Saltasaurus
Nemegtosaurus Isisaurus
Neuquensaurus (=Titanosaurus)
Opisthocoelicaudia colberti
Quaesitosaurus
Upchurch et al. 2004 Alamosaurus
Rapetosaurus
Andesaurus
Saltasaurus
Argentinosaurus
Isisaurus
Austrosaurus
(Titanosaurus)
Gondwanatitan
colberti
Lirainosaurus
Curry 2001 Aegyptosaurus Malawisaurus
Aeolosaurus Opisthocoelicaudia
Alamosaurus Pellegrinisaurus
Ampelosaurus Phuwiangosaurus
Andesaurus Saltasaurus
Antarctosaurus Isisaurus
Argentinosaurus (Titanosaurus)
Argyrosaurus colberti
NOTE: Boldface type indicates genera included in only one analysis.
FIGURE 2.1. Recent hypotheses of titanosaur relationships. (A) Salgado et al. (1997); (B) Upchurch (1998); (C) Sanz et al.
(1999); (D) Curry Rogers and Forster (2001); (E) Wilson (2002); (F) Upchurch et al. (2004). Topologies have been simplified
to reflect genera included in more than one analysis (see table 2.2). (Modified from Wilson and Upchurch 2003.)
unambiguous synapomorphies related to caudal Upchurch defined Titanosauroidea as all taxa

vertebral morphology and included the South more closely related to “true titanosaurids” (e.g.,
American Neuquensaurus and Saltasaurus. Saltasaurus) than they are to brachiosaurids. In
Upchurch (1998) analyzed 205 characters (74 terms of lower-level relationships, Upchurch pro-
cranial, 76 axial, 68 appendicular) among 26 vided the first evidence for inclusion of
sauropods, including 5 titanosaurs (Alamosaurus, Phuwiangosaurus as a basal titanosaur. As in the
Andesaurus, Malawisaurus, Opisthocoelicaudia, analysis by Salgado et al. (1997), Andesaurus and
Saltasaurus) and several other controversial taxa Malawisaurus were successive sister-taxa to other
(Euhelopus, Phuwiangosaurus, Nemegtosaurus, titanosaurs. In spite of the recognized paraphyly
Quaesitosaurus) (fig. 2.1B). Although Upchurch of Andesauridae (Salgado et al. 1997), Upchurch
(1995) originally supported the monophyly of a (1998) retained the name in his systematic clas-
group including titanosaurs and diplodocoids as sification of the Sauropoda to designate a group
sister-taxa, the 1998 revision based on additional of basal titanosaurs including Phuwiangosaurus,
character data and taxa supported the monophyly Andesaurus, and Malawisaurus. Upchurch (1998)
of Titanosauriformes. Within Titanosauriformes, restricted the Titanosauridae to the clade
2.1D). A clade containing Haplocanthosaurus
and Andesaurus was the sister-taxon to
Titanosauroidea, thus revising Upchurch’s
(1998) use of the term to exclude Andesaurus.
Within Titanosauroidea, Sanz et al. (1999) sup-
ported the monophyly of Titanosauria and pro-
posed Eutitanosauria, for a more derived clade
of titanosaurs diagnosed, in part, by the pres-
ence of dermal armor.
Curry Rogers and Forster (2001) analyzed
228 characters (74 cranial and 154 postcranial)
FIGURE 2.2. Adams consensus tree of titanosaur relation- among 16 ingroup taxa and supported the mono-
ships based on the six analyses depicted in figure 2.1.
phyly of Titanosauriformes (fig. 2.1E). Sixteen
(Modified from Wilson and Upchurch 2003.)
additional steps were required to move
Titanosauria into its more traditional position as
including Alamosaurus, Opisthocoelicaudia, and the sister-taxon to Diplodocoidea. The mono-
Saltasaurus. Upchurch (1998, 1999) did not phyly of Titanosauria was strongly supported by
support the inclusion of Nemegtosaurus and 20 characters, 9 of which were unambiguous,
Quaesitosaurus within Titanosauroidea. and two main titanosaur clades were distin-
Wilson and Sereno (1998) analyzed 109 char- guished by the analysis. One clade included
acters (32 cranial, 24 vertebral, 53 appendicular) Rapetosaurus, Nemegtosaurus, Malawisaurus,
for 10 sauropod taxa and obtained a single most Quaesitosaurus, and Antarctosaurus and was diag-
parsimonious tree with 153 steps (fig. 2.1C). nosed by four unambiguous cranial synapomor-
Titanosauria was included as a higher-level ter- phies. Malawisaurus was the sister-taxon of
minal taxon. Wilson and Sereno (1998) corrobo- Quaesitosaurus (Rapetosaurus Nemegtosaurus).
rated Titanosauriformes monophyly, but within Rapetosaurus and Nemegtosaurus were united by
Titanosauriformes they hypothesized a sister- seven ambiguous synapomorphies, five of which
group relationship between Euhelopus and occur in one or more diplodocoids. A second
Titanosauria within Somphospondylii. Relation- clade (Saltasaurinae in their analysis) included
ships among titanosaurs were not discussed in a Alamosaurus, ‘T.’ colberti, Neuquensaurus,
phylogenetic context in this analysis of higher- Saltasaurus, and Opisthocoelicaudia.
level sauropod relationships, but a few points of Wilson (2002) scored 234 morphological
note were mentioned regarding Titanosauria as a characters (76 cranial, 72 axial, 85 appendicular, 1
terminal taxon (Wilson and Sereno 1998). They dermal) for 27 taxa. Three equally parsimonious
suggested that Andesaurus and Malawisaurus are trees resulted, each supporting 26 internal nodes
basal titanosaurs that have an important role in (fig. 2.1F). In this analysis Titanosauriformes,
highlighting derived features among the clade Somphospondylii, and monophyly of Titanosauria
and confirmed that there is no phylogenetic sup- were well supported. Titanosauria united Malaw-
port for Andesauridae as defined by Bonaparte isaurus, Nemegtosauridae (Nemegtosaurus
and Coria (1993). In addition, Wilson and Sereno Rapetosaurus), ‘T.’ colberti (Isisaurus colberti
(1998) included both Nemegtosaurus and [Wilson and Upchurch 2003]), and Saltasauridae
Quaesitosaurus within Titanosauria. (Opisthocoelicaudiinae Saltasaurinae).
Sanz et al. (1999) conducted a phylogenetic Wilson (2002) identified several problematic
analysis of 43 characters (1 cranial, 27 vertebral, areas of the phylogeny, including the weak reso-
15 appendicular) derived from seven titanosaurs lution of ‘T.’ colberti (Wilson and Upchurch 2003)
and the two neosauropods. They obtained a sin- and Nemegtosaurus relative to Saltasauridae. The
gle most parsimonious tree of 70 steps (fig. problems in resolution were hypothesized to
reflect the incompletely known skeletons of these new taxa continue to be named and described at a
taxa and the nonoverlap of preserved anatomy rapid pace (see Introduction), few provide the
among titanosaurs. cranial and postcranial character data required
Upchurch et al. (2004) conducted a phyloge- if we are to address the most pertinent ques-
netic analysis of 309 characters among 41 sauro- tions of titanosaur phylogeny. The only means
pod genera. Characters in this analysis were syn- of improved phylogenetic resolution for
thesized from Wilson and Sereno (1998), Titanosauria is through more detailed analyses
Upchurch (1995, 1998, 1999), and Curry Rogers and addition of key specimens with cranial and
and Forster (2001). They obtained 1,056 most postcranial data.
parsimonious trees and applied reduced consen-
sus methods, which indicated that 36 of the 41
RAPETOSAURUS KRAUSEI, A KEYSTONE
sampled sauropod genera have stable positions
TITANOSAUR
across all 1,056 most parsimonious trees.
Deletion of five taxa, including the titanosaurs The recent discovery and description of
Andesaurus and Argentinosaurus, resulted in a sin- Rapetosaurus krausei from the Upper Cretaceous
gle fully resolved topology (fig. 2.1G). A bootstrap Maevarano Formation of Madagascar (Curry
analysis produced low support values (50%) for Rogers and Forster 2001, 2004; Curry, 2001)
all resolved clades within Titanosauria. Their provide the cranial and postcranial data from
analysis supported Titanosauriformes and juvenile and adult skeletons required for more
Titanosauria monophyly, as well as a mono- detailed, total-evidence analyses of titanosaur
phyletic Lithostrotia (diagnosed by the presence phylogeny.
of osteoderms) and Saltasauridae (the least inclu-
sive clade containing Opisthocoelicaudia and HISTORICAL BACKGROUND
Saltasaurus). Notably, they did not include Fragmentary sauropod remains have been
Nemegtosaurus or Quaesitosaurus within Titano- known from the Late Cretaceous of the
sauria but, instead, placed them within Mahajanga Basin (Maevarano Formation,
Diplodocoidea. Though bootstrap values for this Campanian(?)–Maastrichtian [Rogers et al.
inclusion were low, the clade uniting these two 2000]) for over a century. In 1896, a shipment
Mongolian sauropod skulls was high (60%). of fragmentary dinosaurian remains from the
northwest coast of Madagascar arrived in the
CONSENSUS AND CONTROVERSY office of French paleontologist Charles Depéret.
In spite of the diversity of hypotheses in each of The sauropod remains—namely, two weath-
the analyses outlined above, some resolution on ered procoelous caudal vertebrae (UCB
titanosaur relationships is apparent (figs. 2.1, [Universíté Claude Bernard, Lyon] 92829, UCB
2.2). Most analyses agree that Andesaurus is a ti- 92305; fig. 2.3A, B), a partial humeral diaphysis
tanosaur, as are Malawisaurus, and Isisaurus (UCB 92831; fig. 2.3C), and a large osteoderm
( ‘T.’ colberti [Wilson and Upchurch 2003]). (UCB 92827; fig. 2.3D) (Depéret 1896a,
More derived titanosaurs include Opisthocoelicau- 1896b)—were recovered from several distinct
dia, Alamosaurus, and a group most commonly localities (fig. 2.4A).
uniting Saltasaurus and Neuquensaurus. That Depéret (1896a) assigned the four elements
said, Titanosauria is among the most generic-rich from Madagascar to a new species of titanosaur,
of any sauropod clade, and few of the individual Titanosaurus madagascariensis. At that early date,
specimens include more than isolated (and often two titanosaurs were already known from India
poorly preserved) postcranial elements. Given the (T. indicus Lydekker 1877, T. blanfordi Lydekker
disparity of data and taxa considered, it is no won- 1879), and one taxon was recognized from the
der that general agreement and resolution on ti- Wealden Sandstone of the Isle of Wight
tanosaur ingroup phylogeny are lacking. Though (Macrurosaurus semnus Seeley 1869, 1876). The
FIGURE 2.3. Syntype elements
of ‘Titanosaurus madagascarien-
sis’ (Depéret 1896a). (A) UCB
(Universíte Claude Bernard)
92829, proximal caudal cen-
trum in ventral view. This cen-
trum is now referred to
Rapetosaurus krausei. (B) UCB
92305, midcaudal centrum in
left lateral view. This centrum is
now referred to Malagasy Taxon
B (Curry Rogers 2001). (C) UCB
92831 partial humeral diaphysis
in anterior view. (D) UCB 92827
osteoderm in dorsal view. Scale
bar equals 3 cm.
extremely procoelous caudal centra and proxi- collected titanosaur bones, including proximal
mally positioned neural arches of the Malagasy and midcaudals and fragmentary limb elements
caudal vertebrae indicated a close relationship from the Mahajanga Basin, were referred to ‘T.’
with the Indian titanosaurs, but several distinc- madagascariensis. Huene (1929) later subsumed
tions, including the deeply excavated ventral cen- the ‘T.’ madagascariensis syntype material within
trum and a double chevron articulation, war- the South American genus Laplatasaurus, and
ranted assignment to a new species. Depéret additional remains were reported from India (cf.
(1896a) was prescient in attributing the large Laplatasaurus madagascariensis Huene and
dermal ossification to the new species of Matley 1932). This reassessment stemmed from
titanosaur. He cited the unique pattern of vascu- the more robust “shorter and wider” caudal cen-
larization to rule out other dinosaurian forms trum (UCB 92305) included in the ‘T.’ madagas-
characterized by dermal ossifications (e.g., anky- cariensis syntype. More recent authors ques-
losaurs, stegosaurs), but this interpretation was tioned Huene’s assessment and reinstated the
subsequently called into question (e.g., Russell name ‘T.’ madagascariensis for all Late Cretaceous
et al., 1976; Hoffstetter 1957). In more recent sauropod material from Madagascar (e.g.,
years, dermal ossifications have been recognized Lavocat 1955 Besairie 1972; Russell et al. 1976;
from titanosaurs around the world (e.g., Csiki McIntosh 1990a, 1990b; Ravoavy 1991).
1999; Munyikwa et al., 1998), including new dis-
coveries from the same strata in the Mahajanga A REVISED PERSPECTIVE
Basin of Madagascar (Dodson et al. 1998). Over the last decade, a diverse Late Cretaceous
Collecting efforts continued sporadically in vertebrate assemblage that includes nonavian
the early 1900s (e.g., Thévenin 1907), and all the and avian theropod dinosaurs, fish, crocodiles,
FIGURE 2.4. (A) Map modified from Depéret’s (1896) report of fossil vertebrates from the Late Creta-
ceous of the Mahajanga Basin of northwestern Madagascar. Four localities were recorded by collec-
tors as indicated. (B) Outcrop map of Upper Cretaceous and Tertiary strata in the Mahajanga Basin of
northwestern Madagascar. Berivotra study area is highlighted. (C) Schematic profile of the Anem-
balemba Member (modified from Rogers 2004.) The Anembalema member in the Berivotra area is
underlain by the Masorobe member of the Maevarano Formation (MM) and overlain by the marine
Berivotra Formation (BF).
FIGURE 2.5. Rapetosaurus krausei from
the Upper Cretaceous Maevarano For-
mation of Madagascar. This skull recon-
struction is a composite based on holo-
type (UA 8698) and referred specimens
(FMNH PR 2184–2192, 2194,
2196–2197) and shown in (A) pos-
teroventral view, (B) left lateral view, and
(C) dorsal view. Shaded bones are pre-
served. an, angular; aof, antorbital fenes-
tra; ar, articular; bo, basioccipital; bs, ba-
sisphenoid; d, dentary; en, external
nares; eo, exoccipital; f, frontal; itf, in-
fratemporal fenestra; j, jugal; l, lacrimal;
ls, laterosphenoid–orbitosphenoid; m,
maxilla; n, nasal; orb, orbit; pa, parietal;
pf, prefrontal; pm, premaxilla; po, pos-
torbital; pop, paroccipital process; pt,
pterygoid; q, quadrate; qj, quadratojugal;
sa, surangular; so, supraoccipital; sq,
squamosal; stf, supratemporal fenestra.
Scale bar equals 10 cm. (D) Rapetosaurus
krausei skeletal reconstruction based on
a referred juvenile skeleton (FMNH PR
2209) recovered from a bonebed 2 in
MAD93-18. Scale bar equals 1 m. Bones
in gray are preserved. (Skull and skeletal
reconstructions by Mark Hallett; modi-
fied from Curry Rogers and Forster
2001.)
mammals, frogs, turtles, and snakes has been (namely, Malagasy Taxon B), ‘T.’ madagascariensis
recovered from the Anembalemba Member of must be considered a nomen dubium (Curry
the Maevarano Formation (fig. 2.4B) (e.g., 2001; Curry Rogers and Forster 2001; Wilson
Krause et al. 1994, 1997, 1999; Forster et al. and Upchurch 2003).
1996, 1998; Sampson et al. 1996, 1998, 2001;
Curry 1997; Asher and Krause 1998; Gottfried RAPETOSAURUS KRAUSEI
and Krause 1998; Buckley and Brochu 1999; The holotypic adult Rapetosaurus krausei skull,
Buckley et al. 2000; Rogers et al. 2000; Krause along with a well-preserved, associated juvenile
2001). All told, hundreds of new sauropod fos- postcranial skeleton with a partial skull, provides
sils have been recovered including associated our first look at titanosaur anatomy from head to
and articulated remains. These new finds indi- tail (fig. 2.5) (Curry 2001; Curry Rogers and
cate the presence of two distinct titanosaur taxa Forster, 2001). The disarticulated bones of the
in the Upper Cretaceous strata of Madagascar holotype skull were recovered from a single stra-
(Ravoavy 1991; Curry 2001; Curry Rogers and tum over an area of ~1 m2 at locality MAD96-02
Forster 1999a, 1990b, 2001). The new taxa are (fig. 2.6). Holotypic and the referred juvenile
most readily distinguished on the basis of their skeleton show no duplication of elements and
caudal vertebral morphologies along with other exhibit exact sutural articulations. Additional
postcranial variations. Both caudal morpholo- skull material from a juvenile specimen was
gies are included in the ‘Titanosaurus’ madagas- found in direct association with a nearly com-
cariensis syntype, which calls the validity of ‘T.’ plete juvenile postcranial skeleton at locality
madagascariensis into question. Pending further MAD93-18 (fig. 2.7). As in the adult skull, juve-
study of the sauropod fauna in Madagascar nile material also exhibits sutural articulation and
FIGURE 2.6. Map of quarry MAD96-02, which yielded the holotype skull of Rapetosaurus krausei
(UA; Université d’Antananarivo 8698). Preserved bones exhibit exact sutural articulations and in-
cluded the following: right maxilla with 8 teeth, left maxilla, right lacrimal, left jugal, right and left
nasals, right quadrate, right and left pterygoids, partial basioccipital, right paroccipital process, left
dentary with 11 teeth, right and left angulars, right surangular, and 5 additional teeth. Representa-
tive bones are labeled; scale bar equals 10 cm.
essentially no duplication of elements. Several of directed median parietal processes; a pterygoid

the juvenile skull elements from MAD93-18 with a dorsoventrally expanded rostral process
share autapomorphies with the adult holotype and an extremely shallow basipterygoid articula-
(fig. 2.8), and thus the associated juvenile skele- tion; basiptyergoid processes that diverge only at
ton and skull were both assigned to a new their distal ends; and a dentary with 11 alveoli
titanosaur genus and species: Rapetosaurus krau- extending for two-thirds of the length of this ele-
sei (Curry 2001; Curry Rogers and Forster 2001, ment (Curry 2001, Curry Rogers and Forster
2004; Curry Rogers 2001). 2001, 2004). Several of the appendicular skeletal
Rapetosaurus is diagnosed by an enlarged characters reported as autapomorphies for
antorbital fenestra extending along and anterior Rapetosaurus in Curry Rogers and Forster (2001)
to the tooth row; a preantorbital fenestra posi- are now interpreted to have a wider distribution
tioned caudal to the antorbital fenestra; a rostrally among titanosaurs and are excluded from the
located, dorsoventrally elongated subnarial fora- diagnosis here as in Upchurch et al. (2004).
men; a narrow and caudodorsally elongate maxil- The skull of Rapetosaurus is superficially sim-
lary jugal process; a median dome on the frontal; ilar to skulls of diplodocoid taxa in its elongated,
a V-shaped quadratojugal articulation on the narrow morphology, with external nares retracted
quadrate; a supraoccipital with two rostrally to the level of the orbits (figs. 2.5A–C, 2.8). The
FIGURE 2.7. (A) Cross section through the multiple bone-bearing horizons of quarry MAD93-18. Three discrete bonebeds
(BB1, BB2, BB3) each yield bones referable to Rapetosaurus. (B) BB2 yields the juvenile skull and skeleton described in the
text. The skeleton was closely associated, but only a few vertebrae were articulated. BB1 and BB3 yielded bones of subadult
and adult Rapetosaurus. (C) Juvenile Rapetosaurus skeleton from BB2; cervical vertebrae and ribs, radius in the foreground,
scapulae, metacarpals, and femora in the distance. (D) Subadult Rapetosaurus skeleton from BB3; ribs are articulated in dis-
tance, and forelimb and hindlimb (including left foot) are also closely associated in BB3. (E) Map of BB2, quarry MAD93-18.
Map highlights the skeleton of Rapetosaurus krausei. Scale bar equals 1 m. (Modified from Rogers 2005.)
basicranium of Rapetosaurus is tipped only pterygoids and external nares that are undivided
slightly caudoventrally, and its teeth are peglike and bound, in part, by the ascending process of
with high-angled wear facets. In spite of this the maxilla. The Rapetosaurus jugal has an elon-
general similarity to the skulls of taxa like gate maxillary process that lengthens the antor-
Diplodocus and Apatosaurus, Rapetosaurus differs bital region of the skull, and the infraorbital
from these taxa in the presence of platelike margin of the skull is shortened as the jugal
FIGURE 2.8. Rapetosaurus krausei adult and juvenile skull elements compared. (A) Juvenile
(FMNH PR 2190) and adult (UA 8698) right quadrates in posterior view; (B) juvenile
(FMNH PR 2191) and adult (UA 8698) left pterygoids in medial view, anterior to right;
(C) adult (UA8698) and juvenile (FMNH PR 2184, 2197) basicrania (in posterior view; and
(D) juvenile (FMNH PR 2187) and adult (UA8698) right surangulars in lateral view, ante-
rior to left. Scale bar equals 3 cm.
contribution to the antorbital fenestra boundary insights on cranial character correlation (e.g.,
is reduced. Rapetosaurus frontals are character- narial retraction and braincase rotation) and
ized by prominent midline elevations, and the indicates that these characters are decoupled
caudal crest of the parietal is well defined. The from one another during this phase of sauropod
basicranium includes elongated basipterygoid skull evolution, as hypothesized by Chiappe
processes with a narrow, U-shaped division, and et al. (2001).
elongate, topographically low basal tubera. Teeth The axial skeleton of Rapetosaurus (Figs.
extend the full the length of the upper and lower 2.5D, 2.9A–E) is characterized by at least 15 cer-
jaws and have precise occlusion along their lin- vical vertebrae, 10 dorsal vertebrae, and 6
gual surfaces. Comparisons of Rapetosaurus sacral vertebrae. The total number of caudals is
skull material to that of other known titanosaurs unknown, but 18 middle and posterior caudals
and neosauropods have provided cranial were preserved with FMNH PR 2209 from
synapomorphies for Titanosauriformes and MAD93-18. The anterior cervical vertebrae are
Somphospondylii, which were previously diag- characterized by elongated centra (elongation
nosed only by postcranial features. In addition index centrum length/height of posterior face
to new data for phylogeny reconstruction, the 4.0) with shallow, poorly defined lateral pneu-
Rapetosaurus skull material provides new matic fossae and butterfly-shaped neurocentral
FIGURE 2.9. Representative postcranial elements from a juvenile specimen of Rapetosaurus krausei (FMNH
PR 2209). (A) Spongy bone in presacral vertebrae. Middorsal vertebra is shown in right lateral view. Inset
details somphospondylus bone texture in a middorsal centrum). (B) Right scapula in lateral view. Note the
unexpanded distal blade and medially deflected glenoid. (C) Left humerus in anterior view. Arrow outlines
rectangular proximal end. (D) Anterior dorsal vertebra in left lateral view with posteriorly inclined dorsal
neural spines with rudimentary fossae and lamination. (E) Anterior caudal vertebra in left lateral view is
procoelous with an anteriorly positioned neural arch. (F) Left ulna in anterior view. Arrow points to prominent
olecranon process. (G) Left radius in posterior view. Arrow points to the medially beveled distal end. (H) left
pubis, ischium, and ilium in lateral view. Arrows point to reduced ischium length relative to that of pubis.
(I) Left femur in anterior view. Arrow points to the prominent proximolateral flange. Scale bar equals 3 cm.
articulations (resulting in constriction at the morphology, with centra and neural arches
midpoint of the neural canal). Anterior cervical becoming short and broad, while retaining the
neural arches are relatively low and proximally butterfly-shaped neurocentral articulation. All
divided (though single in their distal extremity), cervical neural arches have lateral fossa con-
with strongly developed lateral lamination and taining deep pneumatic foramina with sharply
fossae. The cervicodorsal transition is marked demarcated borders. Cervical ribs are elongate
by a dramatic departure in standard cervical and gracile, and extend posteriorly the length
of at least three succeeding vertebrae. Dorsal transverse and anteroposterior dimensions.
vertebrae are characterized by strongly opistho- The radius of Rapetosaurus is gracile and dis-
coelous centra with straight, parallel neurocen- tinctively antieorly bowed, with equidimen-
tral articulations and deep lateral pneumatic sional proximal and distal ends. The radius and
fossae. Neural arches are low, with posteriorly ulna articulation is clearly demarcated by the
tipped triangular neural spines and trans- presence of a strong, interosseus ridge that
versely oriented, winglike diapophyses. In pos- extends along the full length of the posterior
terior dorsals the neural spine is more verti- surface of the radius. The ulna is similar to
cally oriented, and by the most posterior dor- those of other titanosaurs, with a low but well-
sals the neural spine does not project posterior defined olecranon process. The Rapetosaurus
to the neural arch facets. Pre- and postspinal metacarpus is vertically oriented, with meta-
laminae are prominent throughout the series, carpals bound to form a semilunate arch with a
as are strongly developed lateral neural arch flat, proximal surface.
laminae and fossae. Dorsal ribs are delicate The hindlimb of Rapetosaurus (fig. 2.9H, I)
and perforated by pneumatic formina proxi- is known from ilium, ischium, and pubis, as
mally. Sacral vertebrae are unfused in FMNH well as femur, tibia, fibula, and metatarsus. The
PR 2209, with opisthocoelous centra anteriorly pelvic girdle of Rapetosaurus is characterized by
and slight procoely in the posteriormost sacral. an ilium with a broad, anterolaterally flaring
Neural arches are more than twice the height preacetabular process. The ischium is only 54%
of the centra, with rugose, sutural surfaces the length of the pubis, and the acetabulum is
anteirorly and posteriorly. The sacral ribs are composed primarily of the ilium and pubis,
short, broad wings that bear a small proximal with only a minor ischial contribution. The
crest on the anterior face. Middle and posterior femur is characterized by a proximolateral
caudal vertebrae are strongly procoelous, with flange and medially deflected head. The tibia is
relatively high neural spines that retain spinal incomplete but has a short cnemial crest, tapers
lamination. Centrum face width and height are middiaphysis, and is slightly anteroposteriorly
subequal in all preserved caudals. expanded in both proximal and distal extremes.
The forelimb of Rapetosaurus (figs. 2.9B, C, F, The fibula is relatively simple with a slightly
G) is known from scapula, coracoid, humerus, sigmoidal outline in lateral view and flared
radius, ulna, and metacarpals. The scapula is proximal and distal ends. The articulated
generally similar to those of other titanosaurs metatarsus is a broad, nearly flat, arched pes
but has a uniform scapular blade width, the strikingly different from the tightly bound
scapular blade articulates with the body of the manus. The length of the metatarsals increases
scapula at an acute angle, and the acromion from I to III, with metatarsal IV slightly shorter
process extends anterior to the coracoid articu- than metatarsal III, and metatarsal V short and
lation. The coracoid has a rounded outline and triangular.
subequal dimensions. The lateral surface of the Rapetosaurus exhibits a suite of features that
coracoid is convex, and it has a larger glenoid confirm its status as a titanosaur, and it is the
surface than it has for articulation with the only titanosaur with a wealth of both cranial
scapula. The sternal plates are laminar and and postcranial data. To date, no other
semilunate with strongly concave lateral bor- titanosaur preserves as much high-quality
ders and a highly convex medial border. A skeletal data as Rapetosaurus krausei. As such,
broad, subrectangular deltopectoral crest char- Rapetosaurus has significant implications for
acterizes the proximal humerus of phylogenetic work within Titanosauria and is of
Rapetosaurus. The humerus has a straight dia- great significance for delimiting variation
physis, with an elliptical cross section at its within the clade. New data derived from the
midpoint. The distal humerus has subequal Rapetosaurus skeleton and skull highlight the
fact that Titanosauria exhibited a wide range of Antarctosaurus was coded on the basis of the
morphologies and were not limited to a narrow holotypic cranium and referred postcrania only.
window of variation. Character polarity was determined by out-
group comparison with outgroup taxa regarded
as paraphyletic with respect to the ingroup
THE INTERRELATIONSHIPS
(Brachiosaurus, Camarasaurus, Dicraeosaurus,
OF TITANOSAURIA
Diplodocus, Euhelopus). Most characters were
Now, with Rapetosaurus as a key, we can begin to binary, and multistate characters were left
delve deeper into titanosaur phylogeny. A unordered and unweighted. The high number
generic-level phylogeny of Titanosauria is pre- of terminal taxa and abundant missing data
sented here. The analysis incorporates charac- (table 2. 4), particularly in the titanosaur sam-
ters utilized by other workers (Calvo and Salgado ple precluded exact treebuilding methods, so
1995; Upchurch 1995, 1998, 1999; Salgado et al. heuristic searches were conducted with TBR
1997; Wilson and Sereno 1998; Wilson 1999; branch swapping, and branches with maximum
Curry Rogers and Forster 2001; Upchurch et al. length zero were collapsed to yield polytomies.
2004), as well as novel characters generated Characters were optimized under ACCTRAN
from research in museum collections and the and trees were rooted with Camarasaurus.
primary literature. Details and implications of The strict consensus tree for 200,000
the analysis are summarized below. The appen- equally parsimonious trees (tree length, 989;
dixes contain a character–taxon matrix (appen- CI 0.56; RI 0.70; RCI 0.39) is shown in
dix 2.1), a list of characters and character states fig. 2.10. Many of the nodes have low decay
(appendix 2.2), and two synapomorphy lists indices, suggesting weak support for the
(appendixes 2.3 and 2.4). hypothesized relationship (fig. 2.10). Specific
phylogenetic hypotheses were more thoroughly
ANALYSIS investigated in MacClade (Maddison and
Thirty-five terminal taxa (including 29 pur- Maddison 2000) by “fitting” data to alternative
ported titanosaurs; table 2.3) were scored for hypotheses (i.e., Nemegtosaurus nested within
364 morphological characters (109 cranial, 139 Diplodocoidea rather than Titanosauria). In all
axial, 113 appendicular, and 2 dermal characters) cases, these alternatives were far less parsimo-
(appendixes 2.1 and 2.2) in MacClade nious than the topologies presented in this
(Maddison and Maddision 2000) and analyzed analysis.
in PAUP* (Swofford 1999). Scoring was based Titanosauriformes monophyly was unam-
on personal observations for all taxa except biguously supported (appendix 2.3). No unam-
Euhelopus and Phuwiangosaurus, which were biguous characters support the monophyly
scored from published illustrations, photo- of Somphospondylii or Titanosauria in the
graphs, and descriptions. The remains used to strict consensus summary of the data, and
score certain genera deserve additional com- most titanosaurs collapse into an unwieldy
ment. Scoring of Isisaurus colberti was based on polytomy. However, two clades of note are
the type specimen (‘T. indicus’ Jain and retained in the strict consensus tree: (1) a clade
Bandyopadhyay 1997). Several Indian titanosaur including Lirainosaurus as the sister-taxon to a
skulls were subsumed within Jainosaurus septen- polytomy comprised of typical saltasaurines
trionalis (Huene and Matley 1932; Berman and (e.g., Neuquensaurus, Saltasaurus, Rocasaurus),
Jain 1983; Hunt et al. 1994), and the skull Quaesitosaurus, Malagasy Taxon B, and the
recently described as Antarctosaurus septentrion- unnamed titanosaur from Jabalpur, India (fig.
alis by Chatterjee and Rudra (1996) is here des- 2.10; nodes A, B); and (2) a clade uniting
ignated “Jabalpur Titanosaur indet.” and is not Rapetosaurus and Nemegtosaurus as sister taxa
definitively associated with postcranial remains. (fig. 2.10; node C). No unambiguous support
TABLE 2.3
Geological Age and Geographical Range for 27 Titanosaur Terminal Taxa Analyzed
CONTINENT
AGE (STAGE) (COUNTRY) MATERIAL ORIGINAL REFERENCE
Aegyptosaurus Late Cretaceous Africa Axial, Stromer 1932

baharijensis (Cenomanian) (Egypt) appendicular
Aeolosaurus Late Cretaceous South Axial, Powell 1986
rionegrus (Campanian) America appendicular
(Argentina)
Agustinia ligabuei Early Cretaceous South Axial, Bonaparte 1998
(Aptian) America appendicular
(Argentina)
Alamosaurus Late Cretaceous North Axial, Gilmore 1922
sanjuanensis (Maastrichtian) America appendicular
(USA)
Ampelosaurus Late Cretaceous Europe Cranial, axial, Le Loueff 1995
atacis (Maastrichtian) (France) appendicular
dermal
Andesaurus Early Cretaceous South Axial, Calvo and
delgadoi (Albian– America appendicular Bonaparte
Cenomanian) (Argentina) 1991
Antartosaurus Late Cretaceous South Cranial, Huene 1929
wichmannianus (Campanian– America appendicular
Maastrictian) (Argentina)
Argentinosaurus Early Cretaceous South Axial, Bonaparte and
huinculensis (Aptian– America appendicular Coria 1993
Coniacian) (Argentina)
Argyrosaurus Early Cretaceous South America Axial, appendicular Lydekker 1893
superbus (Senonian) (Argentina)
Brazil Series B Early Cretaceous South Axial Powell 1986
(Senonian) America
(Brazil)
Epacthosaurus Early Cretaceous South Axial, Powell 1986
scuittoi (Senonian) America appendicular
(Argentina)
Isisaurus Late Cretaceous Asia (India) Axial, Jain and
( Titanosaurus) (Maastrichtian) appendicular Bandyopadhyay
colberti 1997
Jainosaurus Late Cretaceous Asia (India) Cranial Hunt et al. 1994
septentrionalis (Campanian–
Maastrichtian)
Janenschia Late Jurassic Africa appendicular Wild 1991
robusta (Kimmeridgian) (Tanzania)
Japalpur indet. Late Cretaceous Asia (India) Cranial Unpublished
(Maastrictian)
Lirainosaurus Late Cretaceous Europe (Spain) Cranial, axial, Sanz et al. 1999
astibiae (Campanian– appendicular,
Maastrichtian) dermal
Magyarosaurus Late Cretaceous Europe Cranial, axial, Huene 1932
dacus (Maastrichtian) (Romania) appendicular,
dermal
TABLE 2.3 (continued)
CONTINENT
AGE (STAGE) (COUNTRY) MATERIAL ORIGINAL REFERENCE
Malagasy Late Cretaceous Africa Axial, Curry Rogers 2001

Taxon B (Campanian– (Madagas-car) appendicular
Maastrichtian)
Malawisaurus Early Cretaceous Africa Cranial, axial, Jacobs et al. 1993
dixeyi (Aptian) (Malawi) appendicular,
dermal
Nemegtosaurus Late Cretaceous Asia cranial Nowinski 1971
mongoliensis (Maastrichtian) (Mongolia)
Neuquensaurus Late Cretaceous South Axial, Powell 1986
(Campanian– America appendicular
Maastrichtian) (Argentina)
Opisthocoelicaudia Late Cretaceous Asia Axial, Borsuk-Bialynicka
skarzynskii (Maastrichtian) (Mongolia) appendicular 1977
Paralititan Late Cretaceous Africa Axial, Smith et al. 2001
stromeri (Cenomanian) (Egypt) appendicular
Phuwiangosaurus Early Cretaceous Asia Axial, Martin et al. 1994
sirindhornae (Thailand) appendicular
Quaesitosaurus Late Cretaceous Asia Cranial Kurzanov and
orientalis (Maastrichtian) (Mongolia) Bannikov 1983
Rapetosaurus Late Cretaceous Africa Cranial, axial, Curry Rogers and
krausei (Campanian– (Madagascar) appendicular, Forster 2001
Maastrichtian) dermal
Rocasaurus Late Cretaceous South Axial, Salgado and
muniozi (Campanian– America appendicular Azpilicueta
Maastrichtian) (Argentina) 2000
Saltasaurus Late Cretaceous South Cranial, axial, Bonaparte and
loricatus (Campanian– America appendicular, Powell 1980
Maastrichtian) (Argentina) dermal
Santa Rosa Late Cretaceous South Cranial Unpublished
indet. (Campanian– America
Maastrichtian) (Argentina)
exists for the saltasaurine polytomy, and the regarded as close titanosaur relatives are at the
strict consensus tree does not resolve unam- base of the Adams consensus, each in an unre-
biguous support for components of the basal solved dichotomy with another taxon. Euhelopus
polytomy. and Jainosaurus are basal to the dichotomy
An Adams consensus tree highlights more including Phuwiangosaurus and Andesaurus
resolved branches and reveals that several con- within Titanosauria. More derived titanosaur
sistent nestings of taxa are among the most par- relationships are not unambiguously supported
simonious trees (fig. 2.11). Nine additional in this analysis, but six clades of note are con-
unambiguous characters from the cranial and served in the Adams consensus: (1) Titanosauria
postcranial skeleton support Titanosauriformes (fig. 2.11; node A); (2) Saltasauridae (fig. 2.11;
monophyly, and 22 additional steps are node J); (3) Lithostrotia (fig. 2.11; node B); and
required to move Titanosauria into its more tra- (4) Saltasaurinae (fig. 2.11; node G), (5)
ditional position as the sister-group to Opisthocoelicaudiinae (fig. 2.11; node I); and (6)
Diplodocoidea. Several sauropods generally an unnamed clade including Rapetosaurus,
FIGURE 2.10. Strict consensus of 200,000 most parsimonious trees (989 steps, CI = 0.56, RI = 0.70)
generated by a heuristic search in PAUP*. Trees were rooted with Camarasaurus as the outgroup and opti-
mized with accelerated transformations, and all characters were unordered. Bremer support values are
listed at labeled nodes. Labeled nodes (A–C) indicate monophyletic groups that are discussed more fully in
text. No formal names are assigned to these clades at present.
Nemegtosaurus, Malawisaurus, and several other is not calculated for these nodes (e.g., the poly-
derived lithostrotians (fig. 2.11; node E). tomy of Antarctosaurus, Argentinosaurus, and
Unambiguous support for some of the nodes Opisthocoelicaudia).
within these clades is listed in appendix 2.3. This maximum parsimony analysis of the
Other clades recognized in the Adam’s consen- total-evidence data set resulted in poor ingroup
sus are polytomous, and unambiguous support resolution in strict consensus summaries, as
FIGURE 2.11. Adams consensus produced from analysis of 200,000 most parsimonious trees. Node A indi-
cates Titanosauria, which includes a basal polytomy of Phuwiangosaurus and Andesaurus. Node B indicates
Lithostrotia. Node E marks the Rapetosaurus clade. Node G indicates Saltasaurinae. Node I indicates Opistho-
coelicaudiinae. Node J indicates Saltasauridae. Other labeled nodes demarcate monophyletic groups but are not
formally named at present. Unambiguous upport for these unnamed clades, as well as for Titanosauridae,
Saltasauridae, and the Rapetosaurus clade, is listed in appendix 2.3.
well as relatively low Bremer support values. matrix (tables 2.4, 2.5), allowing more resolution
MacClade’s search for redundant taxa identified in the Adams consensus of the data and a reduc-
none in the matrix, and no taxa could be deleted tion in most parsimonious trees. Though the
using Wilkinson’s (1995) rules for “safe” taxo- reduction in tree number allowed for detailed
nomic deletion. Given these constraints, I fol- study of the 44 trees, it did not result in a better-
lowed the precedent set by several recent authors resolved strict consensus tree (fig. 2.12).
(Gauthier 1986; Rowe 1988; Benton 1990; Exploration of each of the 44 equally parsimo-
Novacek 1992; Grande and Bemis 1998), and nious trees indicates that relationships among
removed taxa that were 15% or less completely three major titanosaur clades are stable, with
coded, and ran the matrix a second time. only Quaesitosaurus migrating throughout the
Fourteen titanosaurs were removed from the cladogram. In fact, Quaesitosaurus can be equally
FIGURE 2.12. Cladogram show-
ing phylogenetic hypothesis of
titanosaur relationships, based
on the strict consensus of 44
most parsimonious trees (704
steps, CI = 0.72, RI = 0.74) gen-
erated by a heuristic search in
PAUP*. Trees were rooted with
Camarasaurus as the outgroup,
optimized with accelerated
transformations, and all charac-
ters were unordered. Bremer
support values are listed at la-
beled nodes. Labeled nodes
(A–E) indicate monophyletic
groups that are discussed more
fully in text. No formal names
are assigned to these clades at
present.
parsimoniously placed as the sister-taxon to (fig. 2.13; node D), (3) Saltasaurinae (fig. 2.13;
almost every titanosaur in the analysis and is node E), (4) the Rapetosaurus clade (fig. 2.13; node
responsible for the poor resolution in the strict G), and (5) Opisthocoelicaudiinae (fig. 2.13; node
consensus tree for the taxonomically reduced K). Where calculated, unambiguous support for
data set. That said, the strict consensus of the these nodes is outlined in appendix 2.4.
data does support the higher-level groupings of
Titanosauriformes and Somphospondylii (fig.
DISCUSSION
2.12; nodes A, B), as well as a portion of
Titanosauria (fig. 2.12; nodes C, D). The analysis presented here has several impor-
The Adams consensus tree (fig. 2.13) of the tant implications for titanosaur phylogeny. First
reduced data set indicates that Phuwiangosaurus and foremost, in spite of an enormous amount of
occupies a basal position within Titanosauria. missing data (14 titanosaurs included in this
Unambiguous support for Titanosauria is not cal- study are coded for less than 15% of their data;
culated in the Adams consensus tree. However, tables 2.4, 2.5), several clades are well-supported.
in individual trees several unambiguous postcra- Titanosauriformes monophyly is apparent when
nial characters support Titanosauria (appendix all known titanosaurs are considered, in support
2.4). Within Titanosauria, Ampelosaurus and of the dismantled grouping of titanosaurs and
Lirainosaurus form successive sister-taxa to sev- diplodocoids (e.g., Huene 1929; Romer 1956;
eral major titanosaur clades. These clades were Steel 1970). Titanosauriform monophyly is sup-
also recovered in the Adams consensus for the ported with cranial characters, as is the mono-
complete data set discussed above. (1) phyly of Somphospondylii (titanosauriforms
Saltasauridae (fig. 2.13; node C), (2) Lithostrotia more closely related to Saltasaurus than to
FIGURE 2.13. Adams consen-
sus produced from analysis of
44 MPTs resulting from re-
duced analysis. Node A indi-
cates Somphospondylii. An un-
labeled node demarcates
Titnaosauria, which includes a
basal polytomy of Phuwian-
gosaurus and Andesaurus with
more derived titanosaurs. Node
C indicates Saltasauridae. Node
D indicates Lithostrotia. Node E
marks Saltasaurinae. Node G
indicates the Rapetosaurus clade.
Node K indicates Opisthocoeli-
caudiinae. Other labeled nodes
demarcate monophyletic groups
but are not formally named at
present. Unambiguous upport
for these unnamed clades, as
well as those outlined here, is
listed in appendix 2.4.
Brachiosaurus [Wilson and Sereno 1998]). As nosaurus is in a polytomy with Antarctosaurus and
such, the new, more comprehensive titanosaur Opisthocoelicaudia within Titanosauridae. Salgado
sample adds strength to hypotheses formerly et al. (1997) also hypothesized a close relationship
based only on postcranial data. between Argentinosaurus and Opisthocoelicaudia.
In addition to firm establishment of the Epachthosaurus is the sister-taxon of Lirainosaurus
appropriate titanosaur outgroup, this analysis Saltasauridae, instead of being closely allied
also highlights new lower-level resolution. In with Malawisaurus (Salgado et al. 1997).
individual trees and in Adams consensus sum- The monophyly of Lithostrotia and
maries of the data, several titanosaur clades are Saltasauridae are strongly supported in this
sustained, all of which have been supported in analysis, though Malawisaurus is not considered
some form in previous analyses (e.g., Salgado a basal titanosaur (contra Wilson 2002). Within
et al. 1997; Upchurch 1998; Wilson et al. 1999; each of these major groups, additional resolu-
Curry Rogers and Forster 2001). This analysis tion is possible, and is in general agreement
confirms that Epachthosaurus, Andesaurus, and with the recent analyses by Wilson (2002) and
Argentinosaurus are not closely related (contra Upchurch et al. (2004).
Bonaparte and Coria 1993). Andesauridae Opistocoelicaudiinae is supported in individ-
(Bonaparte and Coria, 1993) is paraphyletic in ual trees, though it is not apparent in the strict
this analysis, and forcing its monophyly requires consensus of the data, and in this analysis it
an addition of 12 steps. While Andesaurus retains includes Isisaurus as the sister-taxon of
its position as the most basal member of Alamosaurus an unresolved polytomy (Opistho-
Titanosauria, Epachthosaurus and Argentinosaurus coelicaudia, Antarctosaurus, and Argentinosaurus).
are more nested within the group. Argenti- When poorly preserved taxa are excluded from
TABLE 2.4 TABLE 2.5
Missing Data for Titanosaur Cranial and Postcranial Skeleton Taxa Excluded from Parsimony Analysis
PERCENTAGE MISSING DATA PERCENTAGE MISSING DATA
CRANIAL POSTCRANIAL CRANIAL POSTCRANIAL
Aegyptosaurus 100.0 89.6 Aegyptosaurus 100.0 89.6

Aeolosaurus 100.0 85.7 Aeolosaurus 100.0 85.7
Augustinia 100.0 89.0 Andesaurus 100.0 86.3
Alamosaurus 100.0 33.3 Augustinia 100.0 89.0
Ampelosaurus 72.4 46.6 Antarctosaurus 89.0 100.0
Andesaurus 100.0 86.3 Argentinosaurus 100.0 85.2
Antarctosaurus 89.0 100.0 Argyrosaurus 100.0 92.0
Argentinosaurus 100.0 85.2 Epacthosaurus 100.0 94.2
Argyrosaurus 100.0 92.0 Jabalpur 96.1 100.0
Brazil Series B 100.0 66.6 Titanosauria indet.
Epacthosaurus 100.0 94.2 Jainosaurus 97.0 100.0
Isisaurus 100.0 36.1 Janenschia 100.0 89.3
( ‘Titanosaurus’) Malagasy Taxon B 100.0 96.4
Jabalpur 96.1 100.0 Paralititan 100.0 93.7
Titanosauria indet. Santa Rosa 92.6 100.0
Jainosaurus 97.0 100.0 Titanosaura indet.
Janenschia 100.0 89.3
Lirainosaurus 93.6 55.3
Magyarosaurus 100.0 47.1 (Mendoza titanosaur [Wilson et al. 1999]), and
Malagasy Taxon B 100.0 96.4 includes several other taxa (Quaesitosaurus,
Malawisaurus 58.7 51.4
Malagasy Taxon B, and Jabalpur indet.), though
Nemegtosaurus 31.2 100.0
Neuquensaurus 100.0 19.6
support for inclusion of Quaesitosaurus and
Opisthocoelicaudia 100.0 21.6 Jabalpur indet. is weak.
Paralititan 100.0 93.7 The Rapetosaurus clade is the last major group
Phuwiangosaurus 100.0 43.1 within Titanosauria recognized by this analysis.
Quaesitosaurus 38.5 100.0 Although only the sister-taxon relationship
Rapetosaurus 18.5 7.7 between Nemegtosaurus and Rapetosaurus is
Rocasaurus 100.0 69.8 retained in the strict consensus trees, in Adams
Saltasaurus 72.5 19.2
consensus trees the Rapetosaurus clade includes a
Santa Rosa 92.6 100.0
Titanosaura indet. polytomy of the South American titanosaurs
Agustinia and Brazil Series B as the sister-taxon to
Malawisaurus (Nemegtosaurus Rapetosaurus).
the analysis, the relationship of Opistho- In previous analyses (e.g., Salgado et al. 1997;
coelicaudia remains unresolved, and it is equally Smith et al. 2001; Wilson 2002) Malawisaurus
parsimoniously placed as the sister- taxon to has occupied a more basal position within
Alamosaurus or Isisaurus. Titanosauria. Here (as in Curry Rogers and
Saltasaurinae is strongly supported in all Forster 2001) Malawisaurus is firmly nested
analyses and contains traditional members of a within the Rapetosaurus clade, and removal
longstanding group of titanosaurs (Saltasaurus, requires 11 additional steps.
Neuquensaurus). Lirainosaurus is the sister-taxon The positions of several controversial
of Saltasauridae. Saltasauridae includes tradi- sauropods are also better understood. Notably, as
tional members of the group (e.g., Saltasaurus, hypothesized by Upchurch (1998), Phuwian-
Neuquensaurus), does not include others gosaurus is the sister-taxon of Titanosauria. Even
more significantly, Mongolian sauropods are well as a wealth of new character data from all
members of Titanosauria, in conflict with the parts of the skeleton.
hypothesis presented by Upchurch (1998, 1999) Through the addition of Rapetosaurus and
and Upchurch et al. (2004), who concluded that recent revision of several other key titanosaur
Nemegtosaurus and Quaesitosaurus were mem- genera (e.g., Nemegtosaurus, Isisaurus), the fol-
bers of Diplodocoidea. Forcing either of the lowing key aspects of titanosaur phylogeny are
Mongolian sauropods into a sister-group rela- addressed and more fully resolved.
tionship with Diplodocoidea results in an
1. Monophyly of Titanosauriformes can
increase of at least 26 steps. This analysis also
now be confirmed with unambiguous
suggests that Nemegtosaurus, Opisthocoelicaudia,
support from the cranial and postcranial
and Quaesitosaurus are likely not conspecific.
skeleton.
Similarly, Paralititan and Aegyptosaurus are also
not likely to be conspecific, or even closely related 2. Monophyly of Somphospondylii is also
at the generic level. Finally, a survey of a sample confirmed with unambiguous support
of the 200,000 most parsimonious trees demon- from the cranial and postcranial skeleton.
strates that the primary lack of resolution in the 3. Lithostrotia, Saltasauridae, Saltasaurinae,
consensus trees results from pervasive migration and Opisthocoelicaudiinae are mono-
of several very incomplete titanosaurs (e.g., phyletic, though this analysis differs from
Quaesitosaurus, Jainosaurus). other recent workers’ (Wilson 2002;
Upchurch et al. 2004) in the placement
CONCLUSION of several well-known titanosaur genera.
4. An unnamed clade uniting Malawisaurus,
Over the last several years, a handful of rigor-
Rapetosaurus krausei, and Nemegtosaurus
ous looks at the anatomy and phylogeny of
is monophyletic and confirms the inclu-
Titanosauria have clarified our view of this
sion of Nemegtosaurus in Titanosauria.
enigmatic sauropod group. Revision of the type
genus and species, Titanosaurus indicus, has 5. Resolution within the clades mentioned
provided us with a working definition of what a above is greatly improved in this analysis,
titanosaur is, and we are better able to diagnose though including the titanosaurs with a
titanosaurs. Discoveries of new fossils from preponderance of missing data complicates
around the globe extend the titanosaur reign resolution in strict consensus trees.
geographically and temporally, and give us a 6. Phuwiangosaurus and Andesaurus are
clearer image of how titanosaurs made their liv- basal titanosaurs, but Andesauridae is
ing. We have even reached some resolution on not monophyletic. The other purported
titanosaur phylogeny, particularly with regard “andesaurids,” Argentinosaurus and
to higher-level relationships. Titanosauria is no Epachthosaurus are nested within the
longer shrouded in so deep a mystery, but we clades mentioned above.
are far from a complete understanding of the
group, particularly with regard to lower-level
relationships. All analyses of titanosaur interre- ACKNOWLEDGMENTS
lationships have lacked a full body view for a Thanks go to Jack McIntosh for inspiring all of us
diverse taxonomic sample due to the sheer lack to search for new sauropods, new approaches,
of skeletal data for almost all titanosaur genera. and new answers. Thanks are also due to C.
Rapetosaurus krausei is a keystone taxon that is Forster, D. Krause, S. Sampson, and J. Gauthier,
most significant because it preserves cranial as well as R. Rogers, J. A. Wilson, M. Carrano, and
and postcranial material that allows across-the- M. O’Leary for their many comments and advice
board comparisons with other titanosaurs, as on preparing this chapter. D.B. Weishampel, J. A.
Wilson, and one anonymous reviewer greatly Late Cretaceous of Madagascar Nature 405:
improved the quality of this chapter. 941–944.
Calvo, J. O., and Bonaparte, J. F. 1991. Andesaurus del-
gadoi gen. et sp. nov. (Saurischia-Sauropoda),
LITERATURE CITED
dinosaurio Titanosauridae de la Formación Río
Asher, R. J., and Krause, D. W. 1998. The first pre- Limay (Albiano-Cenomanio), Neuquén, Argentina.
Holocene (Cretaceous) record of Anura from Ameghiniana 28: 303–310.
Madagascar. J. Vertebr. Paleontol. 18: 696–699. Calvo, J. O., and Salgado, L. 1995. Rebbachisaurus
Benton, M. J. 1990. The species of Rhyncosaurus, a tesonei sp. nov. A new Sauropoda from the Albian-
rhyncosaur (Reptila, Diaprida) from the Middle Cenomanian of Argentina; new evidence of the
Triassic of England. Philosophical Transactions of origin of the Diplodocidae. GAIA 111: 13–33.
the Royal Society of London B. 328: 213–306. Chatterjee, S., and Rudra, D. K. 1996. KT events in
Berman, D. S., and Jain, S. L. 1983. The braincase of India: impact, rifting, volcanism and dinosaur
a small sauropod dinosaur (Reptilia: Saurischia) extinction. Mem. Queensland Mus. 39: 489–532.
from the Upper Cretaceous Lameta Group, Chiappe, L. M., Salgado, L., and Coria, R. A. 2001.
Central India, with a review of Lameta Group Embryonic skulls of titanosaur sauropod
Localities. Ann. Carnegie Mus. 51: 405–422. dinosaurs. Science 293: 2444–2446.
Berman, D. S., and McIntosh, J. S. 1978. Skull and rela- Csiki, Z. 1999. New evidence of armoured
tionships of the Upper Jurassic sauropod titanosaurids in the Late Cretaceous—
Apatosaurus (Reptilia, Saurischia). Bull. Carnegie Magyarosaurus dacus from the Hateg Basin
Mus. Nat. Hist. 8: 1–35. (Romania). Oryctos 2: 93–99.
Besairie, H. 1972. Géologie de Madagascar. I. Les ter- Curry, K. A. 1997. Vertebrate fossils from the Upper
rains sédimentaires. Ann. Geol. Madagascar 35: Cretaceous Ankazomihaboka Sandstones, Maha-
1–463. janga Basin, Madagascar. J. Vertebr. Paleontol. 17:
Bonaparte, J. F. 1986. The early radiation and phylo- 40A.
genetic relationships of sauropod dinosaurs, Curry, K. A. 2001. The Evolutionary History of the
based on vertebral anatomy. In: Padian, K. (ed.). Titanosauria. Unpublished Ph.D. thesis, Stony
The Beginning of the Age of Dinosaurs. Brook University, Stony Brook, NY. 552 pp.
Cambridge University Press, Cambridge, New Curry Rogers, K. 2001. A new sauropod from
York. Pp. 247–258. Madagascar: implications for lower level titanosaur
Bonaparte, J. F. 1998. An armored Sauropod from phylogeny. J. Vertebr. Paleontol. 21: 43A.
the Aptian of northern Patagonia, Argentina. In: Curry Rogers, K., and Forster, C. A. 1999a. New evi-
Second Symposium on Gondwana Dinosaurs, dence of titanosaurian sauropods in the Late
Tokyo. P. 10. Cretaceous of Madagascar. In: Abstracts, VII
Bonaparte, J. F. and Coria, R. 1993. Un neuvo y gigan- International Symposium on Mesozoic Terrestrial
tesco sauropodo titanosaurio de la Formación Río Ecosystems, Buenos Aires. P. 20.
Limay (Albiano-Cenomanio) del a Provincia del ———. 1999b. New sauropods from Madagascar: a
Neuquén, Argentina. Ameghiniana 30: 271–282. glimpse into titanosaur cranial morphology and
Bonaparte, J. F. and Powell, J. 1980. A continental evolution. J. Vertebr. Paleontol. 19: 40A.
assemblage of tetrapods from the Upper ———. 2001. The last of the dinosaur titans: a new
Cretaceous beds of El Brete, Northwestern sauropod from Madagascar. Nature 412: 530–534.
Argentina (Sauropoda-Coelurosauria-Carnosauria- ———. 2004. The skull of Rapetosaurus krausei
Aves). Mem. Soc. Geol. Fr. N. S. 139: 19–28. (Sauropoda: Titanosauria) from the Late Cretaceous
Borsuk-Bialynicka, M. 1977. A new camarasaurid of Madagascar. J. Vertebr. Paleontol. 24: 121–144.
sauropod, Opisthocoelicaudia skarzynskii, gen. n., Day, J. J., Upchurch, P., Norman, D. B., Gale, A. S.,
sp. n. from the Upper Cretaceous of Mongolia. and Powell, H. P. 2002. Sauropod trackways, evo-
Palaeontol. Polonica 37: 1–64. lution, and behavior. Science 296: 1659.
Buckley, G. A., and Brochu, C. 1999. An enigmatic Depéret, C. 1896a. Note sur les Dinosauriens
new crocodiles from the Upper Cretaceous of Sauropodes et Théropodes dans le Crétacé
Madagascar. In: Unwin, D. (ed.). Cretaceous Fossil supérieur de Madagascar. Bull. Soc. Geol. France
Vertebrates: Special Papers in Palaeontology No. 21: 176–194.
60. Paleontological Association, London. Pp. ———. 1896b. Sur L’existence de Dinosauriens,
149–175. Sauropodes et Théropodes dans le Crétacé.
Buckley, G. A., Brochu, C., Krause, D. W., and Pol, supérieur de Madagascar. C.R. Hebdomadiares
D. 2000. A pug-nosed crocodyliform from the Sea. Acad. Sci. (Paris) 122: 483–485.
Dodson, P., Krause, D. W., Forster, C. A., Sampson, Jain, S., and Bandyopadhyay, S. 1997. New
S. D., and Ravoavy, F. 1998. Titanosaurid titanosaurid (Dinosauria: Sauropoda) from the
(Sauropoda) osteoderms from the Late Cretaceous Late Cretaceous of central India. J. Vertebr.
of Madagascar. J. Vertebr. Paleontol. 18: 563–568. Paleontol. 17: 114–136.
Erickson, G. M., Curry Rogers, K., and Yerby, S. 2001. Jianu, C. M., and Weishampel, D. B. 1999. The small-
Dinosaurian growth patterns and rapid avian est of the largest, a new look at possible dwarfing in
growth rates. Nature 412: 429–433. sauropod dinosaurs. Geol. Mijnbouw 78: 335–343.
Falconer, H. 1868. Memorandum on two remark- Krause, D. W. 2001. Fossil molar from a Madagascan
able vertebrae sent by Dr. Oldham from marsupial. Nature 412: 497.
Jubbulpore-Spilsbury bed. In: Paleontological Krause, D. W., Hartman, J. H., and Wells, N. A. 1997.
Memoirs and Notes of the Late Hugh Falconer, Late Cretaceous vertebrates from Madagascar:
Vol. 1. Pp. 418–419. implications for biotic change in deep time. In:
Forster, C. A., Chiappe, L. M., Krause, D. W., and Goodman, S. D., and Patterson, B. D. (eds.). Natural
Sampson, S. D. 1996. The first Cretaceous bird Change and Human Impact in Madagascar.
from Madagascar. Nature 382: 532–534. Smithsonian Institution, Washington, DC. Pp.
Forster, C. A., Sampson, S. D., and Chiappe, L. M. 3–43.
1998. The theropod ancestry of birds: new evi- Krause, D. W., Hartman, J. H., Wells, N. A., Buckley,
dence from the Late Cretaceous of Madagascar. G. A., Lockwood, C. A., Wall, C. E., Wunderlich,
Science 279: 1915–1919. R. E., Rabarison, J. A., and Randriamiaramanana, L.
Gauthier, J. 1986. Saurischian monophyly and the 1994. Late Cretaceous mammals. Nature 368: 298.
origin of birds. In: Padian, K. (ed.). The Origin of Krause, D. W., Rogers, R. R., Forster, C. A., Hartman,
Birds and the Evolution of Flight. Memoirs of the J. H., Buckley, G. A., and Sampson, S. D. 1999.
California Academy of Sciences. Pp. 1–55. The Late Cretaceous vertebrate fauna of
Gilmore, C. W. 1922. A new sauropod dinosaur from Madagascar: Implications for Gondwanan paleo-
the Ojo Alamo Formation of New Mexico. biogeography. GSA Today 9: 1–7.
Smithson. Misc. Collect. 72: 1–11. Kurzanov, S. M., and Bannikov, A. F. 1983. A new
———. 1946. Reptilian fauna of the North Horn sauropod from the Upper Cretaceous of Mongolia.
Formation of central Utah. U.S. Geol. Surv. Prof. Paleontol. J. 2: 90–96.
Paper 210C: 1–52. Lavocat, R. 1955. Etude des gisements de
Gottfried, M. D., and Krause, D. W. 1998. First Dinosauriens de la region de Majunga
record of gars (Ginglymodi, Actinopterygii) on (Madagascar). Travaux Bur. Geol. 69: 1–19.
Madagascar: Late Cretaceous remains from the Le Loueff, J. 1995. Ampelosaurus atacis (nov. gen., nov.
Mahajanga Basis. J. Vertebr. Paleontol. 18: 275–279. sp.), un nouveau Titanosauridae (Dinosauria,
Grande, L., and Bemis, W. E. 1998. A comprehensive Sauropoda) du Crétacé superieur de la Haute
phylogenetic study of amiid fishes (Amiidae) Vallee de L’Aude (France). C. R. Acad. Sci. Paris
based on comparative skeletal anatomy. An (Ser. IIa) 321: 693–699.
empirical search for interconnected patterns of Lydekker, R. 1877. Notices of new and other
natural history. J. Vertebr. Paleontol. Mem. 4, 18: Vertebrata from Indian Tertiary and Secondary
1–690. Rocks. Rec. Geol. Soc. India 10: 30–43.
Hoffstetter, R. 1957. Quelques observations sur le ———. 1879. Indian Pre-Tertiary Vertebrata. Part 3.
stégosaurinés. Bulletin de Musée d’Histoire Fossil Reptilia and Batrachia. Paleontol. Indica I:
Naturelle, Paris 29: 537–547. 1–36.
Huene, F. von. 1929. Los Saurisquios y Ornithisquios ———. 1893. Contributions to the study of the fossil
del Cretaceo Argentino. An. Mus. La Plata 2: 1–196. vertebrates of Argentina. I. The dinosaurs of
Huene, F. von, and Matley, C. A. 1932. The Cretaceous Patagonia. An. Mus. La Plata Sec. Paleontol. 2: 1–14.
Saurischia and Ornithischia of the central Maddison. W. P., and Maddison, D. R. 2000. MacClade,
provinces of India. Mem. Geol. Surv. India 21: Version 4.0. Sinauer Associates, Sunderland, MA.
1–74. Martin, V., Buffetaut, E., and Suteethorn, V. 1994. A
Hunt, A., Lockley, M. D., Lucas, S. G., and Meyer, C. A. new genus of sauropod dinosaur from the Sao
1994. The global sauropod fossil record. GAIA Khua Formation (Late Jurassic to Early
10: 261–279. Cretaceous) of northeastern Thailand. C. R. Acad.
Jacobs, L. L., Winkler, D. A., Downs, W. R., and Sci. Paris 319: 1085–1092.
Gomani, E. M. 1993. New material of an Early Martínez, R. D., Gimenez, O., Rodríguez, J., Luna, M.,
Cretaceous titanosaurid sauropod dinosaur from and Lamanna, M. C. 2004. An articulated speci-
Malawi. Paleontology 36: 523–534. men of the basal titanosaurian Epachthosaurus
sciuttoi (Dinosauria: Sauropoda) from the early Russell, D. A., Russell, D., Taquet, P., and Thomas,
Late Cretaceous Bajo Barreal Formation of H. 1976. Nouvelles récoltes de vertebras dans les
Chubut Province Argentina. J. Vertebr. Paleontol. terrains continentaux du Crétacé supérieur de la
24: 107–120. region de Majunga (Madagascar). Soc. Geol.
McIntosh, J. S. 1989. The sauropod dinosaurs: a brief France C. R. Sommaires 5: 204–207.
survey. In: Padian, K., and Chure, D. (eds.). The Salgado, L., and Azpilicueta, C. 2000. Un nuevo
Age of Dinosaurs. Short Courses in Paleontol N. 2. saltasaurino (Sauropoda, Titanosauridae) de la
University of Knoxville, Knoxville, TN. Pp. 85–99. provincia de Río Negro (Formacíon Allen,
———. 1990a. Sauropoda. In: Weishampel, D. B., Cretácico Superior), Patagonia, Argentina.
Dodson, P., and Osmolska, H. (eds.). The Ameghiniana 37: 259–264.
Dinosauria. University of California Press, Berkeley. Salgado, L., Calvo, J. O., and Coria, R. A. 1997.
Pp. 345–401. Evolution of the titanosaurid sauropods. I.
———. 1990b. Species determination in sauropod Phylogenetic analysis based on the postcranial
dinosaurs with tentative suggestions for their evidence. Ameghiniana 34: 3–32.
classification. In: Carpenter, K. and Curry, P. J. Sampson, S. D., Carrano, M. T., and Forster, C. A.
(eds.). Dinosaur Systematics: Approaches and 2001. A bizarre predatory dinosaur from the
Perspectives. Cambridge University Press, Late Cretaceous of Madagascar. Nature 409:
Cambridge, New York. Pp. 53–71. 504–506.
Munyikwa, D., Sampson, S. D., Rogers, R. R., Forster, Sampson, S. D., Krause, D. W., Dodson, P., and Forster,
C. A., Curry, K. A., and Curtice, B. D. 1998. C. A. 1996. The premaxilla of Majungasaurus
Vertebrate paleontology and geology of the (Dinosauria: Theropoda) with implications for
Gokwe Formation, Zimbabwe. J. Afr. Earth Sci. Gondwanan paleobiogeography. J. Vertebr.
27: 142–143. Paleontol. 16: 601–605.
Novacek, M. J. 1992. Fossils, topologies, missing Sampson, S. D., Witmer, L. M., Forster, C. A.,
data, and the higher level phylogeny of eutherian Krause, D. W., O’Connor, P. M., Dodson, P., and
mammals. Syst. Biol. 41: 58–73. Ravoavy, F. 1998. Predatory dinosaur remains
Nowinski, A. 1971. Nemegtosaurus mongoliensis n. gen. from Madagascar: Implications for the Cretaceous
n. sp. (Sauropoda) from the uppermost Cretaceous biogeography of Gondwana. Science 280:
of Mongolia. Palaeontol. Polonica 25: 57–81. 1048–1051.
Powell, J. E. 2003. Revision of South American Santos, V. F., Lockley, M. G., Meyer, C. A., Carvalho,
Titanosaurid dinosaurs: palaeobiological, palaeo- J., Galopim, A. M., and Moratalla, J. J. 1994. A
biogeographical, and phylogenetic aspects. Rec. new sauropod tracksite from the Middle Jurassic
Queen Victoria Mus. 111: 1–173. of Portugal. GAIA 10: 5–14.
Ravoavy, F. 1991. Identification et mise en catalogue Sanz, J. L., Powell, J. E., Le, Loueff, J. Martinez, R.,
des vertebras fossils récoltes dans le Crétace and Suberbiola, X. P. 1999. Sauropod remains
supérieur continental de la region de Berivotra from the upper Cretaceous of Laño (northcentral
(Majunga) fouille 1987. Univ. Antananarivo Spain). Titanosaur phylogenetic relationships.
Mem. Recherche II: 55–104. In: Astiba, H., Corral, J. C., Murelaga, X., Oue-
Rogers, R. R. 2005, Fine-gravel debris flows and extraor- Extebarria, X., and Pereda-Suberbiola, X. (eds.).
dinary vertebrate burials in the Late Cretaceous of Geology and Palaeontology of the Upper
Madagascar. Geology 33 (4): 297–300. Cretaceous Vertebrate-Bearing Beds of the Laño
Rogers, R. R., Hartman, J. H., and Krause, D. W. Quarry (Basque-Cantabrrian Region, Iberian
2000. Stratigraphic analysis of Upper Cretaceous Peninsula). Estud. Mus. Cie. Nat. Alava 14 (num.
rocks in the Mahajanga Basin, northwestern espec. 1): 235–255.
Madagascar: Implications for ancient and mod- Seeley, H. G. 1869. Index to the fossil remains of
ern faunas. J. Geol. 108: 275–301. Aves, Ornithsauria and Reptilia, from the
Romer, A. S. 1956. Osteology of the Reptiles. Secondary system of strata arranged in the
University of Chicago Press, Chicago. 772 pp. Woodwardian Museum of the University of
———. 1966. Vertebrate Paleontology. University of Cambridge, Cambridge. 143 pp.
Chicago Press, Chicago. 468 pp. ———. 1876. On Macrurosaurus semnus (Seeley), a
———. 1968. Notes and Comments on Vertebrate long-tailed animal with procoelous vertebrae
Paleontology. University of Chicago Press, from the Cambridge Upper Greensand, pre-
Chicago. 304 pp. served in the Woodwardian Museum of the
Rowe, T. 1988. Definition, diagnosis, and origin of University of Cambridge. Q. J. Geol. Soc. London
Mammalia. J. Vertebr. Paleontol. 8: 241–264. 32: 440–444.
Smith, J. B., Lamanna, M. C., Lacovara, K. J., Dodson, Wilkinson, M. 1995. Coping with abundant missing
P., Smith, J. R., Poole, J. C., Giegengack, R., and entries in phylogenetic inference using parsi-
Attia, Y. 2001. A giant sauropod dinosaur from an mony. Syst. Biol. 44: 501–514.
Upper Cretaceous mangrove deposit in Egypt. Wilson, J. A. 1999. The Evolution and Phylogeny of
Science 292: 1704–1706. Sauropod Dinosaurs. Unpublished Ph.D. disser-
Steel, R. 1970. Handbuch der Paläoherpetology, tation, University of Chicago, Chicago. 384 pp.
Saurischia. Gustav Fischer Verlag, Jena. 87 pp. ——. 2002. Sauropod dinosaur phylogeny: critique
Stromer, E. 1932. Wibeltierreste der Baharije-Stufe and cladistic analysis. Zool. J. Linn. Soc. 136:
(untestes Cenoman). 11. Sauropoda. Abh. Bayer. 217–276.
Akad. Wissensch. Math.-Naturwiss. Abt. 10: 1–21. Wilson, J. A., and Carrano, M. T. 1999. Titanosaur
Swofford, D. L. 1999. PAUP*: Phylogenetic Analysis locomotion and the origin of “wide-gauge” track-
Using Parsimony (*and Other Methods), Version ways: a biomechanical and systematic perspec-
4.0b2a. Sinauer Associates, Sunderland, MA. tive on sauropod locomotion. Paleobiology 25:
Thévenin, A. 1907. Dinosauriens (Paléontologie de 252–267.
Madagascar IV). Ann. Paleontol. 2: 121–136. Wilson, J. A., and Sereno, P. C. 1994. Higher-level
Upchurch, P. 1995. Evolutionary history of sauropod phylogeny of sauropod dinosaurs. J. Vertebr.
dinosaurs. Philos. Trans. Roy. Soc. London B 349: Paleontol. 14: 42A.
365–390. ———. 1998. Early evolution and higher-level phy-
———. 1998. The phylogenetic relationships of logeny of sauropod dinosaurs. J. Vertebr.
sauropod dinosaurs. Zool. J. Linn. Soc. 124: Paleontol. (Suppl. to No. 2): 1–68.
43–103. Wilson, J. A., and Upchurch, P. 2003. A revision of
———. 1999. The phylogenetic relationships of the Titanosaurus Lydekker (Dinosauria-Saurpoda),
Nemegtosauridae (Saurischia, Sauropoda). J. the first dinosaur genus with a ‘Gondwanan’ dis-
Vertebr. Paleontol. 19: 106–125. tribution. J. Syst. Palaeont. 1 (3): 125–160.
Upchurch, P., Barrett, P. M., and Dodson, P. 2004. Wilson, J. A., Martinez, R. N., and Alcober, O. 1999.
Sauropoda. In: Weishampel, D. B., Dodson, P., Distal tail segment of a titanosaur (Dinosauria:
and Osmólska, H. (eds.). The Dinosauria, 2nd ed. Sauropoda) from the Upper Cretaceous of
University of California Press, Berkeley. Pp. Mendoza, Argentina. J. Vertebr. Paleontol. 19:
259–324. 591–594.
Wild, R. 1991. Janenschia n. g. robusta (E. Fraas 1908) Yu, C. 1993. The Skull of Diplodocus and the
pro Tornieria robusta (E. Fraas, 1908) (Reptilia, Phylogeny of the Diplodocidae. Unpublished
Saurischia, Sauropodomorpha). Stuttgarter eitr. Ph.D. dissertation, University of Chicago,
Naturk. B 173: 1–4. Chicago. 150 pp.
APPENDIX 2.1. CHARACTER–TAXON MATRIX
1 10 20 30
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Ampelosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 1 1
Andesaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 1 1 1
Apatosaurus 0 0 1 2 1 0 0 1 0 0 1 1 1 1 1 1 1 0 1 1 0 1 0 1 ? 0 1 0 0 0 1 0 0 0 0
Argentinosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 1 1 0 1 1 1 0 1 0 0 0 0 0 1 0 0 0 1 0 0 2 0 0 1 0 1 0 0 1 0 0 0 0 1 0
Camarasaurus 1 1 0 1 1 0 0 1 0 0 0 1 0 1 0 0 0 1 0 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0 0
Dicraeosaurus 0 ? 1 ? ? ? ? 1 ? ? 1 ? ? 1 ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? 1 0 1 0 0 1
Diplodocus 0 0 1 2 1 0 0 1 0 0 1 1 1 1 1 1 1 0 1 1 0 1 0 1 1 0 1 0 0 0 1 0 0 0 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus 0 0 0 0 0 0 0 0 0 ? 1 0 0 0 ? 0 0 ? 0 ? ? 0 0 ? ? 0 ? 0 ? ? ? ? ? ? ?
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Magyarosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malawisaurus 1 ? 0 1 0 ? ? ? ? ? ? ? ? ? ? 0 0 ? 0 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2
Nemegtosaurus 0 1 1 1 ? 1 ? 0 0 1 0 0 1 1 ? 1 1 ? 1 ? ? 0 0 1 ? 1 1 0 0 1 0 1 1 1 1
Neuquensaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Opisthocoelicaudia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Phuwiangosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Quaesitosaurus 1 1 1 0 ? ? ? 0 0 ? ? 0 1 0 ? ? ? ? ? ? ? ? 0 1 1 ? ? ? 0 ? 0 1 0 0 1
Rapetosaurus 0 0 1 0 ? 1 1 ? 1 1 1 1 ? 0 2 1 1 0 0 1 1 0 1 1 ? 1 1 1 ? 1 0 0 1 1 1
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Saltasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 1 0 1
Titanosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 1 1 1
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
APPENDIX 2.1. (continued)
40 50 60 70
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Ampelosaurus 1 1 1 1 0 ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 0 0 ? ? ? ? ? ? ? ? ? ? ?
Andesaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus 1 ? 1 2 1 0 2 0 ? ? ? ? ? ? ? ? ? ? ? 0 0 0 0 0 0 1 0 1 2 0 0 ? 1 1 0
Apatosaurus 0 0 1 1 ? 0 0 0 1 1 0 1 0 1 1 0 1 0 0 0 0 1 1 0 1 ? 1 1 0 0 0 0 0 1 0
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 1 0 0 0 1 0 1 0 0 0 1 0 1 0 0 0 0 0 0 ? 0 1 0 0 0 ? 0 0 0 0 0 1 1 0 1
Camarasaurus 0 0 0 1 0 0 0 0 1 0 0 0 1 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 2 1 1 0 0
Dicraeosaurus 0 0 0 1 1 1 1 0 ? ? ? ? ? 0 0 ? ? ? ? 0 0 0 1 1 1 ? 1 1 1 1 0 1 0 ? 0
Diplodocus 0 0 1 1 0 0 0 0 1 1 0 1 0 1 0 0 1 0 0 0 0 1 1 0 1 ? 1 1 0 0 0 0 0 1 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus ? ? ? ? ? ? ? 0 ? ? 0 0 0 0 0 ? 0 0 0 ? ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ?
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? 0 0 1 0 ? 0 ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 0 ? ? ? ? ? ? ? ? ? ? 0
Magyarosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malawisaurus 1 1 1 1 0 1 0 ? ? ? 0 1 ? ? ? 1 ? ? ? ? ? 0 0 0 0 ? 1 1 0 0 ? ? 1 ? 0
Nemegtosaurus 1 1 1 1 0 2 0 1 1 1 0 1 1 1 1 ? 1 1 ? 1 0 0 1 ? 0 ? 1 1 1 0 1 0 1 ? 1
Neuquensaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Quaesitosaurus 0 0 1 1 0 2 0 1 ? 1 0 0 0 1 1 1 1 0 1 1 1 0 0 ? 0 ? 1 1 1 0 1 1 1 ? 1
Rapetosaurus 1 1 1 1 0 2 0 0 ? ? ? ? ? 1 1 1 1 1 1 1 1 0 0 0 0 1 0 1 1 1 2 0 ? ? 1
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Saltasaurus 1 ? 0 1 1 2 2 1 ? ? ? ? ? ? ? ? ? ? ? ? 1 0 0 0 0 1 0 1 1 0 ? ? ? ? 1
Titanosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Santa Rosa indet. 1 1 0 2 0 0 2 1 ? ? ? ? ? ? ? ? ? ? ? 1 1 0 0 0 0 1 0 1 ? 0 0 1 1 1 0
Jabalpur indet. ? ? 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 0 ? 0 0 ? 1 0 0 ? 0 1 ? ? 1
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
80 90 100
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Ampelosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? 1 0 ? 0 0 0 0 ? ? ? ? ? ? ? ? ? 0 1 1 0 2 0
Andesaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus 0 0 0 ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 0 0 1 1 0 2 1 ? 1 1 1 1 ? ? 1 1 0 ? ? ? ? 1 0 0 1 1 ? ? ? 0 1 1 2 3 2
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 1 1 1 0 1 1 0 2 0 1 0 0 0 0 0 0 1 0 0 0 1 0 1 1 0 0 1 1 1 1 0 1 0 0 0
Camarasaurus 0 1 0 0 1 1 0 0 0 ? 0 0 1 0 0 0 0 0 0 0 1 1 1 0 1 0 1 1 0 0 0 0 0 0 1
Dicraeosaurus 0 0 1 0 1 ? ? 0 ? ? ? ? ? 0 0 0 ? 1 0 0 ? ? ? ? ? ? ? ? ? 0 0 ? 2 3 ?
Diplodocus 0 0 1 1 0 2 1 ? 1 1 1 1 1 0 1 1 0 1 0 0 0 1 0 0 1 1 0 1 1 0 1 1 2 3 2
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus ? ? ? ? ? 1 0 0 0 ? 0 ? ? 1 0 0 0 0 0 0 1 ? ? ? ? ? ? ? ? 0 1 0 0 0 1
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus 0 0 1 2 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 2 2
Magyarosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malawisaurus ? 0 ? 2 ? ? ? ? ? ? ? ? ? 0 ? 0 ? 0 1 0 ? ? ? ? ? ? ? ? ? 0 0 ? 0 1 1
Nemegtosaurus 0 1 1 2 1 ? ? 1 1 0 1 1 0 0 1 0 1 0 0 1 1 0 1 ? 0 0 0 0 0 0 1 0 0 1 2
Neuquensaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Quaesitosaurus 1 1 0 0 1 ? ? 1 ? ? ? ? ? 1 0 0 ? 0 0 1 0 1 1 ? 1 0 0 0 0 1 1 0 ? 2 2
Rapetosaurus 1 0 1 2 1 1 1 1 1 0 1 0 ? ? 0 0 1 ? ? 1 ? 1 1 1 1 0 ? ? ? 0 1 1 1 1 2
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Saltasaurus 1 1 0 2 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Titanosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Santa Rosa indet. 0 0 1 2 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. 0 1 1 ? 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
110 120 130 140
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus ? ? ? ? ? ? ? ? 1 ? ? ? ? ? ? ? 1 0 0 0 ? 1 1 0 ? ? ? ? ? ? ? ? ? ? ?
Ampelosaurus 1 0 0 0 ? 0 ? ? 0 ? 1 ? 0 1 2 ? 0 0 ? ? 1 1 1 0 1 1 1 0 0 0 0 0 3 1 0
Andesaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 1 1 1 1 0 0 1 2 0 1 1 0 0 1 0 0 0 1 0 1 1 1 1 1 1 1 0 1 0 0 0 1 3 0 0
Argentinosaurus ? ? ? ? 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? 1 1 ? ? 1 ? 0 ? ? ? 2 ? ? 0 ? ? ? ? ? ? 1 0 0 0 1 0 0 0 3 1 1
Brachiosaurus 0 0 0 0 0 0 0 1 0 1 1 0 0 1 0 1 0 0 1 1 1 1 1 1 0 1 0 1 0 0 0 0 2 0 0
Camarasaurus 1 0 0 0 0 0 0 0 0 1 1 0 1 1 0 1 0 1 0 0 1 0 1 0 1 1 1 1 0 0 1 1 3 0 0
Dicraeosaurus 0 0 1 ? 0 0 1 0 0 0 1 0 1 1 0 1 0 1 0 0 0 0 0 1 1 1 0 0 0 0 0 1 3 0 0
Diplodocus 1 1 1 1 0 0 1 2 0 1 1 1 0 1 0 1 0 1 1 0 1 1 1 1 0 1 1 1 0 0 0 1 3 0 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus 1 0 0 0 1 0 ? 2 1 ? 0 1 0 1 0 1 0 0 0 1 1 1 1 ? 1 0 1 0 0 0 0 0 3 0 0
Brazil Series B ? ? ? ? 1 1 ? 1 1 ? 1 1 0 1 2 1 ? 0 1 0 1 ? ? 0 ? 1 1 0 ? 0 0 0 3 1 1
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus 1 ? ? ? 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 ? ? ? 0 0 ? ? ?
Magyarosaurus ? ? ? ? 1 1 ? ? 1 ? 1 ? ? ? 2 ? ? 0 ? ? ? ? ? 0 ? 1 1 0 0 0 0 0 3 ? ?
Malawisaurus 1 1 0 0 1 0 ? ? 1 0 0 ? ? ? 1 ? ? 0 ? ? ? ? ? 0 0 0 1 0 ? 0 0 0 0 0 0
Nemegtosaurus 0 1 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus ? ? ? ? 1 0 ? ? 1 0 1 0 0 1 2 ? ? 2 ? 1 ? 1 1 0 1 1 1 0 ? 0 0 0 1 1 0
Opisthocoelicaudia ? ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Phuwiangosaurus ? ? ? ? ? 1 ? ? 1 ? 1 1 0 0 2 ? ? 0 ? ? 1 1 1 0 1 1 1 1 ? 0 0 1 3 0 0
Quaesitosaurus 0 0 ? 0 ? ? ? ? ? ? ? ? ? 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 1 1 1 0 1 1 ? 2 1 0 1 ? ? ? 1 1 1 0 0 1 1 1 1 1 0 1 1 1 1 1 0 1 2 0 0
Rocasaurus ? ? ? ? ? 1 ? ? 0 0 ? ? 0 1 2 1 0 0 0 1 ? 1 1 ? ? ? ? ? ? ? ? ? ? ? ?
Saltasaurus ? ? ? ? 1 1 ? ? 1 0 0 1 0 0 1 0 ? 2 0 ? 0 1 1 0 1 1 1 1 ? 0 0 0 1 0 0
Titanosaurus ? ? ? ? ? 0 ? ? 1 ? 1 0 0 1 1 0 0 0 1 0 1 1 1 0 1 1 1 0 0 0 0 0 3 1 0
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
150 160 170
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus 0 0 ? ? 0 ? ? 0 0 0 1 1 ? 0 0 ? 0 0 1 3 1 0 0 0 ? ? ? 0 0 1 1 ? 1 1 0
Ampelosaurus 0 0 ? 0 0 0 0 0 1 0 0 0 0 0 1 1 0 0 0 ? 1 0 0 0 1 1 0 ? 1 1 1 0 1 1 1
Andesaurus ? ? ? 1 ? 0 0 ? ? ? ? ? 0 0 ? ? ? ? ? ? ? ? ? ? ? 1 0 0 0 ? 0 0 ? ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 1 1 3 1 0 0 1 1 2 1 1 1 0 1 0 0 1 1 0 1 0 1 0 0 1 0 0 1 0 0 0 0 1 0 0
Argentinosaurus 0 0 ? 0 0 0 0 ? 0 0 1 1 0 1 0 1 0 0 1 3 0 1 0 1 ? 1 0 1 0 0 1 ? 0 1 0
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia 0 0 ? ? 0 0 0 1 ? 0 ? ? 0 ? ? ? ? ? ? 3 ? 0 ? ? ? 1 ? ? 1 1 ? ? ? ? ?
Brachiosaurus 0 ? 1 1 0 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 0 1 0 0 1 1 0 0 0 0 0 1 1 0 0
Camarasaurus 1 0 1 0 0 0 1 0 2 1 0 1 0 1 0 0 0 0 0 0 0 1 0 0 1 1 0 1 0 0 0 1 0 0 0
Dicraeosaurus 1 1 1 0 0 1 0 1 ? 1 0 1 1 0 1 0 0 1 ? 0 0 ? 0 0 1 0 0 0 0 0 0 0 0 1 0
Diplodocus 1 1 3 0 0 0 1 1 2 1 0 1 0 1 0 0 1 1 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? 0 0 0 0 ? ? 0 ? ? 0 ?
Euhelopus 1 1 0 0 0 0 0 ? 1 0 1 1 0 0 1 0 ? ? 0 3 2 0 0 0 0 1 1 0 1 0 0 1 1 0 0
Brazil Series B 0 ? 3 ? 0 0 0 1 ? 0 1 1 0 0 1 0 ? ? ? ? ? 0 ? ? 1 1 1 0 1 1 1 0 ? 1 1
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus 0 0 ? 1 1 0 0 1 1 0 1 ? 0 0 0 0 0 0 0 1 ? 0 0 0 1 1 0 0 0 1 0 ? 1 1 0
Magyarosaurus ? ? ? 0 0 0 0 1 0 0 1 ? ? 0 1 0 0 0 1 1 1 0 1 ? 1 1 ? ? 1 1 ? 0 1 ? ?
Malawisaurus 0 0 ? 1 1 0 0 1 2 0 1 1 2 1 0 0 0 0 0 2 1 0 1 0 ? 1 ? ? 1 1 1 ? ? 0 ?
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus 0 0 ? 1 0 0 1 0 2 0 1 1 0 0 1 0 0 0 0 3 2 0 0 0 1 1 0 0 0 1 0 0 0 1 0
Opisthocoelicaudia 1 0 2 1 0 0 ? 0 0 1 0 1 0 1 0 0 0 0 1 1 1 0 0 1 0 1 1 1 1 1 0 1 1 0 1
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Phuwiangosaurus 1 0 ? 0 0 0 0 ? ? ? ? ? 0 0 1 0 ? ? ? ? ? ? ? ? 0 1 0 0 0 1 0 ? ? 0 ?
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 0 0 3 0 0 0 0 1 2 0 1 0 0 1 0 0 0 0 0 3 2 0 1 0 1 1 1 1 1 1 1 1 0 1 1
Rocasaurus 0 0 ? 1 ? 0 0 0 0 0 ? 1 ? 0 1 ? 0 0 1 3 2 0 0 0 ? ? ? ? 0 1 ? 0 ? ? 0
Saltasaurus 0 0 ? 1 0 0 0 1 2 0 1 1 0 0 1 0 0 0 0 3 2 0 0 0 1 1 0 0 1 1 0 0 0 1 ?
Titanosaurus 0 ? ? 0 0 0 0 1 2 0 0 1 0 0 0 0 0 0 0 3 2 0 0 0 1 1 0 0 0 1 0 0 1 0 ?
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
180 190 200 210
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ?
Alamosaurus 1 1 0 0 0 1 1 1 0 3 2 0 1 0 0 0 0 ? ? ? ? ? ? ? 0 0 1 1 ? ? ? 1 1 3 1
Ampelosaurus 0 0 0 ? ? 0 0 0 0 ? 0 1 0 0 0 0 0 1 ? ? 0 ? 0 ? ? ? ? ? 0 ? ? ? ? ? ?
Andesaurus ? 0 ? ? ? 1 ? ? 0 ? ? 0 ? ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 1 0 0 0 1 0 1 1 0 3 3 0 0 0 0 0 0 0 0 ? 0 3 1 1 1 1 0 0 0 0 0 0 0 1 0
Argentinosaurus 1 1 0 0 0 1 1 0 1 3 0 ? 0 1 0 0 0 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? 0 ? ? ? ? 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 0 1 0 0 1 0 0 0 1 2 2 0 0 0 0 0 0 1 0 0 0 0 1 0 1 1 0 1 0 0 0 ? 0 0 0
Camarasaurus 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 1 0 1 1 0 ? 0 0 0 0 0 0 0
Dicraeosaurus 1 0 0 1 0 0 1 1 0 0 0 0 0 ? ? 0 0 0 0 ? 0 3 0 1 1 1 0 ? ? ? 0 ? 0 1 0
Diplodocus 1 0 0 0 1 0 1 1 0 3 3 1 0 0 0 1 0 0 0 0 0 3 1 1 0 0 0 0 1 0 0 0 0 1 0
Epacthosaurus 1 0 ? ? ? ? 1 ? ? 3 ? ? ? 0 0 ? ? ? 0 ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus 0 ? 0 1 1 0 0 0 0 ? ? 1 0 1 0 0 0 1 1 ? 0 ? ? 0 1 1 0 0 0 0 ? ? ? ? ?
Brazil Series B 1 1 0 0 ? ? 1 1 0 3 ? 1 ? 0 0 ? ? 0 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus 0 0 ? 0 ? 1 1 1 0 3 3 1 0 0 0 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? 0 ? ? 1 1
Magyarosaurus ? ? ? ? ? ? 1 1 0 1 1 ? 0 ? ? ? ? ? ? ? 0 ? 0 ? ? ? ? ? ? ? ? ? ? ? ?
Malawisaurus ? 1 0 0 1 ? 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus 1 0 0 0 1 1 1 1 1 3 3 0 1 1 0 0 0 1 1 1 0 1 0 0 0 0 ? 1 1 1 1 ? ? 1 1
Opisthocoelicaudia 0 1 1 0 1 0 1 1 1 1 1 1 0 1 0 0 0 ? 1 ? 1 ? 0 0 1 0 0 ? 0 1 ? 1 1 2 1
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? 3 ?
Phuwiangosaurus ? ? 0 0 1 1 0 0 0 0 0 ? ? ? ? 0 0 1 0 ? 0 ? 0 ? ? ? ? ? 0 ? ? ? ? ? ?
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 1 1 0 0 1 0 1 1 0 3 3 1 0 0 1 1 1 0 1 ? 0 1 1 0 0 0 1 0 1 0 1 1 0 ? ?
Rocasaurus 1 0 ? ? 1 1 1 1 ? 3 3 ? 0 0 1 ? 1 0 ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ?
Saltasaurus 1 0 0 ? 1 1 1 1 0 3 3 1 0 0 0 0 1 0 1 0 0 1 0 0 ? ? ? 0 1 1 1 ? ? 1 0
Titanosaurus 1 0 0 0 0 1 1 1 1 1 0 1 0 1 0 0 0 0 1 1 0 1 0 0 ? ? ? ? 0 ? ? ? ? ? 0
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ?
220 230 240
Aegyptosaurus ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 0 1 ? ? ? ? 0 ? ? ? ? ? ? ? ?
Aeolosaurus 1 0 ? 1 2 0 0 0 ? ? ? 0 0 0 0 1 1 0 0 1 1 2 0 1 1 0 2 ? ? ? ? ? ? ? ?
Alamosaurus 1 0 0 0 0 1 0 0 1 1 0 1 0 0 ? 1 1 1 0 1 1 0 0 1 0 0 ? ? ? ? ? ? 1 1 1
Ampelosaurus 1 0 ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 0 1 1 2 1 1 0 ? ? ? ? 1 ? ? ? ?
Andesaurus 0 0 ? 1 1 1 ? ? ? ? ? 0 0 ? ? ? 0 1 0 0 0 0 1 0 1 0 ? ? ? ? ? ? 1 1 1
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 1 0 0 0 2 0 1 1 1 1 0 1 1 1 1 1 1 1 0 0 0 0 0 1 0 0 1 1 1 1 0 0 0 0 0
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 0 0 0 0 2 1 0 0 1 1 0 0 0 0 0 0 1 1 0 0 0 0 0 1 1 0 ? ? ? 0 ? ? ? ? 0
Camarasaurus 0 0 0 1 1 1 0 0 1 1 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 0 0 ? ? 0 0 ? 0 1 0
Dicraeosaurus 1 0 0 1 0 0 1 0 1 1 0 1 1 0 ? 1 1 ? 0 0 0 0 0 1 0 0 1 1 ? 1 0 0 0 0 0
Diplodocus 1 1 1 1 0 0 1 1 1 1 0 1 1 1 1 1 0 1 1 1 0 1 1 0 0 0 1 1 1 1 0 0 0 0 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 1 ? ? ?
Brazil Series B 1 0 0 1 0 0 0 0 0 0 0 0 ? 0 ? 1 1 0 0 1 1 0 1 0 1 ? ? ? ? 0 ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Lirainosaurus 1 0 ? 1 1 0 ? 1 1 ? ? 0 ? 0 ? 1 0 1 0 1 1 1 2 1 1 0 0 ? ? ? ? ? ? ? ?
Magyarosaurus 1 0 ? 0 0 0 0 0 1 1 0 1 0 0 ? 1 0 0 1 0 1 1 1 1 1 0 0 ? ? ? ? ? ? ? ?
Malawisaurus 1 ? ? ? ? ? ? 0 1 1 1 1 0 ? ? ? ? ? 0 0 0 1 1 1 1 0 ? ? ? 0 1 ? 1 1 1
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus 1 0 ? 0 0 0 1 1 1 1 1 1 0 0 0 1 0 0 0 1 1 0 2 0 1 1 2 1 ? ? ? ? 1 1 1
Opisthocoelicaudia 2 0 0 1 0 1 0 0 1 1 0 1 0 0 ? 0 0 1 0 1 2 0 0 1 1 0 1 0 0 ? 1 1 1 1 1
Paralititan 1 ? ? 1 2 1 ? ? ? 1 ? 0 0 ? ? 0 ? ? ? 1 ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ?
Phuwiangosaurus 0 0 ? 1 2 0 0 0 0 0 0 0 0 0 0 0 0 ? 1 0 0 ? ? ? 1 0 1 0 ? ? ? ? 1 1 1
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 1 0 0 0 2 0 1 0 1 1 1 0 1 0 ? 1 0 1 0 0 1 0 1 1 1 0 0 0 0 0 1 0 1 1 1
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 1 ? ? ? 1 1 2 0 ? ? ? ? ? ? ?
Saltasaurus 1 0 ? 0 0 0 1 1 1 1 1 1 1 0 0 1 1 ? 0 1 1 0 2 0 1 1 2 1 ? ? 1 1 1 1 1
Titanosaurus 1 0 0 0 0 0 0 0 0 1 1 ? 0 0 ? 1 1 0 1 1 1 0 0 1 1 0 ? ? ? 1 0 1 1 1 1
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 0 1 1 2 0 1 1 2 1 ? ? ? ? ? ? ?
250 260 270 280
Aegyptosaurus ? ? ? ? 1 1 ? 1 1 ? ? ? ? ? ? ? ? ? ? 0 0 0 0 0 0 ? 1 0 1 1 1 1 0 1 1
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 1 0 1 1
Alamosaurus 0 1 1 0 1 0 0 1 1 2 1 1 1 1 ? ? 0 1 0 0 0 1 0 1 1 1 0 0 0 1 0 0 1 1 0
Ampelosaurus ? ? 1 0 0 0 2 1 ? ? 1 0 1 0 0 0 1 ? ? 1 1 0 0 0 0 1 ? 0 0 1 1 1 1 0 0
Andesaurus ? ? 1 0 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus ? ? ? ? 0 0 1 1 ? ? 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 0 0 0 0 0 0 1 0 0 ? 1 1 0 0 0 0 0 0 ? 1 1 1 0 0 1 0 0 0 0 1 0 0 0 0 1
Argyrosaurus ? ? ? ? 1 1 1 0 1 ? 1 1 ? ? ? ? ? ? ? 0 1 2 0 0 1 0 ? 0 0 1 0 1 1 1 1
Augustinia ? ? ? ? 1 1 1 2 ? ? 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus ? ? 1 ? 0 1 0 1 0 1 0 1 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 1 1 0 0 1 0 1
Camarasaurus 0 0 0 1 0 1 0 0 0 ? 0 1 1 1 0 1 1 0 0 0 0 0 1 0 0 1 1 1 0 1 1 0 1 0 0
Dicraeosaurus ? ? 0 ? 0 0 0 1 1 2 0 1 0 0 ? 0 ? ? ? 1 0 0 0 0 1 0 ? 0 0 ? 0 0 0 0 ?
Diplodocus 0 0 0 0 0 0 0 0 0 2 0 1 0 0 ? 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 1 0 1
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? 0 ? 1 ? ? ? ? 1 ? 0 1 ? ? ? ? ? ? ? ? ?
Euhelopus ? ? ? ? 0 0 ? ? ? ? 1 0 0 0 ? 0 ? ? ? ? ? ? 0 0 ? 0 ? 0 0 ? ? ? ? ? ?
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 0 1 1 1 0 0 ? 0 0 ? 1 1 1 1 ?
Lirainosaurus ? ? ? ? 0 0 1 1 ? 0 1 ? 2 1 ? 0 1 1 ? 0 1 0 1 0 0 1 0 0 1 ? ? ? ? ? ?
Magyarosaurus ? ? ? ? 0 1 1 1 1 ? 1 0 1 1 1 0 0 ? ? 0 1 1 1 0 0 1 0 0 0 1 1 0 1 1 0
Malawisaurus ? ? 1 1 0 1 ? ? ? ? 1 0 1 1 ? 0 ? 1 ? 0 0 1 0 0 0 1 0 0 1 1 0 ? 1 1 1
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus ? 1 1 0 1 1 0 1 1 2 1 1 1 1 0 1 0 1 ? 0 1 1 1 1 1 1 0 1 1 1 0 1 1 1 1
Opisthocoelicaudia 0 1 1 0 0 0 0 0 1 0 1 1 1 1 ? 1 0 1 1 1 1 1 0 1 1 1 ? 0 0 ? 1 1 1 1 0
Paralititan ? ? ? ? ? ? 0 ? ? ? 1 ? ? ? ? ? ? ? ? 0 1 2 1 0 1 0 ? ? 1 ? ? ? ? ? ?
Phuwiangosaurus ? 1 1 0 1 1 0 0 0 ? 1 ? ? ? ? ? ? ? ? 0 0 ? 0 0 0 0 ? ? ? 1 0 0 1 1 ?
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 1 0 1 1 1 1 0 2 1 ? 1 0 1 0 0 0 1 1 0 1 0 0 1 0 0 0 0 1 0 1 0 0 0 1 0
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Saltasaurus ? ? 1 0 0 1 0 0 1 2 1 1 1 1 1 1 0 1 ? 0 1 1 1 1 0 1 ? 0 ? 1 1 0 1 1 1
Titanosaurus ? 0 1 0 1 0 0 1 1 2 1 0 ? ? ? ? 0 ? ? 0 1 0 0 ? 0 1 ? 0 0 0 1 0 1 1 1
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? 1 1 1 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
320 330 340 350
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 0 ? 1 1 1 0 ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? 0 0 1 0 0 0 0 1 1 0 0 0 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus 0 1 1 0 0 0 1 ? 1 1 1 1 1 0 0 1 0 ? ? 1 1 1 1 0 1 ? ? ? 1 ? ? ? 0 0 ?
Ampelosaurus ? 0 1 ? 0 1 2 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 ? ? ? 0 1 1 1 1 0 ?
Andesaurus ? ? ? ? 0 0 1 1 1 0 1 1 1 1 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 0 0 0 1 1 0 2 0 0 0 0 1 0 0 0 0 0 0 0 0 1 1 0 0 1 1 ? 0 0 1 0 1 0 0 0
Argentinosaurus 0 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 0 0 1 0 1 0 2 2 0 0 0 1 1 0 0 1 1 0 1 1 1 1 0 0 1 0 0 1 0 1 0 1 0 1 0
Camarasaurus 0 1 1 0 1 0 1 0 0 1 1 1 1 0 0 0 1 0 0 0 1 1 0 0 1 1 1 0 0 1 0 1 0 0 0
Dicraeosaurus 0 0 0 1 1 0 2 0 0 0 0 0 0 0 0 1 1 0 0 0 1 1 0 0 1 0 1 0 0 1 1 1 0 0 0
Diplodocus 1 1 1 0 0 0 2 0 0 0 0 0 0 0 0 1 0 1 1 0 1 1 0 1 1 1 1 0 0 1 1 1 0 0 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus 0 0 1 0 0 0 1 0 0 0 0 1 1 0 0 0 1 0 1 1 1 1 0 0 1 1 1 0 0 1 0 1 0 0 ?
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 1 ? ? ? ? ? ? ? ?
Lirainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 0 1 ? 1 1 1 2 1 ? ? 2 1 0 0 0 1 ?
Magyarosaurus ? ? ? ? ? ? 3 ? ? ? ? ? ? ? ? ? ? 0 0 1 1 0 1 1 0 0 1 0 1 1 0 1 0 1 ?
Malawisaurus ? ? ? ? ? ? ? 0 ? 0 0 1 1 1 ? ? ? ? ? 1 ? ? ? ? 0 1 1 0 ? 0 0 ? 0 1 ?
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus 0 1 1 ? 0 ? 1 ? 1 1 1 1 1 1 1 1 0 1 0 1 1 0 1 1 1 0 1 0 0 ? ? ? 1 ? ?
Opisthocoelicaudia 0 1 ? 0 1 2 1 2 1 1 1 1 1 1 0 0 0 1 0 1 2 1 1 0 0 0 0 0 1 1 0 1 1 0 1
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Phuwiangosaurus 0 0 0 ? 0 0 2 0 ? 0 1 1 1 0 0 ? 0 1 0 0 2 1 1 1 0 1 1 1 1 0 0 1 0 1 ?
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 0 1 0 1 0 0 1 1 1 1 ? ? ? 1 0 1 0 0 ?
Rocasaurus 0 0 1 ? 0 0 1 0 1 1 ? 1 1 1 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Saltasaurus 0 1 1 ? 0 0 1 0 1 1 0 1 1 1 0 0 0 1 1 1 1 0 1 0 0 0 1 1 1 1 1 1 1 1 ?
Titanosaurus 0 1 1 0 ? ? 0 2 1 1 1 1 1 1 0 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 ? ?
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
290 300 310
Aegyptosaurus 1 1 ? 1 1 ? 0 0 ? 2 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus 1 0 1 ? 1 ? 0 1 1 ? ? ? ? ? ? ? ? ? ? ? 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Alamosaurus 1 2 ? ? 0 ? 0 1 1 ? 1 0 1 2 0 1 1 ? ? 1 ? 0 0 1 ? 1 0 2 ? ? 1 0 1 2 1
Ampelosaurus 0 2 1 1 0 ? 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 ? 2 ?
Andesaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? 0 ? 2 1 1 1 1 ? ? 0 2 1 1 ? 0 ? ? ? ? ? ? ? ? ?
Apatosaurus 1 0 0 0 0 0 0 0 0 1 0 0 0 1 0 0 0 1 1 1 0 2 0 1 0 1 1 1 1 0 1 0 0 0 0
Argentinosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 1
Argyrosaurus 1 0 1 0 0 ? 0 0 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Brachiosaurus 1 0 0 0 0 2 0 0 0 1 1 0 1 1 0 1 1 1 ? 0 1 ? 1 1 ? 0 0 1 0 0 1 1 1 2 0
Camarasaurus 1 2 0 2 0 2 1 0 1 1 0 0 0 0 0 0 1 1 1 0 0 2 1 1 0 ? ? 1 1 1 1 0 1 1 1
Dicraeosaurus ? ? ? ? 0 ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 0 0 0 1
Diplodocus 1 0 0 0 0 1 ? 0 0 1 0 ? 0 0 1 1 1 0 1 0 1 1 0 1 0 1 0 0 1 1 0 0 0 2 1
Epacthosaurus 0 ? ? ? 1 ? ? ? ? ? ? 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 ? ? ? ? ? ? ?
Euhelopus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 0 1 2 0
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia ? ? ? ? 1 ? 0 ? ? ? ? ? ? 0 0 ? ? 1 1 0 0 2 0 1 1 1 0 ? ? ? ? ? ? ? ?
Lirainosaurus 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Magyarosaurus 1 2 1 0 1 1 0 1 1 ? ? ? ? ? ? ? ? 1 ? ? 0 0 0 0 ? ? ? ? ? ? 1 ? ? ? ?
Malawisaurus ? ? ? ? ? ? ? ? ? 0 ? 0 0 2 0 1 ? 1 0 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus 1 1 1 2 0 1 0 1 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 1 2 1
Opisthocoelicaudia 1 0 1 1 1 0 0 1 1 1 1 1 1 ? 1 1 1 ? ? 1 ? ? 0 ? ? 1 0 2 ? ? 1 1 1 2 1
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 ? ? ? ? ? ? ?
Phuwiangosaurus 1 1 0 0 0 ? ? 1 0 ? ? ? ? ? ? ? ? ? 0 ? ? ? ? ? ? ? ? ? ? ? ? 0 1 2 1
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus 1 2 1 0 0 1 1 1 1 2 0 0 0 1 0 0 0 1 0 0 0 0 0 0 1 1 1 ? ? ? 1 1 1 2 1
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 2 1
Saltasaurus 0 1 0 2 1 1 0 1 1 ? ? ? ? ? ? ? ? ? ? ? 0 0 0 0 0 ? ? ? ? ? 1 1 1 2 0
Titanosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 1 1 1 0 0
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
360 364
Aegyptosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Aeolosaurus ? ? ? ? ? ? ? ? ? ? ? ? 1 0
Alamosaurus ? ? ? ? ? ? ? ? ? ? ? ? 0 ?
Ampelosaurus ? ? ? ? ? ? ? ? ? ? ? ? 1 1
Andesaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Antarctosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Apatosaurus 0 1 0 0 0 0 0 0 0 0 0 ? 0 0
Argentinosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Argyrosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Augustinia ? ? ? ? ? ? ? ? ? ? ? ? 1 ?
Brachiosaurus 0 0 ? ? 0 1 0 0 ? 0 0 ? 0 0
Camarasaurus 0 0 0 0 0 1 0 0 0 0 0 0 0 0
Dicraeosaurus 0 ? ? ? 0 0 0 ? 0 ? ? ? 0 0
Diplodocus ? 1 0 0 0 0 0 0 0 0 0 0 0 0
Epacthosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Euhelopus 0 0 ? 1 0 1 0 1 0 ? ? 0 0 0
Brazil Series B ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jainosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Janenschia 1 ? 1 0 0 1 0 0 0 0 0 0 ? ?
Lirainosaurus ? ? ? ? ? ? ? ? ? ? ? ? 1 ?
Magyarosaurus ? ? ? ? ? ? ? ? ? ? ? ? 1 2
Malawisaurus ? ? ? 0 ? ? 1 ? ? ? ? ? 1 2
Nemegtosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Neuquensaurus ? ? ? ? ? ? ? ? ? ? ? ? 1 1
Opisthocoelicaudia 1 1 1 ? 0 1 0 0 0 ? ? 1 0 ?
Paralititan ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Phuwiangosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Quaesitosaurus ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Rapetosaurus ? 0 1 1 1 1 0 0 1 1 1 0 1 2
Rocasaurus ? ? ? ? ? ? ? ? ? ? ? ? ? 1
Saltasaurus ? ? ? 1 ? 0 0 ? ? ? ? ? 1 1
Titanosaurus ? ? ? ? ? ? ? ? ? ? ? ? 0 ?
Santa Rosa indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Jabalpur indet. ? ? ? ? ? ? ? ? ? ? ? ? ? ?
Malagasy TaxonB ? ? ? ? ? ? ? ? ? ? ? ? ? ?
APPENDIX 2.2. CHARACTERS ORDERED BY lary body (0); maxilla contacts jugal via elon-
ANATOMICAL REGION gate posterior process that extends pos-
terodorsally from maxillary body (1).
The cladistic codings for the 364 characters (109 skull,
10. Maxilla–nasal contact: present (0); absent (1)
139 axial, 112 appendicular, 2 dermal) used in this analy-
(Nowinski 1971).
sis are listed below in anatomical order. Approximately
50% of the characters in this work are new and used in a 11. “Shelflike” area lateral to external naris, extend-
novel manner. Other characters have been used by previ- ing onto rostral end of maxilla: present (0);
ous workers as cited. References cited with each character absent (1) (McIntosh 1990a; Upchurch 1998).
relate to usage of the character in previous phylogenetic 12. Preantorbital fenestra: absent (0); present (1)
analyses and commonly provide a reference for further (Wilson and Sereno 1998).
information or illustrations of the character; they do not
13. Subnarial foramen on premaxilla–maxilla
necessarily provide a comprehensive list of works that
suture: circular (0); slit-shaped, more than
have mentioned or employed the character. In many cases
two times longer than broad (1) (Wilson
the character definitions, polarities, or distributions have
1998).
been modified, as indicated. New characters are discussed
in detail by Curry (2001). 14. Subnarial foramen on premaxilla–maxilla
suture: faces laterally (0); faces dorsally (1)
CRANIAL CHARACTERS (Upchurch 1998).
EXTERNAL FEATURES OF THE SKULL 15. Antorbital fenestra, maximum diameter: much
shorter than (0), subequal to (1), or much
1. Short, deep snout: absent (0); present (1) longer than (2) orbital diameter (Wilson 1998).
(modified from Upchurch 1998; Wilson and
16. External nares: retracted to level of anterior
Sereno 1998).
orbit (0); retracted to position between orbits (1)
2. Posterolateral processes of premaxilla and lat- (Steel 1970; Gauthier 1986; McIntosh 1990a;
eral processes of maxilla: without midline con- Upchurch 1995, 1998; Wilson and Sereno
tact (0); with midline contact forming marked 1998).
narial depression, subnarial foramen not visi-
17. External nares: face laterally or rostrolaterally
ble laterally (1) (Wilson and Sereno 1998).
(0); face dorsally or rostrodorsally (1)
3. Premaxilla: formed from heavy and thick main (Upchurch 1995, 1998).
body with distinct ascending process (0);
18. External nares, maximum diameter: shorter (0);
formed from transversely narrow main body
or longer (1) than orbital maximum diameter
that is greatly elongated rostrocaudally (1)
(McIntosh 1990a; Upchurch 1995, 1998;
(Upchurch 1998).
Wilson and Sereno 1998).
4. Ascending process of premaxilla: directed
19. Rostral rim of external naris: lies in front of ros-
caudodorsally (0); directed dorsally (1); absent
tral margin of antorbital fenestra (0); lies caudal
(2) (Gauthier 1986; Wilson and Sereno 1994;
to rostral margin of antorbital fenestra (1)
Upchurch 1995, 1998).
(McIntosh 1990a; Upchurch 1995, 1998).
5. Caudolateral process of premaxilla: present (0);
20. External naris: divided by contact between max-
absent (1) (Upchurch 1998).
illa and/or premaxilla and nasals (0); undi-
6. Position of maxillary ascending process: cen- vided, nasals do not contact maxilla or premax-
tral or anterior to the center of the maxilla (0); illa (1) (modified from Gauthier 1986; Wilson
posterior to the center of the maxilla (1). and Sereno 1994; Upchurch 1995, 1998).
7. Dorsal ascending process of maxilla: con- 21. Nasal, medial and lateral processes: medial
tributes less than one-half of its maximum nasal process substantially longer than lateral
length to the articulation with the contralat- nasal process (0); medial and lateral nasal
eral maxilla (0); contributes over one-half of processes equivalent in maximum length (1);
its maximum length to articulation with the lateral nasal process longer than medial nasal
contralateral maxilla (1). process (2).
8. Maxillary flanges: do not contact each other 22. Jugal, contribution to antorbital fenestra: very
on midline (0); contact on midline (1). reduced or absent (0); large, bordering approxi-
9. Maxilla, jugal process: maxilla contacts jugal mately one-third of perimeter (1) (Wilson
via a blunt posterior process at level of maxil- 1998).
23. Jugal–maxillary process structure: jugal bluntly (2) transverse diameter of supratemporal fen-
overlaps posterior border of maxilla (0); jugal estra (modified from Wilson 1998).
has tongue-in-groove articulation with elongate 40. Postparietal foramen: absent (0); present (1);
maxillary process (1). (Wilson 1998).
24. Jugal-quadratojugal contact: jugal has broad 41. Supratemporal fenestra, maximum diameter:
articulation with quadratojugal (0); jugal has much longer than (0); subequal to (1); or
narrow articulation only along dorsal portion much less than (2) that of foramen magnum
of quadratojugal (1). (Wilson 1998).
25. Jugal–ectopterygoid contact: present (0); 42. Supratemporal fenestra: faces dorsally or dor-
absent (1) (Wilson and Sereno 1998). solaterally (0); faces laterally (1); faces
26. Lacrimal: posteriorly inclined (0); vertically anterodorsally (2) (Salgado and Calvo 1992;
oriented (1) (Salgado and Calvo 1997). Upchurch 1998).
27. Lacrimal contribution to external naris: 43. Supratemporal fenestra, margin: includes
absent (0); present (1) (Nowinski 1971). squamosal (0); excludes squamosal (1)
(Upchurch 1995, 1998).
28. Postorbital, contact with jugal: postorbital
articulates with jugal in simple overlapping 44. Infratemporal fenestra, rostral margin: lies
contact (0); postorbital articulates with below midpoint of orbit or more caudally (0);
several deep pits on posterior wing of jugal extends as far as, or beyond, rostral margin of
(1). orbit (1) (Gauthier 1986; Upchurch 1995,
1998).
29. Postorbital contact with laterosphenoid:
45. Temporal bar: parallel (0) or perpendicular (1)
absent (0); present (1).
to tooth row (Wilson 1998).
30. Prefrontal anterior process: absent (0); pres-
46. Quadratojugal, rostral process: straight or
ent (1) (Wilson and Sereno 1998).
curves gently upward toward its rostral tip
31. Prefrontal, posterior process size: small, not (0); has “step”-like change of direction at
projecting far posterior of frontal nasal suture midlength, such that rostral half of process
(0); elongate, approaching parietal (1) (Wilson runs rostroventrally (1) (Upchurch 1998).
1998).
47. Quadratojugal, angle between rostral and dor-
32. Frontals, midline contact (symphysis): sal rami: 90 or less (0); at least 130 (1)
sutured (0); fused (1) (Salgado and Calvo (Upchurch 1998).
1992; Upchurch 1998; Wilson 1998).
48. Squamosal–quadratojugal contact: absent (0);
33. Frontal, medial convexity in dorsal view: present (1) (Gauthier 1986; Wilson and
absent (0); present (1). Sereno 1998).
34. Frontal contribution to supratemporal fossa: 49. Squamosal, general morphology: dorsally
absent (o); present (1) (Wilson and Sereno, convex hook in lateral view (0); shaped like
1998). inverted “L” in lateral view (1).
35. Frontoparietal suture: on same plane as 50. Squamosal–postorbital suture: postorbital has
midline suture of each respective bone to its overlapping contact with lateral surface of
counterpart (0); depressed relative to mid- squamosal (0); deep anterodorsal notch in
line suture of each element’s counterpart head of squamosal to receive postorbital (1).
(1); elevated relative to midline sutures (2).
51. Quadrate, posterior fossa depth: shallow (0);
36. Parietal, anterior inclination with wide cau- deeply invaginated (1) (Kurzanov and
dodorsal exposure of crest: absent (0); pres- Bannikov 1983; Wilson and Sereno 1998).
ent (1) (Salgado and Calvo 1997; Wilson and 52. Quadrate shaft, long axis: oriented perpendicu-
Sereno 1998; Upchurch 1998). lar to long axis of skull (0); directed caudodor-
37. Parietal, elongate lateral process: absent (0); sally (1) (modified from Upchurch 1998).
present (1). 53. Quadrate head, orientation: directly dorsal to
38. Parietal, contribution to posttemporal fenes- lateral margin of quadrate body (0); extends
tra: present (0); absent (1) (Wilson 1998). lateral to lateral margin of quadrate body (1).
39. Parietal, distance separating supratemporal 54. Quadrate–quadratojugal articulation: via sin-
fenestrae: less than (0); twice (1); or equal to gle sutural scar (0); via V-shaped bifid scar (1).
55. Supraoccipital–parietal articulation: supraoc- long, slender, strongly laterally compressed
cipital has broad contact with caudal border process that lacks dorsal groove (1) (McIntosh
of parietal (0); supraoccipital underlies pari- 1990a; Upchurch 1995, 1998).
etal via two midline, anterior processes (1).
70. Basal tubera, anteroposterior depth: approxi-
56. Supraoccipital, formation of nuchal crest: mately half dorsoventral height (0); sheetlike,
formed from single midline prominence (0); 20% dorsoventral height (1) (Wilson 1998).
results from joining of two lateral promi- 71. Basal tubera, angle of divergence: wide, greater
nences, separated by median trough (1). than or equal to 60 (0); narrow, less than 45 (1).
BRAINCASE AND OCCIPUT 72. Basal tubera, posterior extent: tubera project
away from basicranium with considerable relief
57. Occipital region of skull: anteroposteriorly (0); tubera lie in same plane as basicranium (1).
deep, paroccipital processes oriented postero-
laterally (0); flat, paroccipital processes ori- 73. Basal tubera, position of divergence: occurs
ented transversely (1) (Wilson 1998). approximately level with occipital condyle or
slightly below (0); occurs well below occipital
58. Occipital condyle; caudoventrally directed (0); condyle (1).
ventrally directed (1).
74. Paroccipital processes, distal end: slightly
59. Crista prootica: rudimentary (0); expanded lat- expanded with a straight distal margin (0);
erally into “dorsolateral process” (1) (modified elongated with a rounded “tongue”-like distal
from Upchurch 1995; 1998; Wilson 1998). margin (1); elongated at distal tip (2) (modi-
60. Exit for cranial nerve (CN) I: faces anteriorly fied from Upchurch 1998).
(0); faces dorsally (1). Character state assign- 75. Paroccipital processes, transverse projection
ments are also determined by vertically orientat- from basicranium: narrow (ratio of space
ing the supraoccipital (Salgado and Calvo 1997). between occipital condyle and lateral extreme
61. Laterosphenoid, openings for trigeminal nerve of proximal paroccipital process to occipital
(CN V): single opening between prootic and condyle height 0.5) (0); wide (ratio of space
laterosphenoid (0); double opening, one between occipital condyle and lateral extreme
between prootic and laterosphenoid, one of proximal paroccipital process to occipital
between laterosphenoid and orbitosphenoid (1). condyle height 0.8) (1).
62. Basipterygoid processes: short, approximately
PALATE
twice basal diameter (0); elongate, at least
four times basal diameter (1) (Wilson 1998). 76. Pterygoid flange, position: posterior to orbit
(0); posterior to antorbital fenestra (1); ante-
63. Basipterygoid processes: perpendicular to (0)
rior to antorbital fenestra (2) (Wilson 1998).
or angled approximately 45 to (1) skull roof
(Wilson 1998). 77. Pterygoid, breadth of main sheet: less than
20% total length of pterygoid (0); approximately
64. Basipterygoid processes, angle of divergence:
30% total length of pterygoid (1) (McIntosh and
approximately 45 (0); less than 30 (1); over
Berman 1975; Upchurch 1998:fig.4).
45 (2) (Wilson 1998).
78. Pterygoid, basipterygoid fossa: deeply exca-
65. Basipterygoid processes, area between bases: vated dorsal fossa with basipterygoid hook
shallowly concave (0); deeply excavated into present on posterodorsal wing of pterygoid (0);
long narrow pit extending caudally beneath excavation shallow and faces medially (1);
rostral part of braincase (1) (Upchurch 1995, shelflike (2) (modified from Upchurch 1998;
1998). Wilson and Sereno 1998).
66. Basipterygoid angle: shallow U-shaped divi- 79. Pterygoid: anterior wing of pterygoid contacts
sion (0); V-shaped division (1); deep U-shaped opposite element in narrow suture (0); ante-
division (2). rior wing of pterygoid contacts opposite ele-
67. Basipterygoid processes, orientation of distal ment in broad sheet (1).
ends: faces dorsolaterally (0); faces ventrally (1). 80. Pterygoid: pterygoids meet at angle of 30
68. Basisphenoid depression: absent (0); present in ventral view (0); pterygoids meet at angle
(1) (Wilson 1998). of 45 in ventral view (1) (Nowinski 1971).
69. Parasphenoid rostrum: triangular in lateral 81. Pterygoid, ectopterygoid process: lies below
view with groove along dorsal midline (0); lacrimal or more posteriorly (0); lies anterior
to lacrimal, under antorbital fenestra (1) 95. Surangular–angular suture: nearly horizontal
(Upchurch 1994, 1998). (0); dorsally convex (1).
82. Pterygoid, ectopterygoid process: robust, proj- 96. Surangular, lateral exposure (depth): surangu-
ects below level of jaw margin (0); reduced, lar exposure is significantly greater than that
does not project below margin of upper jaw (1) of the angular (0); surangular exposure is less
(Berman and McIntosh 1975; Upchurch 1994, than that of the angular (1).
1998). 97. Splenial posterior process: overlaps angular (0);
83. Palatine, shape of lateral ramus: plate-shaped, inserts between anterior portions of prearticu-
long maxillary contact (0); rod-shaped, narrow lar and angular (1) (Wilson 1998).
maxillary contact (1) (Wilson and Sereno 1998). 98. Splenial posterodorsal process: present,
approaches margin of adductor chamber (0);
MANDIBLE AND TEETH
absent (1) (Wilson 1998).
84. Dentary, depth of anterior end of ramus: 99. Splenial, participation in medial wall of
150% minimum depth (0); slightly less than adductor chamber: present (0); absent (1)
that of dentary at midlength (1) (modified (Wilson 1998).
from Wilson and Sereno 1998). 100. Dentary teeth, number: 17 or fewer (0); more
85. Dentary tooth rows: teeth distributed one-half than 20 (1) (Wilson and Sereno 1998).
of dentary’s total length or more (0); teeth 101. Teeth orientation: perpendicular (0) or ori-
restricted to anterior one-third of dentary (1) ented anteriorly (1) relative to jaw margin
(modified from Gauthier 1986; Upchurch (Wilson 1998).
1998). 102. Tooth crowns, arrangement: aligned slightly
86. Tooth rows, shape of anterior maxillary por- anterolingually, tooth crowns overlapping (0);
tions: broadly arched, anterior portion of aligned along jaw axis, crowns not overlapping
tooth rows U-shaped (0); rectangular, tooth- (1) (Wilson and Sereno 1998).
bearing portion of jaw perpendicular to jaw 103. Teeth, shape of wear facets: V-shaped facets
rami (1) (Wilson and Sereno 1998). (interlocking) (0); high-angled planar facets
87. Dentary, relative length of posterior rami: (1); low-angled planar facets (2) (Wilson and
dorsal and ventral rami approximately same Sereno 1998).
length (0); dorsal ramus significantly shorter 104. Teeth, occlusal pattern: spoonlike, spatulate (0);
than ventral (1). peglike with high-angle wear facets on one side
88. Dentary, anteroventral margin: gently only (1); peglike with interlocking V-shaped
rounded (0); sharply projecting triangular wear facets (2); peglike with low-angle (less
“mental” process (1) (Wilson 1998). than 40) wear facets (3) (modified from
Salgado and Calvo 1997; Upchurch 1998;
89. Dentary, shape of posterodorsal groove:
absent (0); present (1) (Wilson 1998).
105. Tooth crowns, cross-sectional shape at mid-
90. Mandible, angle between long axis of mandibu- crown: elliptical (0); D-shaped (1); cylindrical
lar symphysis and long axis of mandible: (2) (Janensch 1929; Wilson and Sereno, 1998).
approximately 45 (0); approximately 90 (1)
(Upchurch 1998; Wilson and Sereno 1998). 106. Teeth, marginal tooth denticles: absent on pos-
terior edge (0); absent on both anterior and
91. Surangular depth: less than twice (0) or more posterior edges (1) (McIntosh 1990a; Upchurch
than two and one-half times (1) maximum 1994, 1998; Wilson and Sereno 1998).
depth of angular (Wilson and Sereno 1998).
107. Tooth crowns, prominent grooves near mesial
92. Surangular, shape: laterally convex (0); almost and distal margins of labial surface: present (0);
vertical (1). absent (1) (Upchurch 1994, 1998).
93. Coronoid process, elevated relative to suran- 108. Teeth, SI values (tooth crown length divided
gular: absent (0); present (1) (Salgado and by maximum mesiodistal width) for teeth:
Calvo 1997). less than 4.0 (0); greater than 5.0 (1)
94. Surangular–dentary suture: nearly horizontal (Upchurch 1994; 1998).
with respect to long axis of mandible (0); con- 109. Replacement teeth per alveolus: two or fewer
cave with respect to long axis of mandible (1). (0); more than four (1).
AXIAL SKELETON 122. Anterior cervicals, neurocentral junction:
straight (0); constricted and butterfly-shaped
GENERAL PRESACRAL VERTEBRAE
in dorsal view (1).
110. Presacral vertebral bone texture: solid (0);
123. Anterior cervical neural spines: single (0);
spongy, with large, open internal cells (som-
distally divided into U- or V-shaped trough
phospondylus) (1) (Wilson and Sereno 1998).
(bifid) (1); proximally divided by formation of
111. Presacral pneumatic fossae, shape: deep lat- pre- and postspinal coels, and distally fused
eral excavations bordered by sharp lip (0); (2) (modified from McIntosh 1990a; Wilson
shallow lateral depressions (1). and Sereno 1998).
124. Anterior cervicals, neural spines: transversely
CERVICAL VERTEBRAE broad, laterally expanded (0); anteroposteriorly
112. Atlantal intercentrum, occipital facet: rectan- expanded and compressed transversely (1).
gular in lateral view, dorsal aspect of inter- 125. Anterior cervical neural spines, distal mor-
centrum anteroposterior length subequal to phology: straight (0); posteriorly inclined (1).
that of ventral aspect (0); expanded
anteroventrally in lateral view, dorsal aspect 126. Anterior cervical prezygapophyses: short, do
of intercentrum anteroposterior length not extend beyond anterior border of the cen-
shorter than that of ventral aspect (1) (Wilson trum (0); elongate, extend well beyond ante-
and Sereno 1998). rior articular facet (1).
113. Cervical vertebrae, number: 12 (0); 13 (1); 15 127. Anterior cervicals, prespinal lamina: present
or more (2) (modified from Upchurch 1995, (0); absent (1).
1998; Wilson and Sereno 1998). 128. Anterior cervicals, postspinal lamina: present
114. Cervical neural arch lamination: well devel- (0); absent (1).
oped, with well-defined pneumatic fossae and 129. Infraprezygapophyseal laminae on middle
laminae (0); rudimentary, diapopohyseal lam- and caudal cervicals: “single” (0); bifurcate
inae only feebly developed if present (1) toward dorsal ends to form medial and lateral
(Wilson and Sereno 1998). branches (with a triangular hollow in
115. Cervical parapophysis, dorsal surface excava- between) (1) (Upchurch 1998).
tion: absent (0); present but separated from 130. Midcervical centra, anteroposterior
pneumatic fossa by longitudinal ridge (1) length/height of posterior face: 4 (0); 2.5–3.0
(Upchurch 1998). (1) (modified from Upchurch 1995, 1998).
116. Anterior cervical centra: height:width ratio of 131. Middle and posterior cervical centra, articular
anterior face: approximately 1.25 (0); 1.0 or face shape: height width (0); height
less (1) (modified from Martin et al. 1999; width (1).
Upchurch 1998). 132. Middle and posterior cervical centra, ventral
117. Anterior cervical centra, anteroposterior surface shape: with midventral keel (0); with-
length divided by width of posterior face: out midventral keel (1).
short (EI 3.0) (0); elongate (EI 3.0) (1) 133. Middle and posterior cervical centra, pneu-
(Upchurch 1995, 1998). matic fossae: shallow excavations lacking
118. Anterior cervical centra, midventral keel: dividing laminae (0); shallow with dividing
absent (0); present (1). laminae (1).
119. Anterior cervical centra, ventral margin: broad 134. Middle and posterior cervicals, neurocentral
and rounded, ventrally convex (0); broadly conjunction: straight (0); constricted and butterfly-
cave with excavation of ventral centrum (1). shaped (1).
120. Anterior cervical pneumatic fossae: complex, 135. Middle and posterior cervicals, neurocentral
divided by bony septa (0); shallow and sim- junction: terminates posterior to anterior artic-
ple, divided by single median ridge (1); sim- ular face (0); extends full length of centrum (1).
ple, undivided (2) (modified from Wilson and 136. Middle and posterior cervicals, neural canal
Sereno 1998; Upchurch 1998). shape: dorsoventral and mediolateral dimen-
121. Anterior cervical parapophysis: arises at mid- sions are subequal (0); dorsoventral
centrum (0); arises in anterior half of cen- dimensions are greater than mediolateral
trum (1). dimensions (1).
137. Middle and posterior cervical neural arches, 152. Anterior dorsal distal neural spines: straight,
development of laminae and pneumatic fos- equal anteroposterior and transverse dimen-
sae: poorly developed shallow fossae bound sions (0); triangular, transversely expanded (1)
by illformed laminae (0); deep lateral fossae (modified from Wilson and Sereno 1998).
bound by robust laminae (1) (modified from 153. Dorsal neural spines, length measured from
Upchurch 1998; Wilson and Sereno 1998). prezygapophyses: twice (0) or four times (1)
138. Middle and posterior cervicals, pre- and post- centrum length (modified from Wilson and
spinal coels: only prespinal coel present (0); Sereno 1998).
only postspinal coel present (1); both pre- and
154. Dorsal transverse processes: directed strongly
postspinal coel present (2); both coels absent
dorsolaterally (0); directed laterally or slightly
(3).
upward (1) (Upchurch 1998).
139. Middle and posterior cervicals, prespinal lam-
155. Anterior dorsal transverse processes, angle of
ina: absent (0); present (1).
articulation with neural spine: 90 more
140. Middle and posterior cervicals, postspinal (0); 70 (1).
lamina: absent (0); present (1)
156. Anterior dorsals, diapophyses: restricted to
141. Posterior cervical and anterior dorsal neural position posterior to anterior articular face (0);
spines: single (0); bifid (1) (Wilson and extend anterior to anterior articular face (1).
Sereno 1998).
157. Anterior dorsal neural arches, infraprezy-
142. Posterior cervical and anterior dorsal bifid gapophyseal lamina: single (0); divided (1)
neural spines, median tubercle: absent (0); (modified from Wilson 1998).
present (1) (Wilson and Sereno 1998).
158. Anterior dorsal neural arches, infrapostzy-
gapophyseal lamina: single (0); divided (1).
DORSAL VERTEBRAE 159. Anterior dorsals, interprezygapophyseal lam-
143. Dorsal vertebrae: 13 (0); 12 (1); 11 (2); 10 or ina: present (0); absent (1).
fewer (3) (modified from Wilson and Sereno 160. Anterior dorsals, prespinal lamina: absent
1998). (0); present, short and restricted to proximal
144. Caudal dorsal centra ratio of centrum length neural spine (1); present, short and restricted
to height of posterior face: 1.0 (0); 3.0 (1) to distal neural spine (2); present along entire
(modified from Upchurch 1998). length of neural spine (3).
145. Anterior dorsal centra, articular face shape: 161. Anterior dorsals, postspinal lamina: absent
opisthocoelous (0); amphiplatyan (1) (0); present, short and restricted to proximal
(McIntosh 1990a; Upchurch 1998; Wilson neural spine (1); present along entire length
and Sereno 1998). of neural spine (2).
146. Pneumatic fossae in dorsal centra: present (0); 162. Anterior dorsals, neural arches, hyposphene-
absent (1) (Upchurch 1998). hypantra: absent (0); present (1) (modified from
147. Pneumatic fossae in dorsal centra: moder- McIntosh 1990a; Bonaparte and Coria 1993;
ately deep but simple pits (0); deep, ramify Upchurch 1998; Wilson and Sereno 1998).
extensively within centrum and enter the 163. Anterior dorsal neural arches, median infra-
base of neural arch (1) (Upchurch, 1998). postzygapophyseal lamina: absent (0);
148. Anterior dorsals, neural canal dimensions: present (1).
height width (0); height width (1). 164. Anterior dorsal neural arches, spinodiapophy-
149. Anterior dorsals, dorsal boundary of neural seal lamina: undivided (0); divided (1).
canal: ventral extension of prespinal fossa (0); 165. Middle and posterior dorsal centra, ventral
interprezygapophyseal lamina (1); bone ven- concavity: present (0); absent (1).
tral to interprezygapophyseal lamina and dor- 166. Posterior dorsal centra, articular face shape:
sal to neural canal (2). amphicoelous (0); opisthocoelous (1) (Salgado
150. Anterior dorsal neural spines: single (0); et al. 1997; Wilson and Sereno 1998).
bifid (1). 167. Middle and posterior dorsal neural spines,
151. Anterior dorsal neural spines: straight, dor- height (measured from prezygapophyses):
sally directed (0); posteriorly inclined (1). comprise 65% total height of vertebra (0);
comprise 50% total height of vertebra (1) 182. Middle and posterior dorsal neural arches,
(modified from Bonaparte 1986; Upchurch prespinal lamina: absent (0); present, antero-
1998). posteriorly expanded and sheetlike (1); pres-
168. Middle and posterior dorsal neural spine– ent, transversely expanded and triangular (2)
transverse process intersection: 70 (0); (Bonaparte 1986; Calvo and Salgado 1995;
90 (1). Upchurch 1998).
169. Middle and posterior dorsal neural spines, 183. Middle and posterior dorsal neural arches,
orientation: vertical (0); posterior, neural postspinal lamina: absent (0); present,
spines approach level of diapophyses (1) anteroposteriorly expanded and sheetlike (1)
(Wilson 1998). (Bonaparte 1986; Calvo and Salgado 1995;
Upchurch 1998).
170. Middle and posterior dorsal neural arches,
hyposphene-hypantrum articulations: present 184. Middle and posterior dorsal zygapophyses:
(0); absent (1) (Wilson 1998). prezygapophyses at same level as postzy-
gapophyses (0); postzygapophyses dorsal to
171. Middle and posterior dorsal neural spines, prezygapophyses (1). This character is best
shape: tapering or not flared distally (0); documented in lateral view.
flared distally, with pendant, triangular lateral
processes (1) (Upchurch 1995, 1998; Wilson 185. Middle and posterior dorsal prespinal lami-
and Sereno 1998). nae: absent (0); present, short, restricted to
proximal one-half of neural spine (1); pres-
172. Middle and posterior dorsal neural arches, ent, short, restricted to distal one-half of
“infradiapophyseal” pneumatic foramen: neural spine (2); present, elongate, extend-
absent (0); present (1) (Wilson 1998). ing full length of neural spine (3).
173. Middle and posterior dorsal neural arches, 186. Middle and posterior dorsal postspinal lami-
fossa dorsal to neural canal and below postzy- nae: absent (0); present, short, restricted to
gapophyses: present (0); absent (1). proximal one-half of neural spine (1); present,
174. Posterior dorsal neural spines: rectangular short, restricted to distal one-half of neural
through most of length (0); “petal”-shaped, spine (2); present, elongate, extending full
expanding transversely through 75% of length of neural spine (3).
length and then tapering (1) (Wilson 1998). 187. Middle and posterior dorsals, anterior centro-
175. Posterior dorsal vertebrae, transverse diapophyseal lamina: vertical (0); anteriorly
processes: lie caudal, or caudodorsal, to para- inclined (1).
pophysis (0); lie vertically above parapophysis 188. Middle and posterior dorsals, intraprezy-
(1) (Upchurch 1998). gapophyseal lamina: undivided (0);
176. Posterior dorsal vertebrae, neural spines: divided (1).
wider craniocaudally than transversely (0); 189. Middle and posterior dorsals, supraprezy-
compressed craniocaudally (1) (Upchurch gapophyseal lamina: present (0); absent (1).
1998).
190. Middle and posterior dorsals, fossa between
177. Dorsal neural spines, triangular processes: divisions of supraprezygapophyseal lamina:
absent (0); present (1) (Upchurch 1995, absent (0); present (1).
1998).
191. Middle and posterior dorsals, median infra-
178. Middle and posterior dorsal neural arches, postzygapophyseal lamina: absent (0); pres-
anterior centroparapophyseal lamina: present ent (1).
(0); absent (1) (Wilson 1998).
192. Middle and posterior dorsals, median
179. Middle and posterior dorsal neural arches, infraprezygapophyseal lamina: absent (0); pres-
prezygoparapophyseal lamina: present (0); ent (1).
absent (1) (Wilson 1998).
193. Middle and posterior dorsal vertebrae, infra-
180. Middle and posterior dorsal neural arches, diapophyseal lamina: single (0); bifurcated (1)
posterior centroparapophyseal lamina: absent (Salgado et al. 1997).
(0); present (1) (Wilson 1998).
181. Middle and posterior dorsal neural arches SACRAL VERTEBRAE
spinopostzygapophyseal lamina: divided (0); 194. Sacral vertebrae, number: 5 (0); 6 (1) (modified
single (1) (Wilson 1998). from McIntosh 1990a; Salgado et al. 1997;
Upchurch 1995, 1998; Wilson and Sereno 211. Anterior caudal centra (excluding the first),
1998). articular face shape: amphiplatyan or platy-
195. Sacral vertebrae, number supporting acetabu- coelous (0); procoelous (1); opisthocoelous (2)
lum: two to four (0); five or more (1) (Wilson (Bonaparte 1986; Powell 1986; Upchurch
1998). 1998; Wilson and Sereno 1998).
196. Sacral vertebrae, ventral margin: rounded (0); 212. Anterior caudal centra, pneumatic fossae:
ventrally concave (1). absent (0); present (1) (modified from
Upchurch 1998; Wilson 1998).
197. Sacral centra: all opisthocoelous (0); ranging
from opisthocoelous to amphiplatyan to pro- 213. Anterior caudal centra: maintain same length
coelous (1); all amphiplatyan (2). (0); double in length (1) over first 20 verte-
brae (Wilson and Sereno 1998).
198. Sacral centra, pneumatic fossae or very deep
depressions: absent (0); present (1) (Upchurch 214. Anterior caudals, ventral margin of centrum:
1998). rounded, convex (0); ventrally concave and
marked by a deep midline groove (1).
199. Sacral neural spines: approximately twice (0)
or more than four times (1) length of cen- 215. Anterior caudal neural spines: reduced with
trum (Wilson 1998). equal anteroposterior and transverse dimen-
sions (0); transversely widened (1); anteropos-
200. Sacral neural spines, prespinal lamina:
teriorly expanded (2).
anteroposteriorly expanded (0); transversely
expanded (1). 216. Anterior caudal neural spines: high (neural
spine height centrum height) (0); low (neu-
201. Sacral neural spines, postspinal lamina: pres-
ral spine height centrum height) (1).
ent (0); absent (1).
217. Anterior caudal neural arches, spinoprezy-
202. Sacral neural spines, prespinal lamina: single
gapophyseal lamina: absent (0); present and
proximally (0); divided proximally, arising
extending onto lateral aspect of neural spine
from prezygapophyses (1).
(1) (Wilson 1998).
203. Sacral neural spine: meets transverse process
218. Anterior caudal neural arches, spinoprezy-
at 90 (0); meets transverse process at
gapophyseal lamina–spinopostzygapophyseal
70(1).
lamina contact: absent (0); present, forming a
204. Sacral ribs, proximal excavation: absent (0); prominent lateral lamina on lateral aspect of
present (1). neural spine (1).
205. Sacral ribs, position of apex: high, extending 219. Anterior caudal neural arches, prespinal lam-
beyond dorsal margin of ilium (0); low, not ina: absent (0); present (1) (Wilson 1998).
projecting beyond dorsal margin of ilium (1)
220. Anterior caudal neural arches, postspinal
(Wilson 1998).
lamina: absent (0); present (1) (Wilson 1998).
CAUDAL VERTEBRAE 221. Anterior caudal neural arches, postspinal
206. Caudal centra, bone texture: solid (0); spongy fossa: absent (0); present (1) (Wilson 1998).
(somphospondlyus), with large internal cells 222. Anterior caudal neural spines, transverse
(1) (Powell 1986; Wilson 1998). breadth: approximately 50% of (0) or greater
207. Caudal vertebrae: more than 50 (0); 35 or than (1) anteroposterior length (Wilson
fewer (1) (Berman and McIntosh, 1978; 1998).
McIntosh 1990a, 1990b; Upchurch 1998). 223. Anterior caudal transverse processes: triangu-
208. Caudal transverse processes: disappear by lar, tapering distally (0); “winglike”, not taper-
caudal 15 (0); only present through caudal 10 ing distally (1) (Berman and McIntosh 1978;
(1) (Upchurch 1998; Wilson 1998). McIntosh 1990a; Calvo and Salgado 1995;
209. First caudal centrum, articular face shape: Upchurch 1995, 1998).
amphiplatyan (0); procoelous (1); opistho- 224. Anterior caudal transverse processes,
coelous (2); biconvex (3) (Powell 1986; Wilson diapophyseal laminae: absent (0); present (1)
1998). (Wilson 1998).
210. First caudal neural arch, fossa on lateral 225. Anterior caudal transverse processes, anterior
aspect of neural spine: absent (0); present (1) centrodiapophyseal lamina: single (0); bifid
(Wilson 1998). (1) (Wilson 1998).
226. Anterior caudal neural arches: anterior mar- RIBS AND CHEVRONS
gin of neural spine extends beyond anterior 240. Cervical ribs, length: much longer than cen-
margin of centrum (0); anterior margin of trum, overlapping as many as three subse-
neural spine does not extend beyond anterior quent vertebrae (0); shorter than centrum, no
margin of centrum (1) (Salgado et al. overlap (1) (McIntosh 1990a; Upchurch 1998;
1997:figs. 3–4). Wilson and Sereno 1998; Wedel et al.,
227. Anterior caudal neural spines, anterior pedi- 2000a, 2000b).
cle morphology: straight (0); anteriorly con- 241. Dorsal ribs, proximal pneumatocoels: absent
cave and curved (1). (0); present (1) (Wilson and Sereno 1998).
228. Anterior caudal vertebrae, prezygpophyses: 242. Anterior dorsal ribs, cross-sectional shape:
face anteriorly (0); face dorsally (1). subcircular (0); planklike, anteroposterior
229. Middle caudal centra, shape: cylindrical (0); breadth more than three times mediolateral
flat ventrally and laterally (1) (Wilson 1998). breadth (1) (Upchurch 1995, 1998; McIntosh
230. Middle caudal centra, ventral longitudinal et al. 1996).
hollow: absent (0); present (1) (Wilson 243. Middle and posterior caudal chevrons, shape
1998). of blade: “forked” with anterior and posterior
231. Middle and posterior caudal centra, anterior projections (0); single spine (1) (Upchurch
articular face shape: flat (0); procoelous (1); 1998; Wilson and Sereno 1998).
opisthocoelous (2) (Bonaparte 1986; Powell 244. Chevrons, proximal “crux” bridging superior
1986; Upchurch 1995, 1998; Wilson and margin of hemal canal: present (0); absent (1)
Sereno 1998). (Powell 1992; Calvo and Salgado 1995;
232. Middle caudal neural spine: angled Upchurch 1995, 1998; McIntosh 1996;
anterodorsally (0); vertical (1); angled pos- Wilson and Sereno 1998).
terodorsally (2) (Wilson 1998). 245. Chevron, hemal canal length: short, approxi-
233. Middle and posterior caudal neural spine: mately 25% (0), or long, at least 50% (1) of
tapering distally (0); rectangular (1); low and chevron length (Wilson 1998).
rounded (2); tapered anteriorly and posteri- 246. Chevrons, distribution: disappearing by 50%
orly (3) in lateral view (modified from Salgado length of tail (0); persisting throughout at
et al. 1997; Upchurch 1998; Wilson 1998). least 80% of tail (1) (Wilson 1998).
234. Midcaudal neural spines: anterodorsal edge 247. Posterior chevrons, distal contact: fused (0);
of neural spine lies posterior to anterior mar- unfused (open) (1) (Powell 1992; Upchurch
gin of postzygapophyses (0); anterodorsal 1995, 1998).
edge of neural spine lies anterior to or level 248. Middle and distal chevron distal blades, ante-
with anterior margin of postzygapophyses (1) riorly directed process: present (0); absent (1)
(Salgado et al. 1997:figs. 3, 4). (Upchurch 1998).
235. Neural arches of middle caudals: situated 249. Chevrons: proximal division V-shaped (0);
over middle of centrum (0); situated on cra- proximal division U-shaped (1)
nial half of centrum (1) (Calvo and Salgado
1995; Upchurch 1995, 1998).
236. Posterior caudal centra: cylindrical (0); APPENDICULAR SKELETON
dorsoventrally flattened, breadth at least twice PECTORAL GIRDLE
height (1) (Powell 1986, 1992; Wilson et al. 250. Scapular blade, relationships to coracoid: per-
1999; Sanz et al. 1999). pendicular to (0) or forming a 45 angle with
237. Distalmost caudal centra, articular face shape: (1) coracoid (Wilson 1998).
platycoelous (0); biconvex (1); procoelous (2) 251. Scapular, acromion process: not expanded
(Wilson 1998). (0); with rounded expansion on acromial side
238. Distalmost caudal centra, length-to-height (1) (modified from Upchurch 1998; Wilson
ratio: 4 (0); 5 (1) (Powell 1986; Upchurch and Sereno 1998).
1998; Wilson 1998). 252. Scapular blade, distal end: flared relative to
239. Distalmost biconvex caudal centra: 10 or proximal end (0); same width as proximal
fewer (0); more than 30 (1) (Wilson 1998). end (1); narrower than proximal end (2).
253. Scapular blade and acromion process, angle (1); prominent, extends well over one-half
of intersection: between 50 and 70 (0); total length of element (2) (modified from
90 (1); 45 (2). Upchurch 1998; Wilson and Sereno 1998).
254. Scapular spine: at right angle to long axis of 268. Humeral deltopectoral attachment: reduced
element (0); gently curved, not perpendicular to a low, rounded crest or ridge (0); promi-
to long axis of element (1). nent, with anterolateral extension (1) (Wilson
255. Scapula glenoid contribution vs. coracoid gle- and Sereno 1998).
noid contribution: 1:1 (0); coracoid has 269. Humeral deltopectoral crest: relatively narrow
greater contribution (1); scapula has greater throughout length (0); markedly expanded
contribution (2). distally (1) (Wilson, 1998).
256. Scapular glenoid, orientation: relatively flat 270. Humerus, head position: level with proximal
(0); strongly beveled medially (1) (Wilson and margin of deltopectoral crest (0); projects
Sereno 1998). above level of proximal margin of deltopec-
toral crest (1).
257. Scapular base, cross-sectional shape: flat or
rectangular (0); D-shaped (1) (Wilson and 271. Humeral proximolateral corner: rounded (0);
Sereno 1998). square (1) (Wilson 1998).
258. Coracoid, proximodistal length: less than (0), 272. Humerus, relative sizes of ulnar and radial
approximately twice (1), or subequal to (2) condyles: ulnar condyle transversely
length of scapular articulation of coracoid expanded (0); radial condyle transversely
(Wilson 1998). expanded (1) (Powell 1986, McIntosh 1990a).
259. Coracoid, anteroventral margin shape: 273. Humeral distal condyles, articular surface
rounded (0); rectangular (1) (Wilson 1998). shape: restricted to distal portion of humerus
(0); expanded onto anterior portion of humeral
260. Coracoid, scapular articular surface: antero-
shaft (1) (Powell 1986; Wilson 1998).
posteriorly elongate (0); mediolaterally
expanded (1). In Magyarosaurus, Malagasy 274. Humeral distal end, development of
Taxon B, and Saltasaurus, the scapular articu- condyles: flat (0); divided (1) (Powell 1986;
lar surface of the coracoid is mediolaterally Wilson 1998).
expanded. 275. Ulna, robusticity: robust (midshaft
261. Coracoid, infraglenoid lip: absent (0); present width:ulnar length 0.25) (0); gracile (mid-
(1) (Wilson 1998). shaft width:ulnar length 0.20) (1).
262. Position of coracoid foramen: deeply inset 276. Ulna, length-to-proximal breadth ratio:
into coracoid body (0); at margin of coracoid gracile, maximum width of proximal end is
body (1). 25% of ulnar length or less (0); stout, maxi-
mum width of proximal end is 33% of ulnar
263. Sternal plate, shape: oval, lacking distinct
length or more (1) (Wilson and Sereno 1998).
concavities (0); crescentic, with strongly con-
cave lateral borders (1) (Salgado et al. 1997; 277. Craniomedial process of proximal ulna, shape
Wilson and Sereno 1998). of articular surface: flat (0); strongly concave
(1) (Upchurch 1995:fig. 14, 1998).
264. Maximum length of sternal plate divided by
humerus length: 0.65 or less (0); 0.75 or 278. Ulnar proximal condylar processes: subequal
more (1) (McIntosh 1990a; Upchurch 1998). (0); unequal, anterior arm longer (1) (modi-
fied from Wilson and Sereno 1998).
279. Ulnar olecranon process: prominent, project-
FORELIMB
ing above proximal articulation (0); rudimen-
265. Humerus, robusticity: gracile (midshaft tary, level with proximal articulation (1)
width:humeral length 0.15) (0); robust (mid- (McIntosh 1990a; Wilson and Sereno 1998).
shaft width:humeral length 0.25) (1). 280. Ulna, distal articular surface: triangular with
266. Humeral diaphysis, shape of lateral margin: anteromedial apex (0); circular (1).
straight (0); concave (1). 281. Radius: robust (midshaft width:radial length
267. Humerus, deltopectoral crest: extends less 0.25) (0); gracile (midshaft width:radial
than one-half total length of element (0); length 0.15) (1) (modified from McIntosh
extends about one-half total length of element 1990a; Upchurch 1995, 1998).
282. Radius: proximal and distal ends equally ex- 296. Metacarpal I, divided trochlea on distal
panded (0); proximal end more transversely condyle: present (0); absent (1) (modified
expanded (1); distal end more transversely from Wilson and Sereno 1998).
expanded (2). 297. Metacarpal I distal condyle, transverse axis
283. Radius, well-defined interosseous ridge: orientation: beveled approximately 20º proxi-
absent (0); present (1). modistally (0) or flat (1) with respect to axis of
284. Radius, proximal articular surface: shaft (modified from Wilson and Sereno
rounded/subrectangular, with subequal trans- 1998).
verse and anteroposterior dimensions (0); tri- 298. Metacarpal II: shaft oval/rounded in cross sec-
angular, with expanded anteroposterior tion (0); shaft triangular in cross section (1).
dimensions (1); transversely expanded (2). 299. Metacarpal II, distal end shape: oval (0); sub-
285. Radius, proximal end: slender, maximum rectangular (1).
width of proximal end 25% of radius length 300. Metacarpal III: longest metacarpal (0);
or less (0); robust, maximum width of proxi- shorter than II or IV (1).
mal end 33% of radius length or more (1)
(McIntosh 1990a; Upchurch 1995, 1998). 301. Metacarpal III, shaft cross section: triangular
(0); oval to circular (1).
286. Radius, distal articular surface: oval, with
subequal anteroposterior and transverse 302. Metacarpal III, proximal end: triangular, with
dimensions (0); transversely expanded, ellip- posterior apex (0); triangular, with medially
tical (1); anteroposteriorly expanded (2). concave edge (1); triangular, with posterior
base (2).
287. Radial distal condyle: subrectangular, flattened
posteriorly and articulating in front of ulna 303. Metacarpal IV, shape in anterior view: proxi-
(0); circular (1) (Wilson and Sereno 1998). mal and distal ends flared (0); proximal and
distal ends not flared (1).
288. Radius, distal breadth: slightly larger than (0)
or approximately twice (1) midshaft breadth 304. Metacarpal IV, proximal end: anteroposteri-
(Wilson 1998). orly elongate oval (0); triangular (1).
289. Distal radius, transverse axis orientation: per- 305. Metacarpal IV, distal surface: rounded (0); tri-
pendicular to (0) or beveled approximately angular (1).
20° proximolaterally (1) relative to long axis of 306. Metacarpal V: gracile (ratio midshaft
shaft (Curry Rogers and Forster 2001). width:metacarpal length 0.10) (0); robust
290. Humerus-to-femur ratio: 0.6 (0); 0.7–0.8 (ratio midshaft width:metacarpal length
(1); 0.9 (2) (modified from McIntosh 0.10) (1).
1990a; Upchurch 1994, 1995, 1998; Wilson 307. Metacarpal V: proximal and distal ends sube-
and Sereno 1998). qually expanded (0); proximal end more
expanded (1).
MANUS 308. Phalangeal number: 2-3-4-3-2 or more (0);
reduced, 2-2-2-2-2 or fewer (1); ossified pha-
291. Longest metacarpal-to-radius ratio: 0.3
langes absent (2) (Wilson and Sereno 1998).
(0);0.45 (1) (modified from Salgado et al.
1997; Wilson and Sereno. 1998). 309. Manual phalanx I.1: rectangular (0); wedge-
shaped (1) (Wilson 1998).
292. Metacarpal I: shorter than metacarpals II
and III (0); longer than metacarpal III and 310. Manual nonungual phalanges: broader trans-
subequal to or longer than metacarpal II (1) versely than long proximodistally (0); longer
(Upchurch 1998). proximodistally than broad transversely (1)
(Wilson and Sereno, 1998).
293. Metacarpal I: shorter than (0) or longer than
(1) metacarpal IV (Wilson and Sereno 1998).
294. Metacarpal I, proximal surface shape: broad PELVIC GIRDLE
quadrangle (0); oval with posterior process 311. Pelvis, shape: narrow, ilia longer anteroposte-
(1); triangular (2). riorly than distance separating preacetabular
295. Metacarpal I, proximal and distal ends: processes (0); broad, distance between preac-
slightly expanded relative to midshaft (0); etabular processes exceeds anteroposterior
widely flared relative to midshaft (1). length of ilia (1) (Wilson and Sereno 1998).
312. Ilium, shape of preacetabular process in lat- 325. Ischial blade: equal to or longer than (0) or
eral view: triangular and tapering to a point much shorter than (1) pubic blade (Wilson
(0); broad, with rounded anterior tip (1) and Sereno 1998).
(Calvo and Salgado 1995; Upchurch 1998). 326. Ischial blade, emargination distal to pubic
313. Ilium, position of highest point of iliac blade: peduncle: present (0); absent (1) (Salgado et
central or posterior to acetabulum (0); cen- al. 1997).
tered over pubic peduncle (1) (Salgado et al. 327. Ischial distal shaft: triangular, depth of ischial
1997: fig.1). shaft increases medially (0); bladelike, medial
314. Ilium, pubic peduncle contribution to acetab- and lateral depths subequal (1) (Wilson and
ulum: shares acetabular border equally with Sereno 1998).
ischial peduncle (0); comprises less than one- 328. Ischial distal shafts, cross-sectional shape of
half acetabular border (1); comprises over articulated ischia: V-shaped, forming an angle
one-half acetabular border (2). of nearly 50º with each other (0); flat, nearly
315. Ilium, pubic peduncle: projects at acute angle parallel (1) (McIntosh 1990a; Upchurch
relative to sacral axis (0); projects at right 1998; Wilson and Sereno 1998).
angle relative to sacral axis (1) (Salgado et al.
329. Ischium, width across ischial shaft at
1997: figs. 8, 9).
midlength divided by length of ischium:
316. Ilium, ischial peduncle size: low, rounded (0); 0.15 (0); 0.20 (1) (Jacobs et al. 1993;
large, prominent (1) (McIntosh 1990a; Upchurch 1998).
330. Ischium, distal end: elongated and narrow
317. Ilium, preacetabular process (when resting (0); short, reduced, and broad, does not
on pubic and ischial peduncles): projects extend beyond body (1).
anterolateraly (0) or perpendicularly (1) to
331. Ischium, length of pubic peduncle:
body axis (Salgado et al., 1997).
dorsoventrally short (distance from upper
318. Ilium, preacetabular process, shape (when corner of ischium’s pubic blade to posterior
resting on pubic and ischial peduncles): border of iliac peduncle is subequal to or
pointed, arching ventrally (0); semicircular, longer than pubic articular surface) (0);
anterodorsally oriented (1) (modified from dorsoventrally extended (distance from
Salgado et al. 1997; Wilson and Sereno upper corner of ischium’s pubic blade to
1998). posterior border iliac peduncle is shorter
319. Pubis, ambiens process development: small, than pubic articular surface) (1) (Salgado et
confluent with (0), or prominent, projecting al. 1997: fig. 5).
anteriorly from (1), anterior margin of pubis 332. Ischium, iliac peduncle: distinctive and well
(Wilson and Sereno 1998). separated from body of ischium (0); low
320. Pubis, relationship of acetabular and ischial and rounded, not separated from ischial
surfaces: perpendicular to one another (0); body (1).
meet at obtuse angle (1).
321. Pubis, ratio of length of acetabular surface to
HINDLIMB
length of ischial surface: ischial surface
longer (0); acetabular surface longer (1); two 333. Femur, robusticity:robust, midshaft
surfaces subequal (2). width:femoral length 0.20 (0); gracile, mid-
shaft width:femoral length 0.10 (1).
322. Pubis, expansion of proximal and distal ends:
unexpanded (0); only proximal end widened 334. Femur, location of head: dorsal to greater
relative to shaft (1); distal end widened rela- trochanter (0); level with greater trochanter (1).
tive to midshaft (2); both proximal and distal 335. Femoral shaft, lateral margin: straight (0);
ends expanded (3). proximal one-third deflected medially (1)
323. Puboischial contact: approximately one-third (Wilson and Sereno 1998; Wilson and
total length (0), one-half total length (1), or Carrano 1999).
entire length of pubis (2) (Wilson and Sereno 336. Femoral midshaft transverse diameter: sube-
1998). qual to (0), 125%–150% of (1), or at least 185%
324. Ischium:pubis length ratio: 0.90 (0); 0.75 of (2) anteroposterior diameter (Wilson
(1) (Calvo and Salgado 1995; Upchurch 1998). 1998).
337. Femoral distal condyles, relative transverse 352. Calcaneum: present (0); ossified calcaneum
breadth: subequal to one another (0); medial absent (1) (McIntosh, 1990a; Upchurch, 1998).
condyle much broader than lateral (1) (Wilson
1998). PES
338. Femoral distal condyles, transverse axis orien- 353. Metatarsal I, length: shortest metatarsal (0);
tation: perpendicular or slightly beveled dor- metatarsal V shorter than metatarsal I (1).
solaterally (0), or beveled dorsomedially 354. Metatarsal I, anterior extension on proximal
approximately 10 (1), relative to femoral shaft condyle: absent (0); present (1).
(Wilson and Carrano 1999).
355. Metatarsal I distal condyle, transverse axis ori-
339. Femoral distal condyles, articular surface: entation: angled dorsomedially to (0) or per-
restricted to distal portion of femur (0); pendicular to (1) axis of shaft (Wilson 1998).
expanded onto anterior portion of femoral
356. Metatarsal I distal condyle, posterolateral pro-
shaft (1) (Wilson and Carrano 1999).
jection: present (0); absent (1) (Wilson 1998).
340. Tibia, proximal articular condyle shape: medi-
357. Metatarsal I and V proximal condyles: subequal
olaterally narrow, long axis anteroposterior
to (0) or smaller than (1) those of metatarsals II
(0); expanded transversely, condyle subcircu-
and IV (Wilson and Sereno 1998).
lar (1); subequal expansions of anterior and
posterior dimensions (2) (Wilson and Sereno 358. Metatarsal III: proximal ends and distal ends
1998). expanded and broad (0); proximal and distal
ends not expanded (1).
341. Tibia, cnemial crest: projects anteriorly (0);
projects laterally (1) (Wilson and Sereno 359. Metatarsals III and IV, minimum transverse
1998). shaft diameters: less than 65% of (0) or sube-
qual to (1) those of metatarsals I or II (Wilson
342. Tibia, shape of distal end: oval (0); triangular (1).
and Sereno 1998).
343. Tibia, distal end: transversely expanded and
360. Metatarsal V, shaft in cross section: circular
anteroposteriorly compressed (0); transverse
(0); triangular (1).
and anteroposterior dimensions subequal (1)
(Salgado et al. 1997). 361. Metatarsal V, proximal articular surface: sin-
gle, flat surface (0); surface divided into two
344. Tibia, distal breadth: approximately 125% of
portions by sharp angle (1).
(0), more than 200% 0f (1), or subequal to
(2) midshaft breadth. 362. Pedal digit I ungual, length relative to pedal
digit II ungual: 25% larger than (0) or subequal
345. Fibula, robusticity: robust, midshaft
to (1) that of digit II (Wilson and Sereno 1998).
width:fibular length 0.25 (0); gracile, mid-
shaft width:fibular length 0.15 (1).
DERMAL
346. Fibula, proximal articular surface shape: flat
363. Osteoderms: absent (0); present (1) (Depéret
to slightly convex (0); steeply angled (1).
1896a, 1896b; Wilson and Sereno 1998).
347. Fibula, lateral trochanter: absent (0); present
364. Osteoderm morphology: absent (0); rounded
(1) (Wilson and Sereno 1998).
and platelike with ventral keels (1); triangular,
348. Fibula, muscle scar on lateral surface at with disklike external surface and flat ventral
midlength: oval (0); elongate ridge running surface, lacking ventral keel (2) (Powell 1986;
subparallel to long axis of shaft (1) (Powell Csiki 1999).
1992; Upchurch 1998).
349. Fibular distal condyle, relative width: subequal
APPENDIX 2.3. LIST OF
to width of shaft (0); expanded transversely,
greater than midshaft width (1) (Wilson SYNAPOMORPHIES FOR THE TOTAL-
1998). EVIDENCE PHYLOGENETIC ANALYSIS
350. Astragalar posterior fossa, shape: divided by WITH NO TAXA REMOVED.
vertical crest (0); undivided (1) (Wilson and This appendix lists character support for each of the
Sereno 1998). nodes documented in figures 2.10 and 2.110. The distri-
351. Astragalus, transverse length: 50% more than bution of apomorphies is based on the Accelerated
(0) or subequal to (1) proximodistal height Transformation option in PAUP* (Swofford 1999).
(Wilson 1998) Character numbers listed below correspond to characters
numbered in the character–taxon matrix (appendix 2.1) NODE E: RAPETOSAURUS CLADE
and in the character list (appendix 2.2). 161, postspinal lamina restricted to proximal neural
spine in anterior dorsals; 169, middle and posterior
dorsal neural spines posteriorly directed; 171, middle
STRICT CONSENSUS TREE
and posterior dorsal neural spines distally flared; 175,
TITANOSAURIFORMES transverse processes of posterior dorsals lie vertically
12, absence of preantorbital fenestra; 36, presence of above parapophysis; 181, middle and posterodorsal
parietal crest; 83, platelike contact between maxilla neural arches with divided spin0diaphyseal lamina;
and platine; 94, surangular-angular contact dorsally 257, scapular basis flat or rectangular; 280, ulnar dis-
concave; 113, presence of 13 cervical vertebrae; 123, tal articular surface triangular with anteromedial
single anterior cervical neural spines; 246, chevron apex; 282; distal end of radius transversely expanded;
blades curve backward and down. 321, pubis acetabular surface longer than ischial
No unambiguous support for Somphospondylii. articular surface.
ADAM’S CONSENSUS TREE NODE F: UNNAMED

TITANOSAURIFORMES 111, shallow presacral pneumatic fossa; 163, presence
12, absence of preantorbital fenestra; 34, frontal con- of median infrapostzygapophyseal lamina in anterior
tributes to boundary of supratemporal fenestra; 36, dorsals; 233, low and rounded midcaudal neural
parietal with wide exposure of posterior crest; 83, spines; 277, craniomedial process of ulna flat proxi-
palatine with plate-shaped maxillary contact; 94, con- mally; 281, presence of gracile radius; 284, radius,
cave surangular–dentary with respect to long axis of proximal articular surface rounded with subequal
mandible; 123, undivided anterior cervical neural anteroposterior and transverse dimensions; 364,
spines; 125, anterior neural spines slightly posteriorly osteoderms triangular and lack ventral keel.
inclined; 137, poorly developed middle and posterior
cervical neural arch pneumatic fossae; 150, single ante- NODE G: SALTASAURINAE
rior dorsal neural spines; 151, anterior dorsal neural 1, absence of short, deep snout; 16, external nares
spines posteriorly inclined; 173, middle and posterior retracted to level of orbits; 64, basipterygoid
dorsal neural arches lack fossa between neural canal processes diverge at 30; 70, basal tubera sheetlike;
and prezygapophyses; 222, anterior caudal neural 108, SI vales for teeth exceed 5.0.
spines transversely thin; 226, anterior border of neu-
ral spine extends anterior to centrum in proximal NODE H: UNNAMED
caudals; 235, mid–caudal neural arches anteriorly 144, posterior dorsal centra length:height ratios
positioned; 248, chevron blades curve backward and exceed 1.0; 218, presence of prominent lateral lamina
downward; 335, proximal one-third of femoral shaft in anterior caudal neural arches.
deflected medially.
NODE I: OPISTHOCOELICAUDIINAE
NODE A: TITANOSAURIA No unambiguous support calculated in Adams con-
Unambiguous support not calculated in Adam’s consensus tree.
sensus.
NODE J: SALTASAURIDAE
NODE B: LITHOSTROTIA No unambiguous support calculated in Adams con-
226, anterior caudal neural spine extends beyond sensus tree.
anterior margin of centrum; 276, stout ulna; 285,
proximal end of radius is robust; 289, distal end of
APPENDIX 2.4. LIST OF
radius is beveled proximolaterally.
SYNAPOMORPHIES FOR THE TOTAL-
EVIDENCE PHYLOGENETIC ANALYSIS
NODE C: UNNAMED
268, prominent humeral deltopectoral attachment; WITH SOME TAXA REMOVED
281, presence of robust radius. This appendix lists character support for each of the
nodes documented in figures 2.12 and 2.13. Taxa less
NODE D: UNNAMED than 15% completely coded have been removed from the
165, absence of ventral concavity in middle and pos- analysis. The distribution of apomorphies is based on the
terior dorsals; 174, posterior dorsal neural spine Accelerated Transformation option in PAUP*.
expands transversely for 75% of length, then tapers. Character numbers listed below correspond to characters
numbered in the character–taxon matrix (appendix 2.1) elongate and oval; 337, subequal breadth of femoral
and in the character list (appendix 2.2). (*) Indicates distal condyles.
synapomorphies also present in the strict consensus tree
of the reduced data set. NODE E: SALTASAURINAE
236, posterior caudal centra dorsoventrally flattened;
NODE A: SOMPHOSPONDYLII 237, distalmost caudal centra procoelous.
*5, presence of caudolateral process of premaxilla;
*8, maxillary flanges do not contact one another on NODE F: UNNAMED
midline; *14, subnarial foramen faces laterally; 84, 123, anterior cervical neural spines proximally
dentary, depth of anterior end of ramus slightly less divided and distally fused; 238, distalmost caudal cen-
than at midlength; *101, teeth oriented perpendicu- tra extremely elongate.
lar relative to jaw margin; *110, spongy presacral
bone texture; 114, rudimentary cervical lamination;
NODE G: RAPETOSAURUS CLADE
117, elongate anterior cervicals; 147, pleurocoels in
167, middle and posterior dorsal neural spines com-
dorsal centra deep but simple pits; 154, dorsal trans-
prise less than 50% of total vertebral height; 169,
verse processes directed dorsolaterally; *160,
middle and posterior dorsal neural spines posteriorly
prespinal lamina present along entire length of ante-
inclined; 171, middle and posterior dorsal neural
rior dorsal neural spine; *162, absence of
spins distally flared with triangular processes; 175,
hyposphene-hypantra in anterior dorsals; 187, ante-
posterior dorsal transverse processes vertical to para-
rior centrodiapophyseal lamina in middle and poste-
pophysis; 177, presence of triangular lateral process
rior dorsals anteriorly inclined; *256, scapular gle-
on dorsal neural spines; 233, rectangular midposte-
noid beveled medially; 257, scapular base flat or rec-
rior caudal neural spines.
tangular; 320, pubis, acetabular and ischial surfaces
meet at obtuse angle.
NODE H: UNNAMED
230, midcaudal centra lack midline, ventral hollow.
NODE B: UNNAMED
182, presence of prespinal lamina in middle and pos-
terior dorsal neural arches; 183, presence of antero- NODE I: UNNAMED
posteriorly expanded postspinal lamina in middle 120, anterior cervical pleurocoels shallow and sim-
and posterior dorsals; 193, single infradiapophyseal ple but divided by single median ridge; 154, dorsal
lamina in midposterior dorsals; 230, presence of transverse processes directed dorsolaterally (rever-
midventral hollow in middle caudals; 259, coracoid sal); 155, anterior dorsal transverse processes meet
with rectangular anteroventral margin. neural spine at an angle of at least 90°; 221, post-
spinal fossa in anterior caudal neural spines; 266,
straight lateral margin of humerus; 341, cnemial
NODE C: SALTASAURIDAE
crest curves laterally.
149, dorsal boundary of neural canal in anterior dor-
sals either interprezygapophyseal lamina or a ventral
extension of the prespinal fossa; 222, anterior caudal NODE J: UNNAMED
neural spines transversely narrow; 262, coracoid *1, absence of short, deep snout; *16, external nares
foramen deeply inset into coracoid body; 341, cnemial retracted to between orbits; *17, external nares face
crest curves anteriorly. dorsally or rostrodorsally; 90, angle between
mandibular symphysis and long axis 90°; *108,
NODE D: LITHOSTROTIA teeth SI that exceeds 5.0.
111, shallow, lateral presacral pneumatic fossae;
204, presence of proximal excavation in sacral ribs; NODE K: OPISTHOCOELICAUDIINAE
251, acromial edge of scapula lacks expansion; 304, No unambiguous support for this clade.
proximal end of metacarpal IV anteroposteriorly
three
Phylogenetic and Taxic Perspectives

on Sauropod Diversity
Paul Upchurch and Paul M. Barrett
factors, the basic pattern of sauropod diversity

T he sauropoda represent one
of the most diverse and geographically
widespread dinosaurian radiations. After their
must be established accurately.
No summary of sauropod evolutionary his-
origin during the Late Triassic, sauropods in- tory and palaeobiology is complete without a
creased rapidly in diversity, and acquired consideration of diversity (e.g., how the number
a nearly global distribution by the Middle of sauropod taxa fluctuated through time). Yet
Jurassic. Diversity appears to peak in the Late this aspect of sauropod history has been rela-
Jurassic, at which point sauropods were very tively neglected. Most studies have been based
abundant and dominated many terrestrial envi- on direct observation of numbers of sauropod
ronments. Although sauropod diversity appar- species and/or genera in the fossil record.
ently declined very rapidly in the Early Creta- Although this approach has the potential to pro-
ceous, several lineages remained as a significant vide an accurate measure of diversity, it can be
component in many faunas, and the titanosaurs seriously affected by non-random sampling
underwent a major radiation during the mid- or biases. In principle, such problems can be par-
Late Cretaceous. tially corrected when the phylogenetic relation-
These changes in diversity took place ships of taxa are taken into account (Smith
against a backdrop of profound geological and 1994), but to date only one study (Weishampel
biotic events. The gradual fragmentation of and Jianu 2000) has applied this method to
Pangaea affected physical environments by sauropod diversity. This neglect is not surpris-
altering sea levels and climatic regimes, which ing given the confused nature of much of
in turn could have had an impact on sauropod sauropod species- and genus-level systematics
diversity. In addition, evolutionary interactions and the lack (until recently) of detailed cladistic
with predators, competing large-bodied herbi- analyses for this group. Fortunately, several
vores, and plants may also have affected sauro- recent studies (Upchurch 1995, 1998, 1999;
pod diversity. Before we can attempt to tease Wilson and Sereno 1998; Curry Rogers and
apart the effects of these physical and biological Forster 2001; Wilson 2002; Upchurch et al.,
104
FIGURE 3.1. Taxic diversity estimates for sauropod genera and species obtained by Hunt et al. (1994). AAL, Aalenian; ALB,
Albian; APT, Aptian; BAJ, Bajocian; BAR, Barremian; BAT, Bathonian; BER, Berriasian; CAL, Callovian; CAM, Campanian;
CAR, Carnian; CEN, Cenomanian; CON, Coniacian; HAU, Hauterivian; HET, Hettangian; KIM, Kimmeridgian; MAA,
Maastrichtian; NOR, Norian; OXF, Oxfordian; PLI, Pliensbachian; RHA, Rhaetian; SAN, Santonian; SIN, Sinemurian; TTH,
Tithonian; TOA, Toarcian; TUR, Turonian; VAL, Valanginian.
2004) have gone a long way toward removing estimation, in which diversity is measured by
the obstacles to detailed analysis of sauropod counting the number of taxa present in the fossil
diversification patterns. Thus, the main aims record at each point in time. The resulting “con-
of the current chapter are (1) to present and sensus” is that sauropods gradually increased in
compare alternative estimates of sauropod lin- diversity throughout the Jurassic, reaching a peak
eage diversity and (2) to evaluate the implica- in the Kimmeridgian–Tithonian. At the Jurassic–
tions of these diversity patterns for our under- Cretaceous boundary, however, sauropods seem
standing of sauropod evolutionary history and to undergo a major extinction event: although
sampling biases. The reader is also referred to many of the familial lineages survive into the
the following companion chapter by Barrett Early Cretaceous, the overall level of genus or
and Upchurch, which examines the relative species diversity is drastically reduced (Bakker
diversity of different sauropodomorph clades 1978). Sauropods maintained a lowered level of
and considers the possible relevance of morpho- diversity until the end of the Cretaceous, at which
logical characters related to feeding strategy. point the number of genera increased as a result
of the titanosaur radiation. In many of these stud-
ies, fluctuations in the numbers of sauropod gen-
PREVIOUS WORK
era or species are accepted as genuine changes in
Few studies have examined sauropod diversity diversity, even though it is possible that such vari-
in detail. This topic has, however, been treated ations actually reflect changes in the quality of the
in a superficial fashion by several studies that fossil record.
have attempted either an “overview” of dinosaur The larger data sets employed by Hunt et al.
evolutionary history (Bakker 1977, 1978; Sereno (1994) and Barrett and Willis (2001) enabled
1997, 1999) or an assessment of the factors them to detect a somewhat more complex pattern
controlling dinosaur diversity (Horner 1983; of peaks and troughs in sauropod diversity (fig.
Weishampel and Horner 1987; Haubold 1990; 3.1). In particular, these studies noted the pres-
Barrett and Willis 2001). All of these studies, as ence of peaks in diversity during the Middle
well as the detailed treatment of sauropod Jurassic (Bajocian), Late Jurassic (Kimmeridgian),
diversity presented by Hunt et al. (1994), early Early Cretaceous (Valanginian–Barremian),
are based on the taxic approach to diversity late Early Cretaceous (Albian), and Late Cretaceous
P H Y L O G E N E T I C A N D TA X I C P E R S P E C T I V E S O N S AU R O P O D D I V E R S I T Y 105
(Campanian–Maastrichtian). The results of these noticeably.” To this, it might be added that the
analyses agree with those for the Dinosauria as a impact of area fragmentation may be more
whole obtained by Haubold (1990). These fluc- complex than often assumed. Although the for-
tuations in diversity appear to correlate with mation of geographic barriers can promote
changes in sea level (Haubold 1990; Hunt et al., increased diversity through allopatric specia-
1994) as determined by Haq et al. (1987). tion, it may also result in extinction events as
Haubold (1990) and Hunt et al. (1994) pro- the size of certain habitats decreases. Neverthe-
posed a non-biotic explanation for this correla- less, if biological factors are responsible for the
tion: during periods of marine transgression, correlation between sea level and sauropod
the remains of terrestrial organisms are more diversity, then fluctuations in the latter can be
likely to reach aquatic environments and are, legitimately regarded as real evolutionary
therefore, more likely to be preserved; con- events rather than artifacts caused by uneven
versely, regressions are predicted to produce a preservation rates.
decrease in preservation. Thus, peaks and Lockley et al. (1994) presented the only
troughs in apparent diversity could reflect a attempt to use a large data set of sauropod
taphonomic effect caused by the impact of sea trackways to assess changes in diversity and
level on preservation potential for terrestrial abundance. Sauropods appear to be most
taxa. This view receives some additional support abundant in the Late Jurassic and Early
from Hunt and coworkers’ (1994) observation Cretaceous. Narrow-gauge trackways (i.e.,
that the peaks in sauropod diversity are posi- non-titanosaur tracks) are most common in
tively correlated with the number of sauropod- the Jurassic, whereas after the Kimmeridgian
bearing formations for each point in time. If the wide-gauge trackways of titanosaurs dom-
this “taphonomic bias” hypothesis is correct, inate Cretaceous biomes (Wilson and Carrano
then important features of sauropod diversity, 1999). Although the trackway data are in
such as the Jurassic–Cretaceous boundary broad agreement with diversity estimates
extinction, could be artifacts rather than real evo- based on body fossils, the usefulness of the
lutionary events. former is severely limited by our inability to
Other authors have noted the same correla- identify the track-makers to lower taxonomic
tion between diversity and sea level but have levels. Nevertheless, trackways can make
developed biological mechanisms to explain some important contributions to our under-
this phenomenon (Bakker 1977; Horner 1983; standing of sauropod diversity, such as the
Weishampel and Horner 1987). For example, recent discovery of wide-gauge titanosaur
during periods of marine transgression, low- trackways in the Middle Jurassic of England,
land areas are likely to become constricted in which extend the temporal range of this clade
size and potentially separated from each other. back by approximately 12 Ma (Day et al. 2002,
This promotes allopatric speciation and may 2004).
therefore result in a peak in diversity. Con- Sereno (1997, 1999) presented clado-
versely, marine regression removes the barriers grams for the Dinosauria that were calibrated
between separate lowland areas, allowing the against time and used these to assess changes
mixing of previously isolated biotas and poten- in lineage diversity. Unfortunately, Sauropoda
tially resulting in extinction events. Haubold was represented by only a small data
(1990:101) considered these mechanisms but set (9 and 11 lineages in Sereno [1997] and
rejected them on the basis that “the ‘habitat bot- Sereno [1999], respectively), so the details
tlenecks’ (Weishampel and Horner, 1987) in of their diversity are poorly resolved.
transgressive phases, or extinctions due to regres- Nevertheless, Sereno noted that sauropod
sions (Bakker, 1977), reduce the fossil record diversity peaked in the Late Jurassic and
regionally, but do not diminish the diversity agreed with Upchurch (1995) that this
106 P H Y L O G E N E T I C A N D TA X I C P E R S P E C T I V E S O N S AU R O P O D D I V E R S I T Y
suggested a Middle Jurassic radiation for the METHODS AND MATERIALS
major neosauropod clades.
TAXIC VERSUS PHYLOGENETIC DIVERSITY
Weishampel and Jianu (2000) presented the
ESTIMATES
first detailed analysis of sauropodomorph diver-
sity that utilized both observed stratigraphic There are many approaches to the estimation of
ranges and phylogenetic relationships. The lineage diversity through time (Smith 1994),
sauropodomorph cladogram employed by though they generally fall into one of two broad
Weishampel and Jianu (2000) was a compo- categories—taxic (or “probabilistic”) and phylo-
site of the trees produced by Galton (1990; genetic. Taxic methods assess diversity by count-
prosauropods) and Upchurch (1995, 1998; ing the number of taxa observed at each point
sauropods). This composite tree was enlarged fur- in time. Ideally, these taxa should be species,
ther by (1) replacing genera by monophyletic, but but it is often necessary to use higher taxo-
polytomous, clusters of species and (2) adding nomic ranks as a proxy for species level diver-
those taxa not considered by Galton (1990) or sity. The main reason for this is that the fossil
Upchurch (1995, 1998), through identification of record is inevitably incomplete: as a result, the
their probable relationships on the basis of representation of lower-level taxa, especially
synapomorphies (i.e., the taxa were added directly species, will be particularly poor. Sampling
to the tree, rather than their relationships being higher-level taxa should help to compensate for
assessed through incorporation in a data matrix such sampling biases (Foote 1996). This is par-
and application of cladistic analysis). Thus, the ticularly noticeable when dealing with
Weishampel and Jianu (2000) sauropodomorph “Lazarus” taxa (i.e., groups that appear, appar-
cladogram is virtually comprehensive in terms of ently go extinct, and then reappear later). Thus,
its species sampling, but this is achieved at the the use of higher taxa, rather than species,
cost of poorer resolution and assumptions regard- allows gaps in stratigraphic ranges to be identi-
ing relationships (see below). The main focus of fied as sampling errors, but this benefit is
Weishampel and Jianu’s study was a comparison achieved at the cost of poorer pattern resolution
of the lineage diversity of Sauropodomorpha and (Smith 1994).
Ornithischia to investigate possible evolutionary A variety of more sophisticated probabilis-
interactions between these groups and with tic approaches has been developed to amelio-
plants. As a result, Weishampel and Jianu rate sampling errors. For example, Alroy
(2000) did not discuss changes in sauropod (2000) introduced the “appearance event ordi-
diversity in detail, though some important points nation” method, which, like many other simi-
emerge from their study. First, Weishampel and lar approaches, uses a large database to infer
Jianu’s diversity curves for Sauropoda show very confidence limits on the stratigraphic ranges
little correspondence between the taxic and the of taxa. Such techniques, however, cannot deal
phylogenetic diversity estimates (hereafter with “singletons” (i.e., taxa known from only a
referred to as TDEs and PDEs, respectively). single point in time). Given that we are dealing
This could be interpreted as indicating that with a relatively small data set for sauropod
sauropods have a particularly poor fossil record. genera (chapter 4, appendix 4.2), and 70%
Second, the PDEs indicate the presence of a of these taxa are only known from single
much higher peak in diversity during the occurrences (or several specimens from a sin-
Middle Jurassic (approximately 75% of that gle horizon), we have not attempted to apply
observed for the Late Jurassic) than had been this technique here.
estimated previously. These results, and other There are obvious difficulties with taxic
aspects of the Weishampel and Jianu study, are approaches, especially when large data sets are
compared with our analyses under “Results and not available. For example, such approaches
Discussion”, below. cannot identify gaps in the fossil record that
FIGURE 3.2. A schematic dia-
gram illustrating the differences
between the taxic and the phylo-
genetic approaches to diversity
estimation. (A) Stratigraphic
ranges for the taxa A and B. (B)
Taxic diversity estimate through
time. (C) Stratigraphic ranges
and inferred ghost range for
taxa A and B. (D) Diversity esti-
mate based on both observed
and inferred ranges.
occur prior to a taxon’s first stratigraphic time period. This simple example also illus-
appearance (Smith 1994). This and other con- trates how the TDEs and PDEs may differ even
cerns prompted the development of the PDE though based on the same stratigraphic range
(Fisher 1982; Paul 1982; Cripps 1991; Norell data. In figure 3.2, the TDE suggests that only
1992; Norell and Novacek 1992; Smith 1994), one taxon (B) is present during time t1, so that
in which both observed stratigraphic ranges the appearance of A in t2 would be interpreted
and inferred “ghost” ranges are used to deter- as an increase in diversity. The PDE, in con-
mine diversity by counting the number of line- trast, suggests that diversity has remained con-
ages that pass through each designated time stant during t0–t2.
period (fig. 3.2). A ghost range is defined as the The apparent “superiority” of PDEs over
minimal portion of a lineage’s stratigraphic TDEs has led several authors to suggest that the
range that is missing but that must have existed former technique should replace the latter
because of the stratigraphic range of its sister- (Novacek and Norell 1982; Norell 1992; Smith
taxon. Thus, consider two taxa, or evolutionary 1994; Rieppel 1997). Some recent studies, how-
lineages, A and B (fig. 3.2). A and B are sister- ever, have demonstrated a number of serious
taxa according to a cladistic analysis, and nei- problems with the PDE approach (Foote 1996;
ther can be the ancestor of the other because Wagner 2000a, 2000b; Wagner and Sidor
each possess autapomorphies. A and B have 2000).
different stratigraphic ranges: in particular, First, practical constraints mean that many
the first appearance of A at time t2 is later than cladistic analyses do not sample all of the taxa
the first appearance of B at time t0. Since A and within a particular clade. This could create
B must have descended from a common ances- errors in PDEs in at least two ways. First, the
tor that existed at or prior to the first appearance taxa included or excluded from the analysis may
of B, there must be a lineage linking this ances- not be randomly distributed in the stratigraphic
tor to the first appearance of A. Thus, in figure record. The earliest taxa may be particularly
3.2, a ghost range is inferred for A, extending fragmentary and workers may therefore choose
from time t0 to time t2. The PDE can then be not to include them. Alternatively, workers may
determined by counting the total number of lin- focus on early “basal” forms because of an a pri-
eages (observed and inferred) for each designated ori belief that these are particularly important
in determining phylogenetic relationships. Any is because the PDE approach produces a back-
systematic bias regarding the temporal distri- ward “smearing” of origination times that tends
butions of taxa could lead to incorrect infer- to diminish the significance of mass extinction
ences of diversity change through time. Second, events. Thus, if a major decrease in diversity is
omission of taxa from a cladistic analysis could suggested by a PDE method, it is legitimate to
increase the probability of errors in the topology regard this as evidence for a genuine evolution-
of the most parsimonious trees: this in turn ary event (i.e., a mass extinction) rather than an
could produce systematic errors in PDEs (see artifact caused by a patchy fossil record.
below). In short, both the TDE and the PDE have
Second, Foote (1996) and Wagner (2000a) advantages and disadvantages. We therefore
have noted that PDEs introduce an asymmetri- employ both approaches in the pluralistic spirit
cal “correction” into diversity estimates: that is, advocated by Foote (1996).
they potentially extend origination times back-
ward but do not offer a corresponding correc- PRESERVATION/SAMPLING BIASES
tion that extends extinction times forward. One of the fundamental difficulties associated
Sampling theory suggests that corrections with the reconstruction of diversity patterns
should be made nearly equally to both origina- concerns the interpretation of “apparent” diver-
tion and extinction times (Foote 1996:5), and it sity curves. Whichever method is applied, we
therefore seems probable that PDEs introduce a ultimately obtain a graph that suggests fluctua-
systematic bias into diversity pattern recon- tions in diversity through time, but we can
structions. never be sure whether these apparent changes
Third, Wagner (2000a) used simulations to in diversity represent “real” evolutionary events
study the impact of various factors (e.g., specia- or differences in preservation or sampling
tion mode, speciation rate, extinction rate, and rates. There are many reasons why sampling
errors in tree topology) on the accuracy of rates might be uneven through time (and
PDEs. This approach demonstrated that PDEs indeed between different clades). These include
tend to overestimate the amount of range exten- geological factors (such as differences in the
sion, with this error typically peaking in the volumes of fossiliferous sediments from differ-
first time interval (Wagner 2000a: 76–77). This ent time periods); biological traits (organisms
problem becomes more severe as phylogenetic have different preservation potentials depend-
trees decrease in their topological accuracy. ing on their size, structure, dispersal strategies,
Although the above three points raise some and so on), and “human” factors (such as the
serious concerns regarding the accuracy of ease or difficulty associated with the logistics of
PDEs, there are ways in which errors can be collection from particular rock units).
minimized. First, it seems that PDEs (and The impact of sampling biases might be
measures of fossil record quality based on phy- evaluated by determining whether our opportu-
logenetic topologies) are most accurate when nities to observe/collect specimens have
the operational taxonomic units (OTUs) have remained approximately evenly distributed
certain characteristics. For example, greater across time. For example, one could plot the vol-
accuracy is achieved when the OTUs are (1) ume of sedimentary rock for a particular facies,
stem-based taxa, (2) species (or at least the low- or the number of fossiliferous formations,
est feasible taxonomic level), and (3) treated as against time. This approach was adopted by
ancestors in the phylogeny if they lack autapo- Hunt et al. (1994), who noted that their TDEs
morphies (Wagner 2000a, 2000b; Wagner and for sauropods seemed to correlate closely with
Sidor 2000). Second, Wagner (2000a) noted the numbers of sauropod-bearing formations
that PDEs should be “conservative” when it from the Late Triassic to the Late Cretaceous
comes to the detection of mass extinction. This (see above). This result suggests that the
apparent fluctuations in sauropod diversity are provided the various sources of error are ran-
artifacts caused by different preservation rates. domly distributed: but it is conceivable that this
If we wish to identify true evolutionary patterns, is not the case. For example, the numbers of for-
therefore, we must demonstrate that apparent mations may be exaggerated in areas where sed-
diversity is not purely controlled by the number imentary analysis began relatively recently,
of opportunities to observe (NOOs). whereas in more thoroughly understood areas, a
Although assessment of preservation rates plethora of formations may be reduced through
is highly desirable, there exist a number of careful correlation and synonymization.
problems with such an approach. Similarly, time periods of particular interest
First, there will be occasions when NOOs (e.g., the Kimmeridgian–Tithonian boundary)
will correlate with genuine rather than artifac- may receive a disproportionate amount of atten-
tual changes in diversity. For example, as a clade tion, resulting in inflation of NOOs.
radiates, in terms of lineage diversity and/or Third, determining what is, or is not, an
geographic distribution, we might expect its opportunity to observe members of a particular
members to appear in increasing numbers of clade is somewhat ambiguous because the deci-
formations or rock units. Similarly, as a group sion depends on a complex combination of fac-
declines toward its extinction, it would be tors, including the paleoecology, physiology,
expected to be found in decreasing numbers of biogeography, and evolutionary history of mem-
formations. Thus, because Hunt et al. (1994) bers of a clade. Wagner (2000b:351–352) dis-
designated their “opportunities to observe” as cussed opportunities to observe with regard to
numbers of sauropod-bearing formations, we lower-level taxa, such as species. Suppose, for
cannot really be sure what the correlations example, that species A and B occur in a partic-
between this and the sauropod TDEs mean. We ular facies type (X) in a geographic region (Y) at
suggest that this problem can be circumvented, time t0. A and B can be assumed to have the
or at least ameliorated, by defining opportuni- same preservation potential because they are
ties to observe in a broader way. Thus, in this similar in terms of their size, structure, and so
study, we consider any dinosaur-bearing forma- on. We should expect to find the same species
tion as an opportunity to observe, since any rock at later points in time provided there is a con-
unit capable of preserving large terrestrial verte- tinuation of the existence of facies X in area Y
brate material is presumably capable of preserv- from t0 to the subsequent time periods con-
ing a sauropod. Thus, if we see, at a particular cerned. If these conditions are fulfilled, but we
point in time, that there are a large number of do not find species A (despite intense sampling
dinosaur-bearing formations but a relatively low and the continued presence of B) during later
sauropod diversity, it seems legitimate to infer time periods, we can conclude that this is prob-
that the latter is a genuine evolutionary signal ably the result of genuine extinction rather than
rather than an artifact of preservation. failure of preservation or sampling. Such con-
Second, estimating NOOs is somewhat arbi- straints, however, do not apply when we are
trary. What, for example, do the units of meas- considering the history of a clade of taxa rather
urement actually mean? Is a formation from the than a single species. This is because changes
Late Jurassic of North America equivalent to one in the behavior, physiology, or structure of
from the Middle Jurassic of China? Should a organisms through their evolutionary history
rock unit that has produced one fragmentary may allow them to cross biogeographic barri-
and indeterminate dinosaur specimen be given ers, invade new facies types, or change their
the same weight as one that has produced mul- preservation potentials.
tiple skeletons of many different diagnosable Despite the difficulties outlined above, we
taxa? These ambiguities may not cause too feel that it is preferable to consider some
much distortion in our estimate of NOOs approximate estimate for the quality of the fossil
record, rather than none at all, when evaluating which the sauropod was found and other
diversity curves. We have, therefore, plotted the deposits whose absolute age has been deter-
number of dinosaur-bearing formations against mined radiometrically. The danger with the use
time (see below) based on the data provided by of an absolute timescale is that it can give a false
Weishampel et al. (2004). sense of precision that, currently, is not obtain-
able. Here, we have used a nonlinear stage-
STRATIGRAPHIC RANGES AND TIMESCALES based scale for the x-axes in figures 3.1, 3.4, and
Both the TDE and the PDE methods require 3.5 but have converted standard European
information on the stratigraphic ranges of taxa. Stages to a linear absolute timescale based on
Previous studies have adopted subtly different Gradstein et al. (1995) in figure 3.6.
approaches in terms of how stratigraphic The second issue, concerning dating,
ranges have been treated. involves the treatment of uncertainty or poor
The first issue concerns whether the age of a stratigraphic constraint. Hunt et al. (1994)
taxon should be represented in terms of its pres- noted that the ages of many sauropod taxa are
ence during a particular stratigraphic unit (e.g., not well constrained. Thus, Barapasaurus, from
assigning the taxon to one or more of the the Kota Formation of India, is dated as
Standard European Stages, such as the “Hettangian–Pliensbachian”, reflecting uncer-
Kimmeridgian) or by applying an absolute age tainty in the dating of the deposits rather than
based on radiometric dating of the relevant the presence of a long-lived genus. Hunt et al.
strata. Hunt et al. (1994) and Barrett and Willis (1994) noted that such uncertainty could distort
(2001) used stratigraphic stages, whereas diversity estimates. For example, if the true age
Weishampel and Jianu (2000) employed of Barapasaurus were Hettangian, then count-
absolute dates. The latter method has one major ing this taxon as present during the
advantage: when diversity curves are plotted Pliensbachian would contribute to an overesti-
against absolute time (or at least a linear scale of mate of diversity. To compensate for this, Hunt
relative time), then the slopes of the curves give et al. (1994) down-weighted the contribution of
a direct estimate of the tempo of radiation and taxa that occur in more than one Stage. Thus,
extinction events. This will not necessarily be Barapasaurus would contribute one-third of a
the case when the “timescale” is a sequence of sauropod genus in each case to the diversity
stratigraphic units because different Standard estimates for the Hettangian, Sinemurian, and
European Stages are of different absolute Pliensbachian. We sympathize with this
lengths (e.g., the Kimmeridgian and Albian attempt to correct for factors that might over-
Stages last for 3.4 and 13.3 Ma, respectively weight diversity estimates, but we do not
[Gradstein et al. 1995]). The use of absolute employ Hunt and coworkers’ strategy for three
ages, however, is an ideal that is difficult to important reasons.
achieve at present. Outside of Europe and North First, the proportional down-weighting
America, the absolute ages of many strati- employed by Hunt et al. (1994) will also distort
graphic units are not well constrained by radio- diversity estimates, albeit in a subtly different
metric dates. Most studies that report the dis- way compared to a simple equally weighted
covery of new sauropod taxa provide informa- approach. Suppose that, a particular genus is
tion only on stratigraphic position, rather than dated as coming from one of two consecutive
absolute age. In many cases, therefore, the stages, X and Y. Hunt and coworkers’ down-
absolute ages for sauropod species, utilized by weighting would yield 0.5 of a genus in each of
Weishampel and Jianu (2000), are unlikely to X and Y. Suppose that, in reality, the true age for
be based on direct radiometric dating: rather, the genus is restricted to X alone. The strategy
many would have to be based on biostrati- of Hunt et al. does reduce the overestimate of
graphic correlations between the deposits in diversity for Y, but also underestimates the
diversity for X. The simpler equally weighted of down-weighting is appropriate when obtain-
strategy would count one genus present in X ing PDEs. In short, one of the advantages of the
and one in Y. This doubles the error for Y but at PDE is that it automatically partially compen-
least there is no underestimate for X. It seems, sates for uncertainties in the ages of taxa, and
therefore, that both down- and equal-weighted therefore no additional down-weighting is
schemes are equally problematic. required.
Second, the Hunt et al. down-weighting In summary, we do not believe that down-
strategy seems somewhat arbitrary because it weighting strategies offer any real benefits, and
does not take into account the absolute duration in some cases such approaches will be more
of stages. Consider genus A from the misleading than a simple equally weighted
Kimmeridgian–Tithonian and genus B from strategy. We therefore apply the latter in the fol-
the Albian. In absolute terms, the age of genus lowing analyses.
A is better constrained than that for B, since the
total duration of the Kimmeridgian Tithonian TAXONOMIC LEVEL OF ANALYSIS
is 9 Ma, whereas that for the Albian is 13.3 Ma The two largest previous studies of sauropod
(Gradstein et al. 1995). Yet because genus A diversity utilized genera and species (Hunt
cannot be constrained to a single stage, its con- et al., 1994) or just species (Weishampel and
tribution to the diversity estimate is halved, Jianu 2000) as their taxonomic units of analy-
whereas that of B remains at its full weight. If sis. We prefer to analyze sauropod diversity at
unequal weighting is to be applied in a rigorous the generic level, and it is therefore worth con-
fashion, therefore, it must be applied in considering the costs and benefits of the alternative
junction with an absolute, or linear relative, approaches.
timescale. The most important advantage associated
Third, down-weighting is partially incom- with diversity estimates at the species level is
patible with the application of a phylogenetic that a more detailed picture of diversity trends
approach to diversity estimation. For example, may be obtained. However, as Weishampel and
suppose the age of genus A cannot be con- Jianu (2000) noted, many sauropod genera are
strained beyond the fact that it comes from one monospecific (85% according to the taxo-
of two consecutive stages, X and Y (where X nomic revision of Upchurch et al. [2004]), and
precedes Y). Suppose also that the age of genus we might not expect there to be major differ-
B, the sister-taxon to A, is known to be con- ences between species- and genus-level diver-
strained to stage X. Under these conditions, sity curves. There are also several disadvantages
there will be occasions when the uncertainty in associated with any attempt to utilize sauropod
the age of A does not affect the diversity esti- species at the current time. First, sauropod
mate and, therefore, down-weighting would be species-level taxonomy is often rather problem-
inappropriate. For example, if the true age of A atic. For example, a recent revision of the
is Y, then the phylogenetic diversity approach Middle Jurassic English genus Cetiosaurus
infers the presence of a ghost range extending (Upchurch and Martin 2002, 2003) found that
back through stage X. In this case, down- only 1 of 13 published species could be con-
weighting the presence of A in X is inappropri- firmed as genuinely belonging to that genus.
ate since the PDE indicates that A, or a lineage Although taxonomic and nomenclatural prob-
leading to A, was present during X. Down- lems also exist at the generic level (Upchurch et
weighting would only be appropriate if the true al. 2004), in general sauropod genera are con-
age of A is X, because the possible presence of siderably more “stable” than the species they
A in Y now results in an overestimate of diver- contain. Second, all major analyses of sauropod
sity for Y. Since we do not know what the true phylogeny (Calvo and Salgado 1995; Upchurch
age of A is, we cannot be sure whether any form 1995, 1998, 1999; Salgado et al. 1997; Wilson
and Sereno 1998; Curry Rogers and Forster known from a separate time period, such that
2001; Wilson 2002; Upchurch et al. 2004) have A1 occurs during t1, A2 occurs during t2 and so
used the genus as the standard OTU. If we wish on (fig. 3.3). When the relationships among the
to analyze sauropod diversity using information five species are unknown, the total amount of
from cladograms, therefore, we are somewhat inferred ghost range equals 10 time units
constrained to consider genera. Weishampel (fig. 3.3A): this is the same result that would be
and Jianu (2000) attempted to circumvent this achieved if the fully resolved species-level
point by using the generic-level cladogram of cladogram happened to be maximally incon-
Upchurch (1998), with each multispecies gruent with the stratigraphic distributions of
genus replaced by a “star burst” (i.e., polyto- the five species (figs 3.3B, C). Figures 3.3D–F
mous) cluster of species. Furthermore, those show the required ghost ranges (4 time units
taxa not considered by Upchurch (1998) were in total) when the fully resolved species-level
added to the sauropod cladogram via determi- cladogram happens to be maximally congruent
nation of possible synapomorphies, rather than with the stratigraphic ranges of the species.
direct cladistic analysis (Weishampel and Jianu This simple example illustrates one of the
2000: 131). As a result, Weishampel and Jianu intrinsic biases introduced into the PDE
(2000) produced a virtually complete phyloge- method by polytomies: the ghost ranges will be
netic tree for sauropod species, although the overestimated because all missing lineages will
topology inevitably contained many polytomies. have to extend back as far as the oldest known
This increased the sample size in the phylogeny taxon within the polytomy. In other words poly-
utilized by Weishampel and Jianu, but this ben- tomies are undesirable because they introduce
efit may have been outweighed by the potential a systematic bias toward inflated PDEs, rather
errors introduced into the analysis. The first than a random error that might be at least par-
potential source of error concerns the relation- tially cancelled out in large enough data sets.
ships of those taxa that were added to the Such errors will be most severe when the strati-
Upchurch (1998) cladogram on the basis of graphic range of a supposed species cluster is
synapomorphies. Many such taxa contain particularly extensive, as occurs with problem-
unique combinations of character states that atic taxa such as Cetiosaurus and Titanosaurus.
mean that they cannot be unequivocally “slot- These conclusions are consistent with the
ted” into a particular position: indeed, more results of the simulation studies carried out by
recent enlarged analyses (Upchurch et al. Wagner (2000a) and Wagner and Sidor
2004) indicate that consideration of additional (2000), in which errors in the estimation of
taxa has some important effects on the rela- tree topology are shown to result in a system-
tionships between the original 26 sauropod atic overestimation of diversity early in a
genera considered by Upchurch (1998). Thus, clade’s history.
it seems likely that the topology utilized by Here, therefore, we prefer to avoid poly-
Weishampel and Jianu (2000) cannot be fully tomies (and therefore sauropod species) via the
justified in terms of cladistic analysis. The sec- use of a generic-level analysis of diversity. Later
ond problem concerns the presence of multi- in this chapter, we examine the similarities and
ple polytomies in the Weishampel and Jianu differences between the PDEs obtained by
tree. The precise effect of such polytomies on Weishampel and Jianu’s analysis of species and
diversity estimation will vary depending on the those for genera obtained by ourselves.
ages and number of species involved. As an
example, however, consider the hypothetical DATA
situation in figure 3.3. Suppose we observe the Both TDEs and PDEs require information on the
fossil record of genus A that is known from five stratigraphic ranges of sauropod genera. The tax-
species (A1, A2, etc.). These species are each onomy and stratigraphic ranges of sauropod
FIGURE 3.3. A schematic diagram showing the effect of polytomies on
phylogenetic diversity estimation. (A) Phylogenetic relationships of the
five taxa A1–A5. Top cladogram shows no resolution; bottom cladogram
is resolved so that it is maximally incongruent with the stratigraphic
ranges of the taxa. (B) Stratigraphic ranges of taxa A1–A5 and ghost
ranges required by the relationships in (A). (C) Diversity through time
based on the observed and inferred ranges shown in (B). (D) Cladogram
for taxa A1–A5 that is maximally congruent with the stratigraphic ranges
of the taxa shown in (E). (E) Stratigraphic ranges of taxa A1–A5 and ghost
ranges required by the relationships in (D). (F) Diversity through time
based on the observed and inferred ranges shown in (E). In the diagrams
showing observed and ghost ranges, solid lines indicate known strati-
graphic range, while dotted lines indicate ghost ranges.
genera have been reviewed and revised recently and ‘Texan Pleurocoelus’) are deleted a posteriori
by Upchurch et al. (2004). These authors con- from the 1,056 original MPTs. The resulting
cluded that there are currently approximately 98 reduced consensus tree therefore contains 36
valid, or potentially valid, sauropod genera. sauropod genera and is shown, plotted against
These taxa, and their stratigraphic ranges, form stratigraphic Stages, in figure 3.4. Our PDE val-
the basis for our TDEs and are summarized in ues were calculated directly from the inferred
chapter 4 (appendix 4.2). ghost ranges and observed stratigraphic ranges
Upchurch et al. (2004) also provide the most in figure. 3.4, with conversion to an absolute
detailed cladogram of sauropod genera cur- timescale using Gradstein et al. (1995).
rently available. The data matrix comprises 309
characters for 41 ingroup sauropod genera and
RESULTS AND DISCUSSION
6 dinosaurian and dinosauromorph outgroups.
The Heuristic search in PAUP 4.0 (Swofford The results of our diversity analyses are shown
2000) produced 1,056 most parsimonious trees in figures 3.5 and 3.6. Figure 3.5 presents a com-
(MPTs), which were then subjected to reduced parison between the phylogenetic estimates of
consensus analysis (Wilkinson 1994). The latter sauropod diversity produced in the current
indicated that a single fully resolved topology chapter (UB) and Weishampel and Jianu
can be obtained if five taxa (Andesaurus, (2000) (WJ). Figure 3.6 compares the TDE and
Argentinosaurus, Lapparentosaurus, Nigersaurus, PDE curves produced by our data alone and
FIGURE 3.4. A time-calibrated version of the reduced consensus cladogram produced by Upchurch et al. (2004) (see text for
details). Stage name abbreviations are listed in the legend to figure 3.1. Solid lines show the observed stratigraphic range for
each sauropod genus; dotted lines indicate inferred ghost ranges.
FIGURE 3.5. Comparison of the phylogenetic diversity estimates obtained by Weishampel and Jianu (2000) and in the cur-
rent chapter. Stage name abbreviations are listed in the legend to figure 3.1.
FIGURE 3.6. Comparison of the taxic and phylogenetic diversity estimates for sauropod genera, and the number of
dinosaur-bearing formations, obtained in the current chapter (see text for details). Stage name abbreviations are listed in the
legend to figure 3.1. DBF, number of dinosaur-bearing formations (i.e., number of opportunities to observe); PDE, phyloge-
netic diversity estimate; TDE, taxic diversity estimate. Numbers of dinosaur-bearing formations were calculated using the
data of Weishampel et al. (2004). Absolute ages are based on Gradstein et al. (1995).
also shows the number of dinosaur-bearing for- ses by constructing PDEs based on the sauro-
mations through time. pod cladograms presented by Pisani et al.
(2002) and Wilson (2002).
COMPARISON WITH WEISHAMPEL There are, however, some subtle differences
AND JIANU (2000) between the WJ and the UB PDE curves that
The PDEs derived by WJ and UB are shown in deserve discussion. Most notably, the Middle
figure 3.5. The absolute diversity estimates Jurassic increase in sauropod diversity com-
often differ between the two studies, presum- mences somewhat earlier in the UB PDE curve
ably because of differences in taxon sampling than in the WJ study. Similarly, the rapid decline
(see above). In general, however, the two curves in diversity, toward or at the Jurassic–Cretaceous
are in close agreement in terms of the relative boundary, happens a little later in the UB curve.
changes in diversity through time. This agree- To some extent these disagreements in timing
ment is important since it indicates that the may reflect the different timescales employed
potential pitfalls associated with the WJ and UB by the two studies: conversion of the absolute
analytical approaches have not led to signifi- dates presented by WJ into stratigraphic stages
cantly different diversity patterns. Thus, the uti- (or even parts of stages) may have introduced
lization of sauropod species, and the introduc- a small distortion into the direct comparison
tion of polytomies, has not adversely affected between the PDE values. We suspect, however,
the WJ study. Similarly, the fact that the current that much of the discrepancy between the UB
chapter is based on a cladogram that does not and the WJ studies, especially regarding the
sample all available taxa does not seem to have Middle Jurassic, represents a real phenome-
led to major errors regarding relative changes non caused by differences in the data sets. The
in diversity, though clearly the absolute diver- new cladogram (fig. 3.4), used here, contains
sity levels have been underestimated. Future several Middle Jurassic forms whose relation-
studies may gain further insights into the ships have important effects on the number
impact of differences in phylogenetic hypothe- and extent of ghost lineages. These crucial
taxa, such as Bellusaurus, have not previously Sauropod diversity started to increase
been incorporated into a formal cladistic noticeably during the Early Jurassic, though
analysis, and hence their precise relationships there may have been a slight decline in the
would not have been available to WJ. These Toarcian. To some extent, apparent diversity,
differences mean that the WJ study generates especially that inferred from the TDEs, seems
fewer early Middle Jurassic ghost ranges than to be affected by preservation rates during the
does the UB analysis. Late Triassic and Early Jurassic. PDEs, TDEs,
and NOOs all gradually decline from a peak at
SAUROPOD DIVERSITY PATTERNS the Triassic–Jurassic boundary to a low point in
The taxic and phylogenetic diversity curves in the Toarcian. Nevertheless, it is interesting to
figure 3.6 agree strongly on many aspects of note that NOOs in the Early Jurassic are gener-
sauropod evolutionary history. In particular, ally no higher than those for the Late Triassic,
both TDEs and PDEs indicate that (1) diversity yet there is a clear increase in sauropod diver-
increased throughout the Jurassic, reaching a sity in the former time period.
peak in the Kimmeridgian–Tithonian; (2) the The most serious discrepancy between the
Jurassic increase in diversity was punctuated diversity curves occurs in the Aalenian to lower
by two short-term decreases, one in the Bathonian interval, with PDEs as much as five
Toarcian and one in the Oxfordian; (3) there times higher than the TDEs. This observation
was a dramatic decrease in diversity at the should be treated with some caution because
Jurassic–Cretaceous boundary; and (4) diver- the large number of ghost lineages extending
sity reached another peak during the back to the Aalenian have been created partly by
Campanian–Maastrichtian. There are, how- the age and relationships of Bellusaurus
ever, discrepancies between the PDE and the (fig. 3.4). The stratigraphic age of this genus is
TDE values at various points in time, and not well constrained: the ‘Wucaiwan’ or
these potentially represent important sam- ‘Shishigou’ Formation, in which Bellusaurus is
pling biases or problems with taxon sampling. found, has been suggested to be somewhere
within the Aalenian–Callovian (see chapter 4,
PATTERNS AND ARTIFACTS appendix 4.2). Furthermore, the cladistic analy-
The earliest sauropod body fossils currently sis by Upchurch et al. (2004) places Bellusaurus
occur in the Late Triassic. These include in a clade along with Atlasaurus and Jobaria,
Blikanasaurus from the Carnian or early representing an unexpected and potentially
Norian of South Africa (Galton and van controversial clustering of taxa that is not
Heerden 1985), which was regarded as a strongly supported by numerous unequivocal
prosauropod until the recent cladistic analysis synapomorphies. Alteration of either the age or
by Galton and Upchurch (2004), and the relationships of Bellusaurus, therefore,
Isanosaurus from the late Norian or Rhaetian could have a profound impact on the mid-
of Thailand (Buffetaut et al. 2000). This pro- Jurassic PDE values. These concerns reflect the
vides confirmation of the suggestion that more general point made by Foote (1996) and
sauropods diverged from their prosauropod Wagner (2000a), that PDEs may overestimate
sister-group at the beginning of the Carnian at diversity, especially toward the beginning of a
the latest (Upchurch 1995). The fossil record clade’s history. Notwithstanding these points,
apparently indicates that sauropod diversity however, several lines of reasoning suggest that
was relatively low during the Late Triassic, the Middle Jurassic increase in sauropod diver-
compared to that for prosauropods, and this sity may reflect a real signal. First, although the
conclusion is not currently contradicted by TDEs lag behind the PDEs, they also reach a
ghost range data (see also Barrett and similar peak in the Middle Jurassic (albeit not
Upchurch, chapter 4). until the Bathonian–Callovian). Second, NOOs
during the Aalenian–Bathonian are similar to the Middle Jurassic, cetiosaurids and basal
those for the Norian and Hettangian, even eusauropods are relatively diverse, whereas
though the latter stages are associated with much neosauropods may have been represented by
lower PDE values. Finally, other lines of evidence only a relatively small number of lineages. The
suggest that the major neosauropod lineages Oxfordian extinction may have been particularly
(brachiosaurs, titanosaurs, diplodocoids) had serious for the cetiosaurid and basal eusauropod
already radiated by the Bathonian. These include taxa, allowing neosauropods to radiate into sev-
the presence of putative diplodocoid remains and eral vacant niches during the Kimmeridgian
titanosaur trackways in the Bathonian of the and Tithonian. At present, however, the fossil
United Kingdom (Upchurch and Martin 2003; record for sauropods is too sparse to allow direct
Day et al. 2002, 2004) and the presence of a vic- and detailed testing of this possibility.
ariance pattern consistent with Pangaean frag- The Late Jurassic has long been regarded as
mentation (Upchurch et al. 2002). Sauropod the pinnacle of sauropod diversity. This view is
diversity may have been considerably higher fully supported by the TDE and PDE values for
during the early Middle Jurassic than has been the Kimmeridgian and Tithonian. During these
proposed previously. For example, the TDE val- stages, there are at least 27 valid or potentially
ues obtained by Hunt et al. (1994) and the UB valid genera. This peak in diversity could have
analysis suggest that Middle Jurassic sauropod been partly caused by a particularly rich fossil
diversity was only approximately 40% of that record at this time, especially the well-preserved
for the Late Jurassic, whereas the WJ and UB and intensively studied Morrison and
PDE values indicate that this relative diversity Tendaguru faunas. However, NOOs are actually
measure is closer to 75%. Thus, the significant lower for these stages than they are for the
increase in sauropod diversity, which includes Oxfordian and Berriasian (fig. 3.6). The UB
the important early radiation of neosauropod PDE values for the Kimmeridgian–Tithonian
lineages, may have commenced somewhat ear- are approximately 67%–75% those obtained
lier and been considerably more important from TDEs, indicating that the cladogram in
than previously recognized from taxic estimates figure 3.4 has somewhat under sampled the
of diversity. The cause of the discrepancy diversity present in the Late Jurassic. It will be
between PDEs and TDEs is not entirely clear, important to add these unsampled taxa to
though it should be noted that the Aalenian rep- future cladistic analyses, to see whether their
resents one of the poorest parts of the dinosaur incorporation alters the shape of the Jurassic
fossil record as judged from NOOs (fig. 3.6), PDE curve.
and therefore sampling error may be at least Both TDEs and PDEs indicate a dramatic
partly responsible. decrease in sauropod diversity across the
Both PDEs and TDEs indicate a dramatic Jurassic–Cretaceous boundary: from the Tithon-
drop in sauropod diversity during the ian to the Berriasian, TDEs drop from 18 to 2,
Oxfordian. This apparent extinction is less while the corresponding PDEs fall from 14 to 6.
marked in the PDEs, but this may represent an This represents a decrease in diversity of
artifact caused by backward “smearing” of origi- 57%–89%, returning sauropod diversity to the
nation times (see above). In the Jurassic, NOOs levels seen briefly during the Oxfordian. This
reach their highest level during the Oxfordian, extinction event is even more marked when the
higher even than in the Kimmeridgian and diversity levels for the Kimmeridgian are com-
Tithonian. Thus, the Oxfordian drop in diversity pared with those of the Berriasian—a decrease
potentially represents a genuine extinction of 65–93% over 5.5 Ma. Two factors suggest that
event rather than an artifact. We hypothesize this diversity decrease represents a genuine
that the Oxfordian may represent an impor- mass extinction. First, NOOs actually increase
tant transition in sauropod evolution. During dramatically (from an average of 60 to 100)
at the Jurassic–Cretaceous boundary. Second, caused by undersampling of taxa in the cladis-
as Wagner (2000a) has noted, PDEs are con- tic analysis. For example, the “comprehensive”
servative with respect to mass extinctions (see WJ PDEs show the same pattern as the UB
above). It is interesting to note that much of PDEs: that is, a gradual decline in diversity
this extinction affected generic-level diversity but from the Valanginian to the Coniacian, with no
left many of the higher taxa with lineages that Albian peak (fig. 3.5). This “true” diversity pat-
survive into the Early Cretaceous. Thus, tern may have been modified to give an Albian
although diplodocids and all non-neosauropods peak in TDEs by the presence of a sampling
die out at the end of the Jurassic, many other dis- bias: NOOs are higher during the Aptian–Albian
tinct clades (nemegtosaurids, rebbachisaurids, than they are in the Early Cretaceous and
dicraeosaurids, brachiosaurids, and titanosaurs) Cenomanian–Santonian. Given the evidence
are represented by body fossils and/or ghost currently available, it is possible that sauropod
ranges in the Berriasian. diversity remained fairly constant, or declined
There is a significant discrepancy between only slightly, from the Berriasian to the Albian
TDE and PDE values in the mid-Cretaceous and then decreased more rapidly during the
(Albian), with the former being up to 2.75 times Cenomanian.
higher than the latter. This could indicate that The presence of a depressed preservation
the PDEs are artificially low because the cladis- rate in the Turonian and Coniacian is obvious
tic analysis did not sample sufficient taxa from from the fact that no diagnosable sauropod gen-
the mid-Cretaceous. The cladistic analysis by era are known from these stages. There are a
Upchurch et al. (2004) originally contained 41 number of fragmentary and indeterminate
sauropod genera, though 5 were deleted a pos- sauropods from these stages, such as material
teriori to obtain the reduced consensus tree from the Turonian of Uzbekistan and
shown in figure 3.4. Four of these five genera Kazakhstan (Weishampel et al. 2004) and the
(Andesaurus, Argentinosaurus, Nigersaurus, and Coniacian of Argentina (Bonaparte 1996), but
‘Texan Pleurocoelus’) are of mid-Cretaceous age even this poor-quality material is scarce. Prior
(see chapter 4, appendix 4.2). Assignment to the application of phylogenetic diversity esti-
of these taxa to their “probable” relationships mation, it might have been predicted that the
(i.e., Andesaurus, Argentinosaurus, and ‘Texan Turonian–Coniacian diversity trough would be
Pleurocoelus’ as basal titanosaurs that are closer at least partially eliminated by the discovery of
to other titanosaurs than to Phuwiangosaurus, numerous ghost ranges extending back from
with Nigersaurus as a rebbachisaurid) would the major titanosaur radiation of the Late
not introduce major new ghost ranges back into Cretaceous. Both the WJ and the UB PDE
the early part of the Cretaceous. For example, curves (fig. 3.5), however, show that this is not
Nigersaurus is the earliest of the three known the case. Clearly a few sauropod lineages
rebbachisaurid genera (see chapter 4, appendix existed during the Turonian–Coniacian, but
4.2). These four additional taxa, therefore, are there is currently no evidence to suggest that
likely to increase the PDE values for the mid- sauropods were diverse at this time. This view
Cretaceous (i.e., Aptian–Cenomanian) without is supported by the NOOs for the Cretaceous
increasing the values for the earliest Cretaceous (fig. 3.6). The number of dinosaur-bearing for-
(Berriasian–Barremian). This would tend to mations decreases in the early Late Cretaceous,
bring the PDE and TDE curves into closer especially during the Turonian, Coniacian, and
agreement, though it would not be enough by early Santonian. This pattern may explain part
itself to produce a major mid-Cretaceous peak of the dip in apparent sauropod diversity
in PDEs. There is also evidence that the dis- between the Albian and the Campanian peaks.
agreement between the UB TDEs and the UB Yet it should also be noted that NOOs during
PDEs for the mid-Cretaceous is not solely the early Late Cretaceous are substantially
higher than those for portions of the Jurassic These explanations fall into two main cate-
that have produced numerous sauropod gen- gories, so that peaks and troughs in apparent
era. Thus, the scarcity of sauropod material diversity are regarded as (1) artifacts produced
from the Turonian and Coniacian certainly rep- by geological processes or (2) real patterns
resents some form of sampling bias, but this caused by the impact of physical and biological
appears to have been exacerbated by a genuine processes on evolutionary history.
rarity of sauropods. The principal artifact-based explanation
The TDE and PDE curves agree that a major argues that peaks and troughs in diversity
radiation of sauropods (essentially titanosaurs) reflect the impact of sea level change on the
occurred during the Santonian–Maastrichtian. preservation potential of terrestrial taxa
The diversity levels achieved at this time rival (Haubold 1990; Hunt et al., 1994). The appar-
those observed in the Middle Jurassic, though not ent close correlation between peaks in diversity
approaching the peak seen in the Late Jurassic. and sea level high-stands, and between troughs
The TDEs exceed the PDEs by approximately in diversity and sea level low-stands, is impres-
1.5–3 times, reflecting the latter’s underestima- sive. Furthermore, there is a plausible causal
tion of diversity due to taxon sampling problems. link between sea level and observed taxic diver-
At least part of the Campanian–Maastrichtian sity in the fossil record, since the preservation
pattern can be interpreted as resulting from vari- potential of terrestrial organisms could be
ation in preservation rates. Figure 3.6 suggests a higher when shallow seas spread into lowland
close correspondence between fluctuations in areas. However, this “taphonomic bias” expla-
NOOs and PDE or TDE values (e.g., peaks in the nation for fluctuations in sauropod apparent
early Campanian and late Campanian–early diversity was proposed at a time when phyloge-
Maastrichtian). However, relative to the rest of netic diversity estimates were not available:
the Cretaceous, NOOs are only moderately ele- PDE values provide an important opportunity
vated (except for the extreme peak in the latest to test this hypothesis rigorously because they
Maastrichtian), suggesting that the Campanian can suggest which diversity peaks and troughs
peak in sauropod diversity also partly reflects a are real and which are probably artifacts.
genuine radiation of new taxa. Haubold (1990) and Hunt et al. (1994) iden-
The diversity curves in figure 3.6 also indi- tified five major peaks in taxic diversity that cor-
cate that sauropods went into decline prior relate with sea level high-stands: (1) Bajocian,
to the final K–T extinction, rather than disap- (2) Kimmeridgian, (3) Valanginian–Barremian,
pearing “suddenly” at the latter boundary. (4) Albian, and (5) Campanian–Maastrichtian.
Whether this decline represents a real phe- Four of these peaks (Nos. 1–3 and 5) are seen in
nomenon, or is an artifact caused by the cur- the WJ and UB PDEs (figs. 3.5 and 3.6), sug-
rent uncertainty surrounding the stratigraphic gesting that they represent real increases in
ages of many sauropods, is not entirely clear. diversity and cannot be explained simply as arti-
Nevertheless, the most detailed evidence cur- facts of preservation. There are also four
rently available suggests that, as with several notable troughs in diversity that correlate with
other dinosaur groups, sauropods were sea level low-stands: (1) Toarcian–Aalenian, (2)
undergoing a gradual decrease in diversity Oxfordian, (3) Berriasian, and (4) Turonian–
immediately prior to the final “mass” extinc- Coniacian. Two of these (Nos. 2 and 4) appear
tion event. to be genuine low points in diversity, although
the Turonian–Coniacian trough must also rep-
PROCESSES AFFECTING SAUROPOD DIVERSITY resent some form of sampling bias. Thus, of
Most studies of sauropod, or even dinosaur, the nine major points of comparison between
diversity have attempted to explain the observed sauropod diversity and sea level, only one peak
fluctuations in terms of one or more processes. (Albian) and two troughs (Toarcian–Aalenian
and Berriasian) provide support for the disrupt some of the correlation between real
“taphonomic bias” hypothesis. The remaining peaks and troughs in diversity and highs and
peaks and troughs in diversity, although still lows in sea level. For example, the Toarcian–
correlated with sea level, seem to represent Aalenian sea level low-stand seems to coincide
real evolutionary phenomena. This result with a major increase in PDE values, while the
casts some doubt on the validity of the tapho- Albian increase in sea level is not currently cor-
nomic bias hypothesis, at least in terms of its related with a corresponding increase in PDE
being the main control on observed taxic values. Furthermore, we may wish to question
diversity. Furthermore, the graph of dinosaur- the validity of the correlation between overall
bearing formations through time (fig. 3.6) also dinosaur diversity and sea level. Although the
suggests that there is not a simple relationship sauropod diversity curves match sea level fluc-
between sea level and preservation rate. For tuations reasonably well, this may not be the
example, NOOs do not peak in the Bajocian, case for theropods and ornithischians. So far,
Kimmeridgian–Tithonian, or Valanginian– phylogenetic diversity estimation has not been
Barremian relative to adjacent stages. This tends applied to theropods, and the PDEs for
to undermine the mechanism proposed by Ornithischia obtained by Weishampel and
Haubold (1990) and Hunt et al. (1994), which Jianu (2000:fig. 5.8) show major peaks and
supposedly links apparent diversity directly to troughs in diversity that do not match those for
preservation rates. This is a complex issue, sauropods.
however, since “number of dinosaur-bearing While major extrinsic events, such as sea
formations” is not the only measure by which level fluctuations, may have played some role in
the quality of the fossil record can be measured. controlling diversity levels, it is also conceivable
Certain facies may preserve a small quantity of that the main constraints on sauropod diversity
high-quality and therefore diagnosable mate- were factors relating to the particular ecology
rial, while others may produce large quantities and evolutionary potential of this clade. Much
of indeterminate fragments. This may explain of sauropod diversity seems to be related to
why, despite the lower NOO values, the Late modifications to the skull, neck, and appendic-
Jurassic has produced considerably more valid ular skeleton, which would have had a direct or
sauropod species than the apparently well-rep- indirect effect on the feeding mechanisms the
resented Turonian–Coniacian. animals could employ. Upchurch and Barrett
The above analysis suggests that other (2000) noted that, when the number of possi-
processes, apart from uneven sampling rates, ble feeding strategies is plotted against time,
may have been responsible for the correlation there is a positive correlation with diversity.
between sea level and diversity. Several studies This issue is explored in detail by Barrett and
have proposed possible mechanisms by which Upchurch (chapter 4).
sea level changes could affect the evolution of
clades in a manner that would modify diversity.
CONCLUSION
In particular, increases in sea level could pro-
mote the separation and isolation of terrestrial The taxic and phylogenetic diversity estimation
areas, resulting in allopatric speciation. methods are sometimes portrayed as rival tech-
Similarly, decreases in sea level could cause pre- niques. While each approach has its advan-
viously isolated areas to come into contact, tages and disadvantages, the current chapter
allowing biotic mixing and perhaps extinction demonstrates the valuable insights that can be
(Bakker 1977, 1978; Weishampel and Horner obtained when both methods are applied to the
1987). Thus, sea level fluctuations can be plau- same data set. Sauropod diversity fluctuated
sibly linked to real, as well as artifactual, considerably during their evolution, with
changes in diversity. Our analyses, however, major radiations occurring during the Middle
and Late Jurassic. These were followed by a ACKNOWLEDGMENTS
dramatic extinction at the Jurassic–Cretaceous P.U.’s research is funded by a Natural
boundary, then a gradual decline throughout Environment Research Council Fellowship
most of the Cretaceous, and, finally, a brief (No. GT59906ES). P.U. also wishes to thank
but important radiation of titanosaurs just the Department of Earth Sciences, Cambridge
prior to the K–T mass extinction. Although University, for supporting this work. We wish to
much of this pattern has been suspected from thank Dan Chure, Kristi Curry Rogers, and Jeff
previous studies, the current analysis indi- Wilson for inviting us to contribute to this
cates that the Middle Jurassic radiation com- book. Jack McIntosh has provided encourage-
menced earlier, and was more important, ment and advice on many occasions, and has
than has been recognized hitherto. The tim- always been supportive and interested in our
ing and magnitude of this Middle Jurassic work on sauropods—Thanks, Jack.
radiation have been obscured by a severe sam-
pling bias affecting the Aalenian and early LITERATURE CITED
Bajocian.
Alroy, J. 2000. New methods for quantifying macroevo-
The step from “pattern identification” to
lutionary patterns and processes. Paleobiology 26:
“process inference,” regarding the detection 707–733.
and interpretation of diversity, is inevitably dif- Bakker, R. T. 1977. Tetrapod mass extinctions—Model
ficult: not only are we potentially dealing with of the regulation of speciation rate and immigra-
“one–off” historical events, but also it seems tion by cycles of topographic diversity. In:
likely that several processes interacted in a com- Hallam, A. (ed.). Patterns of Evolution. Elsevier,
Amsterdam. Pp. 439–468.
plex manner. There appears to be some correla-
———. 1978. Dinosaur feeding behaviour and the
tion between sea level and TDE or PDE values, origin of flowering plants. Nature 274: 661–663.
suggesting an impact on some aspect of evolu- Barrett, P. M., and Willis, K. J. 2001. Did dinosaurs
tion, such as speciation and extinction rates. It invent flowers? Dinosaur–angiosperm coevolu-
is also clear that the sauropod fossil record has tion revisited. Biol. Rev. 76(3): 411–447.
been affected by sampling biases, most notably Bonaparte, J. F. 1996. Dinosaurios de America del
Sur. Museo Argentino de Ciencias Naturales,
in the early part of the Middle Jurassic and the
Buenos Aries. 174 pp.
Turonian–Coniacian. Buffetaut, E., Suteethorn, V., Cuny, G., Tong, H., Le
The results of the current chapter should be Loeuff, J., Khansubha, S., and Jongautchariyakul,
treated with some caution since they depend so S. 2000. The earliest known sauropod dinosaur.
heavily on the accuracy of stratigraphic dating, Nature 407: 72–74.
Calvo, J. O., and Salgado, L. 1995. Rebbachisaurus
taxonomic identifications, phylogenetic recon-
tessonei sp. nov. A new Sauropoda from the
struction, and the cladistic topology. All of these Albian–Cenomanian of Argentina; new evi-
components inevitably contain errors, and dence on the origin of the Diplodocidae. GAIA
future work could result in some major shifts 11: 13–33.
in the magnitude and timing of the diversity Cripps, A. P. 1991. A cladistic analysis of the cornutes
peaks and troughs. Nevertheless, it is interest- (stem chordates). Zool. J. Linn. Soc. London 102:
333–366.
ing to note that the taxic and phylogenetic diver-
Curry Rogers, K. A., and Forster, C. A. 2001. The last
sity estimates obtained by Hunt et al. (1994), by of the dinosaur titans: a new sauropod from
Weishampel and Jianu (2000), and in the cur- Madagascar. Nature 412: 530–534.
rent analysis generally agree quite closely. Day, J. J., Upchurch, P., Norman, D. B., Gale, A. S.,
Although this agreement must partly reflect the and Powell, H. P. 2002. Sauropod trackways: evo-
lution and behavior. Science 296: 1659.
overlapping and therefore non-independent
Day, J. J., Norman, D. B., Gale, A. S., Upchurch, P.,
nature of the data sets, it seems that a consen- and Powell, H. P., 2004. A Middle Jurassic
sus on sauropod diversity history is beginning dinosaur trackway site from Oxfordshire, UK.
to emerge. Palaeontology 47: 319–348.
Fisher, D. C. 1982. Phylogenetic and macroevolutionary Pisani, D., Yates, A. M., Langer, M. C., and Benton, M.J.
patterns in the Xiphosurida. North Am. Paleontol. 2002. A genus-level supertree of the Dinosauria.
Conv. Proc. 3: 175–180. Proc. Roy. Soc. of London Ser. B 269: 915–921.
Foote, M. 1996. Perspective: evolutionary patterns in Rieppel, O. 1997. Falsificationist versus verification-
the fossil record. Evolution 50: 1–11. ist approaches to history. J. Vertebr. Paleontol.
Galton, P. M. 1990. Basal Sauropodomorpha— 17(3-suppl): 71A.
Prosauropoda. In: Weishampel, D. B., Osmólska, H., Salgado, L., Coria, R. A., and Calvo, J. O. 1997. Evolution
and Dodson, P. (eds.). The Dinosauria, 1st ed. of titanosaurid sauropods. I. Phylogenetic analysis
University of California Press, Berkeley. pp.320–344. based on the postcranial evidence. Ameghiniana
Galton, P. M., and Upchurch, P. 2004. Basal 34(1): 3–32.
Sauropodomorpha—Prosauropoda. In: Weisham- Sereno, P. C. 1997. The origin and evolution of
pel, D. B., Osmólska H., and Dodson, P. (eds.). dinosaurs. Annu. Rev. Earth Planet. Sci. 25: 435–489.
The Dinosauria, 2nd ed. University of California Sereno, P. C. 1999. The evolution of dinosaurs. Science
Press, Berkeley. Pp. 232–258. 284: 2137–2147.
Galton,P. M., and van Heerden, J. 1985. Partial Smith, A. B. 1994. Systematics and the Fossil Record.
hindlimb of Blikanasaurus cromptoni n. gen. and Blackwell Scientific, Oxford, 223 pp.
n. sp., representing a new family of prosauropod Swofford, D. L. 2000. PAUP: phylogenetic analysis
dinosaurs from the Upper Triassic of South using parsimony, version 4.0b4a. MacMillan,
Africa. Géobios 18: 509–516. London.
Gradstein, F. M., Agterberg, F. P., Ogg, J. G., Upchurch, P. 1995. The evolutionary history of sauro-
Hardenbol, J., van Veen, P., Thierry, J., and Huang, pod dinosaurs. Philos. Trans. Roy. Soc. London
Z. 1995. A Triassic, Jurassic and Cretaceous time Ser. B 349: 365–390.
scale. SEPM (Soc. Sediment. Geol.) Special Publ. ———. 1998. The phylogenetic relationships of sauro-
54: 95–126. pod dinosaurs. Zool. J. Linn. Soc. London 124:
Haq, B. U., Hardenbol, J., and Vail, P. R. 1987. 43–103.
Chronology of fluctuating sea levels from the ———. 1999. The phylogenetic relationships of the
Triassic. Science 235: 1156–1167. Nemegtosauridae (Saurischia, Sauropoda). J.
Haubold, H. 1990. Dinosaurs and fluctuating sea Vertebr. Paleontol. 19(1): 106–125.
levels during the Mesozoic. Hist. Biol. 4: 75–106. Upchurch, P., and Barrett, P. M. 2000. The evolution
Horner, J. R. 1983. Cranial osteology and morphology of sauropod feeding mechanisms. In: Sues, H.-
of the type specimen of Maiasaura peeblesorum D. (ed.). The Evolution of Herbivory in Terrestrial
(Ornithischia: Hadrosauridae), with discussion of Vertebrates, Perspectives from the Fossil
its phylogenetic position. J. Vertebr. Paleontol. 3: Record. Cambridge University Press, Cambridge.
29–38. Pp. 79–122.
Hunt, A. P., Lockley, M. G., Lucas, S. G., and Meyer, Upchurch, P., and Martin, J. 2002, The Rutland
C. A., 1994. The global sauropod fossil record. Cetiosaurus: the anatomy and relationships of a
GAIA 10: 261–279. Middle Jurassic British sauropod dinosaur.
Lockley, M. G., Meyer, C. A., Hunt, A. P., and Lucas, Palaeontology 45(6): 1049–1074.
S. G. 1994. The distribution of sauropod tracks ———. 2003. The anatomy and taxonomy of
and trackmakers. GAIA 10: 233–248. Cetiosaurus (Saurischia, Sauropoda) from the
Norell, M. A. 1992. Taxic origin and temporal diversity: Middle Jurassic of England. J. Vertebr. Paleontol.
the effect of phylogeny In: Novacek, M. J., and 23(1): 208–231.
Wheeler, Q. D. (eds.). Extinction and Phylogeny. Upchurch, P., Hunn, C. A., and Norman, D. B. 2002.
Columbia University Press, New York. Pp. 89–118. An analysis of dinosaurian biogeography: evi-
Norell. M. A., and Novacek, M. J. 1992. Congruence dence for the existence of vicariance and dispersal
between suprapositional phylogenetic patterns: patterns caused by geological events. Proc. Roy.
comparing cladistic patterns with the fossil Soc. London Ser. B 269: 613–622.
record. Cladistics 8: 319–337. Upchurch, P., Barrett, P. M., and Dodson, P., 2004.
Novacek, M. J., and Norell, M. A. 1982. Fossils, phy- Sauropoda; In: Weishampel, D. B., Osmólska, H.,
logeny and taxonomic rates of evolution. Syst. and Dodson, P., (eds.). The Dinosauria, 2nd
Zool. 31: 266–275. Ed. University of California Press, Berkeley.
Paul, C. R. C. 1982. The adequacy of the fossil record. Pp. 259–322.
In: Joysey, K. A., and Friday, A. E. (eds.). Problems Wagner, P. J. 2000a. The quality of the fossil record
of Phylogenetic Reconstruction. Academic Press, and the accuracy of phylogenetic inferences about
London. Pp. 75–117. sampling and diversity. Syst. Biol. 49: 65–86.
———. 2000b. Phylogenetic analyses and the fossil Weishampel, D. B., Barrett, P. M., Coria, R. A., Le
record: tests and inferences, hypotheses and Loeuff, J., Xu, X., Zhao, X.-J., Sahni, A., Gomani,
models. Paleobiology 26: 341–371. E. M.P. and Noto, C. R. 2004. Dinosaur distribution.
Wagner, P. J., and Sidor, C. A., 2000. Age rank/clade In: Weishampel, D. B., Osmólska, H., and Dodson,
rank metrics-sampling, taxonomy, and the mean- P. (eds.). The Dinosauria, 2nd ed. University of
ing of “stratigraphic consistency.” Syst. Biol. 49: California Press, Berkeley. Pp. 517–606.
463–479. Wilkinson, M. 1994. Common cladistic information
Weishampel, D. B., and Horner, J. R., 1987. and its consensus representation: Reduced Adams
Dinosaurs, habitat bottlenecks, and the St. and cladistic consensus trees and profiles. Syst. Biol.
Mary River Formation. In: Currie, P. J., and 43: 343–368.
Koster, E. H., (eds.). Fourth Symposium on Wilson, J. A. 2002. Sauropod dinosaur phylogeny:
Mesozoic Terrestrial Ecosystems, Short critique and cladistic analysis. Zool. J. Linn. Soc.
Papers. Royal Tyrell Museum, Drumheller. London 136: 217–276.
Pp. 224– 229. Wilson, J. A., and Carrano. M. T., 1999. Titanosaurs
Weishampel, D. B., and Jianu. C.-M., 2000. Plant- and the origin of ‘wide gauge’ trackways: a biome-
eaters and ghost lineages: dinosaurian herbivory chanical and systematic perspective on sauropod
revisited. In: Sues, H.-D., (ed.), The Evolution of locomotion. Paleobiology 25: 252–267.
Herbivory in Terrestrial Vertebrates. Perspectives Wilson, J. A., and Sereno, P. C., 1998. Early evolution
from the Fossil Record. Cambridge University and higher-level phylogeny of the sauropod
Press, Cambridge. Pp. 123–143. dinosaurs. Mem. Soc. Vertebr. Paleontol. 5: 1–68.
FOUR
Sauropodomorph Diversity through Time
PALEOECOLOGICAL AND MACROEVOLUTIONARY IMPLICATIONS
Paul M. Barrett and Paul Upchurch
sus is that both groups are monophyletic

S auropodomorph dinosaurs
were incredibly successful animals, by
any standard. They attained high levels of
(Galton 1990; Gauffre 1993; Upchurch 1995,
1998; Sereno 1997, 1999; Wilson and Sereno
alpha-taxonomic diversity, with more than 100 1998; Benton et al. 2000; Pisani et al. 2002;
valid genera known currently (Galton 1990; Galton and Upchurch 2004; Upchurch et al.
McIntosh 1990; Galton and Upchurch 2004; 2004). Prosauropods attained a global distribu-
Upchurch et al. 2004) and were the dominant tion during the Late Triassic and Early Jurassic,
animals, in terms of numerical abundance and before disappearing close to the Early–Middle
biomass, in many Middle and Late Mesozoic Jurassic boundary (Galton 1990). Sauropods
terrestrial biomes (e.g., Young 1951; Dodson first appeared in the Late Triassic (Buffetaut
et al. 1980; Russell et al. 1980; Galton 1986; et al. 2000) and achieved a global distribution
Dong 1992). by the Middle Jurassic, persisting thereafter on
Phylogenetic evidence, and the known almost all of the major continental landmasses
stratigraphic distributions of these animals, until the latest Cretaceous (McIntosh 1990;
suggests that the Sauropodomorpha originated Weishampel 1990).
in the early Late Triassic, though there are also One of the reasons for this success may be
recent reports of Middle Triassic prosauropod the early adoption of facultative or obligate her-
material (Flynn et al. 1999). Sauropodomorph bivory in the Sauropodomorpha. In analyses on
monophyly is well established (Charig et al. living amniotes, Moore and Brooks (1995,
1965; Gauthier 1986; Benton 1990; Sereno 1996) demonstrated an evolutionary association
1997, 1999; Galton and Upchurch 2004), and between the origin of herbivory and amniote
early in its history the clade split into two major diversification by showing that clades of extant
lineages: the Prosauropoda and Sauropoda. herbivorous amniotes (principally mammals)
Although some authors regard the prosauropods are, on average, 16.8 times more speciose than
as paraphyletic to the Sauropoda (Gauthier 1986; their respective carnivorous sister-groups.
Benton 1990; Yates 2003), the current consen- Comparisons between taxic diversity estimates
125
for herbivorous and carnivorous dinosaurs (Fiorillo 1991, 1998; Barrett and Upchurch 1994,
throughout the Mesozoic have yielded similar 1995; Calvo 1994; Christiansen 1999, 2000;
results, with herbivores being between 3 and 10 Upchurch and Barrett 2000). All sauropods
times more diverse than carnivores at any one appear to have been obligate herbivores, and
time (Barrett 1998). Such analyses, and the they experimented with a wide range of tooth
apparent rapid diversification of amniotes after types, occlusal patterns, and jaw actions; in addi-
the origin of herbivory in this clade during the tion, many taxa had specialised craniodental
Late Carboniferous, provide support for the sug- structures for food gathering, such as the pres-
gestion that herbivory can be regarded as a ence of tooth combs. Sauropod postcranial
“major adaptive zone” or a “key innovation” (e.g., adaptations were likewise variable, including
Moore and Brooks 1995, 1996; Hotton et al. differences in body size, absolute and relative
1997; Reisz and Sues 2000). Adaptive zones rep- neck length, neck function (lateral sweeping,
resent particular discrete ways of life that, dorsoventral flexion, or alternation between
although they may be difficult to enter evolution- these modes), and limb proportions (related to
arily, allow the diversification of a group once the specialization for high or low browsing) (Barrett
zone has been entered, whereas key innovations and Upchurch 1994, 1995; Martin et al. 1998;
represent the acquisition of a particular character Stevens and Parrish 1999; Upchurch and
state or states that promote the radiation of a Barrett 2000). Use of gastric mills seems to
clade (e.g., Brooks and McLennan 1991). have been important in some taxa, and fermen-
Prosauropod feeding mechanisms appear to tative digestion was also likely to have been
have been relatively uniform, with little varia- important (Farlow 1987). Sauropod feeding
tion in either the craniodental or the postcranial mechanisms have a strong phylogenetic compo-
characters involved in the collection and pro- nent, and different clades can be characterized
cessing of plant food (Galton 1985, 1986; Barrett in part by specific characteristics, or combina-
1998). Jaw actions were simple orthal move- tions of various characteristics, of the feeding
ments; there was little or no tooth–tooth contact; apparatus (Upchurch and Barrett 2000; contra
oral processing of food was probably negligible; Sereno and Wilson 2001).
and fermentative digestion in elongate digestive Several studies have noted various broad-
tracts and/or the mechanical action of gastric scale changes in sauropodomorph diversity
mills seem to have been the dominant methods through time (Hunt et al. 1994; Barrett 1998;
for breaking down plant food (Galton 1986; Weishampel and Jianu 2000; Barrett and Willis
Farlow 1987; Norman and Weishampel 1991; 2001; Upchurch and Barrett, chapter 3). Here
Barrett 2000). Some prosauropods may have we consider a number of potential macroevolu-
been omnivorous, while others were facultative tionary mechanisms that may account for these
or obligate herbivores (Barrett 2000). The long patterns, including various potential clade–clade
necks, large body size, and bipedal gaits of these interactions (coevolution and competition). We
animals further suggest that some were capable explore the possibility that the observed
of browsing up to 5 m above ground level changes may be at least partially accounted for
(Barrett 1998). by fluctuations in the frequencies of different
Until recently, sauropod feeding was also sauropodomorph feeding mechanisms through
viewed as simple and stereotyped, with little time.
variation in the feeding apparatus between taxa
(e.g., Bakker 1971; Coombs 1975; Dodson 1990).
MATERIALS AND METHODS
However, detailed reconsideration of sauropod
anatomy, phylogeny, and paleoecology has There are various techniques for estimating
revealed a surprising diversity of form and func- diversity through time (Smith 1994). Taxic diver-
tion in the feeding mechanisms of these animals sity estimates (TDEs) for Prosauropoda and
126 SAUROPODOMORPH DIVERSITY THROUGH TIME

Sauropoda were obtained by plotting strati- TDEs (on the condition that as many taxa as pos-
graphic range data for all known valid genera, as sible have been included in the latter), we have
listed in the recent reviews by Galton and shown elsewhere that the PDEs and TDEs for
Upchurch (2004) and Upchurch et al. (2004), sauropods are generally in close agreement
and summing the number of valid taxa for each (Upchurch and Barrett, chapter 3). We believe,
European standard stage (see appendixes 4.1 and therefore, that it is possible to use TDEs in the
4.2). These data are presented as a series of sim- current chapter without the ensuing diversity
ple histograms. Genera were used, rather than patterns being distorted by major preservational
species, as the majority of sauropodomorph taxa biases that are also not expressed in the PDEs.
are monospecific (approximately 85% of sauro- Nevertheless, it must be remembered that there
pod genera and 75% of prosauropod genera are are a number of problems inherent to the TDE
monotypic [Galton and Upchurch 2004; approach (dependency on accurate taxonomy
Upchurch et al. 2004] and there is, therefore, lit- and stratigraphic ranges, problems with preser-
tle difference between the species-level and the vational biases, etc.) (see Upchurch and Barrett,
genus-level diversity curves. Moreover, the desig- this chapter 3).
nation of fossil species is beset by a number of Sauropod taxa were classified according to
taxonomic problems—sexual dimorphism, indi- feeding mechanism, tooth morphology, and
vidual variation, ontogenetic differences, and systematic position to explore changes in the
ecological differences being paramount among frequencies of these categories through time.
them—whereas differences between fossil gen- Sauropod feeding mechanisms are extremely
era are more substantial, with larger discontinu- varied and 11 functional complexes were recog-
ities in morphology between genera than nized in this study. Many of these functional
between species. As a result, genera tend to be complexes are defined in Barrett and Upchurch
more stable than species and may be more (1995) and Upchurch and Barrett (2000),
amenable to diversity analysis, though it should though additional research has resulted in the
be acknowledged that this represents only an identification of several novel feeding mecha-
approximation of species-level patterns (Raup nisms.
and Boyajian 1988; King 1990).
1. Cetiosaurid type: Broad-crowned (BC)
TDEs have several advantages over phyloge-
teeth; interdigitating occlusion (mesial
netic diversity estimates (PDEs) in an analysis of
and distal wear facets); orthal jaw action;
this kind. TDEs reflect “real” empirical data and
cropping of vegetation; intermediate
are not as dependent on the assumption of a cor-
FL/HL (forelimb length:hindlimb length)
rect phylogeny for any particular clade as PDEs.
ratio (ca. 0.84); short (12 or 13 cervical
Moreover, PDEs often include fewer taxa for
vertebrae), laterally inflexible, dorsoven-
operational reasons: unstable taxa (e.g., those
trally flexible neck; low- to medium-level
with large amounts of missing data or character
browser (up to 5 m).
conflict) are often pruned from analyses a poste-
riori to increase resolution, and computational 2. Shunosaurus type: BC teeth; interdigitat-
requirements can limit the number of taxa ing occlusion (mesial and distal wear
included in an analysis. Conversely, TDEs can facets, occasional apical wear); orthal jaw
include all valid genera, even if these taxa are too action; cropping of vegetation; very low
poorly known to be included in a cladogram: FL/HL ratio (0.64); short (12 cervical ver-
poorly known taxa can be referred to clades on tebrae), laterally and dorsoventrally flexi-
the basis of synapomorphies, even if their posi- ble neck; low- to medium-level browser
tion cannot be fully resolved in the phylogeny. (up to 4.5 m).
Finally, although PDEs can extract more infor- 3. Higher eusauropod type: BC teeth; inter-
mation from the fossil record on diversity than digitating occlusion (mesial and distal
SAUROPODOMORPH DIVERSITY THROUGH TIME 127

wear facets); orthal jaw action; cropping vegetation; intermediate FL/HL ratio
of vegetation; intermediate Fl/HL ratios (0.79–0.87); short (12 cervical vertebrae),
(0.85–0.88); elongate (up to 17 cervical laterally inflexible and dorsoventrally flex-
vertebrae), dorsoventrally flexible, later- ible neck; medium to high browsing (up
ally inflexible necks; high browsers (up to to 9 m).
11 m). 10. Brachiosaurid type: BC teeth (but
4. Nemegtosaurid type: Narrow-crowned approaching the NC condition); precision-
(NC) teeth; precision-shearing occlusion shearing occlusion (apical wear facets
(mainly apical wear facets, occasional with occasional mesial and distal wear
mesial and distal wear); ?propalinal jaw facets); orthal jaw action; cropping and
action; “nipping” of vegetation; postcra- “nipping” of vegetation; high FL/HL ratio
nia unknown at present. (1.1); long (12 or 13 extremely elongate cer-
5. Nigersaurus type: NC teeth; ? precision- vicals), laterally inflexible, dorsoventrally
shearing occlusion (apical wear facets); flexible neck; high browsing (up to 14 m).
transversely oriented tooth rows; postcra- 11. Titanosaur type (additional postcranial
nial material undescribed (Sereno et al. data from Curry Rogers and Forster
1999; see Sereno and Wilson, chapter 5). 2001; see Curry Rogers, chapter 2): NC
6. Dicraeosaurid type: NC teeth; ?propalinal teeth; precision-shearing occlusion (api-
jaw action; precision-shearing occlusion cal wear facets); orthal jaw action; “nip-
(mainly apical wear facets, occasional ping” of vegetation; intermediate FL/HL
mesial and distal wear); “nipping” of ratios (0.75–0.83); 13 to 16 cervical verte-
vegetation; very low FL/HL ratio brae present, suggesting some variation
(0.65–0.70); short (11 or 12 cervical verte- in neck function within the group.
brae), laterally flexible, dorsoventrally Titanosaur anatomy is very poorly known
inflexible necks; specialist low browser at present: it is likely that new discoveries
(up to 4 m). will significantly alter our interpretations
7. Apatosaurus type: NC teeth; propalinal of the feeding mechanisms utilised by
jaw action; occlusion (apical wear facets); this clade.
branch-stripping/nipping of vegetation; Taxa were assigned to these categories on the
low FL/HL ratio (ca. 0.70); elongate (15 basis of the characters listed above or on phylo-
cervical vertebrae), laterally and dorsoven- genetic grounds, following mapping of these
trally flexible neck; tripodal ability; low characters on to the sauropod cladogram
and medium browsing (up to 8 m) (cf. Upchurch and Barrett 2000; phylogeny after
(Connely and Hawley 1998; Stevens and Upchurch et al. 2004). In cases where character
Parrish 1999; P.M.B. and P.U., pers. optimization was ambiguous (i.e., where posses-
obs.). sion of one of two different feeding systems was
8. Diplodocus type: NC teeth; propalinal jaw equally likely), taxa were not assigned a feeding
action; lack of occlusion; branch-stripping; mechanism and were excluded from the feeding
low FL/HL ratio (ca. 0.70); elongate (15 or related parts of the analysis.
16 cervical vertebrae), laterally and BC teeth are defined as those with slender-
dorsoventrally flexible neck; tripodal abil- ness indices (SIs) of 4.0, whereas NC teeth
ity; low, medium, and high browsing (up have a SI 4.0. SI is calculated by dividing the
to 12 m). apicobasal length of a tooth crown by its
9. Camarasaurus type: BC teeth; interdigitat- mesiodistal width (Upchurch 1998). In addition,
ing occlusion (mesial and distal wear BC teeth have a “spatulate” (D-shaped) cross sec-
facets); propalinal jaw action; cropping of tion; NC teeth, in contrast, have a cylindrical or

subcylindrical cross section. This division Although the dates for many European and
reflects a fundamental difference in the feeding South American forms are apparently well con-
mechanism: BC genera generally crop vegeta- strained, those from southern Africa, North
tion, whereas those taxa with NC teeth rake or America and China can often be referred only
“nip” vegetation (though there is great variety in to broad stratigraphic intervals. This problem is
the exact mechanism; see above). Where tooth particularly acute in the Early Jurassic. For
morphology is unknown, as a result of poor example, the age of the North American East
preservation, genera were assigned to one of Coast prosauropod fauna (Galton 1976) is
these categories on the basis of their phyloge- poorly constrained by radiometric and biostrati-
netic position (see appendix 4.2). graphical data and is currently considered to
Prosauropoda was not subdivided due to the extend over the duration of the Hettangian–
apparent uniformity of the feeding apparatus Pliensbachian. Similarly, many other Early
throughout the group. Jurassic prosauropods have ranges that extend
A full list of the genera included in the over two or more stratigraphic stages. This
TDEs, along with information on the strati- imprecision may have artificially “smeared” the
graphic range, systematic position and feeding tail end of the TDE, obscuring its true shape.
mechanism of each genus, is included in Furthermore, it is unclear, at present, whether
appendices 4.1 and 4.2. the Rhaetian reduction in prosauropod diversity
(fig. 4.1A) represents a “real” evolutionary event
or a preservational bias: fewer prosauropod-bear-
RESULTS: SAUROPODOMORPH DIVERSITY ing formations are known from the Rhaetian
THROUGH TIME than from either the Norian or the lowermost
Jurassic (see Upchurch and Barrett, chapter 3).
SAUROPODOMORPH DIVERSITY PATTERNS A PDE, which could help to test these differing
The earliest prosauropod (Saturnalia) appears in hypotheses, has yet to be calculated for
the middle Carnian (Langer et al. 1999). prosauropods. For the time being, we cautiously
Thereafter, the group radiates rapidly and seven accept the results of the TDE but are mindful
genera are present during the Norian (fig. 4.1A). that it contains a number of potential pitfalls.
The Rhaetian witnesses a rapid drop in the TDE, The sauropod TDE and its potential short-
to three genera, but following this trough comings are discussed in detail in chapter 3 so
prosauropod diversity apparently increases sud- are not dwelt on here. In outline, sauropods
denly again and reaches its acme (a peak com- first appear in the late Carnian (Blikanasaurus
posed of nine genera) in the early Sinemurian. [Galton and Upchurch 2004; Upchurch et al.
During the remainder of the Early Jurassic, 2004]) and are rare in Late Triassic and Early
prosauropod taxic diversity declines in a stepwise Jurassic deposits, though they exhibit a slight
manner to four genera in the late Pliensbachian. increase in diversity during this interval (fig
Prosauropods apparently became extinct close 4.1B). Sauropod diversification accelerates in
to the Early/Middle Jurassic boundary: their the Middle Jurassic, reaching a peak of 12 gen-
remains have not been identified in any deposits era in the middle Callovian before falling off in
younger than Pliensbachian (Weishampel et al. the Oxfordian. A second peak in the TDE occurs
2004). If the TDE reflects the “real” diversity of during the Kimmeridgian and Tithonian, repre-
these animals, it suggests that prosauropod senting the acme of sauropod diversity, with 26
decline was relatively gradual until the late valid genera. An apparent extinction event at the
Pliensbachian, at which time all remaining taxa Jurassic/Cretaceous boundary results in a mas-
went extinct (fig. 4.1A). sive drop in diversity (though the PDE suggests
Potential problems with this TDE include that this drop is not as marked as that suggested
uncertainties in the dating of several key genera. by the TDE; see chapter 3), which gradually

FIGURE 4.1. Sauropodomorph taxic diversity estimates (TDEs). (A) TDE for the Prosauropoda (based on data from Galton
and Upchurch [2004] and Weishampel et al. [2004]; see appendix 4.1). N 20. (B) TDE for the Sauropoda (based on data
from Upchurch et al. [2004] and Weishampel et al. [2004]; see appendix 4.2). N 98. (C) Plot of the relative abundances
(expressed as proportions of the total number of sauropodomorph genera) of prosauropods and sauropods in the Late
Triassic and Early Jurassic. See text for description and further details. Aal, Aalenian; Alb, Albian; Apt, Aptian; Baj, Bajocian;
Barr, Barremian; Bath, Bathonian; Berr, Berriasian; Call, Callovian; Camp, Campanian; Carn, Carnian; Ceno, Cenomanian;
Con, Coniacian; Hauv, Hauterivian; Hett, Hettangian; Kimm, Kimmeridgian; L, lower; M, middle; Maas, Maastrichtian; Nor,
Norian; Oxf, Oxfordian; Plie, Pliensbachian; Rhae, Rhaetian; Sant, Santonian; Sine, Sinemurian; Tith, Tithonian; Toar,
Toarcian; Turo, Turonian; U, upper; Val, Valanginian.
FIGURE 4.1. (Continued)
recovers through the Early Cretaceous to reach Triassic, early Middle Jurassic (Toarcian–middle
a third peak (of 12 genera) in the Albian. This is Bathonian), and basal Cretaceous (Berriasian–
followed by another apparent crash in diversity middle Valanginian) (fig. 4.2A), though the TDE
in the early Late Cretaceous (also reflected in and PDE clearly demonstrate the continuity of
the PDE; see chapter 3) before a final radiation several sauropod lineages through each of these
in the Campanian and Maastrichtian that intervals (Upchurch and Barrett, chapter 3).
resulted in a peak of 16 genera in the lower Absence of valid genera from the Turonian and
Maastrichtian (fig. 4.1B). Coniacian stages accounts for the zero value of
There was considerable temporal overlap the frequency histogram at this point. These
between prosauropods and sauropods, extend- deficits have been partially compensated for by
ing from the late Carnian to the end of the extrapolating the stratigraphic ranges of the var-
Pliensbachian. This is not surprising, as phyloge- ious feeding mechanisms so that they are con-
nies that recovered a monophyletic Prosauropoda tinuously present from their earliest to their
predict that sauropod origin must be at least as latest known occurrences. For example, the
far back in time as the first appearance of presence of titanosaurs in the Turonian can be
prosauropods. inferred from their presence in the
Cenomanian and Campanian, and their feeding
FEEDING MECHANISM DIVERSITY mechanism can be added to the frequency
Lack of direct information on the feeding mech- curve shown in figure 4.2A. Future work on
anisms of various taxa (e.g., basal sauropods, sauropod phylogeny (with the inclusion of
Bellusaurus) results in the zero frequency values more taxa) and the resulting PDEs for compo-
for sections of the histogram in the Late nent sauropod clades (e.g., Titanosauria,

FIGURE 4.2. Diversity of sauropod feeding mechanisms through time (based on data given in the text and appendix 4.2).
(A) Frequency of different sauropod feeding mechanisms through time, with a partial correction to account for missing data
(see text for details). (B) Plot of the relative abundances (as proportions of the total number of sauropod genera present) of
the various sauropod feeding mechanisms through time. Total number of sauropod genera represented 61 (this reflects
the lack of information, anatomical or phylogenetic, on the feeding mechanisms of many taxa). Abbreviations as in the leg-
end to figure 4.1.
Cetiosauridae) may also help to close these tem- observed on empirical evidence generally
poral gaps. remains below this level until the Aptian–
In general terms, the frequency histogram Albian. The extinction of dicraeosaurids at the
for different sauropod feeding mechanisms end of the Hauterivian causes a drop in predicted
through time is similar in shape to the sauro- diversity during the Barremian, but the appear-
pod TDE (fig. 4.2A). There is a strong, statisti- ance of the Nigersaurus-type feeding mechanism
cally significant, positive correlation between in the Aptian restores feeding diversity levels to
sauropod taxic diversity and the number of three categories.
feeding mechanisms at any one point in time Late Cretaceous feeding diversity does
(r2 0.759, p 0.0001). Feeding mechanism not track the sauropod TDE. Cenomanian–
diversity is low in the Early Jurassic, being Santonian sauropod faunas decrease in feed-
represented by only a single category ing mechanism diversity, with the loss of the
(Cetiosaurus type), though this paucity may brachiosaurid-type and then the Nigersaurus-type
simply reflect our ignorance of early sauropod categories. A small increase in feeding diver-
craniodental adaptations. The Middle Jurassic sity occurs in the late Campanian–middle
witnesses an increase in feeding mechanism Maastrichtian, reflecting the appearance of the
diversity, to three categories (Cetiosaurus type, nemegtosaurid-type feeding mechanism. It is
Shunosaurus type, and higher eusauropod type), possible that the latter had an extensive history
but this is followed by a sharp decrease in prior to its first appearance, as a long ghost lin-
diversity during the Oxfordian. Both the TDE eage is inferred if Nemegtosauridae occupies a
and the PDE indicate that the Oxfordian may basal position within Diplodocoidea (Upchurch
have been a genuine low point in sauropod 1999; Upchurch et al. 2004). Only the
diversity (Upchurch and Barrett, chapter 3), so titanosaur-type feeding mechanism continues
this corresponding drop in the number of dif- through to the end of the Cretaceous. Although
ferent feeding mechanisms at this time is not the sauropod TDE is high in the Late Cretaceous,
surprising. Feeding diversity reached its maxi- almost all of this diversity is confined to a sin-
mum in the Late Jurassic, at which time up gle clade (Titanosauria), accounting for the low
to seven different mechanisms are present diversity of feeding categories at this time. It
(higher eusauropod type, Camarasaurus type, should be noted that alternative phylogenies
brachiosaurid type, titanosaur type, Diplodocus place Nemegtosauridae within Titanosauria
type, Apatosaurus type, and dicraeosaurid type). (Salgado and Calvo 1997; Wilson and Sereno
The basal Cretaceous is marked by an apparent 1998; Curry Rogers and Forster 2001; Wilson
crash in diversity that is exacerbated by the cur- 2002), but this does not currently affect the
rent lack of information on the feeding mecha- number of sauropod feeding categories in the
nism of the only valid Berriasian genus Late Cretaceous. Existing data on nemegtosaurid
(Jobaria). Nevertheless, both the sauropod PDE and titanosaur feeding mechanisms warrant
(Upchurch and Barrett, chapter 3) and the separation between these categories for the
extrapolation of feeding mechanism strati- time being, though it is recognized that new
graphic ranges indicates that several sauropod material from either clade may significantly
lineages were present at this time and that a alter this conclusion.
minimum of three feeding categories was prob- The relative abundances of each feeding
ably in place (brachiosaurid type, titanosaur category also vary through time (fig. 4.2B).
type, and dicraeosaurid type). The extrapolated The Shunosaurus-type, Apatosaurus-type, and
frequency curve suggests that the same level of Camarasaurus-type feeding categories are all
feeding mechanism diversity was maintained restricted in stratigraphic range and are currently
throughout much of the Early Cretaceous, limited to a single genus, but each of these cate-
though the actual number of feeding categories gories may gain more members when more is

known about the feeding mechanisms in various relative and absolute diversity, with the appear-
basal sauropod, diplodocoid, and basal macronar- ance of several diplodocoid and titanosaur gen-
ian lineages. Cetiosaurid-type feeding mecha- era in the late Middle Jurassic and Late Jurassic.
nisms are the most abundant in the Early and In the same interval, BC sauropods (non-
early Middle Jurassic, but these are “replaced” neosauropod sauropods, basal macronarians,
by Shunosaurus-type and “higher” eusauropod- basal titanosauriforms) decline in relative diver-
type feeding mechanisms in the late Middle sity, though their absolute diversity actually
Jurassic and early Late Jurassic. (Diplodocoid increases slightly, and by the Kimmeridgian, NC
[‘Cetiosaurus’ glymptonensis] and basal titanosauri- and BC sauropods each account for approxi-
form (Lapparentosaurus) lineages were also pres- mately 50% of the total sauropod diversity (fig.
ent at this time, but their feeding mechanisms 4.3A, B). BC and NC sauropods decline in diver-
cannot be assessed with confidence on the basis sity from the Tithonian to the Barremian (reflect-
of the criteria outlined above and, so, do not ing the overall TDE), but the BC sauropods suf-
appear in the graph). During the Late Jurassic, fer a greater loss of genera (e.g., the extinction of
taxa with Diplodocus-type and brachiosaurid-type all non-neosauropod eusauropods) than the NC
feeding mechanisms are the most abundant: forms, so that in relative terms NC sauropods
the former disappear in the basal Cretaceous, increase in importance during this period (apart
however, whereas the latter continue to be from an isolated peak in BC diversity in the
moderately important until the end of the Barremian, at which time NC sauropods appear
Albian. Sauropods with titanosaur-type and to be rare). The increase in the TDE during the
dicraeosaurid-type mechanisms are also impor- Aptian–Albian is composed mainly of NC taxa
tant in the Early Cretaceous, though it should be (rebbachisaurids, titanosaurs), which account for
remembered that all feeding categories in this approximately 75% of the sauropod fauna. The
interval are based on a very small sample of valid last BC sauropods (all of which are basal
genera. Titanosaur-type feeding mechanisms titanosauriforms) disappear by the end of the
increase in relative diversity through the latter Early Cretaceous, and the Late Cretaceous peak
part of the Early Cretaceous, becoming the in the TDE is composed solely of NC taxa
dominant category in the Albian. Nigersaurus- (fig. 4.3A, B).
type sauropods also make a moderate contribu- Thus, sauropod faunas are initially com-
tion to late Early and early Late Cretaceous feed- posed of BC genera; there is an extended period
ing diversity. Most of the Late Cretaceous is of overlap between NC and BC genera during
dominated by titanosaur-type feeding mecha- the late Middle Jurassic–Early Cretaceous; BC
nisms, though there is a minor contribution forms gradually decline in importance in the
from nemegtosaurid-type genera in the late Early Cretaceous; and this trend culminates in
Campanian–middle Maastrichtian. Late Cretaceous faunas that contain only NC
genera.
OCCURRENCE OF TAXA WITH EITHER
NARROW-CROWNED OR COMPARISON BETWEEN NARROW-CROWNED
BROAD-CROWNED TEETH CLADES (DIPLODOCOIDEA AND
During the Late Triassic to middle Bathonian the TITANOSAURIA)
sauropod TDE is composed exclusively of taxa The radiation of diplodocoids begins in the late
with spatulate, BC teeth (basal sauropods, Bathonian, but the group remains poorly
cetiosaurids, and other non-neosauropod represented in the Callovian–Oxfordian interval
eusauropod lineages) (fig. 4.3A). The earliest (fig. 4.4A). The Kimmeridgian heralds a
probable NC form appears in the late Bathonian dramatic radiation of diplodocoid taxa, includ-
(the diplodocoid ‘Cetiosaurus’ glymptonensis). ing diplodocids (e.g., Diplodocus, Tornieria),
Thereafter, NC forms rapidly increase in both dicraeosaurids (Dicraeosaurus), and a variety of

FIGURE 4.3. Comparison of the diversities of sauropods with broad-crowned (BC) versus narrow-crowned (NC) teeth (raw
data given in appendix 4.2). (A) Abundances of BC and NC sauropods through time. Total number of genera included 92
(reflecting lack of information, both anatomical and phylogenetic, on tooth morphology in a number of genera). (B) Plot of
the relative abundances (expressed as proportions of the total number of sauropod genera) of sauropod taxa with either BC
or NC teeth through time. Abbreviations as in the legend to figure 4.1.
FIGURE 4.4. Diplodocoid and titanosaur TDEs through time. (A) Frequencies of diplodocoid (N 19) and titanosaur
(N 36) genera (see appendix 4.2). (B) Plot of the relative abundances of the two clades (expressed as proportions of the to-
tal number of diplodocoid titanosaur genera) through time. Abbreviations as in the legend to figure 4.1.
diplodocoids of uncertain systematic position Jablonski et al. 1996). Possible controlling fac-
(e.g., Amphicoelias), with up to 11 valid genera tors include large-scale environmental and tec-
present at this time. This represents the peak of tonic perturbations (e.g., climate change, conti-
diplodocoid diversity: thereafter, diversity nental fragmentation, changes in atmospheric
declines steeply in the basal Cretaceous. The late composition) or modifications to community
Early Cretaceous sees a moderate recovery, with structure and dynamics (e.g., appearance of new
the appearance of rebbachisaurids, but only four predators, loss of prey species, competitive exclu-
valid diplodocoid genera are present in the sion). Sauropodomorph diversity patterns have
Albian interval. Diplodocoids are apparently been largely exempt from such consideration,
absent in the early Late Cretaceous, though the apart from examination of the role of
appearance of the Nemegtosauridae in the late prosauropods in the initial stages of the dinosaur
Campanian indicates the presence of an exten- radiation during the Late Triassic (e.g., Benton
sive ghost lineage (Upchurch 1999; chapter 3; 1983, 1986). Nevertheless, several recent studies
but see comments above relating to the possible have attempted to rectify this deficit (Haubold
systematic position of nemegtosaurids). 1990; Hunt et al. 1994; Weishampel and Jianu
The earliest confirmed titanosaur is 2000; Barrett and Willis 2001; Upchurch and
Janenschia from the Kimmeridgian of Tanzania. Barrett, chapter 3).
Titanosaur diversity apparently remains very low Haubold (1990) and Hunt et al. (1994) have
through the Late Jurassic and basal Cretaceous: proposed that the fluctuations in sauropod TDEs
few valid genera are known at this time, though are not a direct product of evolutionary or envi-
phylogeny predicts the continuity of one or two ronmental processes, but nothing more than
lineages through this interval (Upchurch and taphonomic artifacts under the control of
Barrett, chapter 3). The group has a modest radi- changes in global sea level that affect the preser-
ation in the late Early Cretaceous, reaching a vation potential of sauropod remains. It has also
maximum of five genera in the Albian, but dwin- been suggested that sea level had a more direct
dles rapidly during the early Late Cretaceous. An influence on dinosaur diversity in terms of the
extensive, and apparently rapid, radiation occurs consequent geographical separation or conflu-
in the Campanian and Maastrichtian that is com- ence of terrestrial habitats, either by promoting
posed almost exclusively of derived titanosaurs allopatric speciation or as a result of biotic mix-
(fig. 4.4A). ing (e.g., Bakker, 1977; Weishampel and Horner
Comparison of the TDEs for each of these 1987). However, many of the suggested correla-
clades demonstrates that they coexisted for most tions between sauropod diversity patterns and
of their respective histories but that diplodocoids sea level changes collapse when PDEs are con-
were more diverse than titanosaurs (in both rel- sidered (Upchurch and Barrett, chapter 3). Here
ative and absolute terms) during the Late we evaluate the possible roles of several biologi-
Jurassic, whereas the converse is true in the Late cal processes that may have influenced sauropod
Cretaceous. Only in the late Early Cretaceous do diversity through time.
the two clades closely approach each other in
diversity (fig. 4.4B). FEEDING MECHANISMS AND DIVERSITY
The strong positive correlation between sauropod
taxic diversity and the number of different feed-
DISCUSSION
ing mechanisms present at any one time con-
Fluctuations in clade diversity are often assumed firms previous suggestions that feeding ecology
to have some macroevolutionary significance, played a central role in sauropod evolutionary
and many studies have attempted to identify abi- history (figs. 4.1B, 4.2A, B) (Barrett 1998;
otic or biotic factors that may influence or con- Upchurch and Barrett 2000). The strong
trol such phenomena (e.g., Valentine 1985; linkage between the Late Jurassic radiation of

sauropods and the diversification of feeding the skull and the postcrania of many genera may
mechanisms, and the coincident decline in the be obscuring a diversity of feeding mechanisms
sauropod TDE and in feeding mechanism within the group. For example, complete cervical
diversity during the Early Cretaceous, makes it series are unknown in many titanosaurs, so
tempting to speculate that there was a direct basic comparative data on neck flexibility and
causal link between these various phenomena. length are wanting. Moreover, the dearth of par-
These observations are also concordant with tially complete crania confounds the identifica-
the suggestion that niche partitioning, on the tion of any variation in the feeding apparatus.
basis of differences in feeding ecology, was Alternatively, a number of other possible mech-
important in faunas that contained a variety of anisms may permit niche partitioning among
sympatric sauropod taxa (Fiorillo 1991, 1998; sympatric titanosaur genera, including differ-
Barrett and Upchurch 1995; Upchurch and ences in adult body mass (which would affect
Barrett 2000). The peaks in the sauropod TDE browse height, assimilation rates, mobility,
and in feeding mechanism diversity coincide in population sizes, etc.) and differing dietary
the Late Jurassic at a time when the majority of preferences.
well-known dinosaur faunas, such as those of the
Morrison Formation (USA), the Tendaguru Beds COMPETITIVE REPLACEMENTS:
(Tanzania), and the Upper Shaximiao Formation PREVIOUS WORK
(China), contain a melange of sauropods from a Although the role of competitive replacement in
number of clades, which possess a number of macroevolution has been the subject of intense
different feeding mechanisms. The decline in debate, there is currently little consensus on its
the TDE and the reduction in feeding mecha- frequency or on its importance (relative to other
nism diversity in the Early Cretaceous are con- biological and abiological processes) in control-
current with a change in sauropod faunas at this ling clade diversity over geological time scales
time: with rare exceptions these faunas now con- (e.g., Benton 1996; Sepkoski 1996). A recent
tain only one or two sauropod genera and a cor- attempt to quantify the prevalence of competitive
respondingly low diversity of feeding ecologies replacements in the evolutionary history of
(Weishampel et al. 2004). tetrapods recognized a number of “candidate
The correlation between generic-level diver- competitive replacements” (CCRs) and com-
sity and feeding mechanism variety does not pared the number of familial originations that
continue into the Late Cretaceous, however could have been the result of a competitive
(Figs. 4.1B, 4.2A, B). During this period there is replacement with the total number of origina-
a significant radiation of titanosaurs, culminat- tion events through time (Benton 1996). CCRs
ing in a maximum of 15 genera, which is were identified where two tetrapod families dis-
supplemented by the appearance of the played (1) an overlap in geographical and strati-
Nemegtosauridae. This represents the second graphic ranges and (2) a similar mode of life (in
highest peak in sauropod diversity: however, it terms of body mass, diet, and habitat). Similar
appears that only two feeding mechanisms criteria for recognizing potential competitive
underpin this radiation. The co-occurrence of replacements were also advocated by Sepkoski
several titanosaur genera in South American, (1996).
Indian, and Madagascan faunas (Curry Rogers Of the CCRs proposed by Benton (1996),
and Forster 2001; Weishampel et al. 2004) sug- three involve sauropodomorph clades:
gests that niche partitioning has occurred in replacement of Melanorosauridae by the
these instances despite the paucity of different Vulcanodontidae and Massospondylidae in the
feeding mechanisms. This may simply reflect Hettangian; replacement of Camarasauridae by
our ignorance of the variation in the titanosaur the Cetiosauridae in the Oxfordian; and replace-
feeding apparatus: lack of information on both ment of Cetiosauridae by the Titanosauridae in

the Kimmeridgian. However, the identifications demonstrated to be artificial groupings of
of these CCRs are founded on a database sauropods with diverse taxonomic affiliations
(Benton 1993) that contains a large number of (Upchurch 1995, 1998; Wilson and Sereno
taxonomic and systematic errors. According to 1998; Wilson 2002; Upchurch et al. 2004). For
Benton (1993), data on sauropodomorph fami- example, Euhelopus and Opisthocoelicaudia were
lies were taken from Weishampel (1990), whose once considered to be “camarasaurids” but have
taxonomy was based in turn on the compila- since been realised to be only distantly related
tions of Galton (1990; on prosauropods) and to Camarasaurus: Euhelopus is either a non-
McIntosh (1990; on sauropods). Subsequent neosauropod or a titanosauriform, whereas
work on sauropodomorph taxonomy and sys- Opisthocoelicaudia is a titanosaur (Upchurch
tematics has resulted in a large number of 1995; Wilson 2002). Similarly, the ‘cetiosaurid’
changes to the classifications proposed by Haplocanthosaurus is currently considered to be
Galton and McIntosh (e.g., Upchurch 1995, a basal macronarian or diplodocoid. As both
1998; Wilson and Sereno 1998; Wilson 2002; groups were so poorly defined and character-
Galton and Upchurch 2004; Upchurch et al. ized, any comparison between their geographi-
2004) that impact directly on Benton’s (1996) cal and their stratigraphic ranges becomes
proposed CCRs. Each CCR is dealt with in turn, somewhat meaningless. Furthermore, as these
below. groups (sensu McIntosh 1990; Benton 1993)
contain animals with different feeding mecha-
MELANOROSAURIDAE VS. VULCANODONTIDAE/ nisms, they cannot be regarded as ecologically
MASSOSPONDYLIDAE homogeneous and so mode of life comparisons
‘Vulcanodontidae’ is a paraphyletic grouping of between them are also undermined.
basal sauropods from the Early Jurassic Current classifications, which reflect the
(Upchurch 1995). As such, its geographical and topology of sauropod phylogenetic trees, restrict
stratigraphic ranges cannot be compared to membership of the Camarasauridae to the Late
those of a monophyletic clade, as these attrib- Jurassic North American genus Camarasaurus.
utes of a paraphyletic taxon are dependent on a Use of the term Cetiosauridae has also been lim-
subjective choice on which taxa to include ited, so that it refers only to a monophyletic
within it. Melanorosauridae is currently consid- clade that includes the Early and Middle Jurassic
ered to be a monophyletic clade, but it does not genera Barapasaurus (India), Cetiosaurus (UK),
include the Chinese form Lufengosaurus (Galton and Patagosaurus (Argentina) (Upchurch et al.
1990; Galton and Upchurch 2004; contra 2004). As the spatial and temporal distributions
Benton 1993). As a result, this clade ranges in of these families did not coincide, it is unlikely
time from the ?late Carnian to the Rhaetian but that any competitive interactions could have
does not extend into the Early Jurassic as sug- occurred between them and this CCR is
gested by Benton (1993). rejected.
Therefore, melanorosaurids could not have
interacted with the Early Jurassic genus CETIOSAURIDAE VS. TITANOSAURIDAE
Massospondylus (the only member of the mono- The more restrictive definition of Cetiosauridae
typic Massospondylidae) or with any of the basal (see above) removes the temporal and spatial over-
sauropods included in the ‘Vulcanodontidae’ lap between this clade and the Titanosauridae.
(sensu Benton 1993). The validity of this CCR is, The last cetiosaurid (Patagosaurus: Argentina)
therefore, highly doubtful. becomes extinct at the end of the Callovian, well
before the first appearance of the first titanosaur
CAMARASAURIDAE VS. CETIOSAURIDAE ( Janenschia: East Africa). Consequently, the condi-
The Camarasauridae and Cetiosauridae of tions for a potential competitive replacement
McIntosh (1990) and Benton (1993) have been event are not met.

There has been some debate regarding the variety of early sauropods (Dong et al. 1983; He
effects of taxonomic revision on estimates of et al. 1998; Barrett 1999, 2000; P.M.B. and P.U.,
clade diversity (e.g., Patterson and Smith 1987; pers. obs.). In southern Africa, the prosauropod
Sepkoski and Kendrick 1993; Robeck et al. Massospondylus was a contemporary of the sauro-
2000). The foregoing examples from sauropods pod Vulcanodon, though remains of both animals
demonstrate that revision of a group and have yet to be recovered from the same locality
removal of para- and polyphyletic taxa can (Weishampel et al. 2004). Furthermore, there
indeed have a major impact on the interpreta- is overlap between the modes of life of the vari-
tion of the diversity patterns obtained from such ous members of these two clades. Sauropods
studies. Examination of sauropod diversity were obligate herbivores, while many
within a rigorous phylogenetic framework has prosauropods were at least facultative herbi-
fatally undermined several perceived, but erro- vores (Farlow 1987; Barrett 2000; Upchurch
neous, patterns that were based on more “tradi- and Barrett 2000) and members of both clades
tional” noncladistic classifications. Although are found in the same deposits, indicating that
assessment of the above-mentioned CCRs sug- they shared the same habitats (see above).
gests that competitive exclusion was not an There is also significant overlap in body size:
important process in sauropodomorph evolu- prosauropods range from approximately 1.5 m
tionary history (see also Benton 1996), several (Saturnalia [Langer et al. 1999]) to 10 m
patterns revealed by the prosauropod and sauro- (Riojasaurus [Bonaparte 1972]) in length,
pod TDEs are suggestive of possible competitive whereas the earliest sauropods occupied the
replacement events (see below). size range from 5 m (Blikanasaurus [Galton
1990]) to 14 m (Gongxianosaurus [He et al.
COMPETITION BETWEEN PROSAUROPODS 1998]).
AND SAUROPODS? Prosauropods were the dominant vertebrates
A growing body of evidence indicates that of Late Triassic and Early Jurassic terrestrial fau-
there was considerable overlap between the nas in terms of biomass, and it has been esti-
stratigraphic and geographical ranges of mated that in some faunas they accounted for up
prosauropods and sauropods during the to 95% of the standing crop (Galton 1985). In
Late Triassic and Early Jurassic (fig. 4.1). In contrast, sauropods were rare during this time
the Late Triassic, remains of the sauropod interval and the majority of genera are repre-
Isanosaurus (late Norian–Rhaetian, Thailand sented only by fragmentary remains. However,
[Buffetaut et al. 2000]) were recovered from the comparison of prosauropod and sauropod TDEs
same unit as indeterminate prosauropod mate- demonstrates that during the Early Jurassic the
rial (Buffetaut et al. 1995); similarly, the sauro- number of prosauropod genera steadily declined
pod Blikanasaurus (?late Carnian–?early Norian, while sauropod diversity remained roughly con-
South Africa [Upchurch et al. 2004]) was a stant (fig. 4.1). In terms of relative diversity,
member of the diverse Lower Elliot Formation prosauropods gradually decreased in importance
dinosaur fauna, which also included a minimum from the Norian (when they made up close to
of two prosauropod genera (Euskelosaurus and 100% of the sauropodomorph fauna) to the
Melanorosaurus [Weishampel et al. 2004]). point immediately before their final extinction at
The association between prosauropods and the Pliensbachian–Toarcian boundary (at which
sauropods continues in the Early Jurassic, par- time they constituted approximately 60% of
ticularly in China, where the prosauropods of sauropodomorph taxa). Conversely, sauropods
the Lower Lufeng Formation and adjacent increased in relative diversity during the same
penecontemporaneous faunas (Hettangian– interval, from close to 0% to approximately
Pliensbachian [Young 1951; Dong 1992; Luo 40% of the sauropodomorph fauna. Such a
and Wu 1994; Barrett 2000]) lived alongside a pattern is strongly reminiscent of the classic

“double- wedge” (cf. Benton 1983) that is the sine presence or absence of occlusion. However,
qua non of competitive replacements (fig. 4.1C). other basal sauropods did acquire novel fea-
The evidence presented above demonstrates tures in the feeding apparatus that are absent
that prosauropods and sauropods fulfill all of from prosauropods, including the development
the criteria proposed by Benton (1996) for the of a lateral plate and various modifications
recognition of a CCR: therefore, it is possible to tooth crown morphology (Barrett 1999;
that the early stages of the sauropod radiation Upchurch and Barrett 2000): these indicate
may have driven prosauropod tabefaction (see that sauropod feeding mechanisms were some-
also Barrett 1998). The small number of taxa what different from those of prosauropods.
involved makes it difficult to assign this CCR to Several authors have suggested that
either a “full competitive replacement” or an prosauropods may have been paraphyletic with
“incumbent replacement” model (cf. Benton respect to sauropods (Gauthier 1986; Benton
1996). Nevertheless, both of these models 1990; Yates 2003), but this does not significantly
assume that possession of a “key innovation” affect the proposed prosauropod/sauropod inter-
gives the “successful” incoming group a com- actions outlined above. As prosauropod feeding
petitive advantage over the “unsuccessful” out- mechanisms are so conservative, they can all be
going group. In this case, convergence in mode accommodated within a single feeding category:
of life does suggest that prosauropods and consequently, the only correction that would
sauropods may have been competing directly need to be applied to the discussion above would
for the same resources over an extended time be to regard prosauropods as a paraphyletic
scale. Although the modes of life of these ani- grade, rather than a monophyletic clade.
mals apparently overlapped considerably, the Although differences in the feeding appara-
development of a systematic occlusion (exten- tus may have provided sauropods with an evo-
sive tooth–tooth contact) may have been a “key lutionary advantage, these suggestions must be
innovation” that could have given sauropods a regarded as tentative: the small number of taxa
competitive advantage over prosauropods in the involved precludes rigorous statistical testing
Early Jurassic. No prosauropod is known to and our knowledge of Late Triassic and Early
have employed tooth–tooth contact in the col- Jurassic dinosaurs is poor. Future work on basal
lection or mechanical breakdown of plant food: sauropodomorph phylogeny and functional
although such a mechanism had been proposed morphology, discovery of new material, and
for Yunnanosaurus (Galton 1985), reexamina- more fine-grained comparisons of prosauropod
tion of the pertinent specimens demonstrated and sauropod TDEs and PDEs (by component
that this was not the case (Salgado and Calvo monophyletic lineages) will permit more rigor-
1997; Wilson and Sereno 1998; Barrett 1999, ous evaluation of this hypothesis. Finally, it
2000; P.M.B. and P.U., pers. obs.). In contrast, must also be remembered that other factors
one unnamed basal sauropod from the Lower could also have influenced sauropodomorph
Lufeng Formation of China possesses teeth that diversity during this time interval (see below).
have well-developed high-angled mesial and
COMPETITION BETWEEN SAUROPOD GRADES
distal wear facets (Barrett 1999, 2000);
AND CLADES?
Gongxianosaurus also appears to display
tooth–tooth wear (He et al. 1998). Occlusion BROAD- AND NARROW-CROWNED CLADES
would have permitted a higher degree of oral The BC and NC sauropod groupings (see above)
processing of plant food and/or may have are artificial constructs in a phylogenetic sense:
opened up food sources that were unavailable BC clades are paraphyletic with respect to NC
to animals that lacked it. Unfortunately, the clades, and the NC group has diphyletic origins.
craniodental material of other basal sauropods Nevertheless, comparisons between them may
is not sufficiently well preserved to confirm the be justified, as each group can be defined as a

distinct ecomorphotype or grade (on the basis of These independent lines of evidence sug-
general resemblance in tooth morphology). gest that factors other than competitive replace-
Plotting the frequency of sauropods with BC ment may need to be sought to account for the
teeth relative to that of NC forms reveals an decline in BC sauropod diversity during the
apparent negative correlation between the diver- Early Cretaceous. Furthermore, it might be
sity of these two groups, forming another “dou- unrealistic to expect competition between BC
ble-wedge” pattern that may be indicative of a and NC sauropods, the apparent “double-wedge”
competitive replacement (fig. 4.3B). However, notwithstanding: the varied feeding mecha-
while the stratigraphic ranges of the groups nisms of the component taxa in each grouping
exhibit considerable overlap (as do various probably helped to minimize competition, allow-
mode of life traits, such as body size), compari- ing niche partitioning in diverse sauropod fau-
son of their geographical ranges suggests that nas, such as those of the Late Jurassic Morrison
competition was unlikely, at least on the basis of Formation and Tendaguru Beds (Fiorillo 1991,
current data. 1998; Barrett and Upchurch 1995; Upchurch
Although BC and NC sauropods coexisted in and Barrett 2000).
several late Middle and Late Jurassic faunas
from Europe, North America, and Africa, there DIPLODOCOIDS AND TITANOSAURS
were no NC forms in Asia or South America Plotting the frequencies of diplodocoid and
during this interval (appendix 4.2 [Weishampel titanosaur taxa against each other through time
et al. 2004]). In the Early Cretaceous most NC reveals an apparent negative correlation
forms are found in Africa and South America: between the relative diversities of these two
both are areas in which BC taxa were either rare clades (fig. 4.4B). However, this pattern cannot
or absent. Examples of Early Cretaceous faunas be regarded as strong evidence for competitive
that contain both NC and BC taxa (such as the interactions between these two groups, for
Wessex Formation [Barremian] of the Isle of several reasons. During the Late Jurassic
Wight) are rare: most faunas of this age do not diplodocoids peak in diversity and are found in
contain both BC and NC taxa (appendix 4.2) Europe (Dinheirosaurus), Africa (Tornieria,
(Weishampel et al. 2004). Consequently, it is Dicraeosaurus), and North America (Diplodocus,
unlikely that there was significant competition Barosaurus, and many others). In contrast,
for resources between these two groups in the titanosaurs are extremely rare at this time and
Early Cretaceous due to their limited spatial are apparently confined to Africa (Janenschia):
overlap; competition may have been more consequently, there would have been little inter-
important in Late Jurassic faunas but would action between these two clades. This period is
have been limited to several distinct regions. followed by a significant drop in diplodocoid
Furthermore, patterns arising from assess- diversity at the Jurassic/Cretaceous boundary,
ing the relative diversity of taxa are misleading with the extinction of diplodocids and several
in this case. Although BC sauropods apparently related genera of uncertain systematic position
lose ground to NC genera in the Late Jurassic in (e.g., Amphicoelias): however, the extremely lim-
plots of relative diversity, this change in propor- ited interactions between diplodocoids and
tions coincides with a substantial increase in titanosaurs before this event suggests that com-
the total number of sauropod taxa (due largely petition between these clades was not an impor-
to the rapid radiation of NC forms, particularly tant factor in diplodocoid tabefaction. Although
diplodocoids). Thus, while BC sauropods titanosaurs increase in relative abundance
decline in relative terms in the Late Jurassic, the during the basal Cretaceous, this pattern results
actual number of BC taxa remains high, and is from a combination of extinctions among
equal to the number of NC taxa, until the Early diplodocoid lineages and the small sample size
Cretaceous (compare fig. 4.3A and B). for valid sauropod genera during this interval.

During the Berriasian–Barremian there are only tosaurid genera are known it is currently impos-
one or two genera representing each of these sible to demonstrate the presence of a “double-
clades at any one time: indeed, for much of the wedge” in diversity. Finally, the long stratigraphic
Early Cretaceous a single genus represents the overlap between titanosaurs and nemegtosaurids
diplodocoid clade (Losillasaurus, from Spain). (which potentially extends back into the latest
The exact age of Losillasaurus is uncertain and it Jurassic; see chapter 3) suggests that competition
is currently unknown whether it pertains to between these clades may have been negligible.
either the Late Jurassic or the Early Cretaceous As a result, it is extremely difficult to decide
(Casanovas et al. 2001): hence, more precise whether competition, or some other factor, was
dating of this single genus could result in a the driving force behind the final extinction of the
major change to the relative abundances of Nemegtosauridae.
diplodocoids and titanosaurs during this interval. The case for a competitive replacement is
Moreover, geographical overlap between stronger in the late Early Cretaceous and early
Diplodocoidea and Titanosauria is restricted dur- Late Cretaceous, at a time when titanosaurs and
ing the basal Cretaceous: no faunas of this age rebbachisaurid diplodocoids coexist in several
contain valid genera from both clades (see African and South American faunas (see appen-
appendix 4.2). Several faunas apparently contain dix 4.2). Diversity of both groups in this interval
identifiable elements that provide evidence for is relatively low, with a maximum of five
some spatial overlap between these groups titanosaur and four diplodocoid genera present.
(Weishampel et al. 2004), but many of these Titanosaurs are consistently more diverse than
identifications may be suspect, as they are based diplodocoids at this time, but the TDEs for each
on very fragmentary material, such as isolated group display coincident rises in diversity
teeth or vertebral centra. As a result, the evidence through the Aptian and Albian stages
for competition between these clades during this (fig. 4.4A). In the Cenomanian, both groups
time interval is weak. decrease in diversity, but diplodocoids suffer
The TDE appears to indicate that diplodocoids a much greater loss of genera, with reb-
were absent for much of the Late Cretaceous, but bachisaurids becoming extinct at the end of this
PDEs demonstrate that at least one lineage stage. The geographical and stratigraphic overlap
(Nemegtosauridae) was present throughout this of the two clades suggests that competition with
interval (Upchurch and Barrett, chapter 3). This titanosaurs may have been a factor in reb-
suggests that competition between titanosaurs bachisaurid decline. Modes of life may also have
and nemegtosaurids was possible, as the two been similar due to the observed convergence in
clades overlapped geographically and stratigraph- skull form between titanosaurs and diplodocoids
ically. Moreover, mode of life comparisons are in (see above). However, more information is
accord with this suggestion: recent work has needed on the feeding mechanisms (both cranial
demonstrated extensive homoplasy in the skull and postcranial features) of rebbachisaurids and
morphology of titanosaurs and diplodocoids titanosaurs to confirm or reject this proposition.
(Upchurch 1999; Coria and Chiappe 2001) so it Additional data from PDEs may also help to clar-
is possible that they were competing for the same ify the patterns of titanosaur and diplodocoid
food resources in their shared habitats. radiation and tabefaction, particularly for the
Nevertheless, the extremely limited diversity of poorly known basal Cretaceous and early Late
nemegtosaurids, coupled with their very Cretaceous intervals.
restricted geographical distribution, suggests that
they would have been very vulnerable to extinc- POTENTIAL INTERACTIONS BETWEEN
tion and that a large number of potential mecha- SAUROPODOMORPHS AND ORNITHISCHIANS
nisms could have been responsible for their final Galton (1973, 1976) suggested that the decline of
disappearance. Moreover, as so few nemeg- prosauropods in the Early Jurassic may have

been linked to the appearance of ornithischian faunas and did not account for a significant pro-
dinosaurs with superior food processing and portion of the biomass (Galton 1985), even in
locomotor abilities. As mentioned above, those faunas that contain a diversity of taxa.
prosauropod feeding systems were simple and For example, although three species of het-
stereotyped, and did not permit extensive chew- erodontosaurid are known from the Upper
ing of plant material (Galton 1985, 1986; Elliot and Clarens Formations of South Africa,
Crompton and Attridge 1986; Norman and each species is known from only a handful of
Weishampel 1991). In contrast, the development specimens (Weishampel and Witmer 1990),
of occlusion and the inferred presence of exten- whereas it has been estimated that the con-
sive fleshy cheeks in ornithischians were viewed temporaneous prosauropod Massospondylus is
as key adaptations that greatly increased the effi- represented by at least 80 partial skeletons
ciency of mastication (Galton 1973, 1976). (Galton 1990). Third, most Early Jurassic
A brief survey of prosauropod and ornithis- ornithischians were small, with maximum
chian distribution patterns demonstrates that body lengths ranging between 1 and 2 m and
the two groups did coexist in several Early maximum browse heights of 0.5 to 1 m (Barrett
Jurassic faunas (Weishampel et al. 2004), 1998). In contrast, Early Jurassic prosauropods
including those of the Kayenta Formation in the ranged from 2.5 to 7 m in length, with maxi-
western United States (containing the prosauro- mum browse heights of 3 to 5 m (Barrett 1998;
pod Massospondylus and at least two ornithis- P.M.B. and P.U., pers. obs.). Consequently,
chian taxa—Scutellosaurus and an undescribed Early Jurassic prosauropods and ornithischians
heterodontosaurid ornithopod) and the Upper do not fulfill Benton’s (1996) criteria for the
Elliot and Clarens Formations of South Africa recognition of a CCR. The differences in body
(which include Massospondylus, a variety of het- size and maximum browse height indicate that
erodontosaurids, and the basal ornithischian these animals foraged at different levels: it is
Lesothosaurus). Moreover, ornithischian diversity likely that the food requirements of the two
was increasing in the Early Jurassic, with the groups also differed significantly. Moreover, the
radiation of heterodontosaurids, and basal low biomass of ornithischians and their patchy
thyreophorans, at a time when prosauropod geographical distribution suggest that their
diversity was declining (Weishampel and Jianu interactions with prosauropods and contempo-
2000; Barrett and Willis 2001). Nevertheless, rary vegetation would have been limited in
several lines of evidence suggest that competi- extent.
tion with ornithischians was not an important Although the geographical and stratigraphic
factor in prosauropod extinction. overlaps between sauropod and ornithischian
First, the geographical overlap between taxa are more extensive and numerous than
prosauropods and ornithischians in the Early those between the latter and prosauropods (see
Jurassic was limited. Ornithischians are not listings in Weishampel et al. 2004), there is no
present (or are only represented by rare and evidence for the competitive replacement of any
fragmentary material) in many faunas that con- sauropod clade by ornithischians, or vice versa.
tain abundant prosauropods, such as those of Substantial differences in the modes of life of
the Lower Lufeng Formation of China, the these two groups probably precluded competi-
McCoy Brook Formation of Canada, and the tion in most cases. Many sauropod taxa pos-
Portland Formation of the eastern United sessed adaptations for high browsing (up to a
States (Weishampel et al. 2004). Conversely, potential 14 m above ground level) that would
prosauropods are absent from several faunas have enabled them to gather vegetation that was
that yield Early Jurassic ornithischians inaccessible to ornithischians (with maximum
(Weishampel et al. 2004). Second, ornithischi- browse heights of between 1 and 5 m [e.g.,
ans were rare components of Early Jurassic Bakker 1978]). Order of magnitude variance in

the modal body masses of the two groups, cou- this period. (3) The major radiations of
pled with fundamental differences between the Cretaceous ornithischians occurred in the late
masticatory apparatuses of the various Early Cretaceous (iguanodontian ornithopods,
ornithischian and sauropod clades, may also nodosaurid ankylosaurs) and Late Cretaceous
have resulted in radically different food plant (hadrosaurian ornithopods, ceratopians, anky-
preferences or foraging strategies that would losaurid ankylosaurs) and, on the basis of cur-
minimize competition for resources. rent data, cannot be implicated in sauropod
Sauropods that were specialised for low brows- decline in the basal Cretaceous.
ing (e.g., dicraeosaurids) may have competed Nevertheless, it is interesting to note that fau-
with large ornithischians (e.g., iguanodontian nas dominated by sauropods, such as those from
ornithopods), but problems with regional sam- the Late Jurassic of North America (Dodson et al.
pling biases (see below) in combination with the 1980), China (e.g., Dong 1992), and East
low diversity of low-browsing sauropod clades Africa (Russell et al. 1980) and from the Late
make it difficult to test this hypothesis. In partic- Cretaceous of South America (Bonaparte 1996)
ular, it is difficult to determine whether the coin- and Madagascar (Curry Rogers and Forster
cident diversity changes that are necessary to sat- 2001), contain relatively few ornithischians.
isfy the CCR criteria occurred (i.e., did low- Conversely, ornithischian-dominated faunas,
browsing sauropods become less diverse as such as those of Early Cretaceous Europe (e.g.,
ornithischians increased in diversity in the same Martill and Naish 2001) and Late Cretaceous
region?). North America (e.g., Russell 1989) do not yield
The global TDEs and PDEs of the two abundant sauropod remains. However, strong
groups are broadly similar for much of the sampling biases currently prevent analysis of
Mesozoic, with peaks in diversity during the the fine-grained changes in regional dinosaur
Late Jurassic, late Early Cretaceous, and latest faunas through time that would permit the
Cretaceous and times of lower diversity in the identification (or robust refutation) of CCRs.
Middle Jurassic, basal Cretaceous, and early For example, our knowledge of North American
Late Cretaceous (see Weishampel and Jianu basal Cretaceous faunas is extremely limited:
2000; Barrett and Willis 2001; Upchurch and consequently, there are simply not enough data
Barrett, chapter 3). Consequently, there are no to determine whether the change from the
periods in which global ornithischian diversity sauropod communities of Morrison times to
increases coincide with significant decreases in the ornithischian communities of the late Early
sauropod diversity. For example, the extinction Cretaceous was gradual and incremental (and
of many sauropod lineages at the Jurassic/ thus potentially a series of competitive replace-
Cretaceous boundary cannot be linked to the ments) or whether the sauropod decline was
perceived increase in ornithischian abundance abrupt and not linked to the diversity or abun-
(cf. Bakker 1978), for three reasons. (1) dance patterns of any ornithischian group.
Ornithischian taxa were less diverse and abun- Similarly, dinosaur faunas from the pre-Late
dant than sauropods in the Late Jurassic (e.g., Cretaceous of South America and Madagascar
Bakker 1978), so competition between the are rare, so it is not possible to determine the
groups was probably not intense at this time. sequence of events that led to the dominance of
(2) Both sauropod and ornithischian lineages titanosaurs in these regions, at least on the
suffered extinction at the Jurassic/Cretaceous basis of current data.
boundary, resulting in approximately equal As the criteria for sauropod/ornithischian
diversities for the two clades in basal CCRs are not met (see above), reasons other
Cretaceous faunas (Barrett and Willis 2001)— than competition need to be sought to explain
ornithischian diversity did not markedly the Cretaceous decline of sauropods in some
increase relative to that of sauropods during regions (principally in Laurasia) and their

persistence/diversification in others. One pos- extinction of the “Dicroidium flora”/rhynchosaur
sibility is that regional extinction of specific fauna in the late Norian permitted a global exten-
sauropod clades was responsible for creating sion of the northern conifer flora that promoted
the observed regional diversity and abundance the rise to dominance, and further diversifica-
patterns, though evaluating the evidence for tion, of dinosaurs (largely prosauropods) on the
(and the causes of) such extinctions is beyond southern continents. However, examination of
the scope of this chapter. the prosauropod TDE, and consideration of
prosauropod distribution and palaeoecology,
SAUROPODOMORPH–PLANT INTERACTIONS indicates that this hypothesis is flawed in two
Observations on extant ecosystems have respects. First, Gondwanan prosauropods are not
demonstrated that large herbivores have a con- rare in the Norian: indeed, this is the time at
siderable impact on the growth, structure, and which they reach their peak diversity, with abun-
diversity of vegetation: consequently, we might dant material of several African and South
expect that extinct herbivores had similar American genera (e.g., Galton 1990; Bonaparte
effects on coeval floras (e.g., Weishampel 1984; 1996). Second, the late Norian extinction of the
Wing and Tiffney 1987; Weishampel and “Dicroidium flora” did not correlate with an
Norman 1989; Tiffney 1992; Barrett and Willis increase in Gondwanan prosauropod diversity,
2001). With the exception of a few small but with an abrupt decline, from a maximum of
prosauropods, almost all sauropodomorphs six taxa in the Norian to two taxa in the Early
had body masses in excess of 1,000 kg and can Jurassic (appendix 4.1). Thus, if the record is read
be considered to be “megaherbivores” (sensu literally, this might suggest that the floral replace-
Owen-Smith 1988). It is likely that populations ment drove the tabefaction of Gondwanan
of such large herbivores would have had a sig- prosauropods, rather than their radiation (contra
nificant impact on Mesozoic floras, through the Benton 1983). However, although the drop in
agencies of browsing and other forms of habitat prosauropod diversity coincides with the disap-
disturbance (such as trampling). Conversely, it pearance of the “Dicroidium flora,” the large
would also be expected that large-scale floral number of prosauropod specimens from the
replacements, or the evolution of novel plant Early Jurassic strata of southern Africa (Cooper
defences (chemical or physical), would have 1981; Gow et al. 1990; Galton 1990) indicates
affected the diversity and/or distributions of that prosauropod biomass remained high and
contemporary herbivores. that they continued to be the dominant terrestrial
During the early Late Triassic (Carnian– vertebrates (though it should be noted that
early Norian), Gondwanan lowland floras were remains of prosauropods are extremely rare in
dominated by the seed fern Dicroidium (with other Early Jurassic Gondwanan localities,
smaller numbers of conifers and other plants), although this may be due to poor sampling).
whereas floras in Europe and North America These contradictory observations make it diffi-
lacked seed ferns and were composed largely of cult to evaluate the significance of this floral
conifers, ferns, and cycadophytes (Benton 1983). replacement with respect to the contemporary
Investigation of contemporaneous animal com- fauna: more detailed study of floral and faunal
munities suggested that the southern distributions, and of potential plant–animal
“Dicroidium flora” was associated with a verte- interactions, is needed. Nevertheless, extinction
brate herbivore fauna dominated by rhyn- of the “Dicroidium flora” does not seem to
chosaurs (in terms of both diversity and have promoted the radiation of Gondwanan
biomass), with rare dinosaurs, while the prosauropods.
northern “conifer flora” coexisted with a diverse, There were no major floral replacement
abundant dinosaur fauna consisting mainly of events in the Jurassic and no major plant clades
prosauropods. Benton (1983) proposed that originated during this period (Niklas et al. 1985;

Wing et al. 1992). Moreover, the diversity of vas- early Late Cretaceous, is only weakly supported
cular plants remained relatively constant at present. Mode of life differences and dispari-
throughout this interval, with very low species ties in geographical distribution suggest that
turnover rates (Niklas et al. 1985; Niklas 1997). competition between sauropodomorphs and
Consequently, it is unlikely that the major ornithischians was restricted or unimportant,
changes in sauropod diversity that occurred dur- at least in macroevolutionary terms: evidence
ing the Middle and Late Jurassic resulted from for the supposed CCRs between these clades is
any major restructuring of contemporary plant extremely weak. There are stronger indications
communities. The late Early Cretaceous peak in that competition from sauropods may have
the sauropod TDE and PDE coincides with the been a significant factor in prosauropod decline
initial stages of the angiosperm radiation, but during the Early Jurassic, as both of these
the low biomass and restricted geographical dis- clades fulfil the requirements for a CCR in this
tribution of angiosperms at this time suggest interval. However, this suggestion is made ten-
that sauropod–angiosperm interactions would tatively, given the limited nature of the evi-
have been extremely limited (Barrett and Willis dence. It is difficult to quantify the role of com-
2001). There is limited circumstantial evidence petition in sauropodomorph evolution at pres-
for the coevolution of titanosaurs and angio- ent: if the prosauropod/sauropod CCR is sup-
sperms in the latest Cretaceous: the radiation of ported by additional data, it is arguably an
titanosaurs coincides with the angiosperms rise important process; if not, competition would
to ecological dominance (Barrett and Willis appear to be of local paleoecological signifi-
2001). However, the support for this hypothesis cance only. Further analysis of interactions
is weak and direct evidence for specific among sauropodomorph clades is possible
titanosaur–angiosperm interactions (coprolites, (e.g., detailed clade-by-clade comparisons
enterolites, and faunal/floral associations) is between different sauropod groups) but is
lacking: more work is needed on titanosaur feed- beyond the scope of this chapter. Continued
ing mechanisms, Late Cretaceous angiosperm testing of all of these CCRs is necessary in the
physiognomy, and Late Cretaceous floristics. light of mutable phylogenetic relationships and
new discoveries.
Comparisons of sauropodomorph diversity
CONCLUSIONS
with information on Mesozoic floras reveal one
It is difficult to infer the various processes possible incidence of coevolution, between
that controlled the observed patterns of titanosaurs and angiosperms, in the Late
sauropodomorph diversity. Rigorous analysis is Cretaceous. However, the evidence for coevolu-
hampered by small sample sizes, significant tion is circumstantial and weak: much more
biases in the record, and the likelihood that sev- information is needed to either confirm or refute
eral processes operate concurrently, thereby this hypothesis. Examination of the potential
obfuscating the various signals generated by interactions between prosauropods and contem-
each process independently. Nevertheless, com- porary floras in the Late Triassic does not sup-
bination of diversity data (from both TDEs and port existing hypotheses of faunal succession,
PDEs) with information on organismic distri- but neither does it provide a clear unambiguous
butions, paleoecology, and modes of life allows macroevolutionary signal. The decline of the
some tentative conclusions to be drawn. “Dicroidium flora” may have driven the tabefac-
Current data do not support the validity of tion of Gondwanan prosauropods (not their radi-
many of the CCRs that have been proposed to ation [contra Benton 1983]), but the apparent
account for the extinction of various sauropod abundance of these animals (albeit in paucispe-
clades. One potential CCR, the replacement of cific faunas) after the floral replacement is cur-
the rebbachisaurid clade by titanosaurs in the rently paradoxical.

The strong correlation between diversity and ———. 1977. Tetrapod mass extinctions—Model of
feeding ecology demonstrates that herbivory and the regulation of speciation rate and immigration
by cycles of topographic diversity. In: Hallam, A.
the various adaptive complexes that arose to deal
(ed.). Patterns of Evolution. Elsevier, Amsterdam.
with a diet of plants were central to the radiation Pp. 439–468.
of sauropodomorphs in the Middle and Late ———. 1978. Dinosaur feeding behaviour and the
Jurassic. Similarly, the acquisition of novel feed- origin of flowering plants. Nature 274: 661–663.
ing adaptations may have given basal sauropods Barrett, P. M. 1998. Herbivory in the Non-avian
a competitive advantage over sympatric Dinosauria. Unpublished Ph.D. dissertation,
University of Cambridge, Cambridge. 308 pp.
prosauropods. Finally, these analyses are consis-
———. 1999. A sauropod dinosaur from the Lower
tent with the suggestion that the evolution of var- Lufeng Formation (Lower Jurassic) of Yunnan
ied feeding mechanisms allowed niche partition- Province, People’s Republic of China. J. Vertebr.
ing, which, in turn, would have promoted sauro- Paleontol. 19: 785–787.
pod diversification. Nevertheless, knowledge of ———. 2000. Prosauropods and iguanas: specula-
tion on the diets of extinct reptiles; In: Sues,
sauropod feeding mechanisms cannot currently
H.-D. (ed.). The Evolution of Herbivory in
account for the high diversity of titanosaurs in Terrestrial Vertebrates. Perspectives from the
the Late Cretaceous. Fossil Record. Cambridge University Press,
Much functional morphological, phyloge- Cambridge. Pp. 42–78.
netic, and paleoecological work remains to be Barrett, P. M. and Upchurch, P. 1994. Feeding
done on sauropodomorphs, but preliminary mechanisms in Diplodocus. GAIA 10: 195–203.
———. 1995. Sauropod feeding mechanisms: their
studies such as this one demonstrate that they
bearing on palaeoecology. In: Sun, A.-L., and
have huge potential for contributing to more Wang, Y.-Q. (eds.). Sixth Symposium on
general macroevolutionary and paleoecological Mesozoic Terrestrial Ecosystems and Biota, Short
debates. Papers. China Ocean Press, Beijing. Pp. 107–110.
Barrett, P. M., and Willis, K. J. 2001. Did dinosaurs
invent flowers? Dinosaur–angiosperm coevolu-
ACKNOWLEDGMENTS
tion revisited. Biol. Rev. 76: 411–447.
We dedicate this contribution to Professor. J. S. Benton, M. J. 1983. Dinosaur success in the Triassic:
McIntosh in recognition of the many important a non-competitive ecological model. Q. Rev. Biol.
58: 29–55.
contributions he has made to the study of
———. 1986. The Late Triassic tetrapod extinction
sauropods. Jack has always been ready with a events. In: Padian, K. (ed.). The Beginning of the
smile, a sympathetic ear, and bucket-loads of Age of Dinosaurs: Faunal Change Across the
good advice, even when he clearly thought that Triassic–Jurassic Boundary. Cambridge University
we were talking complete nonsense. We are Press, Cambridge. Pp. 303–320.
grateful to Dr. Chris Brochu for providing use- ———. 1990. Origin and interrelationships of
dinosaurs. In: Weishampel, D. B., Dodson, P., and
ful comments on a draft of this chapter. Many
Osmólska, H. (eds.). The Dinosauria, 1st ed.
thanks go to Drs. Kristina Curry Rogers and Jeff University of California Press, Berkeley. Pp. 11–30.
Wilson for their invitation to contribute to this ———. 1993. Reptilia. In: Benton, M. J. (ed.). The
book. S. M. Feerick provided discussion on the Fossil Record 2. Chapman and Hall, London. Pp.
statistical testability (or otherwise) of the vari- 681–715.
———. 1996. On the nonprevalence of competitive
ous ideas presented herein. P.U. is supported
replacement in the evolution of tetrapods. In:
by a NERC Research Fellowship (NERC/GT/ Jablonski, D., Erwin, D. H., and Lipps, J. H. (eds.).
59906/ES). Evolutionary Paleobiology. University of Chicago
Press, Chicago. Pp. 185–210.
Benton, M. J., Juul, L., Storrs, G. W., and Galton,
LITERATURE CITED P. M. 2000. Anatomy and systematics of the
prosauropod dinosaur Thecodontosaurus antiquus
Bakker, R. T. 1971. Ecology of the brontosaurs. Nature from the Upper Triassic of southwest England.
229: 172–174. J. Vertebr. Paleontol. 20: 77–108.

Bonaparte, J. F. 1996. Dinosaurios de America del Curry Rogers, K., and Forster, C. A. 2001. The last of
Sur. Museo Argentino de Ciencias Naturales, the dinosaur titans: A new sauropod from
Buenos Aires. 174 pp. Madagascar. Nature 412: 530–534.
Brooks, D. R., and McLennan, D. A. 1991. Phylogeny, Dodson, P. 1990. Sauropod paleoecology. In:
ecology, and behaviour: a research programme in Weishampel, D. B., Dodson, P., and Osmólska,
comparative biology. University of Chicago Press, H. (eds.). The Dinosauria, 1st ed. University of
Chicago. 434 pp. California Press, Berkeley. Pp. 402–407.
Buffetaut, E., Martin, V., Sattayarak, N., and Dodson, P., Behrensmeyer, A. K., Bakker, R. T., and
Suteethorn, V. 1995. The oldest known dinosaur McIntosh, J. S. 1980. Taphonomy and paleoecol-
from southeast Asia: a prosauropod from the ogy of the Upper Jurassic Morrison Formation.
Nam Phong Formation (Late Triassic) of north- Paleobiology 6: 208–232.
eastern Thailand. Geol. Mag. 132: 739–742. Dong, Z.-M. 1992. Dinosaurian Faunas of China.
Buffetaut, E., Suteethorn, V., Cuny, G., Tong, H., Le Springer-Verlag, New York. 188 pp.
Loeuff, J., Khansubha, S., and Jongautchariyakul, Dong, Z.-M., Zhou, S., and Zhang, Y. 1983. [The
S. 2000. The earliest known sauropod dinosaur. dinosaurian remains from Sichuan Basin, China].
Nature 407: 72–74. Palaeontol. Sinica Ser. C 23: 139–145. (In Chinese
Calvo, J. O. 1994. Jaw mechanics in sauropod with English summary.)
dinosaurs. GAIA 10: 183–194. Farlow, J. O. 1987. Speculations about the diet and
Casanovas, M. L., Santafé, J. V., and Sanz, J. L. 2001. digestive physiology of herbivorous dinosaurs.
Losillasaurus giganteus, un nuevo saurópodo del Paleobiology 13: 60–72.
tránsito Jurásico–Cretácico de la cuenca de ‘Los Fiorillo, A. R. 1991. Dental microwear on the teeth of
Serranos’ (Valencia, España). Paleontol. Evol. Camarasaurus and Diplodocus: implications for
32–33:99–122. sauropod palaeoecology. In: Kielan-Jaworowska,
Charig, A. J., Attridge, J., and Crompton, A. W. 1965. Z., Heintz, N., and Nakrem, H. A. (eds.). Fifth
On the origin of the sauropods and the classifica- Symposium on Mesozoic Terrestrial Ecosystems
tion of the Saurischia. Proc. Linn. Soc. London and Biota, Extended Abstracts. Contributions
176: 197–221. from the Palaeontological Museum, University of
Christiansen, P. 1999. On the head size of Oslo, 364. Pp. 23–24.
sauropodomorph dinosaurs: implications for ———. 1998. Dental microwear patterns of the sauro-
ecology and physiology. Hist. Biol. 13: 269–297. pod dinosaurs Camarasaurus and Diplodocus:
———. 2000. Feeding mechanisms of the sauropod evidence for resource partitioning in the Late
dinosaurs Brachiosaurus, Camarasaurus, Diplodocus Jurassic of North America. Hist. Biol. 13: 1–16.
and Dicraeosaurus. Hist. Biol. 14: 137–152. Flynn, J. J., Parrish, J. M., Rakotosamimanana, B.,
Connely, M. V., and Hawley, R. 1998. A proposed Simpson, W. F., Whatley, R. L., and Wyss, A. R.
reconstruction of the jaw musculature and 1999. A Triassic fauna from Madagascar, includ-
other soft cranial tissues of Apatosaurus. J. ing early dinosaurs. Science 286: 763–765.
Vertebr. Paleontol. 18: 35A. Galton, P. M. 1973. The cheeks of ornithischian
Coombs, W. P., Jr. 1975. Sauropod habits and dinosaurs. Lethaia 6: 67–89.
habitats. Palaeogeogr. Palaeoclimatol. Palaeoecol. ———. 1976. Prosauropod dinosaurs of North
17: 1–33. America. Postilla 169: 1–98.
Cooper, M. R. 1981. The prosauropod dinosaur ———. 1985. Diet of prosauropods from the Late
Massospondylus carinatus Owen from Zimbabwe: Triassic and Early Jurassic. Lethaia 18: 105–123.
its biology, mode of life and phylogenetic signifi- ———. 1986. Herbivorous adaptations of Late Triassic
cance. Occas. Papers Natl. Mus. Monuments and Early Jurassic dinosaurs. In: Padian, K. (ed.).
Rhodesia Ser. B Nat. Sci. 6: 689–840. The Beginning of the Age of the Dinosaurs.
Coria, R. A., and Chiappe, L. M. 2001. Tooth replace- Cambridge University Press, Cambridge.
ment in a sauropod premaxilla from the Upper Pp. 203–221.
Cretaceous of Patagonia, Argentina. Ameghiniana ———. 1990. Basal Sauropodomorpha—
38: 463–466. Prosauropoda. In: Weishampel, D. B., Dodson,
Crompton, A. W., and Attridge, J. 1986. Masticatory P., and Osmólska, H., (eds.). The Dinosauria,
apparatus of the larger herbivores during Late 1st ed. University of California Press, Berkeley.
Triassic and Early Jurassic times. In: Padian, K. Pp. 320–344.
(ed.). The Beginning of the Age of Dinosaurs. Galton, P. M., and Upchurch, P. 2004. Basal
Cambridge University Press, Cambridge. Pp. Sauropodomorpha—Prosauropoda. In: Weisham-
223–236. pel, D. B., Dodson, P., and Osmólska, H. (eds.).

The Dinosauria 2nd ed. University of California ———. 1996. A comparative analysis of herbivory
Press, Berkeley. Pp. 232–258. and amniote diversification in recent terrestrial
Gauffre, F.-X. 1993. The prosauropod dinosaur ecosystems. Spec. Publ. Paleontol. Soc. 8: 280.
Azendohsaurus laarousii from the Upper Triassic Niklas, K. J. 1997. The Evolutionary Biology of
of Morocco. Palaeontology 36: 897–908. Plants. University of Chicago Press, Chicago.
Gauthier, J. A. 1986. Saurischian monophyly and the 449 pp.
origin of birds. Mem. Calif. Acad. Sci. 8: 1–55. Niklas, K. J., Tiffney, B. H., and Knoll, A. H. 1985.
Gow, C. E., Kitching, J. W., and Raath, M. A. 1990. Patterns in vascular land plant diversification: an
Skulls of the prosauropod dinosaur Massospondylus analysis at the species level. In: Valentine, J. W.
carinatus Owen in the collections of the Bernard (ed.). Phanerozoic Diversity Patterns: Profiles in
Price Institute for Palaeontological Research. Macroevolution. Princeton University Press and
Palaeontol. Afr. 27: 45–58. American Association for the Advancement of
Haubold, H. 1990. Dinosaurs and fluctuating sea Science, Princeton, NJ, and San Francisco. Pp.
levels during the Mesozoic. Hist. Biol. 4: 75–106. 97–128.
He, X.-L., Wang, C.-S., Liu, S.-Z., Zhou, F.-Y., Liu, T.- Norman, D. B., and Weishampel, D. B. 1991. Feeding
Q., Cai, K.-J., and Dai, B. 1998. A new species of mechanisms in some small herbivorous dinosaurs:
sauropod from the Early Jurassic of Gongxian processes and patterns. In: Rayner, J. M. V., and
County, Sichuan. Acta Geol. Sichuan 18: 1–7 (In Wootton, R. J. (eds.). Biomechanics in Evolution.
Chinese, with English abstract.) Cambridge University Press, Cambridge. Pp.
Hotton, N., Olson, E. C., and Beerbower, R. 1997. 161–181.
Amniote origins and the discovery of herbivory. Owen-Smith, R. N. 1988. Megaherbivores: The
In: Sumida, S. S., and Martin, K. L. M. (eds.). Influence of Very Large Body Size on Ecology.
Amniote Origins: Completing the Transition to Cambridge University Press, Cambridge. 369 pp.
Land. Academic Press, San Diego. Pp. 207–264. Patterson, C., and Smith, A. B. 1987. Is periodicity of
Hunt, A. P., Lockley, M. G., Lucas, S. G., and Meyer, mass extinctions a taxonomic artifact? Nature 330:
C. A. 1994. The global sauropod fossil record. 248–251.
GAIA 10: 261–279. Pisani, D., Yates, A. M., Langer, M. C., and Benton, M. J.
Jablonski, D., Erwin, D. H., and Lipps, J. H. (eds.). 1996. 2002. A genus-level supertree of the Dinosauria.
Evolutionary Paleobiology. University Chicago Proc. Roy. Soc. London Ser. B 269: 915–921.
Press, Chicago. 484 pp. Raup, D. M., and Boyajian, G. E. 1988. Patterns of
King, G. M. 1990. Dicynodonts and the end-Permian generic extinction in the fossil record. Paleobiology
event. Palaeontol. Afr. 27: 31–39. 14: 109–123.
Langer, M. C., Abdala, F., Richter, M., and Benton, Reisz, R. R., and Sues, H.-D. 2000. Herbivory in late
M. J. 1999. A sauropodomorph dinosaur from the Palaeozoic and Triassic terrestrial vertebrates. In:
Upper Triassic (Carnian) of southern Brazil. C. R. Sues, H.-D. (ed.). The Evolution of Herbivory in
Acad. Sci. Paris Sci. Terre Planettes 329: 511–517. Terrestrial Vertebrates. Perspectives from the Fossil
Luo, Z.-X., and Wu, X.-C., 1994. The small tetrapods Record. Cambridge University Press, Cambridge.
of the Lower Lufeng Formation, Yunnan, China. Pp. 9–41.
In: Fraser, N. C., and Sues, H.-D. (eds.). In the Robeck, H. E., Maley, C. C., and Donoghue, M. J.
Shadow of the Dinosaurs. Cambridge University 2000. Taxonomy and temporal diversity patterns.
Press, Cambridge. Pp. 251–270. Paleobiology 26: 171–187.
Martill, D. M., and Naish, D. (eds.). 2001. Dinosaurs Russell, D. A. 1989. An odyssey in time: the dinosaurs
of the Isle of Wight. Palaeontological Association, of North America. University of Toronto Press/
London. 433 pp. National Museum of Natural Sciences, Toronto.
Martin, J., Martin-Rolland, V., and Frey, E. 1998. Not 240 pp.
cranes or masts, but beams: the biomechanics of Russell, D. A., Béland, P., and McIntosh, J. S. 1980.
sauropod necks. Oryctos 1: 113–120. Paleoecology of the dinosaurs of Tendaguru
McIntosh, J. S. 1990. Sauropoda. In: Weishampel, (Tanzania). Mem. Soc. Geol. France 139:
D. B., Dodson, P., and Osmólska, H. (eds.). The 169–175.
Dinosauria, 1st ed. University of California Press, Salgado, L., and Calvo, J. O. 1997. Evolution of
Berkeley. Pp. 345–401. titanosaurid sauropods. II. The cranial evidence.
Moore, B. R., and Brooks, D. R. 1995. A comparative Ameghiniana 34: 33–48.
analysis of herbivory and amniote diversification in Sepkoski, J. J., Jr. 1996. Competition in macroevolu-
recent terrestrial ecosystems. J. Vertebr. Paleontol. tion: The double wedge revisited. In: Jablonski,
15: 45A. D., Erwin, D. H., and Lipps, J. H. (eds.).

Evolutionary Paleobiology. University of Chicago ———. 1990. Dinosaurian distribution. In:
Press, Chicago. Pp. 211–255. Weishampel, D. B., Dodson, P., and Osmólska,
Sepkoski, J. J., Jr. and Kendrick, D. C. 1993. Numerical H. (eds.). The Dinosauria, 1st ed. University of
experiments with model monophyletic and para- California Press, Berkeley. Pp. 63–139.
phyletic taxa. Paleobiology 19: 168–184. Weishampel, D. B. and Horner, J. R. 1987. Dinosaurs,
Sereno, P. C. 1997. The origin and evolution of di- habitat bottlenecks, and the St. Mary River
nosaurs. Annu. Rev. Earth Planet. Sci. 25: 435–489. Formation. In: Currie, P. J., and Koster, E. H. (eds.).
———. 1999. The evolution of dinosaurs. Science Fourth Symposium on Mesozoic Terrestrial
284: 2137–2147. Ecosystems, Short Papers. Royal Tyrell Museum,
Sereno, P. C., and Wilson, J. A. 2001. A sauropod Drumheller. Pp. 224–229.
tooth battery: Structure and function. J. Vertebr. Weishampel, D. B. and Jianu, C.-M. 2000. Plant-
Paleontol. 21: 100A–101A. eaters and ghost lineages: dinosaurian herbivory
Sereno, P. C., Beck, A. L., Dutheil, D. B., Larsson, revisited. In: Sues, H.-D. (ed.). Evolution of
H. C. E., Lyon, G. H., Moussa, B., Sadleir, R. W., Herbivory in Terrestrial Vertebrates: Perspectives
Sidor, C. A., Varricchio, D. J., Wilson, G. P., and from the Fossil Record. Cambridge University
Wilson, J. A. 1999. Cretaceous sauropods from the Press, Cambridge. Pp. 123–143.
Sahara and the uneven rate of skeletal evolution Weishampel, D. B., and Norman, D. B. 1989.
among dinosaurs. Science 286: 1342–1347. Vertebrate herbivory in the Mesozoic; jaws, plants
Smith, A. B. 1994. Systematics and the Fossil Record. and evolutionary metrics. Spec. Papers Geol. Soc.
Blackwell Scientific, Oxford. 223 pp. Am. 238: 87–100.
Stevens, K. A., and Parrish, J. M. 1999. Neck posture Weishampel, D. B., and Witmer. L. M. 1990.
and feeding habits of two Jurassic sauropod Heterodontosauridae. In: Weishampel, D. B.,
dinosaurs. Science 284: 798–800. Dodson, P., and Osmólska, H. (eds.). The
Tiffney, B. H. 1992. The role of vertebrate herbivory Dinosauria, 1st ed. University of California Press,
in the evolution of land plants. Palaeobotanist 41: Berkeley. Pp. 486–497.
87–97. Weishampel, D. B., Barrett, P. M., Coria, R. A., Le Loeuff,
Upchurch, P. 1995. The evolutionary history of J., Xu, X., Zhao, X. -J., Sahni, A. Gomani, E. M.P.,
sauropod dinosaurs. Philos. Trans. Roy. Soc. and Noto, C.R. 2004. Dinosaur distribution. In:
London Ser. B 349: 365–390. Weishampel, D. B., Dodson, P., and Osmólska, H.
———. 1998. The phylogenetic relationships of (eds.). The Dinosauria, 2nd ed. University of
sauropod dinosaurs. Zool. J. Linn. Soc. London California Press, Berkeley. Pp. 517–606.
124: 43–103. Wilson, J. A. 2002. Sauropod dinosaur phylogeny:
———. 1999. The phylogenetic relationships of critique and cladistic analysis. Zool. J. Linn. Soc.
the Nemegtosauridae (Saurischia, Sauropoda). London 136: 217–276.
J. Vertebr. Paleontol. 19: 106–125. Wilson, J. A. and Sereno, P. C. 1998. Early evolution
Upchurch, P., and Barrett, P. M. 2000. The evolution and higher-level phylogeny of the sauropod
of sauropod feeding mechanisms. In: Sues, H.-D. dinosaurs. Mem. Soc. Vertebr. Paleontol. 5: 1–68.
(ed.). Evolution of Herbivory in Terrestrial Wing, S. L., and Tiffney, B. H. 1987. The reciprocal
Vertebrates: Perspectives from the Fossil Record. interaction of angiosperm evolution and tetrapod
Cambridge University Press, Cambridge. Pp. herbivory. Rev. Palaeobot. Palynol. 50: 179–210.
79–122. Wing, S. L., Sues, H.-D., Tiffney, B. H., Stucky, R. K.,
Upchurch, P., Barrett, P. M., and Dodson, P. 2004. Weishampel, D. B., Spicer, R. A., Jablonski, D., Bad-
Sauropoda. In: Weishampel, D. B., Dodson, P., gley, C. E., Wilson, M. V. H., and Kovach, W.L. 1992.
and Osmólska, H. (eds.). The Dinosauria, 2nd ed. Mesozoic and Early Cenozoic terrestrial ecosys-
University of California Press, Berkeley. Pp. tems. In: Behrensmeyer, A. K., Damuth, J. D.,
259–322. DiMichele, W. A., Potts, R., Sues, H.-D., and Wing,
Valentine, J. W. (ed.). 1985. Phanerozoic Diversity S. L. (eds.). Terrestrial Ecosystems through Time:
Patterns: Profiles in Macroevolution. Princeton Evolutionary Paleoecology of Plants and Animals.
University Press and American Association for University of Chicago Press, Chicago. Pp. 327–416.
the Advancement of Science, Princeton, NJ, and Yates, A. M. 2003. A new species of the primitive
San Francisco. 441 pp. dinosaur Thecodontosaurus (Saurischia: Sauropodo-
Weishampel, D. B. 1984. Interactions between morpha) and its implications for the systematics of
Mesozoic plants and vertebrates: fructifications early dinosaurs. J. Syst. Palaeontol. 1: 1–42.
and seed predation. Neues Jahrb. Geol. Palaontol. Young, C. C. 1951. The Lufeng saurischian fauna.
Abhandlungen 167: 224–250. Palaeontol. Sinica Ser. C 13: 1–96.

APPENDIX 4.1. PROSAUROPOD GENERA INCLUDED IN THE TAXIC DIVERSITY
ESTIMATES
After Galton and Upchurch (2004), Weishampel et al. (2004), and Barrett and Upchurch (pers. obs.).
GENUS STRATIGRAPHIC RANGE
Ammosaurus Hettangian–Pliensbachian
Anchisaurus Hettangian–Pliensbachian
Azendohsaurus late Carnian
Camelotia Rhaetian
Coloradisaurus Norian
Euskelosaurus late Carnian–early Norian
“Gyposaurus” sinensis Hettangian–early Sinemurian
Jingshanosaurus Hettangian–early Sinemurian
Lessemsaurus Norian
Lufengosaurus Hettangian–early Sinemurian
Massospondylus Hettangian–Pliensbachian
Melanorosaurus late Carnian–Pliensbachian
Mussaurus Norian
Plateosaurus late Norian
Riojasaurus Norian
Saturnalia middle Carnian
Sellosaurus middle Norian
Thecodontosaurus late Carnian–Rhaetian
Yimenosaurus Hettangian–early Sinemurian
Yunnanosaurus Sinemurian–early Pliensbachian

APPENDIX 4.2. SAUROPOD GENERA INCLUDED IN THE TAXIC DIVERSITY ESTIMATES
After Upchurch et al. (2004), Weishampel et al. (2004), and Barrett and Upchurch (pers. obs). A question mark (?) denotes genera for which tooth morphology and feeding mecha-
nism could not be predicted due to unknown phylogenetic position; these genera are excluded from tooth morphology counts. Abbreviations for tooth type: NC, narrow-crowned (cylin-
drical or subcylindrical; SI 4.0); BC, broad-crowned (spatulate; SI 4.0); inferred, tooth morphology predicted on the basis of phylogenetic position (actual teeth unknown).
Abbreviations for feeding mechanisms: APATO, Apatosaurus type; BRACH, brachiosaurid type; CAMARA, Camarasaurus type; CETIO, cetiosaurid type; DICRA, dicraeosaurid
type; DIPLO, Diplodocus type; HIGHEU, “higher eusauropod” type; NEMEGT, nemegtosaurid type; NIGER, Nigersaurus type; SHUNO, Shunosaurus type; TITAN,
titanosaurian type.
GENUS STRATIGRAPHIC RANGE CLASSIFICATION TOOTH TYPE MECHANISM
Abrosaurus Bathonian–Callovian Basal macronarian BC ?

Aegyptosaurus Cenomanian Basal titanosaurian Inferred NC TITAN
Aeolosaurus Campanian Titanosauroidea Inferred NC TITAN
Agustinia Aptian Titanosauria incertae sedis Inferred NC TITAN
Alamosaurus Maastrichtian Titanosauroidea NC TITAN
Amargasaurus Hauterivian Diplodocoidea: Dicraeosauridae Inferred NC DICRAE
Ampelosaurus early Maastrichtian Titanosauroidea NC TITAN
Amphicoelias Kimmeridgian–Tithonian Diplodocoidea incertae sedis Inferred NC ?
Amygdalodon Bajocian Eusauropoda incertae sedis BC ?
Andesaurus Albian–Cenomanian Basal titanosaurian Inferred NC TITAN
Antarctosaurus Campanian–Maastrichtian Titanosauroidea incertae sedis NC TITAN
Apatosaurus Kimmeridgian–Tithonian Diplodocoidea: Diplodocidae NC APATO
“Apatosaurus” minimus Kimmeridgian–Tithonian Sauropoda incertae sedis ? ?
Aragosaurus early Barremian Eusauropoda incertae sedis BC ?
Argentinosaurus Albian–Cenomanian Basal titanosaurian Inferred NC TITAN
Argyrosaurus Campanian–Maastrichtian Titanosauria incertae sedis Inferred NC TITAN
Astrodon middle–late Aptian Basal titanosauriform BC ?
Atlasaurus Bathonian–Callovian Basal macronarian BC ?
Austrosaurus Albian Basal titanosaurian Inferred NC TITAN
Barapasaurus Hettangian–Pliensbachian Cetiosauridae BC CETIO
Barosaurus Kimmeridgian–Tithonian Diplodocoidea: Diplodocidae Inferred NC DIPLO
Bellusaurus Aalenian–Callovian Basal macronarian BC ?
Blikanasaurus ?late Carnian or ?early Norian Basal sauropod Inferred BC ?
Brachiosaurus Kimmeridgian–Tithonian Brachiosauridae BC BRACH

“Brachiosaurus” atalaiensis Kimmeridgian Brachiosauridae Inferred BC BRACH
“Brachiosaurus” nougaredi Albian Brachiosauridae Inferred BC BRACH
Camarasaurus Kimmeridgian–Tithonian Basal camarasauromorph BC CAMARA
Cardiodon late Bathonian Sauropoda incertae sedis BC ?
Cedarosaurus Barremian Brachiosauridae Inferred BC BRACH
Cetiosauriscus middle Callovian Diplodocoidea incertae sedis Inferred NC ?
Cetiosaurus late Bajocian–Bathonian Cetiosauridae Inferred BC CETIO
“Cetiosaurus” glymptonensis late Bathonian Diplodocoidea incertae sedis Inferred NC ?
“Cetiosaurus” humerocristatus Kimmeridgian Brachiosauridae Inferred BC BRACH
Chubutisaurus Albian Basal titanosaurian Inferred NC TITAN
Datousaurus Bathonian–Callovian Sauropoda incertae sedis BC ?
Dicraeosaurus Kimmeridgian Diplodocoidea: Dicraeosauridae NC DICRA
Dinheirosaurus late Kimmeridgian Diplodocoidea incertae sedis Inferred NC ?
Diplodocus Kimmeridgian–Tithonian Diplodocoidea: Diplodocidae NC DIPLO
Dyslocosaurus Kimmeridgian–Tithonian (or Maas) Diplodocoidea: Diplodocidae Inferred NC DIPLO
Dystrophaeus Callovian–Oxfordian Sauropoda incertae sedis ? ?
Dystylosaurus Kimmeridgian–Tithonian Brachiosauridae Inferred BC BRACH
Epachthosaurus Campanian–Maastrichtian Basal titanosaurian Inferred NC TITAN
Euhelopus Kimmeridgian Non-neosauropod eusauropod BC HIGHEU
Gondwanatitan Santonian–Maastrichtian Titanosauroidea Inferred NC TITAN
Gongxianosaurus Hettangian–Toarcian Basal sauropod BC ?
Haplocanthosaurus Kimmeridgian–Tithonian Basal camarasauromorph Inferred BC ?
Huabeisaurus ?Campanian Basal titanosaurian NC TITAN
Hudiesaurus Kimmeridgian–Tithonian Eusauropoda incertae sedis BC ?
Isanosaurus late Norian–Rhaetian Basal sauropod Inferred BC ?
Jainosaurus Maastrichtian Titanosauroidea Inferred NC TITAN
Janenschia Kimmeridgian Basal titanosaurian Inferred NC TITAN
Jobaria Berriasian–Hauterivian Basal macronarian BC ?
Klamelisaurus Aalenian–Callovian Sauropoda incertae sedis BC ?
Kotasaurus Hettangian–Pliensbachian Basal sauropod BC ?
Laplatasaurus Campanian–Maastrichtian Titanosauroidea Inferred NC TITAN
Lapparentosaurus Bathonian Titanosauriformes incertae sedis BC ?
Lirainosaurus late Campanian–early Maastrichtian Titanosauroidea NC TITAN
Losillasaurus Late Jurassic–Early Cretaceous Diplodocoidea incertae sedis Inferred NC ?
Lourinhasaurus late Kimmeridgian–early Tithonian Non-neosauropod eusauropod Inferred BC ?
Magyarosaurus late Maastrichtian Titanosauroidea Inferred NC TITAN
Malawisaurus Aptian Titanosauroidea NC TITAN
Mamenchisaurus Oxfordian–Tithonian Non-neosauropod eusauropod BC HIGHEU
“Morosaurus” agilis Kimmeridgian–Tithonian Sauropoda incertae sedis ? ?
Nemegtosaurus middle Maastrichtian Diplodocoidea: Nemegtosauridae NC NEMEGT
Neuquensaurus Campanian–early Maastrichtian Titanosauroidea Inferred NC TITAN
Nigersaurus Aptian–Albian Diplodocoidea: Rebbachisauridae NC NIGER
Ohmdenosaurus Middle Toarcian Basal sauropod Inferred BC ?
Omeisaurus Bathonian–Tithonian Non-neosauropod eusauropod BC HIGHEU
Opisthocoelicaudia middle Maastrichtian Titanosauroidea Inferred NC TITAN
Oplosaurus Barremian Sauropoda incertae sedis BC ?
Ornithopsis Barremian Titanosauriformes incertae sedis Inferred BC ?
Paralititan Cenomanian Titanosauroidea Inferred NC TITAN
Patagosaurus Callovian Cetiosauridae BC CETIO
Pellegrinisaurus Early Campanian Titanosauroidea Inferred NC TITAN
“Pelorosaurus” conybearei late Berriasian–Valanginian Titanosauriformes incertae sedis Inferred BC ?
Phuwiangosaurus Valanginian–Hauterivian Basal titanosaurian NC TITAN
Pleurocoelus middle–late Aptian Titanosauriformes incertae sedis BC ?
“Pleurocoelus” Texas Middle Albian Titanosauriformes incertae sedis BC ?
Quaesitosaurus late Campanian–early Maastrichtian Diplodocoidea: Nemegtosauridae Inferred NC NEMEGT
Rapetosaurus Maastrichtian Titanosauroidea NC TITAN
Rayososaurus Albian–Cenomanian Diplodocoidea: Rebbachisauridae NC NIGER
Rebbachisaurus Albian Diplodocoidea: Rebbachisauridae Inferred NC ?
Rhoetosaurus Bajocian Sauropoda incertae sedis ? ?
Rocasaurus Campanian–Maastrichtian Titanosauroidea Inferred NC TITAN
Saltasaurus late Campanian–Maastrichtian Titanosauroidea Inferred NC TITAN
Sauroposeidon Aptian–Albian Brachiosauridae Inferred BC BRACH
Seismosaurus Kimmeridgian–Tithonian Diplodocoidea: Diplodocidae Inferred NC DIPLO
Shunosaurus Bathonian–Callovian Non-neosauropod eusauropod BC SHUNO

Supersaurus Kimmeridgian–Tithonian Diplodocoidea: Diplodocidae Inferred NC DIPLO
Tangvayosaurus Aptian–Albian Basal titanosaurian Inferred NC TITAN
Tehuelchesaurus Callovian Non-neosauropod eusauropod Inferred BC HIGHEU
Tendaguria Kimmeridgian Sauropoda incertae sedis ? ?
Titanosaurus Maastrichtian Titanosauroidea NC TITAN
“Titanosaurus” colberti Maastrichtian Titanosauroidea Inferred NC TITAN
Tornieria Kimmeridgian Diplodocoidea: Diplodocidae NC DIPLO
Venenosaurus Barremian Basal titanosaurian Inferred NC TITAN
Volkheimeria Callovian Sauropoda incertae sedis ? ?
Vulcanodon Hettangian Basal sauropod Inferred BC ?
five
Structure and Evolution

of a Sauropod Tooth Battery
Paul C. Sereno and Jeffrey A. Wilson
uring the jurassic, sauropod chapter, we describe the tooth-bearing bones

D dinosaurs rose to predominance among
vertebrate herbivores, in terms of both species di-
and dental battery of Nigersaurus taqueti and pro-
vide an initial cranial reconstruction. We outline
versity and biomass (e.g., Romer 1966; McIntosh the feeding specializations common to diplo-
1990). Their perceived decline on northern land- docoids and how these were modified within
masses during the Cretaceous has been linked to rebbachisaurids.
the evolution of tooth batteries in ornithischian Institutional Abbreviation: MNN, Musée
herbivores (e.g., Lull and Wright 1942; Ostrom National du Niger, Niamey.
1961; Bakker 1978; Lucas and Hunt 1989). On
southern landmasses, in contrast, sauropod di-
SYSTEMATIC PALEONTOLOGY
versity increased during the Cretaceous
(Weishampel 1990; Hunt et al. 1994), and a
SAUROPODA MARSH 1878
newly discovered southern sauropod, the reb-
bachisaurid Nigersaurus taqueti, is now known to EUSAUROPODA UPCHURCH 1995
have evolved a complex tooth battery (Sereno et al. NEOSAUROPODA BONAPARTE 1986
1999). DIPLODOCOIDEA MARSH 1884
Rebbachisaurids are a poorly known sauro-
REBBACHISAURIDAE BONAPARTE 1997
pod clade, reported thus far only from Cretaceous
rocks in South America (Calvo and Salgado NIGERSAURUS SERENO ET AL. 1999
1995; Bonaparte 1996; Apestiguía et al. 2001; TYPE SPECIES : Nigersaurus taqueti Sereno et
Lamanna et al. 2001), Africa (Lavocat 1954; al. 1999.
Taquet 1976; Sereno et al. 1999), and Europe AGE : Early Cretaceous (Aptian–Albian).
(Dalla Vecchia 1998; Pereda-Suberbiola et al. DISCUSSION : Nigersaurus taqueti is the most
2003). Many of these are fragmentary finds, common sauropod and one of the most common
leaving much of the skeletal anatomy of this species recovered in the rich vertebrate fauna
group in question, especially the skull. In this described from Gadoufaoua, Niger Republic
157
FIGURE 5.1. Preliminary reconstruction of the skull of Nigersaurus taqueti based on MNN GDF512. (A) Skull in left lateral
view. (B) Left lower jaw in dorsal view with tooth batteries removed. 1–5, accessory cranial fenestrae; a, angular; altr, alveolar
trough; antfe, antorbital fenestra; ar, articular, br, buccal ridge; cp, coronoid process; d, dentary; d1–d34, replacement foram-
ina for dentary tooth positions 1–34; emf, external mandibular fenestra; en, external naris; f, frontal, fo, foramen; j, jugal; ltf,
laterotemporal fenestra; m, maxilla; mc, Meckel’s canal; nf, narial fossa; pm, premaxilla; po, postorbital; popr, paraoccipital
process; ps, parasphenoid; q, quadrate; qfo, quadrate fossa; qj, quadratojugal; sa, surangular; saf, surangular foramen, so,
supraoccipital; sq, squamosal; sym, symphyseal surface; vc, vascular canal.
(Taquet 1976; Sereno et al. 1998, 1999, 2001, cannot be reliably associated with the holotypic
2003; Larsson and Gado 2000). Nevertheless, specimen. The partial skeleton described by
because the skull and skeleton are delicately con- Taquet (1976) also was discovered in the vicin-
structed and highly pneumatic, there are no com- ity of the holotype and may pertain to the same
plete skulls and only a few partially articulated species.
skeletons. Nigersaurus taqueti was named and LOCALITY AND HORIZON : Gadoufaoua (16°
identified as a rebbachisaurid by Sereno et al. 27´N, 9°8´´E), eastern edge of the Ténéré Desert,
(1999). An earlier report from the same beds of a Niger Republic; Elrhaz Formation.
dicraeosaurid allied with titanosaurs (Taquet REFERRED MATERIAL : MNN GDF513, worn
1976:53) very likely pertains to the same species. crown. Additional skeletal and dental material
is described elsewhere.
NIGERSAURUS TAQUETI SERENO ET AL.1999 REVISED DIAGNOSIS : Rebbachisaurid sauro-
(FIGS. 5.1, 5.3–5.8) pod characterized by five accessory fenestrae in
HOLOTYPE : MNN GDF512, partial disarticu- the jugal, surangular, and angular; tooth num-
lated skull and partially articulated neck pre- ber increased to 20 and to 34 in the maxilla and
served in close association on 1 m2 of sandstone dentary, respectively; tooth replacement
outcrop. Sereno et al. (1999:1346) also list a increased to as many as 10–12 in a single col-
“scapula, forelimbs, and hind limbs” as part of umn; premaxilla and dentary lacking alveolar
the holotype. These and other bones, which we septa; maxilla with oval (vertically elongated)
regard as referable to this species, were found replacement foramina; extension of the dentary
at some distance from the skull and neck and tooth row lateral to the sagittal plane of the
158 S T R U C T U R E A N D E V O L U T I O N O F A S AU R O P O D T O O T H B AT T E R Y
lower jaw; subcircular mandibular symphysis; pronounced narial fossa. As in other diplo-
crowns with prominent mesial (medial) and docoids, the jaw articulation and laterotemporal
distal (lateral) ridges; scapula with prominent fenestra are shifted anteriorly under the orbit
rugosity on the medial aspect of the proximal (figs. 5.1, 5.2). The supratemporal fenestra is
end of the blade. very reduced or absent altogether, in contrast to
those in other diplodocoids (Holland 1924) or
other sauropods.
DESCRIPTION The transversely expanded form of the alveo-
lar ramus of the dentary is unique among
SKULL AND DENTITION dinosaurs (fig. 5.1B). No other dinosaur has a
The skull and neck of the holotypic specimen tooth row that extends lateral to the longitudinal
of Nigersaurus taqueti were found in close asso- plane of the lower jaw. The maximal width
ciation. Most of the dorsal skull roof is pre- across the anterior end of the paired dentaries
served (fig. 5.1A). The braincase is intact, with slightly exceeds the length of the entire lower
the proximal end of the stapes in place in the jaw. Despite the gaping alveolar trough that
fenestra ovalis. The quadrate is the only palatal housed hundreds of slender teeth packed
bone preserved. All of these bones, with the together as a tooth battery, the dentary and other
exception of the frontal and braincase, are bones of the lower jaw in Nigersaurus are as
composed of thin laminae or narrow struts lightly built as those in Diplodocus and much
and are extremely delicate. Five unique acces- more slender than those in Camarasaurus (figs.
sory fenestrae are present, two in the jugal and 5.1, 5.2). The description below is limited to the
three in the surangular and angular (figs. tooth-bearing bones and the teeth.
5.1–5.5). These accessory openings are bor-
dered by bone that tapers gradually in width to PREMAXILLA
a paper-thin edge that has a smooth margin. It The premaxilla is a slender bone, the ventral
is highly unlikely, therefore, that they repre- third of which houses a battery of teeth aligned
sent lesions or some other kind of bone in four columns (fig. 5.3, table 5.1). The dorsal
pathology, like the healed openings reported in two-thirds of the premaxilla extends as a thin
aged individuals of other dinosaurian species lamina appressed to its opposite in the mid-
(e.g., Brochu 2003). line. In lateral view, these two parts of the pre-
The snout is proportionately much shorter maxilla meet at an angle of approximately 30º,
and the dental arcade is less prognathous than greater than that in Diplodocus (figs. 5.1A,
in diplodocids or dicraeosaurids (figs. 5.1, 5.2). 5.2B). Four replacement foramina and a con-
As mentioned above, the external naris is not necting groove for the dental lamina are visi-
retracted above the orbit in Nigersaurus. A ble on the posterior aspect of the bone. Unlike
statement to the contrary—that the external those in other diplodocoids, the alveolar mar-
naris is positioned as in diplodocids—was gin and space within the premaxilla housing
made on the basis of the maxilla before the pre- replacement teeth are not divided by bony
maxilla was exposed (Sereno et al. 1999:1344). septa into discrete alveoli. Rather, they are
Although the nasal and thus the posterior mar- developed as an open trough, the arched ante-
gin of the external naris are not known, the rior wall of which is thin and extends more
anterior portion of the border is far anterior to than 1 cm farther ventrally than the posterior
that in any known diplodocid. The external wall (figs. 5.3, 5.4).
naris is large, laterally facing, separated in the Although the active, or functioning, teeth
midline from its opposite by a vertical premax- have fallen from the trough, a battery of replace-
illa–nasal septum, and surrounded by a more ment teeth is visible in computed tomography
S T R U C T U R E A N D E V O L U T I O N O F A S AU R O P O D T O O T H B AT T E R Y 159
FIGURE 5.2. Skull reconstructions in lateral view of (A) the basal neosauropod Camarasaurus lentus and
(B) the diplodocid Diplodocus longus (from Wilson and Sereno 1998). a, angular; amf, anterior maxillary fora-
men; aofe, antorbital fenestra; ar, articular; asf, anterior surangular foramen; bo, basioccipital; d, dentary; f,
frontal; j, jugal; l, lacrimal; lf, lacrimal foramen; m, maxilla; n, nasal; nf, narial fossa; p, parietal; paofe, prean-
torbital fenestra; pl, palatine; pm, premaxilla; po, postorbital; popr, paraoccipital processes; pra, prearticular;
prf, prefrontal; ps, parasphenoid; psf, posterior surangular foramen; pt, pterygoid; q, quadrate; qj, quadratoju-
gal; sa, surangular; sf, surangular foramen; sq, squamosal; so, supraoccipital; v, vomer.
FIGURE 5.3. (A) Stereopair and (B) matching drawing of the right premaxilla of Nigersaurus taqueti (MNN GDF512) in
lateral view. am, articular surface for the maxilla; altr, alveolar trough; en, external naris; fo, foramen; nf, narial fossa;
plpr, palatal processes. Scale bars equal 5 cm.
TABLE 5.1
Measurements of the Tooth-Bearing Bones and Teeth of Nigersaurus taqueti
Premaxilla
Maximum length 209
Maximum width of maxillary articular surface, midheight 18
Alveolar trough, anteroposterior width 10
Posterodorsal ramus, length 77
Maxilla
Alveolar ramus
Preserved width 94
Depth (to ventral rim of antorbital fenestra) 73
Replacement foramen, height 11
Replacement foramen, width 3
Dentary
Alveolar ramus, width 115
Alveolar trough, anteroposterior width of medial end 12
Symphysis
Dorsoventral height 21
Anteroposterior depth 23
Teeth
Maxillary tooth
Crown height (25)
Basal crown width 4
Dentary tooth
Crown height (20)
Basal crown width 3
NOTE: Measurements are for the holotypic skull of Nigersaurus taqueti (MNN GDF512). Paren-
theses indicate estimated measurement. Measurements are in millimeters.
FIGURE 5.4. High-resolution

computed tomographic sec-
tions through the right premax-
illa of Nigersaurus taqueti
(MNN GDF512) in (A) longitu-
dinal and (B) cross-sectional
views. The packing pattern of
the tooth battery and the asym-
metrical distribution of enamel
on individual crowns are visi-
ble. altr, alveolar trough; mt,
mature tooth; nt, new tooth; rf,
replacement foramen.
(CT) scans in the upper half of the alveolar ramus In dorsal view, the T-shaped bone is divided
(fig. 5.4B). The premaxillary teeth are self-sup- into a mandibular ramus, oriented anteroposte-
porting and in mutual contact along the lengths riorly, and an alveolar ramus, oriented trans-
of their crowns, which presumably erupted and versely (figs. 5.1B, 5.6). The posterior end of the
wore as a unit. The teeth originate just anterior to mandibular ramus, as in other dinosaurs, is ori-
the replacement foramina. As the teeth increase ented in a parasagittal plane. It is divided into a
in length, their roots extend into the upper half of slender dorsal process, which contacts the suran-
the alveolar ramus before migrating ventrally gular, and a broader, tongue-shaped ventral
toward the alveolar trough (fig. 5.4A). process. The anterior end of the mandibular
ramus, unlike that in any other sauropod, is
MAXILLA dorsoventrally compressed and flares in trans-
The ventral one-fifth of the maxilla houses a verse width before joining the alveolar ramus
battery of teeth arranged in columns internal to (fig. 5.1B). As a consequence, this portion of the
each oval replacement foramen (Sereno et al. mandibular ramus appears unusually slender in
1999:fig.2D; fig. 5.5, table 5.1). The dorsal four- lateral view (fig. 5.1A). In dorsal view, this portion
fifths of the bone is developed proximally as a of the mandibular ramus is subtriangular with a
thin plate and distally as a narrow and delicate concave dorsal surface. Laterally it is bounded by
strut that divides the external naris and antor- a sharp, upturned edge that joins the lateral
bital fenestra (fig. 5.1). As in the premaxilla, the extremity of the alveolar ramus. A large oval fora-
alveolar margin is developed as an undivided men opens in the center of this portion of the
trough, as seen in posterior view, with the labial mandibular ramus. This foramen lies dorsal to a
(anterior) wall extending approximately 1 cm sizable vascular canal that passes anteriorly
farther ventrally than the lingual (posterior) within the dentary toward the alveolar trough.
wall. Unlike the premaxilla and dentary, bony Shallow grooves also exit the foramen and pass
septa are present within the body of the alveolar anteriorly toward the row of replacement foram-
ramus. These septa join the posterior and ante- ina (Sereno et al. 1999:fig. 2C; fig. 5.6).
rior walls, enclosing crypts for each replacement The breadth of the alveolar ramus exceeds
series that are approximately twice as deep labi- the length of the mandibular ramus, a
olingually (anteroposteriorly) as mesiodistally remarkable proportion that results in lower
(transversely). jaws that are as broad as they are long (figs.
Although all worn maxillary teeth have 5.1B, 5.6, table 5.1). The extraordinary width of
fallen from the alveolar trough, columns of the anterior end of the lower jaws is the result
replacement teeth are visible in the two most of the lateral extension of the alveolar ramus,
mesial (medial) tooth positions (Fig. 5.5). As in approximately 30% of which is positioned lat-
the premaxilla and dentary, the teeth originate eral to the sagittal plane of the mandibular
just labially (anteriorly) to each replacement fora- ramus. The medial portion of the alveolar
men, extend with growth dorsally and ventrally ramus is subcylindrical. The mandibular sym-
within the alveolar ramus, and then migrate in physis, as a result, is subcircular (table 5.1).
series toward the alveolar trough. The symphysial surface is smooth and gently
concave, suggesting that a small amount of
DENTARY flexion about the sagittal plane may have been
The dentary has a very unusual structure in possible.
Nigersaurus taqueti (figs. 5.1, 5.6, table 5.1). Other The alveolar trough is nearly straight in
rebbachisaurids may eventually be shown to dorsal view (fig. 5.1B). It narrows in width
share some or all of these apomorphies. Several toward its lateral end. Other changes toward
aspects of the dentary, however, currently have the lateral side of the trough include a decrease
no parallel among reptiles. in the size of the replacement foramina and a
FIGURE 5.5. (A) Stereopair and (B) matching drawing of the left maxilla of Nigersaurus taqueti (MNN GDF512) in poste-
rior view with two replacement tooth columns exposed by erosion. al, alveolus; antfe, antorbital fenestra; apm, articular
surface for the premaxilla; apl, articular surface for the palatine; m1– m20, maxillary tooth positions 1–20; mfr, missing
fragment; rf, replacement foramen; sf, subnarial foramen. Scale bars equal 5 cm.
FIGURE 5.6. (A) Stereopair and (B) matching drawing of the left dentary of Nigersaurus taqueti (MNN GDF512) in dorsal
view. altr, alveolar trough; br, buccal ridge; d1–d34, replacement foramina for dentary tooth positions 1–34; mc, Meckel’s
canal; mfr, missing fragment; pdpr, posterodorsal process; pvpr, posteroventral process; sym, symphysis; t, tooth; vc, vascu-
lar canal. Scale bars equal 5 cm.
FIGURE 5.7. Isolated tooth of Nigersaurus taqueti (MNN GDF513) in (A) lingual (internal), (B) mesial or distal (medial or
lateral), and (C) labial (external) views. de, dentine; e, enamel; ee, enamel edge; ewf, external wear facet; iwf, internal wear
facet. Scale bar equals 1 cm.
decrease in the distance between the replace- maxillary, maxillary, and dentary teeth in situ
ment foramina and the edge of the alveolar in MNN GDF513. Upper and lower teeth differ
trough. This gradual decrease in the width of only in size; dentary teeth are smaller, perhaps
the trough and the size and position of the by as much as 20%–30%. This is difficult to
replacement foramina suggests that the estimate more accurately because the avail-
replacement rate and/or tooth size decreased able dentary teeth, a few of which remain
toward the lateral end of the tooth row. Several within the alveolar trough, are of unknown
loose teeth are preserved in the bottom of the maturity. A similar size differential, with
alveolar trough and confirm the gradual upper teeth larger and longer than the lowers,
decrease in tooth size laterally along the bat- has been observed in Diplodocus (Holland
tery. There are 34 replacement foramina and, 1924:389) and the titanosaur Nemegtosaurus
thus, 34 columns of teeth, which is 20 more (Wilson 2005: 305).
than are typically present in the dentary of the The numbers of premaxillary, maxillary, and
diplodocid Diplodocus (Holland 1924; fig. dentary tooth positions are 4, at least 24, and
5.2B). As there are no septa at any depth within 34, respectively. Maxillary and dentary tooth
the alveolar trough, dentary teeth would have counts are approximately double those in other
erupted and migrated toward the functioning diplodocoids such as Diplodocus (Holland,
wear surface as a self-supporting dental battery. 1924; fig. 5.2B). Although the maxillary tooth
No other sauropod has a dentary with an alveo- row is incomplete laterally, there are probably
lar ramus of similar form lacking septa. only a few missing tooth positions, because the
length of the combined premaxillary and maxil-
TEETH lary tooth rows in MNN GDF513 is only slightly
The upper and lower teeth are virtually identi- less than that for the dentary. The higher num-
cal in form (Sereno et al. 1999: fig. 2C; fig. ber of tooth positions in the dentary (34 versus
5.7). These similarities are based on the pre- approximately 30 for the premaxilla plus
maxilla) is consistent with the smaller size of the product of nonocclusal abrasion (Holland
the dentary teeth. 1924; Upchurch and Barrett 2000; fig. 5.7B, C).
Crown shape and structure are uniform The elevated, rounded rim of thick enamel sug-
among upper and lower teeth. Unworn crowns gests that the wear facet was produced by
are gently curved along their length and taper ingested plant matter rather than the harder
gradually to a narrow, rounded tip, which enameled edge of an opposing crown. The
shows subtle wrinkles on its thickly enameled second—a lingual (internal) facet—resembles
labial (external) surface. Teeth in place in the the internal facet on some narrow-crowned
premaxilla and maxilla and near their natural titanosaurians; it cuts the crown at a low angle
position in the dentary confirm that the crowns and appears to be the product of occlusal abra-
are convex labially (anteriorly) in both the upper sion (fig. 5.7B, C).
and the lower jaws.
At midlength, the crown has a trapezoidal
TOOTH BATTERY
cross section, with smooth enamel and a longi-
tudinal groove on each side that accommodates STRUCTURE
the edges of adjacent crowns (Sereno et al. Archosaur teeth, primitively, are anchored by
1999:fig. 2D; fig. 5.7). Toward its tip, the crown alveoli that are separated from adjacent teeth by
has an oval cross section. The enamel is approx- bony septa. They erupt en echelon along the jaw
imately eight times thicker on the labial (exter- rather than in unison as a single unit. In most
nal), as opposed to the lingual (internal), side of dinosaurian herbivores, wear facets develop
the crown in both upper and lower teeth. from tooth-to-tooth occlusion at several points
Whereas other diplodocids and some titanosauri- along the tooth row rather than developing as a
ans have similar-shaped, narrow, cylindrical continuous wear surface.
crowns, markedly asymmetrical enamel has not A tooth battery, by contrast, is here defined as
been reported previously among sauropods and a tooth composite composed of self-supporting
is absent in other diplodocoids (Dicraeosaurus teeth that erupt and wear in unison. The teeth
[Janensch 1935–1936:pl. 12, fig. 16], Diplodocus are supported by adjacent teeth and erupt as a
[Holland 1924:fig. 6]. unit with wear surfaces that are continuous
from one tooth to the next. In Nigersaurus part
TOOTH WEAR of the dentition fulfills these criteria—the pre-
Although wear facets are not present in teeth maxilla and dentary tooth rows. These bones do
preserved in the holotypic jaw bones of not have alveoli, intervening septa, or even well-
Nigersaurus taqueti, worn teeth referable to this developed grooves to guide erupting tooth
species have been recovered from many sites in columns. Rather, the teeth are packed into an
the Gadoufaoua beds and show a distinctive open alveolar trough.
pattern of wear. There is little reason to doubt Although all functioning (worn) teeth have
the reference of these teeth to Nigersaurus fallen away from the jaws of the holotypic spec-
taqueti, despite the presence of an unnamed imen, the tooth battery is intact within the body
titanosaurian in the same horizons; the crowns of the premaxilla (fig. 5.3), as visualized in cross
have narrow cylindrical proportions and highly and longitudinal sections with high-resolution
asymmetrical enamel. CT (fig. 5.4). Embryonic teeth migrate into the
Unlike the teeth in nearly all other alveolar trough via replacement foramina. As
sauropods, those in Nigersaurus taqueti have a they grow in length and diameter, they migrate
pair of wear facets located on opposite sides of deeper into the body of the premaxilla before
the crown. The first—a labial (external) facet— passing ventrally out of the alveolar trough (fig.
is typical of dicraeosaurids and diplodocids; it 5.4A). There appear to be as many as 10 teeth at
cuts the crown at a high angle and appears to be a single tooth position, from embryonic tooth to
wearing crown. The four tooth columns in the (anteriorly). Many mammals, such as rodents,
premaxilla are arranged en echelon so that have analogous anteriorly positioned, self-
the widest portion of a given crown contacts the sharpening lower and upper incisors with lin-
more tapered portion of adjacent crowns, with gual (internal) wear facets (Taylor and Butcher
the prominent edge of enamel lodged in a 1951). As in Nigersaurus, both lower and upper
groove on the side of adjacent crowns (fig. 5.4B). incisors are externally convex, with thickened
Less is known about the structure of the enamel on their labial (external) sides and wear
tooth battery of the dentary, although it was facets facing lingually (internally). This is the
probably very similar. The crowns have the closest analogy to the structure and orientation
same shape, structure, and orientation. The of the individual teeth in Nigersaurus. Mam-
position of the replacement foramina near the malian incisors like these, however, are much
margin of the alveolar trough indicates that more robust and are sharpened in a manner
embryonic teeth began their trajectory closer to very different from that in any dinosaur. The
the open end of the trough and grew deep into chisel-shaped edge is maintained by breakage
the alveolar trough before emerging at the (preferential chipping of the softer dentine) and
functional end of the tooth battery. The most by tooth-to-tooth abrasion. Unlike Nigersaurus,
distal (lateral) tooth columns in the dental bat- the facets on lower and upper incisors are not
tery of the dentary must have angled anterolat- symmetrical (the uppers typically have higher-
erally as shown by the extension of the trough angle, stepped facets), and the shearing break-
lateral to the lateral most replacement foramen age that keeps the leading edge sharp rarely pro-
(fig. 5.1B). duces facets that are uniformly planar.
The body of the maxilla has septa separating We strongly suspect that the low-angle, lin-
columns of teeth (fig. 5.5). Closer to the ventral gual (internal) wear facet was produced by tooth-
end of the maxilla, the septa give way to an open to-tooth abrasion, because the facet is extremely
trough with a groove for each tooth column on flat and cuts smoothly across the external mar-
anterior and posterior walls. As in the premax- gin of enamel. The high-angle labial (external)
illa, there appear to have been at least 10 teeth wear facet, in contrast, appears to have been pro-
to a column medially and fewer laterally; an duced by tooth-to-plant abrasion, because the
incomplete sequence of 8 teeth is visible in the facet is concave, with a rounded, polished rim of
first column (fig. 5.5). The maxillary teeth, pre- enamel along its trailing labial (external) edge
sumably, emerged from the alveolar trough as a (fig. 5.7).
self-supported tooth battery. Because all of the isolated teeth from Niger,
England (Naish and Martill 2001), and Brazil
FUNCTION (Kellner 1996) show a similar pattern of wear—
Exactly how the tooth batteries of Nigersaurus with the low-angle facet first to appear—it is
functioned to produce the wear facets evident on likely that both lower and upper teeth are rep-
isolated teeth remains unsolved, despite knowl- resented and wear in a similar manner. Yet it is
edge of the general structure of the tooth batter- not clear how this is functionally possible.
ies. What we can outline at this point are several Hadrosaurid and ceratopsian tooth batteries
observations that provide some insight into the have opposing, mirrored patterns of tooth wear:
mystery. the enamel is thickest on opposing sides of the
We know that teeth in lower and upper tooth crown (medially in maxillary teeth, laterally in
rows do not match one-to-one given their dif- dentary teeth); the crowns curve in opposite
fering sizes and numbers. We also know that directions (laterally in maxillary teeth, medially
both upper and lower tooth batteries were con- in dentary teeth); and the wear facets occur in
structed in a similar manner, with the convex, opposing orientations (medially or ventromedi-
thickly enameled crown surface facing labially ally in maxillary teeth, laterally or dorsolaterally
in dentary teeth). Although an isolated ornithis- must have been positioned at a good distance
chian tooth from a tooth battery may be difficult from the cutting edge of the functional crown.
to assign to either the upper or the lower tooth Only one crown in each tooth column could
row (when upper and lower crowns have a sim- have been functional. In ornithischian denti-
ilar shape and ornamentation), these teeth have tions, higher-angle wear facets allow more than
mirrored positions when they are found in a single crown in the same tooth column to par-
place. One possible explanation for the uniform ticipate in the active, cutting surface of the den-
pattern of wear in Nigersaurus is that the lingual tal battery.
(internal), low-angle facet is produced by the In the future, we plan to digitally define and
lower crowns passing lingually (internally) to prototype the intact portion of the tooth battery
the upper crowns (the usual tetrapod condition) within the premaxilla in the hope that it will
but that the lower crowns are worn away and shed further light on how the battery func-
eventually obliterated in the process. If this tioned during mastication.
were true, the entire sample of isolated teeth of
Nigersaurus and related taxa found now on sev-
DISCUSSION
eral continents would include only premaxillary
and maxillary teeth from the upper tooth bat-
COMPARISONS
teries. If not, and if we are correct that the low-
angle facet is produced by tooth-to-tooth oc- SKULL
clusion, there must have existed an occlusal In general proportions, the skull of Nigersaurus
mechanism unlike any described to date among taqueti bears little resemblance to that of
tetrapods—a mechanism capable of producing diplodocids (figs. 5.1, 5.2B). It has a more abbre-
low-angle, lingual (internal) wear facets on both viate, less prognathous snout; the depth of the
lower and upper crowns. cranium is approximately 90% of its length (as
Another conundrum involves the cutting measured from the snout to the quadrate
edge of the crown. In ornithischians with tooth condyle). This cranial proportion is even more
batteries and in mammalian herbivores, thick- abbreviate than that of Camarasaurus (fig. 5.2A)
ened enamel is always positioned along the cut- and Jobaria (Sereno et al. 1999), which have a
ting edge. This is not the case in Nigersaurus. cranial depth between 50% and 60% of its
Not only are the lower and upper teeth oriented length. The external naris in Nigersaurus is large
with their curvature and thickened enamel on and parasagittal in position as in Camarasaurus
the same (external) side, but the leading wedge- and Jobaria. In diplodocids, in contrast, the exter-
shaped edge of the crown is formed entirely of nal naris is smaller, dorsally facing, and retracted
dentine. It is difficult to understand how this to a position anterodorsal to the orbit (the condi-
edge, which is perfectly straight, is maintained tion in dicraeosaurids is as yet unknown). The lat-
in the course of wear without the protection of erotemporal fenestra in Nigersaurus is propor-
enamel (fig. 5.7). Even if one envisages adjacent tionately elongate and extends anteriorly as far as
crowns, it is hard to understand how the soft the antorbital fenestra and external nares, farther
leading edge of dentine would not be concave than in any sauropod described to date. There is
from wear rather than straight. no development of a preantorbital fenestra as
A final observation suggests that the tooth occurs in other neosauropods (fig. 5.2B). The der-
battery of Nigersaurus is quite different from mal bones of the skull roof in Nigersaurus are
that in ornithischians. To produce an elongate, remarkably slender and delicate compared to
low-angle wear facet on the lingual (internal) those of other sauropods including diplodocids.
side of the crown (fig. 5.7A), most of this side of The posterodorsal ramus of the maxilla, which
the crown must have been exposed. The suc- separates the external naris and antorbital fossa,
ceeding replacement crown, in other words, is reduced to an extremely delicate, strap-shaped
lamina 1 mm thick and a few millimeters wide and Martill 2001:pl. 36) and from the Upper
(fig. 5.1). Cretaceous Bauru Group in Brazil (Kellner
The lower jaw is easy to distinguish from that 1996:fig. 7). The crowns are narrow and sub-
in diplodocids (figs. 5.1, 5.2B). The coronoid cylindrical and, at midlength, have a trapezoidal
process on the surangular is prominent and cross section. The size, form, and angle of the
deep in lateral view, more closely resembling pair of wear facets are exactly like those
that in the titanosaurian Rapetosaurus (Curry described above for Nigersaurus taqueti. As
Rogers and Forster 2004) than the low profile noted by Kellner (1996:619), the low-angle lin-
jaws of diplodocids (fig. 5.2B). In Nigersaurus the gual (internal) facet is the first to appear and is
teeth are restricted to the transverse portion of always more elongate than the labial (external)
the anterior end of the skull, and at least the facet, as preserved in progressively worn
lower tooth row extends lateral to the parasagittal crowns. The enamel may have an asymmetri-
plane of the lower jaw. In both of these attrib- cal distribution on the crown, but this needs
utes, Nigersaurus is unique among dinosaurs. firsthand verification. Kellner provisionally
Diplodocids show an incipient condition in these regarded these teeth as titanosaurian, because
regards; all but the lateral extremities of the tooth of the predominance of titanosaurian postcra-
rows are positioned along the anterior, trans- nial bones from the same beds. It is very prob-
verse margin of the skull, and the dentary tooth able, however, that these Brazilian teeth belong
row flares just beyond the parasagittal plane of to a rebbachisaurid diplodocoid that lived on
the posterior portion of the lower jaw. South America in the Late Cretaceous. (e.g.,
A dentary from Upper Cretaceous rocks in Limaysaurus; Salgado et al. 2004).
South America referred to Antarctosaurus wich- In other diplodocoids, the teeth are larger
mannianus (Huene 1929:69, pl. 29, fig. 5) is sim- relative to the jaw bones and anchored in indi-
ilar to that in Nigersaurus taqueti in the extreme vidual alveoli (fig. 2B; Dicraeosaurus [Janensch
breadth of the transverse portion of the ramus 1935–1936:figs. 107, 111]; fig. 2B). The crowns
and the concomitant increase in tooth count. have a circular cross section, symmetrical
There are at least 24 teeth in the dentary, with the enamel, and a single low-angle, labial (external)
majority (approximately 18) located in the broad, wear facet (Holland 1924). Despite some varia-
transverse portion of the ramus. The dentary of tion, isolated teeth of Dicraeosaurus show the
this South American form, however, is not as same external facet (Janensch 1935–1936:pl. 12,
derived as that in Nigersaurus. The teeth are set in figs. 23, 25). This facet is characterized by a
sockets rather than an undivided alveolar trough, rounded lip of enamel along its trailing (external)
the tooth row is L-shaped rather than straight and edge (fig. 5.7; ewf) and scratches that course
restricted to the transverse portion of the ramus, across the dentine from the internal to the exter-
and the symphysial surface is narrow rather than nal sides (Fiorillo 1991; Calvo 1994b; Upchurch
circular. Despite these similarities, the phyloge- and Barrett 2000). How this wear facet formed
netic affinity of Antarctosaurus is not yet resolved. has remained a mystery ever since Holland
There is a possibility that this taxon, which was (1924) summarized early speculation, which
found together with titanosaur cranial and post- included scraping cycad trunks (Holland),
cranial remains, may have acquired these fea- procuring fish hiding in stream beds (Tornier),
tures convergently with Nigersaurus. and munching on freshwater bivalves (Sternfeld).
Because Holland (1924:fig. 4) depicted “Diplod-
TEETH AND TOOTH WEAR ocus seizing a mussel,” that hypothesis gained
Teeth that closely match those of Nigersaurus the upper hand, although Holland maintained
taqueti in form, structure, and wear have no personal preference. More recent proposals
been described recently from slightly older include stripping leaves from branches (Dodson
Barremian-age beds on the Isle of Wight (Naish 1990; Barrett and Upchurch 1994) and raking
FIGURE 5.8. Fifth cervical vertebra and co-ossified rib of Nigersaurus taqueti (MNN GDF512) in lateral view. asp, anterior
spine; ns, neural spine; pl, pleurocoel; poz, postzygapophysis; prz, prezygapophysis; rs, rib shaft. Scale bar equals 5 cm.
bark (Bakker 1986). Nigersaurus adds a new wrin- 2005), has continued to confuse discussions of
kle. Here we see the same kind of external facet tooth form and masticatory style among
in teeth set into a tooth battery. These batteries, in diplodocoids (e.g., Upchurch and Barrett 2000).
turn, are housed in a skull with a very different Nemegtosaurus and the conspecific, or closely
shape and even more delicate construction than allied, Quaesitosaurus from the Late Cretaceous
in diplodocids (fig. 5.2). of Asia are better understood as titanosaurians
Some titanosaurians have narrow-crowned rather than diplodocoids (Calvo 1994a; Wilson
teeth that bear a general resemblance to those of and Sereno 1998; Wilson 2002, 2005).
Nigersaurus taqueti, but these teeth have more
robust proportions (e.g., Kellner and Mader EVOLUTION OF A SAUROPOD TOOTH BATTERY
1997:fig.2; Rapetosaurus [Curry Rogers and Nigersaurus taqueti must be placed within the
Forster 2004:fig. 32]). The internal (lingual) context of diplodocoid phylogeny to better under-
facet—the only one present—cuts the crown at stand how its novel tooth battery evolved
a low angle. An external (labial) facet has (fig. 5.9). Diplodocoids include diplodocids,
never been described outside Diplodocoidea. dicraeosaurids, and rebbachisaurids. Diplodocids
Nemegtosaurus, which has internal V-shaped and are best known from the Late Jurassic of North
apical wear facets (Nowinski 1971; Wilson America and include Diplodocus, Apatosaurus,
db 1
34 ewf
iwf
65
Maa
71
DICRAEOSAURIDAE
Cmp
L REBBACHISAURIDAE 4 REBBACHISAURIDAE
84
San
15
2
MACRONARIA
Cen
CRETACEOUS
99
Alb
ns
3 DIPLODOCIDAE
Apt
DIPLODOCIDAE
E
121 2 DICRAEOSAURIDAE
Brm +
DIPLODOCIDAE
Hau
Vlg
Ber srm
144 ewf
Tth
L
159
1 DIPLODOCOIDEA
FIGURE 5.9. Phylogram showing the relationships and recorded durations of the three principal diplodocoid clades (Reb-
bachisauridae, Dicraeosauridae, Diplodocidae) and some of their feeding specializations. Node 1 (Diplodocoidea): slender
cylindrical crowns, low-angle external (labial) wear facet, teeth restricted anteriorly in subrectangular muzzle. Node 2 (Di-
craeosauidae + Diplodocidae): bifid cervical neural spines. Node 3 (Diplodocidae): nares retracted, 15 cervical vertebrae. Node 4
(Rebbachisauridae): low-angle, internal wear facet, increase in tooth number, asymmetrical enamel, dental batteries in some
taxa. 1–34, dentary teeth 1 through 34; 2–15, cervical vertebrae 2–15; db, dental battery; en, external naris; ewf, external wear
facet; iwf, internal wear facet; ns, neural spine; srm, subrectangular muzzle.
Barosaurus, and others (Hatcher 1901; Holland to the end of the Cretaceous at least in South
1924; Berman and McIntosh 1978; Ostrom America. As a sister taxon to dicraeosaurids and
and McIntosh 1999). Dicraeosaurids include diplodocids, rebbachisaurids must have diverged
Dicraeosaurus (Janensch 1935–1936) from the from other diplodocoids before the end of the
Late Jurassic of Africa and the long-spined Late Jurassic, although no trace of the group has
Amargasaurus from the Early Cretaceous of South yet been found from this period (fig. 5.9).
America. Rebbachisaurids include Limaysaurus Although the teeth of Nigersaurus are partic-
(Calvo and Salgado 1995; Salgado et al. 2004) ularly slender, all diplodocoids are characterized
from the Early Cretaceous of South America by proportionately narrow, subcylindrical
and Nigersaurus (Sereno et al. 1999) and crowns that are only weakly expanded and
Rebbachisaurus (Lavocat 1954) from the mid- and flattened (fig. 5.9, node 1). The rate of tooth
Late Cretaceous of Africa. Rebbachisaurid teeth replacement, in addition, appears to have been
and other fragmentary remains indicate that the accelerated across the group, although this is
group was also present in Europe and persisted most apparent in Nigersaurus. In this regard, we
assume that the number of teeth in a single col- COMPARISON TO ORNITHISCHIAN
umn is related to the replacement rate. TOOTH BATTERIES
Diplodocus, for example, has as many as six Tooth batteries evolved twice among ornithis-
teeth in a single column in the maxilla (Marsh chian dinosaurs from a hypothetical common
1884:pl. 4, fig. 3). Worn diplodocoid teeth have ancestor with a simple dentition characterized
a characteristic external, low-angle, wear facet by a suite of ornithischian features related to
on both upper and lower teeth (fig. 5.7 ewf) herbivory—a predentary, cheek embayments on
that must have formed from tooth-to-plant the dentary and maxilla, asymmetrical enamel
abrasion (contra Calvo 1994b; Barrett and in dentary and maxillary crowns, and wear
Upchurch 1994; Upchurch and Barrett 2000). facets from tooth-to-tooth occlusion on the buc-
Most diplodocoid teeth are positioned in a cal (lateral) and lingual (medial) sides of dentary
transverse row along the squared anterior mar- and maxillary teeth, respectively. Evolution of
gin of the snout (fig. 5.9, node 1). tooth-supported batteries occurred first among
Nigersaurus has further developed a number ornithopods, with initial changes toward a den-
of these features (fig. 5.9, node 2). Tooth size tition occurring before the close of the Jurassic,
decreases relative to the size of the jaw bones, and later among ceratopsian dinosaurs, where
and tooth number increases to 30 or more in all structural changes occurred during the Late
lower and upper tooth rows. In Nigersaurus the Cretaceous (Sereno 1997, 2000).
teeth are restricted to the anterior margin of the The fully developed tooth batteries in
snout and extend lateral to the sagittal plane of hadrosaurids and ceratopsids are structurally
the lower jaw, features unknown elsewhere very similar, evolved in comparable stages, and
among dinosaurs. More teeth are present in a involved an increase in body size of approxi-
single column, and the rate of replacement, pre- mately an order of magnitude (Sereno 1997: fig.
sumably, has increased as well. Most of the teeth 7; 2000: fig. 25.9). Independent but similar
in the jaws are pressed so close to one another structural changes include a relative decrease in
that there are no intervening septa and, for the tooth size and increase in tooth columns and
first time among sauropods, batteries of self- replacement rate, the loss of alveolar septa,
supporting teeth erupt as a single unit. Although restriction of the enamel to opposing sides of the
still a functional conundrum, an additional low- crown in maxillary versus dentary teeth, an
angle wear facet appears on the lingual (internal) increase in the prominence of a ridge on the
side of the active crowns in both lower and upper enameled side of the crown, adjustment of the
tooth batteries. crown shape for efficient packing, an increase in
Dicraeosaurids and diplodocids evolved the volume of supporting bone in the maxilla
other features that likely impacted food procure- and dentary, a reduction of postdentary elements
ment but are not present in Nigersaurus or other in the lower jaw, and the development of a coro-
rebbachisaurids. These include a reduction in noid process with an expanded process for mus-
the number of teeth, a projecting “chin” on the cular attachment.
dentary, a relative lengthening of the snout, The circumstances surrounding the evolu-
retraction of the external nares, and an increase tion of tooth batteries in rebbachisaurid
in the length and number of cervical vertebrae sauropods bear a few similarities to and many
(fig. 5.9, nodes 3, 4). In dicraeosaurids and striking differences from those in ornithischi-
diplodocids, in addition, the resting curvature of ans. Similarities include the reduction of tooth
the neck arches anteroventrally (Janensch 1929; size, increase in number of tooth columns,
Stevens and Parrish 1999). This is not the case increase in replacement rate (or at least the
in Nigersaurus, as exemplified by the fifth cervi- number of teeth per column), loss of alveolar
cal vertebra (fig. 5.8). septa between tooth columns, thickened
enamel on one side of the crown and near-loss Gastroliths are present in rebbachisaurids (Calvo
of enamel on the other, and adjustment of crown 1994b), dicraeosaurids (Janensch 1929), and
shape for efficient packing. These similarities, diplodocids (Cannon 1906; Brown 1941; Gillette
thus, are confined to the size, shape, number, 1990). The absence of gastroliths among
and rate of replacement of the teeth and the nondiplodocoid sauropods is based on many
asymmetrical distribution of enamel. articulated skeletons of Shunosaurus,
Fundamental differences begin with timing. Camarasaurus, Jobaria, and Opisthocoelicaudia.
Available rebbachisaurid fossils suggest that a Cedarosaurus, a macronarian sauropod of uncer-
tooth battery among sauropods had evolved some- tain affinity, is thus far the only nondiplodocoid
time during the Early Cretaceous; the teeth from sauropod with gastroliths (Sanders et al. 2001).
the Isle of Wight (Naish and Martill 2001) look The pattern of wear in the rebbachisaurid
very similar to those of Nigersaurus and are tooth battery is completely different from that
Barremian (ca. 125 Ma; middle Early Cretaceous) in ornithischians, which uses thickened enamel
in age. This postdates the establishment of tooth- and tooth-to-tooth occlusion to form a self-
supported dentitions in ornithopods (Late sharpening cutting margin. In rebbachisaurids,
Jurassic) but predates the appearance of ceratop- only one of a pair of wear facets is formed by
sian tooth batteries (Late Cretaceous). Another tooth-to-tooth occlusion, the sharp leading edge
basic difference involves body size. There was no of worn crowns is formed in dentine rather
increase in body size among sauropods concomi- than enamel, the thickened enamel is located
tant with the development of tooth batteries. on the same (labial) side of lower and upper
Although Rebbachisaurus ranks among the largest crowns, and lower and upper crowns appar-
of sauropods, Nigersaurus ranks among the small- ently have identical wear patterns.
est, with a body length of approximately 15 m. The locus of the most rapid replacement and
The orientations of the tooth batteries are wear is different in ornithischian and reb-
diametrically opposed. The tooth batteries have bachisaurid tooth batteries. In ornithischians,
an anteroposterior, rather than a transverse, ori- the crown size, the number of teeth in a col-
entation in ornithischians and rebbachisaurids, umn, and the distance of the replacement fora-
respectively. The ornithischian tooth battery is men from the alveolar margin are all greatest in
located posteriorly within the jaws and used for the middle of the tooth battery. In Nigersaurus,
food processing; cropping is a function of an in contrast, the crown size, the number of teeth
expanded, toothless bill. The sauropod dental in a column, and, to a lesser degree, the distance
battery, in contrast, is located anteriorly and may of the replacement foramen from the alveolar
have been used primarily in cropping. margin are all greatest toward the midline.
The presence or absence of gastroliths as an Finally, the tooth-bearing bones are con-
accessory means to break down plant matter may structed differently in ornithischians and
be correlated with the aforementioned funda- sauropods with dental batteries. In ornithischi-
mental functional differences. The absence of ans with tooth batteries, the dentary, in particu-
gastroliths among ornithischians with advanced lar, is robustly constructed, with a thick and
dentitions (euornithopods, neoceratopsians) prominent coronoid process for attachment of
including dental batteries suggests that more effi- substantial adductor musculature. Postdentary
cient oral processing of plant matter has replaced elements are greatly reduced in size. This pat-
gut-processing by gastroliths (Sereno 1997:473). tern of change in the lower jaw closely mirrors
In sauropods, in contrast, the group in which changes that occurred earlier in the evolution
dental batteries evolved (diplodocoids) has the of the mammalian masticatory apparatus
greatest proven incidence of gastroliths, suggest- (Allin 1975). In Nigersaurus, in contrast, the
ing that their derived dental features did not func- tooth-bearing elements are constructed of thin
tion primarily in the breakdown of plant matter. laminae, the dentary does not gain in relative
length in the lower jaw, the coronoid process is ing. Although the rebbachisaurid dental battery
developed as a thin plate of bone, and the is preserved only in Nigersaurus, isolated teeth
supratemporal fenestra (the usual origin of from Lower Cretaceous horizons on the Isle of
adductor musculature) is closed by approxima- Wight and rocks of Late Cretaceous age in Brazil
tion of surrounding bones. suggest that related forms with potentially a sim-
ilar degree of dental complexity were present on
other continents. It is highly unlikely, however,
CONCLUSIONS
that rebbachisaurids with dental batteries ever
Early in their evolution, sauropods adopted achieved the taxonomic diversity of ornithischi-
tooth-to-tooth occlusion and, in consequence, ans with dental batteries (hadrosaurids, ceratop-
evolved lower and upper tooth rows of equivalent sids) in Late Cretaceous faunas of North America
length, characteristic patterns of wear facets, a and Asia.
more substantial coronoid process, and a robust Dental batteries evolved three times inde-
mandibular symphysis (McIntosh 1990; Calvo pendently within Dinosauria—in euornithopod
1994a; Wilson and Sereno 1998; Upchurch and and neoceratopsian ornithischians and in reb-
Barrett 2000). Among sauropods, it is now bachisaurid sauropods. Fundamental functional
apparent that diplodocoids evolved complex den- differences coupled with their diachronous
titions during the Cretaceous, as exemplified by appearance suggest that dinosaurian dental bat-
the dental batteries of a recently named African teries did not evolve in response to a single envi-
rebbachisaurid, Nigersaurus taqueti (Taquet 1976; ronmental cue, such as the rise of angiosperms
Sereno et al. 1999). during the mid-Cretaceous (Sereno 1997: Sereno
The dental battery on each side of the 1999; Barrett and Willis 2001).
upper and lower jaws is composed of more
than 30 columns of teeth that are packed into ACKNOWLEDGMENTS
a tight self-supporting unit in the premaxilla We thank members of the 1997 and 2000
and dentary. Individual teeth have slender rod- expeditions for discovering the material, C.
shaped crowns characterized by highly asym- Abraczinskas for drawing from specimens and
metrical enamel. Dentary teeth are somewhat executing final drafts of reconstructions, E.
smaller than but otherwise similar to premax- Dong, T. Keillor, and R. Masek for preparing
illary and maxillary teeth. The crowns in both fossils, and B. Gado (Institut de Recherche en
lower and upper jaws have thickened enamel Science Humaine), and I. Kouada (Ministère de
on their convex labial (external) side. Wear L’Enseignement Supérieur de la Recherche et
produces two stereotypical facets, the first de la Technologie) for granting permission to
appearing as a low-angle, lingual (internal) conduct fieldwork. This research was funded by
facet produced by tooth-to-tooth occlusion and The David and Lucile Packard Foundation, the
the second as a high-angle, labial (external) National Geographic Society, the Comer
facet produced by tooth-to-plant abrasion. Science and Education Foundation, and Nathan
Both facets are well developed on crowns with Myhrvold.
significant wear, resulting in a straight, sharp
apical wedge of dentine where the facets inter-
LITERATURE CITED
sect. How either of these facets was produced
remains a significant, and largely unanswered, Allin, E.F. 1975. Evolution of the mammalian middle
question. ear. J. Morphol. 147: 403–438.
Apesteguía, S., Valais, S. D., Gonzales, J. A., Gallina,
Unlike the parasagittal dental batteries of
P. A., and Agnolin, F. L. 2001. The tetrapod fauna
ornithischians, the dental battery in Nigersaurus of ‘La Buitrera,’ new locality from the basal Late
is oriented transversely and may have been used Cretaceous of North Patagonia, Argentina. J.
for cropping rather than prolonged oral process- Vertebr. Paleontol. 21: 29A.
Bakker, R. T. 1978. Dinosaur feeding behavior and the Gillette, D. 1990. Gastoliths of a saurpod dinosaur
origin of flowering plants. Nature 274: 661–663. from New Mexico. J. Vertebr. Paleontol. 10: 24A.
——. 1986. The Dinosaur Heresies. Bath Press, Hatcher, J. B. 1901. Diplodocus (Marsh): its osteology,
Avon. taxonomy, and probable habits, with a restoration
Barrett, P. M., and Upchurch, P. 1994. Feeding of the skeleton. 1: 1–63.
mechanisms of Diplodocus. GAIA 10: 195–204. Holland, W. J. 1924. The skull of Diplodocus. Mem.
Barrett, P. M., and Willis, K. J. 2001. Did dinosaurs Carnegie Mus. 9: 379–403.
invent flowers? Dinosaur-angiosperm coevolu- Huene, F.von. 1929. Los Saurisquios y Ornithisquios
tion revisited. Biol. Rev. 76: 411–447. de Cretacéo Argentino. Ann. Mus. La Plata 3: 1–196.
Berman, D. S., and McIntosh, J. S. 1978. Skull and Hunt, A. P., Lockley, M. G., Lucas, S. G., and Meyer,
relationships of the Upper Jurassic sauropod C. A. 1994. The global sauropod record. In:
Apatosaurus (Reptilia: Saurischia). Bull. Carnegie Lockley, M. G., dos Santos, V. F., Meyer, C. A., and
Mus. Nat. Hist. Pittsburgh 8: 1–35. Hunt, A. (eds.). Aspects of Sauropod Paleobiology.
Bonaparte, J. F. 1996. Dinosaurios de America del GAIA 10: 261–279.
Sur. Impreso en Artes Gráficas Sagitario Iturri, Janensch, W. 1929. Die Wirbelsäule der Gattung
Buenos Aires. Dicraeosaurus. Palaeontographica 2: 39–133.
——. 1997. Rayososaurus agrioensis Bonaparte 1995. ——.1935–1936. Die Schädel der Sauropoden
Ameghiniana 34: 116. Brachiosaurus, Barosaurus und Dicraeosaurus aus
Brochu, C. A. 2003. Osteology of Tyrannosaurus rex: den Tendaguruschichten Deutsch-Ostrafrikas.
insights from a nearly complete skeleton and high- Palaeontographica 2(Suppl. 7): 147–298.
resolution computed tomographic analysis of the Kellner, A.W.A. 1996. Remarks on Brazilian
skull. J. Vertebr. Paleontol. Suppl. 22: 1–138. dinosaurs. Mem. Queensland Mus. 39: 611–626.
Brown, B. 1941. The last dinosaurs. Nat. Hist. 48: Kellner, A.W.A. and Mader, B.J. 1997. Archosaur
290–295. teeth from the Cretaceous of Morocco. Journal of
Calvo, J. O. 1994a. Jaw mechanics in sauropod Paleontology 71: 525–527.
dinosaurs. GAIA 10: 183–193. Lamanna, M.C., Martinez, R.D., Luna, M., Casal, G.,
——. 1994b. Feeding Mechanisms in Some Dodson, P., and Smith, J.B. 2001. Sauropod fau-
Sauropod Dinosaurs. Master’s thesis. University nal transition through the Cretaceous Chubut
of Illinois at Chicago, Chicago. Group of central Patagonia. Journal of Vertebrate
Calvo, J. O., and Salgado, L. 1995. Rebbachisaurus tes- Paleontology 21: 71A.
sonei sp. nov. a new Sauropoda from the Albian- Larsson, H.C.E., and Gado, B. 2000. A new Early
Cenomanian of Argentina; new evidence on the Cretaceous crocodyliform from Niger. Neues
origin of Diplodocidae. GAIA 11: 13–33. Jahrb. Geol. Palaontol. Abhandlungen 217: 131–141.
Cannon, G. L. 1906. Sauropodan gastroliths. Science Lavocat, R. 1954. Sure les dinosauriens du
24: 116. Continental Intercalaire des kem-Kem de la
Curry Rogers, K., and Forster, C. A. 2004. The skull of Daoura. Comptes rendus de la Dix–Neuviéme
Rapetosaurus krausei (Sauropoda: Titanosauria) Session, Congrès Géologique International,
from the Late Cretaceous of Madagascar. J. Alger, ASGA, fasc. 21. Pp. 65–68.
Vertebr. Paleontol. 24: 121–144. Lucas, S. G., and Hunt, A. P. 1989. Alamosaurus and
Dalla Vecchia, F. M. 1998. Remains of Sauropoda the sauropod hiatus in the Cretaceous of the
(Reptilia, Saurischia) in the Lower Cretaceous North American Western Interior. In: Farlow,
(upper Hauterivianl/lower Baremian) limestones J. O. (eds.). Paleobiology of the Dinosaurs.
of SW Istria (Croatia). Geol. Croatia 51: 105–134. Geological Society of America, Special Papers.
Dodson, P. 1990. Sauropod paleoecology. In: Pp. 75–85.
Weishampel, D. B., Dodson, P., and Osmólska, Lull, R. S., and Wright, N. E. 1942. Hadrosaurian
H. (eds.). The Dinosauria. University of California dinosaurs of North America. Geol. Soc. Am. Spec.
Press. Pp. 402–407. Papers 40: 1–242.
Fiorillo, A. R. 1991. Dental microwear on the teeth of Marsh, O.C. 1884. Principal characters of American
Camarasaurus and Diplodocus: implications for Jurassic dinosaurs. Part VII. On the Diplodocidae,
sauropod paleoecology. In: Kielan-Jaworowska, a new family of the Sauropoda, American Journal
Z., Heintz, N., and Nakren, N. A. (eds.). Fifth of Science (series 3) 27: 161–168.
Symposium on Mesozoic Ecosystems and Biota, McIntosh, J. S. 1990. Sauropoda. In: Weishampel,
Extended Abstracts. Contributions of the D. B., Dodson, P., and Osmólska, H. (eds.). The
Paleontological Museum, University of Oslo, Dinosauria. University of California Press. Pp.
Oslo. Pp. 23–24. 345–401.
Naish, D., and Martill, D. M. 2001. Saurischian Sereno, P. C., Beck, A. L., Dutheil, D. B., Gado, B.,
dinosaurs I: Sauropods. In: Martill, D. M., and Larsson, H. C. E., Lyon, G. H., Marcot, J. D.,
Naish, D. (eds.). Dinosaurs of the Isle of Wight. Rauhut, O. W. M., Sadlier, R. W., Sidor, C. A.,
Palaeontological Association, London. Pp. 185–211. Varricchio, D. J., Wilson, G. P., Wilson, J. A. 1998.
Nowinski, A. 1971, Nemegtosaurus mongoliensis n. A long-snouted predatory dinosaur from Africa
gen., s. sp, (Sauropoda) from the Uppermost and the evolution of spinosaurids. Science 282:
Cretaceous of Mongolia. Palaeontologia Polonica 1298–1302.
25: 57–81. Sereno, P. C. Beck, A. L., Dutheil, D. B., Larsson,
Ostrom, J. H. 1961. Cranial morphology of the H. C. E., Lyon, G. H., Moussa, B., Sadlier, R. W.,
hadrosaurian dinosaurs of North America. Bull. Sidor, C. A., Varricchio, D. J., Wilson, G. P., and
Am. Mus. Nat. Hist. 122: 33–186. Wilson, J. A. 1999. Cretaceous sauropods from
Ostrom, J. H., and McIntosh, J. S. 1999. Marsh’s the Sahara and the uneven rate of skeletal evolu-
Dinosaurs: The Collections from Como Bluff. tion among dinosaurs. Science 286: 1342–1347.
Yale University Press, New Haven, CT, London. Stevens, K. A., and Parrish, J. M. 1999. Neck posture
Pereda-Suberbiola, X., Torcida, F., Izquierdo, L. A., and feeding habits of two Jurassic sauropod
Huerta, P., Montero, D., and Perez, G. 2003. dinosaurs. Science 284: 798–800.
First rebbachisaurid dinosaur (Sauropoda, Taquet, P. 1976. Géologie et paléontologie du gise-
Diplodocoidea) from the early Cretaceous of ment de Gadoufaoua (Aptian du Niger). Cahiers
Spain: paleobiological implications. Bull. Soc. Paleontol. 1976: 1–191.
Geol. France 174: 471–479. Taylor, A.C., and Butcher, E.C. 1951. The regulation of
Romer, A. S. 1966. Vertebrate Paleontology. eruption rate in the incisor teeth of the white rat.
University of Chicago Press, Chicago. Journal of Experimental Zoology 117: 165–188.
Salgado, L., A., Garrido, S. Cocca and J.R. Cocca. Upchurch, P., and Barrett, P.M. 2000. The evolution
2004. Lower Cretaceous rebbachisaurids from of sauropod feeding mechanisms. In: Sues, H.-D.
Cerro Aguada del León (Lohan Cura Formation). (eds.). Evolution of Herbivory in Terrestrial
Neuquén Province, northwestern Patagonia, Vertebrates: Perspectives from the Fossil Record.
Argentina. Journal of Vertebrate Paleontology 24: Cambridge University Press, Cambridge. Pp.
903–912. 79–122.
Sanders, F., Manley, K., and Carpenter, K. 2001. Weishampel, D.B. 1990. Dinosaur distributions. In:
Gastroliths from the Lower Cretaceous sauropod D.B. Weishampel, P. Dodson, and H. Osmólska
Cedarosaurus weiskpfae. In: Tanke, D. H., and (eds.). The Dinosauria. University of California
Carpenter, K. (eds.). Mesozoic Vertebrate Press, Berkeley. Pp. 63–139.
Life. Indiana University Press, Bloomington. Wilson, J. A. 2003. Sauropod dinosaur phylogeny:
Pp. 166–180. critique and cladistic analysis. Zool. J. Linn. Soc.
Sereno, P.C. 1997. The origin and evolution of 136: 217–276.
dinosaurs. Annu. Rev. Earth Planet. Sci. 25: Wilson, J. A., and Sereno, P. C. 1998. Higher-level
435–489. phylogeny of sauropod dinosaurs. J. Vertebr.
——. 2000. The fossil record, systematics and evo- Paleontol. 18 (Suppl.): 1–68.
lution of pachycephalosaurs and ceratopsians Wilson, J. A. 2005. Redescription of the Mongolian
from Asia. In: Benton, M., Kurochkin, E., sauropod Nemegtosaurus mongoliensis Nowinski
Shishkin, M., and Unwin, D. (eds.). The Age of (Dinosauria: Saurischia) and comments on Late
Dinosaurs in Russia and Mongolia. Cambridge Cretaceous sauropod diversity. Journal of
University Press, Cambridge. Pp. 480–516. Systematic Palaeontology 3: 283–318.
SIX
Digital Reconstructions of Sauropod

Dinosaurs and Implications for Feeding
Kent A. Stevens and J. Michael Parrish
n recent years, sauropods HAVE key point for analyzing variation in feeding be-
I been interpreted primarily as quadrupedal
herbivores, with sympatric taxa differentiated in
haviors across sauropods, another being varia-
tion in dentition; both are set against the back-
their feeding behavior presumably according to drop of available fodder.
their dentition and feeding height in a What was the relationship between a sauro-
quadrupedal stance (e.g., Fiorillo 1998; Up- pod’s preferred feeding height and the height at
church and Barrett 2000). In order to generate which its head was held when the neck was
detailed hypotheses concerning sauropod pale- undeflected? A bridging assumption is neces-
oecology, it is essential to start with as accurate sary to relate these two parameters for an
a reconstruction of their body plans as can be af- extinct species. The habitual feeding posture of
forded from their fossils. An accurate rendering a terrestrial herbivore can relate to the neutral
of the life posture of a sauropod is necessary in position of its neck in three ways: (1) the herbi-
order to determine the feeding envelope for vore can deflect its neck ventrally relative to the
each taxon in its conventional quadrupedal neutral position for browsing, or “browse by
stance. We review here the body plan of several ventriflexion” (BV); it can raise its neck relative
major sauropod groups, emphasizing the use of to the neutral position, or “browse by dorsiflex-
whole-body reconstructions to determine the ion” (BD); or it can feed at or near the neutral
approximate head height when the animal was position of the neck, or “browse neutrally”(BN).
standing quadrupedally, supporting its weight As browsing behavior is not directly preserved
symmetrically from left to right, and holding in the fossil record, it is necessary to consider
the axial skeleton in an undeflected state. The the phylogenetic and functional distribution of
undeflected state is termed the “neutral pose,” these three feeding models among extant
defined geometrically, and analyzed on the ba- tetrapods.
sis of osteological determinants in extant verte- The form of browsing (BV, BD, or BN) is, in
brates as a guide to their reconstruction for principle, independent of the neutral pose head
sauropods. Neutral position head height is one height (relative to shoulder height) of a given
178
herbivore. Many extant low browsers and graz- in giraffes, and that neck elongation secondar-
ers, such as Thomson’s gazelle (e.g., Leuthold ily provided the ability to drink and reach low
1977:table 2), deer, and horses, engage in BV fodder.
feeding primarily, but not exclusively. In a neu- There are many other modern examples of
tral pose, their heads are held in a high posi- herbivores with neutral head height taller than
tion, presumably for vigilance during periods of the shoulders that primarily engage in browsing
inactivity (Walther 1969). Muscular effort is by ventriflexion, including many cervids,
expended to lower the head to feed, increasing rhinoceratids, and equids (e.g., Leuthold 1972;
the tension on the epaxial nuchal ligaments. Owen-Smith 1988). There are also modern her-
One argument posed against low browsing in bivores that have the head situated much lower
sauropods, even in diplodocids, is that it would in neutral position, again relative to shoulder
leave them vulnerable to attack (Paul 2000). height, which engage in BN, such as some large
While the high head heights that gazelles bovids and other cervids such as the greater and
and other fleet-footed herbivores maintain lesser kudu (e.g., Leuthold 1972; Owen-Smith
when inactive allow them to detect and flee 1988). We are, however, unable to find examples
approaching predators, flight from predation of extant herbivores with heads well below shoul-
was not a practical option for sauropods, the der height in neutral position that predomi-
speeds of which are generally estimated as nantly feed quadrupedally in a BD position.
being much slower than those of their most If vigilance were not a factor for sauropods,
likely predators (e.g., Alexander 1989; Thul- then it might be inferred that the neutral pose
born 1990). Today, large BN to BV megaherbi- would be related closely to their preferred feed-
vores respond to the presence of carnivores by ing heights. Indeed, as we show, the neutral
charging (e.g., rhinos, hippos, elephants pose for some sauropods places the head very
[Owen-Smith 1988]) or indifference (e.g., ele- low to the ground. A sauropod with shorter
phants, hippos [Owen-Smith 1988]). forelimbs than hindlimbs and a steadily
The giraffe is of particular importance to descending neck that brings the head to near-
this chapter, as it has been cited as an extant ground level certainly appears well suited for
model for those sauropods, including bra- BN (and BV) browsing, without compromise to
chiosaurids, euhelopids, and camarasaurids, vigilance, and less adapted to BD. Large extant
that are sometimes reconstructed as giraffe-like grazing herbivores, such as bison, that hold
(e.g., Paul 1987:figs. 16, 17; Currie 1987:figs. 2, 3; their head closer to feeding height (e.g.,
Christian and Heinrich 1998), effortlessly feed- Leuthold 1977:table 2) are perhaps better ana-
ing while in a cervical neutral position (BN logues for those sauropods with low head
browsing) with the head held high above the height in neutral position.
shoulders. We review the osteological basis for Osteologically based reconstruction of neu-
the elevation of the giraffe neck, then examine tral head height provides some refinement on
the validity of proposing such a posture for any the question of sauropod feeding behavior,
sauropod. It should first be noted that, perhaps when combined with consideration of whether
surprisingly, giraffes frequently browse by ven- a given taxon browsed by ventriflexion, by dor-
triflexion, with the head at or below shoulder siflexion, or near the cervical neutral position.
height (Leuthold and Leuthold 1972; Pellew We suggest that careful analysis of the osteo-
1984; Young and Isbell 1991; Woolnough and logically defined neutral pose along the axial
du Toit 2001). The elongate neck of the giraffe skeleton is an indicator of the mean feeding
is not a simple consequence of vertical niche height of the sauropods, particularly as they do
partitioning (Simmons and Scheepers 1996). not, in any of the sauropod taxa we have exam-
Pincher (1949) proposes that predation pro- ined, place the necks in a giraffelike high slope
vided the selection pressure for limb elongation with heads held high, a position consistent with
D I G I TA L R E C O N S T R U C T I O N S O F S AU R O P O D S , I M P L I C AT I O N S F O R F E E D I N G 179
a predominant state of vigilance. As reviewed alized as the extremes of head reach allowed
below, neutral pose reconstructions suggest by the flexibility of its neck (e.g., Martin
that most sauropods would have their necks 1987:fig. 3; Stevens and Parrish 1999). Note
held horizontally or subhorizontally when not that the envelope, thus defined, reflects only
actively feeding or otherwise raising their variation in head position allowed by neck
heads. mobility, not the contributions of trunk and
forelimb movements that undoubtedly broad-
ened this envelope.
THE NEUTRAL POSE AS A BASIS FOR
ESTIMATING FEEDING HEIGHT THE NEUTRAL POSE OF THE APPENDICULAR
In describing the sauropod body plan, it is use- SKELETON
ful to start by establishing the height of the The nature of the articulations between the ele-
acetabular axis above ground level. Fortunately, ments of the sauropod appendicular skeleton
maximum hindlimb length can often be recon- cannot be inferred entirely from osteology,
structed with some confidence, with the pri- although the hindlimb, because of the solid
mary unknowns being the precise amount of joints between sacrum and pelvis, is better con-
cartilage separating the limb elements and the strained than the forelimb and pectoral girdles,
degree of flexure at the knee (e.g., Paul 1987; which have no osteological connection to the
Bonnan 2001). With the femoral heads inserted axial column.
into their associated acetabula, the acetabular Whereas hindlimb length determines the
axis constitutes a pivot point for the whole-body height of the caudal end of the dorsal vertebral
reconstruction—a fulcrum about which the column, forelimb length only defines the
axial skeleton tilts depending on a given recon- height of the glenoid; the inferred height and
struction of the trunk. The literature presents a slope of the anterior part of the trunk depend
range of interpretations regarding the arch to on the position and orientation of the scapulae
the span of vertebral column that supported the on the ribcage and the degree to which the dor-
trunk. The degree of flexure of the arch is sal vertebral column is arched.
important here, for the greater the curvature, In extant quadrupeds, pectoral girdles do
the lower the resulting head height. A given not have bony articulations with the trunk but,
reconstruction of the dorsal vertebrae, with instead, are suspended from the thorax by mus-
associated ribcage, forms an armature on culature and soft tissues. While the presence of
which to place the pectoral girdles and fore- subtle depressions in the dorsal ribs in some
limbs. In creating a skeletal reconstruction to sauropods such as Apatosaurus and Diplodocus
quantitatively estimate mean feeding height, it provide some indication of the position and the
is important to determine potential sources of alignment of the scapulocoracoid (Parrish and
variability in the appendicular and axial skele- Stevens 2002b), determining its precise loca-
ton. These are reviewed below. tion requires considering the scapulocoracoids
For modern quadrupedal herbivores, the within the context of the overall function of the
characteristic posture associated with a stand- girdles, ribcage, and forelimbs.
ing pose is the starting point for exploring the Changing the relative height and inclination
range of feeding movements achieved by of the scapulocoracoid on the trunk can result in
movements of the head and neck. A skeletal significant differences in the height of the cranial
reconstruction of an extinct form, such as a end of the thorax, with a corresponding change in
sauropod, must similarly be posed in a neutral head height (Stevens and Parrish 2005a). Early
position as a basis for examining its character- sauropod mounts differed in the orientation of
istic feeding envelope. As considered here, the the pectoral girdles relative to the ground, and
feeding envelope for a given taxon can be visu- thus, indirectly, in the angle of the scapula
180 D I G I TA L R E C O N S T R U C T I O N S O F S AU R O P O D S , I M P L I C AT I O N S F O R F E E D I N G
relative to the cranial part of the dorsal vertebral and right pairs of zygapophyses. Dorsoventral
column. For instance, Gilmore (1936:pl. 34) and mediolateral angular deflection results in
placed the scapulae of Apatosaurus in a subhori- gliding contact between the zygapophyses that
zontal orientation, nearly parallel to the anterior limit axial rotation (Stevens and Wills 2001).
dorsal column, whereas Osborn and Mook (1921) The postzygapophyses are displaced posteriorly
placed the scapulae of Camarasaurus in a far relative to their associated prezygapophyses
more vertical orientation. The effect of shoulder during dorsiflexion and anteriorly during ven-
girdle placement on the overall reconstruction triflexion. The zygapophyseal facets are super-
can be appreciated by comparing the illustrations imposed and centered in a state of neutral
of the Apatosaurus in Gilmore (1936:pl. 34) versus deflection.
McIntosh et al. (1997) or those of the Diplodocus The state of null deflection is also apparent
mounted at the Carnegie Museum of Natural at the central synovial capsule. The condyle,
History (e.g., McGinnis 1982:68–69) versus that inserted deeply within cotyle, is surrounded by
at the Senckenberg Museum in Frankfurt (e.g., a broad circumferential ligamentous capsule.
Beasley 1907). The effect of variation in shoulder Within the capsule, the articular facets are
girdle placement is also apparent below in digital closely spaced in modern vertebrates with
reconstructions of Brachiosaurus brancai, in com- opisthocoelous centra. In a rhino, for instance,
bination with variation in dorsal vertebral column the separation between cotyle and condyle is
curvature. only a few millimeters, and in the giraffe,
A range of interpretations has also been pro- whose cervical vertebrae are also strongly
posed for the articulation of the elements of the opisthocoelous, the intervertebral gaps are only
sauropod forelimbs, ranging from a vertical, slightly wider, again, of the order of millimeters
columnar arrangement (e.g., Christian et al. (B. Curtice, pers. comm. 2001, pers. obs.). Note
1999) to partial flexure of the limbs (e.g., that this close intervertebral separation may not
Janensch 1950b), resulting in a range of heights be apparent from cursory observation. The cap-
and locations for the humeral head. Most recent sule is substantially wider than the interverte-
investigators (e.g., Bonnan 2001; Christian et al. bral separation, of course, to accommodate the
1999; Wilhite 2003) predict minimal flexion of displacement undertaken by the cotyle during
the elbow during standing, in keeping with the deflection. Articulated sauropod cervical verte-
columnar forelimb posture of extant graviportal bral series are likewise very closely spaced, in all
animals, but some reconstructions (e.g., the instances we have examined.
Humboldt Museum mount of Brachiosaurus In modern vertebrates, dorsiflexion separa-
[Christian et al. 1999:fig. 1]) depict a more reptil- tion between the cotylar and the condylar mar-
ian sprawling pose (e.g., Christian et al. 1999). gins of the capsule increases ventrally (and
diminishes dorsally), placing the surrounding
THE NEUTRAL POSE IN THE AXIAL SKELETON ligaments in progressively greater tension ven-
The neutral pose of the axial skeleton is trally. Conversely, in ventriflexion, the separation
strongly constrained. The presacral vertebral and ligament tension increases dorsally (and
column in sauropods is characterized by diminishes ventrally). In a state of neutral deflec-
opisthocoelous central articulations, a character tion the margins of the associated cotyle and
shared with many large herbivorous modern condyle are parallel; the gap is uniform around
quadrupeds such as rhino, giraffe, horse, and the perimeter of the synovial capsule, thus pro-
camel. Opisthocoely, fortunately, provides par- viding a second osteological indictor of neutral
ticularly strong osteological clues to the state of deflection. When undeflected, the posterior edge
neutral position. of the cotyle is parallel to the attachment scar
Presacral intervertebral articulation involves surrounding the condyle, corresponding to the
synovial joints between centra and between left state of minimal stress on the synovial capsule.
FIGURE 6.1. The cervical vertebral columns of articulated skeletons of (A) crocodilian (Gavialis gangeticus), (B) turkey (Melea-
gris gallopavo), (C) horse (Equus przevalskii), and (D) camel (Camelus dromedarius), each mounted in the undeflected, or “neu-
tral,” pose wherein apposed pre- and postzygapophyses are aligned and centered, and simultaneously, the margins of cotyle
and condyle at each intervertebral joint are parallel. In each case, the neck in this undeflected state is naturally curved in the
manner characteristic of that animal.
When the central articulation is undeflected When the vertebrae of extant mammals are
according to the above criterion, the associated placed in neutral pose, they replicate their
pre- and postzygapophyses are also undeflected habitual, characteristic posture (figs. 6.1, 6.2).
(i.e., superimposed and centered). That is, the For instance, in neutral pose, the neck of the
two criteria are satisfied simultaneously (fig. 6.1) camel exhibits its familiar catenary shape, and
in all extant vertebrates that we have observed. the sharp change in angulation observed at
Their redundancy is particularly useful in the base of the neck of the giraffe is clearly
reconstructing the neutral pose for vertebrae visible (fig. 6.2). Likewise, the sigmoid curves
that are missing their zygapophyses (see the in the necks of theropod dinosaurs, including
Brachiosaurus brancai reconstruction below). extant birds (fig. 6.1A), are associated with an
When successive vertebrae are placed in neutral undeflected neck and derive from the geome-
position, joint by joint, the vertebral column try of the vertebrae. Therefore, the neutral
forms the intrinsic curvature characteristic of pose of the cervical vertebral column, and the
the given extant animal. cranial end of the dorsal vertebral column,
FIGURE 6.2. The steeply as-
cending neck characteristic of
the giraffe arises in the unde-
flected state, not by bending of
the neck. Photographs of verte-
brae C4–T1 of an adult giraffe
are placed in neutral position.
Transparency is utilized to show
the insertion of each central
condyle within the cotyle and
the overlap and alignment of
the zygapophyses. The charac-
teristic neck elevation derives
from wedge- or keystone-shaped
centra, especially apparent in
the sixth and seventh cervical
vertebrae. (Individual
photographs courtesy of Brian
Curtice.)
can be shown to be consistent with the pre- RECONSTRUCTING SAUROPOD

ferred head height in a variety of extant CERVICAL COLUMNS
quadrupedal herbivores, and thus we consider
it a robust predictor of preferred head position METHOD
in sauropods as well. The neutral pose of a vertebral column can be
The cervical and anterior dorsal vertebrae of reconstructed by creating a composite of lateral
sauropods are strongly opisthocoelous, with view illustrations or photographs, all depicted at
prominent condyles associated with deep the same scale, placed such that each successive
cotyles (figs. 6.3, 6.4). Articulation of axial ele- pair of vertebrae is in neutral deflection. This
ments in sauropods is geometrically similar to method has long been used in traditional axial
that in many extant vertebrates, allowing infer- skeleton reconstructions (e.g., Osborn and
ence of their neutral or undeflected state to be Mook 1921:fig. 28). To introduce the technique,
made with some confidence. consider a pair of platycoelous centra. The
Given the commonality in morphology of apposed central facets are parallel to one
the articular facets of sauropod cervical verte- another when the intervertebral joint is unde-
brae and those of a wide range of extant verte- flected, hence their margins also appear parallel
brates, these criteria are presumed to hold for when viewed laterally. This neutral alignment
the cervical vertebrae of sauropods as well, per- is readily reconstructed graphically by two-
mitting the reconstruction of their undeflected dimensional rotation, translation of images of the
state and, hence, their intrinsic curvature. two vertebrae (e.g., using layers in Photoshop)
FIGURE 6.3. Intervertebral articulation in the neutral pose for the sauropod Brachiosaurus brancai (Top, cervical vertebrae
C4 and C5 of the Humboldt Museum specimen SI) and for a giraffe (Bottom, cervical vertebra 7 and first thoracic vertebra),
shown at the same scale. Note the similarity in articulation geometry, both being strongly opisthocoelous with condyles of
circular profile making the center of rotation for dorsoventral flexion unambiguous. The relatively larger zygapophyses of
the giraffe are closer to the center of rotation and thereby permit a greater angular range of motion. Scale bar equals 10 cm.
(Brachiosaurus and giraffe photographs courtesy of Christopher McGowan and Brian Curtice, respectively.)
FIGURE 6.4. Cervical vertebrae C4 and C5 of Brachiosaurus brancai specimen SI (Janensch 1950a: figs. 34, 37). In (A) the
left postzygapophysis of C4 is shaded. In (B) the two vertebrae are composited in neutral position such that the pre- and
postzygapophyses are centered and the margins of the cotyle and condyle at the central articulation are parallel.
until the gap between centra is uniform and the plan, with the morphology of individual ele-
zygapophyses are centered (i.e., both criteria of ments ranging from detailed to schematic. A
neutral position are met). When an entire verte- hybrid approach is also used for the recon-
bral series is composed from individual illustra- struction of the cervical vertebral series: two-
tions by this means, an intrinsic curve is often dimensional reconstructions based on original
revealed, not due to flexion, but from centra that artwork provide an estimate of the intrinsic
are sometimes subtly keystone- or wedge-shaped. curvature of the undeflected cervical vertebral
In sauropods, the articulations of the centra column, which is then used to pose the neck of
throughout the cervical vertebral column are the digital three-dimensional model (Stevens
strongly opisthocoelous. Fortunately, since the and Parrish 2005a).
curvature of the sauropod neck is critical to
RECONSTRUCTING NECK CURVATURE
understanding their paleobiology, opisthocoely
helps, rather than hinders, the reconstruction of DIPLODOCIDS
the neutral state of deflection. With condyle Apatosaurus and Diplodocus were initially
inserted into cotyle and placed in an undeflected reconstructed with necks that were quite
state, the postzygapophyses are centered above straight and extended from an arched back so
prezygapophyses, and the posterior margin of that the necks descended gently from the
the cotyle is parallel to the attachment scar of the shoulders (e.g., Holland 1906:fig. 2). Usually
synovial capsule surrounding the central the necks are depicted in a state of mild dorsi-
condyle. In figure 6.4, line drawings of two cer- flexion at the base, which raises the heads to
vical vertebrae of Brachiosaurus brancai have about shoulder height. Recent renditions of
been composited into neutral position accord- these two diplodocids often provide their necks
ing to these criteria. Note that transparency with a more pronounced sigmoid curve, dorsi-
reveals the insertion of the condyle within cotyle flexed caudally and ventriflexed cranially (e.g.,
and the centering of the zygapophyses. McIntosh et al. 1997:figs. 20.11, 20.12; Wilson
A second method is conceptually similar but and Sereno 1998:foldout 1). Similarly, in the
performed three-dimensionally. Neutral pose original skeletal reconstruction of the
can be determined by direct manipulation, as diplodocid Dicraeosaurus the neck was shown
part of a process of exploring the range of in a pronounced sigmoid curve, abruptly dorsi-
motion in the axial skeleton. Although a pair of flexed at the base and more gradually deflected
vertebrae, or even a complete neck, of an extant downward cranially (Janensch 1929:pl. 16).
animal can be articulated and manipulated, Unfortunately, it is difficult to determine from
sauropod vertebrae are not only unwieldy, but drawings to what extent the curvature was
usually too distorted to allow proper rearticula- intended to depict active flexion versus a shape
tion and manipulation. An alternative is to cre- intrinsic to the neck.
ate digital representations of the vertebrae, The vertebrae in the original descriptions of
which can then be articulated and posed virtu- the above diplodocids were rendered with excel-
ally in three dimensions. A parametric skeletal lent dimensional accuracy as detailed line draw-
modeling approach, DinoMorph (Stevens ings, and the original material was, in most
2002), has been developed and used to esti- cases, sufficiently undistorted to permit recon-
mate the feeding envelopes of two diplodocids structions of their neutral position by graphical
(e.g., Stevens and Parrish 1999). For the pres- compositing (Gilmore 1936:pl. 24, 25; Hatcher
ent study, dimensionally- accurate digital skele- 1901:pl. 4, 6; Janensch 1929:pl. 1). These
tons of three sauropods are posed in neutral images were digitally scanned, then composited
position to estimate mean feeding heights. In (fig. 6.5) to create reconstructions of the axial
each case published dimensional data provide columns in neutral pose. The composites reveal
the basis for reconstructing the overall body that these diplodocids had remarkably straight
FIGURE 6.5. The presacral vertebrae of the diplodocids (A) Apatosaurus louisae, (B) Diplodocus carnegii, and (C) Dicraeosaurus
hansemanni, composited from the original figures, placing each successive pair in neutral deflection (from Stevens and
Parrish 2005b). Apatosaurus composite mirrored left-for-right for uniformity; all vertebral columns at the same scale. Scale
bar equals 1 m.
cervicodorsal transitions; their necks were of its tall limbs and ascending dorsal column
straight extensions of their backs. The cervical but, especially, as a result of the neck that rises
columns of Apatosaurus and Dicraeosaurus steeply at its base. The osteological basis for
show a gentle degree of ventral curvature in the such a posture is reviewed here. This giraffe-like
neutral pose, which brings the head to a posi- posture is even more dramatic in some recon-
tion well adapted to low browsing. The familiar structions of Euhelopus and Camarasaurus,
sigmoid curve attributed to the neck of which are also considered.
Dicraeosaurus (Janensch 1929:pl. 16) was cer- Most reconstructions of Brachiosaurus pro-
tainly within the limits achievable by dorsiflex- vide the sauropod with a steeply upturned neck
ion at the base and ventriflexion more cranially, and depict the cervicodorsal vertebrae as
but this pose was not reflected in the osteology wedge-shaped, with centra longer ventrally
of the neck. But presuming that the osteologi- than dorsally, much as those at the base of the
cally determined neutral pose was also the neck in the giraffe. Figure 6.6 shows in detail
habitual posture for Dicraeosaurus, it usually the cervicodorsal region as originally recon-
held its head close to ground level, in common structed, plus three subsequent depictions, all
with other diplodocids, but at a steeper angle of which can be compared with a detailed fig-
due to its relatively shorter neck. ure of the original fossil material in this critical
region.
BRACHIOSAURUS AND OTHER SAUROPODS OFTEN Accurate line drawings of the available ver-
DEPICTED AS GIRAFFE-LIKE tebrae of Brachiosaurus brancai were provided
The original reconstruction of Brachiosaurus by Janensch (1950a:figs. 14–49). The lateral
(Janensch 1950b:pl. 6–8) shows remarkable views are amenable to composition into a
similarity to the modern giraffe, in part because reconstructed column, for both specimen SI
FIGURE 6.6. Details from four depictions of Brachiosaurus brancai: (A) Czerkas and Czerkas (1991:132); (B) Wilson and
Sereno (1998:foldout 1); (C) Janensch (1950b:pl. 8); (D) McIntosh et al. (1997 fig. 20.16). Note the differing degrees to which
the cervicodorsal region is depicted as upcurved between the posterior of D2 and the posterior of C10. This region, in (A) is
curved merely 5, while the same region is curved 68 in (D) largely due to illustrating the centra as if they were distinctly
wedge- or keystone-shaped (longer ventrally than dorsally). In (A) approximately 2 of the curvature is due to keystoning, the
rest presumably resulting from dorsiflexion. In (D) approximately 48 is accumulated due to the shape of the centra, espe-
cially C12 and C13, and approximately 20 reflects dorsiflexion, which further contributes to the near-vertical posture favored
by that illustrator. Compare with original illustration (E; from Janensch 1950a:fig. 49) of cervicals C10–C13 plus first two
dorsals, where no wedge shape is apparent.
FIGURE 6.7. Neutral pose reconstructions of Brachiosaurus brancai specimens SII (Top, C3–D2) and SI (Bottom, C2–C7)
from individual line drawings (Janensch 1950a:figs. 14–49). SI is mirrored left-for-right to facilitate comparison with SII
(from Stevens and Parrish in 2005b). The slight ventral curvature in neutral position appears intrinsic to Brachiosaurus and
is consistent with head-down feeding. Note that vertebrae C10–D2 of SII were collinear in situ, and while the neural arches
were ablated, the centra show no keystoning or other osteological evidence to support a giraffe-neck interpretation (see also
fig. 6.6). In fact, the partially exposed ventral surface of the central condyle of D2 is evidence of some dorsiflexion between
D1 and D2. Scale bar equals 1 m.
(vertebrae C2 through C7) and specimen SII not change its undisputed role as a high
(C3 through D2). Cranially, both exhibit the browser; see figure 6.8.
gradual ventral curvature in Dicraeosaurus Euhelopus zdanskyi (Wiman 1929:fig. 3) is
(fig. 6.5). In specimen SII, although the neu- another sauropod traditionally depicted as
tral arches were not preserved caudal to C9, giraffelike, with the life pose originally drawn
the central articulations provide clear evidence having the same upturned neck as found in situ
for the neutral deflection between subsequent (fig. 6.9). The dorsiflexion at the base of the
pairs of vertebrae from the midneck to the sec- neck (primarily between C16 and C17 and
ond dorsal (Fig. 6.7). The result is a remark- between C17 and D1), however, reflects a “death
ably straight neck at the base, quite contrary to pose” resulting from shrinkage of the nuchal
most restorations of this taxon (but see ligaments. This conclusion is drawn from
Czerkas and Czerkas 1991:132). Properly observing that the degree of dorsiflexion as
restoring the neck of Brachiosaurus as extend- measured between successive centra equals the
ing straight from the shoulders, however, does angle of dorsiflexion at the zygapophyses
FIGURE 6.8. A DinoMorph model of Brachiosaurus brancai shows the effect on head height of pectoral girdle placement
and dorsal column curvature. The neck curvature is held constant in the neutral pose derived from digital composites of the
original steel engravings (fig. 6.7). The postcervical skeleton is based on specimen SII (Janensch 1950a, 1950b, 1961). Four
alternative head heights result from combinations of high versus low arch to the dorsal column and high versus low place-
ment of the pectoral girdles on the ribcages. The high arch and low pectoral girdle placement case is shown in full contrast,
while the others are depicted at lower contrast. The highest head height is associated with the combination of low arch and
low girdles. The lowest head height derives from the combination of high arch and high girdles, which brings the head only
to shoulder height. Vertical scale bar equals 5 m.
FIGURE 6.9. The original life reconstruction of Euhelopus zdanskyi (A) was depicted with a giraffe like neck, ascending with
about 38° of slope (from Wiman 1929: fig. 3). The sharp curvature at the base of the neck replicated that of the original spec-
imen as found in situ (B; from Wiman 1929:pl. 3). In (C) (from Paul 2000: appendix A), the slope is increased to about 65.
With removal of the “death pose” dorsiflexion that is localized to the base of the neck (D; from Stevens and Parrish 2005b),
the neck of Euhelopus emerged from the shoulder as a straight extension of the dorsals, sloping more or less downward de-
pending on the arch of the dorsal column and the relative height of the (unknown) forelimb. The resultant low head height
was similar to that of the diplodocoids and consistent with low browsing.
(Stevens and Parrish 2005b). Removing this ical soft tissue constraints that might be vio-
degree of dorsiflexion restores the neutral pose lated taphonomically. In estimating the range
and Euhelopus is revealed to be a low browser, of deflection that might have been achieved
not at all giraffelike. about the osteologically defined neutral posi-
Similarly, the juvenile Camarasaurus at the tion, we turn to extant vertebrates.
Carnegie Museum (CMNH11338 [Gilmore
1925:fig. 14]) also exhibits the extreme curva- ESTIMATING NECK FLEXIBILITY
ture of a death pose. The neck was preserved IN SAUROPODS
with the swanlike curvature that is now popu- In seeking modern analogues on which to base
larly portrayed in life reconstructions of this estimates of sauropod neck flexibility, potential
taxon. Examination of the original specimen candidates include long-necked vertebrates
reveals that many of the postzygapophyses are such as the ratite, giraffe, and camel. All have
completely disarticulated from their associated considerable mediolateral flexibility. A giraffe,
prezygapophyses (Parrish and Stevens 1998). for instance, can reach insects biting the base of
Even if this condition could have been toler- its own neck. Some mammals achieve greater
ated in life, such an extreme of dorsiflexion lateral angular deflection per vertebral joint
certainly did not constitute the neutral pose than avians, having only 7, compared to 11 to 25,
for the neck. Furthermore, quantitatively, cervical vertebrae over which to distribute the
unlike in Euhelopus, the angular deflection at curvature. Dorsoventrally, however, these long
the zygapophyses was generally greater than necks vary considerably in flexibility, with the
that at the centra, suggesting that the neck in giraffe showing little flexibility except at the cer-
neutral pose was at least partially ventriflexed, vical–dorsal transition, whereas the camel can
in common with our observation of other touch the back of its head to its shoulder
sauropod taxa. In summary, we have yet to (Gauthier-Pilters and Dagg 1981:fig. 29).
find any sauropod with evidence of osteologi- With such a wide range of observed flexibil-
cal adaptations for an upraised neck in the ity, there might appear to be little ground for
undeflected, neutral pose. Dorsiflexion could postulating any particular range of neck flexi-
undoubtedly greatly increase the head height bility in the sauropods. Specifically, it has been
for feeding many meters above the ground, suggested (Sereno et al. 1999) that estimating
but the popular rendition of sauropods such as sauropod neck mobility based on the observed
Brachiosaurus, Euhelopus, and Camarasaurus intervertebral limitations of avian cervical verte-
as giraffe-like is unwarranted. brae (Stevens and Parrish 1999; fig. 6.10)
Finally, we note that while the neutral pose might be too conservative in light of the camel’s
represents the approximate center of a tetra- remarkable flexibility. In order to test the valid-
pod’s potential feeding envelope, the predomi- ity of the assumptions made regarding the neu-
nant feeding position is not necessarily the tral pose and estimates of the range of mobility
neutral position. As noted, many open habitat in sauropods, we performed several compar-
ungulates assume close to the neutral position isons between the observed ranges of move-
while inactive yet frequently engage in BV ment in extant long-necked mammals and the
feeding, often far from the neutral position. amount of mobility produced by manually
There are well-defined limits, fortunately, to manipulating skeletonized and disarticulated
how far the neck may deviate from the neutral specimens of the cervical columns of individu-
position. Based on behavioral observations als of the same species (Parrish and Stevens
and manipulation of living and preserved 2002a). Independent studies involving the
necks of extant birds and mammals, we are manipulation of extant ostrich are also consis-
confident that the feeding positions of tent with our estimates regarding the angular
sauropods in life would be limited by mechan- constraints imposed by zygapophyseal synovial
FIGURE 6.10. DinoMorph model of Apatosaurus louisae showing the extremes of dorsoventral motion and lateral flexibility
estimated by Stevens and Parrish (1999). The three-dimensional skeletal model is based on Gilmore (1936) and data from
Philip R. Platt (pers. comm. 2003). The neck of Apatosaurus was found to be capable of a substantial range of motion (see
also fig. 6.15), with more lateral and dorsal flexibility than the longer-necked Diplodocus, but it shared with Diplodocus the
theoretical ability to lower its head below mean ground level, a potential adaptation for feeding on subaqueous plants from
shore.
capsules (Wedel and Sanders 1999). In avians, dorsiflexion is combined with lateral deflection;
these synovial capsules limit displacement for example, adding left lateral deflection
between associated pre- and postzygapophyses causes the left postzygapophysis to eventually
such that they remain overlapping, with a safety make contact, and at this limit both the left and
margin, throughout the range of motion. The the right zygapophyseal pairs preserve a safe
remarkable dorsal flexibility of the camel neck degree of overlap. The osteological stop is pres-
is likewise achieved without zygapophyseal dis- ent along most of the neck of the domestic
articulation (fig. 6.11). The elongate zygapophy- turkey and the rhino, but in the horse, for
seal facets provide sufficient travel to permit example, it is apparent only at the base of the
this extraordinary flexibility. neck. The extreme dorsal flexibility of the camel
Neck flexibility in some extant vertebrates is is likewise limited by bony contact that prevents
delimited by osteological stops that prevent the neck from disarticulation (fig. 6.12), despite
excessive displacement. In the domestic turkey, the external appearance of disarticulation in the
for instance, dorsiflexion causes the postzy- behaving animal.
gapophyses to translate caudally until they con- Although a camel’s death pose is achievable
tact the ascending base of the dorsal spine of in life (Gauthier-Pilters and Dagg 1981:fig. 60
the next vertebra, often nesting into a matching vs. 33), it bears reminding that the camel neck
depression just posterior to the associated permits such flexibility by design and it is oste-
prezygapophyses medial to the paradiopophy- ologically prevented from disarticulation in dor-
seal lamina. The zygapophyseal pair is still siflexion. The flexion of the camel neck, while
articulated, that is, overlapping, at this limit. considerable at its limits, is delimited by the
This mechanism operates unilaterally when same biomechanical means as in avians and, by
FIGURE 6.11. The remarkable dorsal flexibility of the camel’s neck (inset; after Gauthier-Pilters and
Daag 1981:fig. 29) might suggest that sauropod necks were also more flexible than might be pre-
dicted based on birds, wherein the zygapophyseal synovial capsules limit the travel of the enclosed ar-
ticular facets prior to disarticulation (cf. Stevens and Parrish 1999; Sereno et al. 1999). The camel’s
flexibility, however, can be replicated without disarticulation between associated pre- and postzy-
gapophyses. Dorsiflexion is in fact limited osteologically prior to disarticulation (see fig. 6.12). The
large angular deflection achieved per joint in the camel’s neck is a geometric consequence of its elon-
gate zygapophyses set close to the center of dorsoventral rotation.
analogy, in sauropods. However, the osteologi- taut at a limit of displacement between

cal features that limit dorsiflexion in many zygapophyses. In avian cervical vertebrae, at the
highly flexible extant taxa are not apparent, as limits of travel the zygapophyses maintain
far as we have observed, in sauropods. roughly 25%–50% overlap (Stevens and Parrish
The giraffe’s restricted dorsoventral motion is 1999; Wedel and Sanders 1999).
also limited osteologically, both dorsally, as just The maximum displacement between
described, and ventrally (B. Curtice, pers. comm. zygapophyses permitted by the surrounding
2001; Parrish and Stevens, pers. obs.). It should capsules translates to a maximum angular
be noted that the extreme lateral flexibility of the deflection. The cervical flexibility of the giraffe,
giraffe is “at the edge,” with minimal overlap at the camel, and various avians has been found to
the zygapophyses, and not protected by the oste- be predictably related to the geometry of the
ological stops present in avian vertebrae. As in zygapophyseal facets and their placement rela-
many vertebrates, the range of motion is further tive to the axes of rotation (figs. 6.11, 6.13).
delimited by the ligamentous synovial capsules Given the morphological similarities between
surrounding the zygapophyses, which become avian and sauropod cervical vertebrae, Stevens
FIGURE 6.12. Dorsiflexion at the base of the neck in the camel is arrested osteologically prior to disarticulation at the
zygapophyses. Postzygapophyses lock into depressions just posterior to the associated prezygapophyses at the limit of
dorsiflexion.
and Parrish (1999) estimated the flexibility of sal vertebrae of most sauropods is as yet
two diplodocids, using dimensionally accurate unknown, and the placement of the pectoral
models for the articular facets and centra and girdles is still a matter of some uncertainty as
assuming that the zygapophyses were limited well. Nonetheless, the maximum effect of these
to a maximum displacement (or minimum tol- factors on mean feeding height can be esti-
erated overlap) based on that observed in avians mated for various taxa that have been modeled
(figs. 6.13, 6.14). The result was less overall three-dimensionally.
range of motion than commonly expected, pri- The variation in head height resulting from
marily in those regions of the neck where the differing expectations for pelvic girdle place-
relatively small zygapophyses would have ment can be envisioned by similar triangles.
allowed little excursion. Holding other factors constant, the head can be
thought of as cantilevered at the end of the pre-
sacral column ahead, pivoted about the acetab-
ESTIMATING FEEDING HEIGHTS
ular axis, and supported by the glenoid. The
In neutral pose, the neck of every sauropod we variation in head height is proportional to the
have studied thus far has negligible curvature at ratio of rostrum–acetabulum distance to
the base and, at its cranial extent, a tendency to glenoid–acetabulum distance, typically about
droop or curve ventrally. Determining the mean three times the variation at the shoulder. While
head height requires understanding the height there is significant uncertainty in the reconstruc-
and slope of the base of the neck, which in turn tion of the pectoral girdles, this effect is com-
is determined by the reconstruction of the pounded by the degree of the arch to the dorsal
trunk. The precise degree of arching of the dor- column (fig. 6.8). Clearly these differences do not
FIGURE 6.13. Giraffe dorsoventral range of motion for cervical vertebra C7 articulating on T1. The center of rotation, as in-
dicated, was estimated by a circular fit to the profile of the condyle of T1. In (A) dorsiflexion of approximately 9 (relative to
indicated neutral position slope) is limited osteologically when the postzygapophyses of C7 contact T1. In (B) ventriflexion
of roughly 30 is permitted while preserving substantial overlap (about 50%) between zygapophyseal facets.
FIGURE 6.14. Diplodocus

carnegii cervical vertebrae C13
and C14 (after Hatcher 1901: pl.
4) showing estimated limits of
the dorsoventral range of mo-
tion. In (A) the right postzy-
gapophysis of C13 is shaded,
and the midline of prezygapoph-
ysis of C14 is indicated along
with the center of rotation for
dorsoventral flexion at the cen-
ter of curvature of the condyle of
C14. In (B) the two vertebrae are
placed in articulation and the
dorsiflexion limit of 6 is illus-
trated, defined as the angular
deflection that reduces the zy-
gapophyseal overlap to 50%
(note that the anterior margin of
the postzygapophysis of C13 is
displaced anteriorly to the indi-
cated midline). In (C) ventriflex-
ion is limited to about 8 (at
which point the posterior mar-
gin of the postzygapophysis is
displaced to the prezygapoph-
ysis midline).
FIGURE 6.15. DinoMorph models of Brachiosaurus in neutral pose and Apatosaurus and Diplodocus maximally dorsiflexed,
showing that the vertical feeding range of these three sympatric sauropods overlapped significantly: Apatosaurus could
reach the neutral pose head height of Brachiosaurus, and Diplodocus, while less dorsally flexible, was capable of 4-m
elevation. Vertical scale bar represents 5 m.
categorically change the expectation for browse Sauropod feeding is of interest for several rea-
height, but they do have some consequences sons. First, as has been widely noted (e.g.,
regarding conventional views on niche partition- Stevens and Parrish 1999), sympatry among
ing among sauropods. Based on DinoMorph sauropod genera is widespread, particularly in
simulations, Diplodocus was capable of reaching the Late Jurassic (although the recent study by
vegetation at least 4 m high, and Apatosaurus had Curry Rogers and Forster [2001] indicates that
sufficient dorsal flexibility to reach what we esti- two or more sympatric species of sauropods
mate to be the mean head height of Brachiosaurus may have occurred in Madagascar as late as the
(fig. 6.15). On the other hand, Brachiosaurus Late Cretaceous). Second, because sauropods
could readily reach down to ground level (without are among the largest terrestrial herbivores that
the need to splay its legs giraffe-style), provided it ever lived, understanding their bioenergetics is
had the modest ability to flex about 8 ventrally in useful both in considering the scaling of metab-
the proximal cervical vertebrae. Thus these two olism in vertebrates and in examining the like-
clades shared 4–6 m of vertical feeding range, lihood that sauropods, and other dinosaurs,
which also appears to have overlapped those of were endotherms, were ectotherms, or had a
camarasaurids (fig. 6.15). unique metabolic physiology.
It is clear from the cranial and dental studies
that significant differences existed among sauro-
FEEDING STRATEGIES AND
pod feeding mechanisms. Brachiosaurids and
PALEOECOLOGY
camarasaurids had broad, spatulate teeth that
In the last decade, many studies have addressed appear to have been optimal for biting off resist-
sauropod feeding, mostly from the standpoint ant vegetation, whereas the teeth of diplodocids
of tooth morphology and microwear (Fiorillo and titanosaurs were more peglike in shape and
1998; Christiansen 2000; Upchurch and probably functioned best in cropping or, per-
Barrett 2000), jaw muscle reconstruction and haps, in stripping leaves from branches (Barrett
cranial mechanics (Calvo 1994; Barrett and and Upchurch 1994). The recent feeding studies
Upchurch 1994; Upchurch and Barrett 2000), are unanimous in their assertion that no signifi-
or bioenergetics (Farlow 1987; Dodson 1990). cant processing of food occurred in the mouths
of any known sauropods, although the spatulate- estimating the additional height that might
toothed forms probably sliced their food when have been provided by a sauropod rearing up on
grasping it with their mouths rather than just its hindlimbs. The maximum feeding height,
pulling it off the branches. however, would be far harder to estimate,
Another topic of debate is the range of because the acetabulum height would no
heights and, in particular, the maximum height longer serve as a definitive anchor point.
at which a given sauropod taxon might have fed. The hindlimb stance would probably need
Different workers have based their estimates of to be widened to achieve stability, which in turn
maximum feeding height on their assumptions would lower the center of mass and the fulcrum
regarding cervical posture. Upchurch and about which the presacral trunk would pivot.
Barrett (2000) estimated the maximum browse Unless the back were flexed dorsally to com-
height as the sum of shoulder height and neck pensate for the bending moment induced by
length. In other words, their maximum browse the massive gut, the animal’s vertical reach
height assumed a completely vertical neck. The would be less than sometimes optimistically
reconstructions in McIntosh et al. (1997) depicted. The height of the head at the maxi-
approach this assumption as well. As discussed mum vertical reach of a tripodal pose, and the
above, the results of our cervical studies suggest width of the feeding envelope in this posture,
substantially lower maximum browsing heights. would depend on how nearly erect the sauropod
Head elevation was far from vertical, reaching could stand, on whether it settled some of its
only approximately 42° above the horizontal in weight on the tail kangaroo-style, and on its sta-
Apatosaurus and about 15° in Diplodocus (fig. bility, which would ultimately depend on the
6.15). Since the curvature associated with dorsi- acuity of its neuromuscular coordination. The
flexion shortens the effective neck length, maxi- sauropod would need to integrate visual and
mum browse height would be more accurately vestibular signals to movement and to coordi-
estimated as the sum of shoulder height plus nate compensatory movements in order to pre-
roughly 40%–60% of the neck length. vent catastrophic instability, especially when
Maximum browse height might be of lesser eco- attempting to extend the neck laterally as well
logical importance than mean browse height, as dorsally.
however, given the prevalence of BV feeding in The relationship between sauropod feeding
modern herbivores, where larger and taller taxa strategies and Mesozoic floral evolution is of
may capitalize on their ability to ventriflex (i.e., interest in at least two ways. First, establishing
to reach down from their neutral feeding a correspondence between sauropod tooth,
height) to obtain food resources when their pre- skull, and neck forms and particular plant diets
ferred fodder becomes scarce (Daag and Foster would clarify one of the most important types of
1976; Leuthold 1977). energy flow among organisms in Jurassic and
Occasionally some sauropods (usually Cretaceous ecosystems. Second, at least some
diplodocids [e.g., Bakker 1986; Paul 2000]) paleobotanists (Wing and Tiffney 1987) and
have been depicted as standing tripodally, bal- vertebrate paleontologists (Bakker 1978) have
anced on their hindlimbs and the proximal end suggested that “clear-cutting” of Jurassic and
of the caudal vertebral column, and thus as hav- Early Cretaceous forests by sauropod herds may
ing the potential to reach far higher vegetation. have been instrumental in creating ecological
If some sauropods were capable of assuming conditions that favored the origin of flowering
such a tripodal feeding posture, a far greater plants.
range of vertical niche partitioning would have The lithological unit containing the greatest
been possible than if all sauropods were obliga- diversity and abundance of sauropod fossils is
tory quadrupeds. Again, an accurate body the Upper Jurassic (and possibly Lower
reconstruction is a necessary prerequisite to Cretaceous) Morrison Formation of western
North America. Although floras of the and lower teeth of Diplodocus suggested that it
Morrison Formation are not nearly as well pre- utilized both low and high browsing.
served and understood as its sauropod fauna, Diplodocids, along with the titanosaur clade
they have been the object of many studies. Nemegtosauridae, have distinctive adaptations
Miller (1987) surveyed Morrison compression (including dorsally placed nostrils, a tooth comb
floras, mostly from Montana, and noted that restricted to the front of the jaws that was suit-
cycadophytes, seed ferns, and ferns were gener- able for sieving or cropping, and a ventral incli-
ally more abundant than conifers and ginkgo- nation of the head relative to the horizontal axis
phytes. In their survey of Morrison taphonomy of the braincase) that may have facilitated low
and paleoecology, Dodson et al. (1980) inter- browsing and, perhaps, even placement of the
preted the Morrison depositional basin as a greater part of the heads underwater during
mostly arid, strongly seasonal, alluvial plain dis- feeding on aquatic plants (Parrish 2003).
sected by stream channels and occasional One ecological paradox that has long been
lakes. Rees et al. (2004) surveyed fossil plant posited regarding the Jurassic sauropod-
distribution and taphonomy in the Late Jurassic dominated ecosystems is how they could sup-
Morrison Formation and concluded that the port such a diverse and large assemblage of
Morrison landscapes were savanna-like, domi- giant herbivores. One common model pro-
nated by herbaceous plants and short trees such posed involves herds of sauropods moving
as ginkgos and cycads, most of which were con- through forests and denuding them of vegeta-
centrated near rivers and lakes. Although taller tion as they trampled the trunks (e.g., Wing and
conifers were part of the landscape, they appear Tiffney 1987). Given that the default inclination
to have been rare and confined to riparian cor- for most sauropod necks appears to be one that
ridors and small, isolated patches of forest positions the head close to the ground, perhaps
(Rees et al. 2004). it is more plausible that they were also feeding
Comparison of estimated browsing heights mostly at or near ground level. Lycopods and
of various sauropod taxa with heights of extant ferns appear to have been abundant in
members of plant clades that were abundant Mesozoic terrestrial ecosystems and, also,
during the Jurassic and Cretaceous (fig. 6.16) would have served as a rapidly growing, readily
shows a striking correspondence between the renewable food source, much like today’s
mean browse heights of the sauropods and angiospermous grasses. This is not a new spec-
these generalized vegetation dimensions. The ulation, having been proposed for diplodocids
only sauropod taxa that appear to have had the (Krasilov 1981; Chatterjee and Zheng 1997),
capability of grazing on tall gymnosperms were but it does potentially dovetail current thinking
Brachiosaurus and perhaps Camarasaurus. about sauropod functional morphology, late
Studies of cranial anatomy and tooth form in Mesozoic paleoecology, and plant physiology.
both genera, and of tooth wear in Camarasaurus,
suggest that they fed on more resistant plant
CONCLUSIONS
material and possibly utilized more oral process-
ing than was the case for diplodocids (Calvo Since sauropods were first described, it has
1994; Fiorillo 1998, Upchurch and Barrett gradually become standard to reconstruct most
1990). A survey of relative abundances of of these dinosaurs as high browsers. The
sauropods in the Morrison and Tendaguru sites neutral pose of the presacral vertebral column
(Rees et al. 2004) indicates that brachiosaurids can be reconstructed from the osteology of the
are more abundant in the conifer-rich articular facets of its component vertebrae.
Tendaguru beds than the Morrison deposits. However, for all sauropods thus studied our
Upchurch and Barrett (1990) also put forth the analysis shows the following conditions: the
hypothesis that differential wear on the upper dorsal portion of the vertebral column is
FIGURE 6.16. The bulk of the potential plant biomass in the middle Mesozoic, as exemplified here by the Morrison
flora, appears to have been medium to low browse. Even the tallest conifers had the greatest amount of their foliage
located in their bottom halves, and some of the fastest-growing and potentially most nutritious plants were
ground cover (e.g., lycopods and small ferns). There was little vertical stratification in the flora to drive potential
diversification among these herbivores. “Head-down” feeding by sauropods would be particularly apt for accessing
the foliage and fructifications of groups like ferns, cycads, and ground cover. Even without reconstructing Bra-
chiosaurus with a giraffelike neck, this sauropod was capable of BN feeding at about 5–6 m, while BD feeding by
dorsiflexion of merely 3 per joint would potentially raise the head 9 m above ground level. Ventriflexion by about
8 per joint proximally would place the head at ground level, permitting BV feeding that overlaps the mean feeding
heights of camarasaurids and diplodocids. The diplodocids, in occasional BD feeding, could have overlapped the ca-
marasaurid vertical range but were unique (with the possible exception of the later, nonsympatric Euhelopus) in their
ability to feed in BV below mean surface level, a potential adaptation for aquatic feeding (see fig. 6.9). Contrary to
the common expectation, vertical differentiation in sauropod feeding envelopes is not evident among these taxa,
many of which were coeval and sympatric. The predominantly low mean feeding heights of these sauropods corre-
spond closely to the height of the bulk of available browse. Although dental adaptations suggest clear diversity in
feeding preferences among sauropods, there was little vertical stratification of floras to drive potential diversification
among these herbivores.
arched, the cervical–dorsal transition is gymnosperms. Recent paleobotanical studies

straight, negligible intrinsic curvature occurs in (e.g., Rees et al. 2004; J. T. Parrish et al., in
the posterior cervical vertebrae, and a shallow, press) suggest that the greatest abundance of
concave-downward curvature occurs cranially. forage in the sauropod-rich, Late Jurassic
Most sauropods were medium to low browsers, Morrison Formation was low- to medium-
with generally a “head-down” feeding angle. height ground cover, which is consistent with
Although Brachiosaurus, possibly the tallest the concept of most Morrison sauropods being
sauropod, is shown to have a nearly horizontal low browsers.
neck, it nonetheless held its head about 6 m
above ground level. Therefore, Brachiosaurus ACKNOWLEDGMENTS
and, perhaps, Camarasaurus (at the upward The authors are deeply appreciative of the con-
extent of its feeding envelope) appear to have tributions Dr. Jack McIntosh has made to the
been the only Jurassic sauropods clearly capable field of sauropod paleontology. Jack’s devotion
of feeding as “high browsers” on arborescent to sauropods, his encyclopedic familiarity with
their fossil record, and his willingness to share Christian, A., and Heinrich, W. -D. 1998. The neck
that knowledge have enriched our research, posture of Brachiosaurus brancai. Mittelungen
Mus. Naturk. Berlin Geowissenschaften 1: 73–80.
and our lives, immensely. Jack has always
Christian, A., Heinrich, W. -D. and Golder, W. 1999.
expressed energetic support for our efforts to Posture and mechanics of the forelimbs
define the pose and movements of sauropods of Brachiosaurus brancai (Dinosauria: Sauropoda).
quantitatively. Seeing Jack dance a sponta- Mittelungen Mus. Naturk. Berlin Geowissenschaften
neous Scottish jig in response to our poster at 2: 63–73.
the Pittsburgh SVP quickened our hearts. Christiansen, P. 2000. Feeding mechanisms of the
sauropod dinosaurs Brachiosaurus, Camarasaurus,
Since then our work has been driven largely by
Diplodocus, and Dicraeosaurus. Hist. Biol. 14:
a desire to induce a repeat performance. Ron 137–152.
Toth and Judy Parrish provided advice and pho- Currie, P. J. 1987. New approaches to studying
tographs that were instrumental in the estima- dinosaurs in Dinosaur Provincial Park. In:
tion of plant height estimates for figure 6.16. Czerkas, S. J., and Olson, E. C. (eds.). Dinosaurs
Past and Present. Vol. II. University of
Brian Curtice and Christopher McGowan pro-
Washington Press, Seattle. Pp. 100–117.
vided photos of the Humboldt Museum Curry Rogers, K., and Forster, C. A. 2001. The last of
sauropods that augmented those taken by the dinosaur titans: a new sauropod from
J.M.P. Eric Wills provided important contribu- Madagascar. Nature 412: 530–534
tions to the programming of DinoMorph. Czerkas, S. A., and Czerkas, S. J. 1991. Dinosaurs: A
Thanks go to Jim Madsen for providing us with Global View. Mallard Press, New York. 247 pp.
Daag, A. I., and Foster, J. B. 1976. The Giraffe: Its
the opportunity to manipulate a cast of
Biology, Behavior, and Ecology. Van Nostrand
Diplodocus carnegii. Heinrich Mallison pro- Reinhold, New York. 210 pp.
vided three-dimensional data for the modeling Dodson, P. M. 1990. Sauropod paleoecology. In:
of B. brancai. Matt Bonnan and Phil Platt pro- Weishampel, D. B., Dodson P., and Osmolska, H.
vided valuable insights into sauropod limb pos- (eds.). The Dinosauria. University of California
Press, Berkeley. Pp. 402–407.
ture. This work was supported in part by
Dodson, P. M., Behernsmeyer, A. K., Bakker, R. T.,
National Science Foundation Grant 0093929. and McIntosh, J. T. 1980. Taphonomy and paleoe-
cology of the dinosaur beds of the Jurassic
LITERATURE CITED Morrison Formation. Paleobiology 6: 208–232.
Farlow, J. O. 1987. Speculations about the diet and
Alexander, R. M. 1989. Dynamics of Dinosaurs and digestive physiology of herbivorous dinosaurs.
Other Extinct Giants. Columbia University Press, Paleobiology 13: 60–72.
New York. 167 pp. Fiorillo, A. R. 1998. Dental microwear patterns of
Bakker, R. T. 1978. Dinosaur feeding behavior and sauropod dinosaurs Camarasaurus and Diplodocus:
the origin of flowering plants. Nature 274: evidence for resource partitioning in the Late
661–663. Jurassic of North America. Hist. Biol. 13: 1–16.
———. 1986. Dinosaur Heresies. William Morrow, Gauthier-Pilters, H., and Daag, A. I. 1981. The Camel,
New York. 481 pp. Its Ecology, Behavior and Relationship to Man.
Barrett, P. M., and Upchurch, P. 1994. Feeding University of Chicago Press, Chicago. 208 pp.
mechanisms of Diplodocus. GAIA 10: 195–204. Gilmore, C. W. 1925. A nearly complete articulated
Beasley, Walter. 1907. Diplodocus: the greatest of all skeleton of Camarasaurus, a saurischian dinosaur
earthly creatures. Scientific American 96(24): from the Dinosaur National Monument. Mem.
491–492. Carnegie Mus. 10: 347–384.
Bonnan, M. F. 2001. The Evolution and Functional Gilmore, C. W. 1936. The osteology of Apatosaurus
Morphology of Sauropod Dinosaur Locomotion. with special reference to specimens in the
Ph.D. dissertation. Northern Illinois University, Carnegie Museum. Mem. Carnegie Mus. 11:
DeKalb, IL. 722 pp. 175–300.
Calvo, J. O. 1994. Jaw mechanics in sauropod Hatcher, J. B. 1901. Diplodocus (Marsh): its osteology,
dinosaurs. GAIA 10: 183–194. taxonomy, and probable habits, with a recon-
Chatterjee, S., and Zheng, Z. 1997. The feeding strate- struction of the skeleton. Mem. Carnegie Mus. 1:
gies in sauropods. J. Vertebr. Paleontol. 17: 37A. 1–63.
Holland, W. J. 1906. The osteology of Diplodocus Paul, G. S. 1987. The science and art of restoring the
Marsh. Mem. Carnegie Mus. 2: 225–278. life appearance of dinosaurs and their relatives: a
Janensch, W. 1929. Die Wirbelsäule der Gattung rigorous how-to guide. In: Czerkas, S. J., and Olson,
Dicraeosaurus hausemanni. Palaeontographica 3 E. C., (eds.). Dinosaurs Past and Present. Vol. II.
(Suppl. 7): 39–133. University of Washington Press, Seattle. Pp. 5–49.
Janensch, W. 1950a. Die Wirbelsäule von Brachiosaurus ———. 2000. Restoring the life appearance of
brancai. Palaeontographica (Suppl. 7) 3: 27–93. dinosaurs. In: Paul, G. S. (ed.). The Scientific
Janensch, W. 1950b. Die Skelettrekonstrucktion von American Book of Dinosaurs. Bryon Press and
Brachiosaurus brancai. Palaeontographica 3 (Suppl. Scientific American, New York. Pp. 78–106.
7): 95–103. Pellew, R. 1984. The feeding ecology of a selective
Janensch, W. 1961. Die Gliedmaszen und browser, the giraffe (Giraffa camelopardalis tippel-
Gliedmaszengürtel der Sauropoden der Tendaguru- skirchi). J. Zool. 202: 57–81.
Schichten. Palaeontographica (Suppl. 7)3: 177–235. Pincher, C. 1949. Evolution of the giraffe. Nature
Krasilov, V. A. 1981. Changes of Mesozoic vegetation 164: 29–30.
and the extinction of the dinosaurs. Paleogeog. Rees, P. M., Noto, C. R., Parrish, J. M., and Parrish,
Paleoclimatol. Palaeoecol. 34: 207–224. J. T. 2004. Late Jurassic climates, vegetation, and
Leuthold, B. 1977. African Ungulates: A Comparative dinosaur distributions. J. Geol. 112: 643–653.
Review of Their Ethology and Behavioral Ecology. Sereno, P. C., Beck, A. L., Moussa, B., Dutheil, D.,
Springer Verlag, Berlin. 307 pp. Larsson, H. C. E., Lyon, G. H., Sadlier, R. W., Sidor,
Leuthold, B., and Leuthold, W. 1972. Food habits of C. A., Varrichio, D. J., Wilson, G. P., and Wilson,
giraffe in Tsavo National Park, Kenya. East Afr J. A. 1999. Cretaceous sauropods from the Sahara
Wildl. J. 10: 129–141. and the uneven rate of skeletal evolution among
Martin, J. 1987. Mobility and feeding of Cetiosaurus: dinosaurs. Science 286: 1342–1347.
Why the long neck? Occas. Paper Tyrell Mus. Simmons, R., and Scheepers, L. 1996. Winning by a
Paleontol. 3: 150–155. neck: sexual selection in the evolution of the
McGinnis, H. J. 1982. Carnegie’s Dinosaurs. giraffe. Am. Nat. 148: 771–786.
Carnegie Museum of Natural History, Pittsburgh, Stevens, K. A. 2002. DinoMorph: parametric model-
PA. 129 pp. ing of skeletal structures. Senckenbergiana
McIntosh, J., M. K. Brett-Surman and J. O. Farlow. Lethaea 82(1): 23–34.
1997. Sauropods. In: Farlow, J. O., and Brett- Stevens, K. A., and Parrish, J. M. 1999. Neck posture
Surman, M. K. (eds.). The Complete Dinosaur. and feeding habits of two Jurassic sauropod
Indiana University Press, Bloomington. dinosaurs. Science, 284: 798–800.
Pp. 264–290. ———. 2005a. Neck posture, dentition, and feeding
Miller, C. M. 1987. Land plants of the northern Rocky strategies in Jurassic sauropod dinosaurs. In:
Mountains before the appearance of flowering Carpenter, K., and Tidwell, V. (eds.). Thunder
plants. Ann. Ma. Bot. Garden 74: 692–706. Lizards: The Sauropodomorph Dinosaurs.
Osborn, H. F., and Mook, C. C. 1921. Camarasaurus, University of Indiana Press, Bloomington (in
Amphicoelias, and other sauropods of Cope. Mem. press).
Am. Mus. Nat. Hist. 3: 247–287. ———. 2005b. The intrinsic curvature of sauropod
Owen-Smith, R. N. 1988. Megaherbivores: The necks (Saurischia: Dinosauria) (in preparation).
Influence of Very Large Body Size on Ecology. Stevens, K. A., and Wills, E. D. 2001. Gracile versus
Cambridge University Press, Cambridge. 369 pp. robust cervical vertebral designs in sauropods. J.
Parrish, J. M. 2003. Mapping ecomorphs onto sauro- Vertebr. Paleontol. Suppl. 21(3): 104A.
pod phylogeny. J. Vertebr. Paleontol. Suppl. 23(3): Thulborn, R. A. 1990. Dinosaur Tracks. Chapman
85A–86A. and Hall, London. 430 pp.
Parrish, J. M., and Stevens, K. A. 1998. Undoing the Upchurch, P., and Barrett, P. M., 2000. The evolu-
death pose: Using computer imaging to restore tion of sauropod feeding mechanisms. In: Sues,
the posture of articulated dinosaur skeletons. J. H. D. (ed.). Evolution of Herbivory in Terrestrial
Vertebr. Paleontol. Suppl.18: 69A. Vertebrates. Cambridge University Press,
———. 2002a. Neck mobility in long-necked verte- Cambridge. pp. 79–122.
brates: from modern mammals to sauropods. J. Walther, F. 1969. Flight behavior and avoidance of
Morphol. 248: 270. predators in Thompson’s gazelle (Gazella thom-
———. 2002b. Rib angulation, scapular position, soni Gunther, 1884). Behaviour 34: 184–221.
and body profiles in sauropod dinosaurs. J. Wedel, M. J., and Sanders, R. K. 1999. Comparative
Vertebr. Paleontol. Suppl. 22(3): 95A. morphology and functional morphology of the
cervical series in Aves and Sauropoda. J. Vertebr. Wing, S. L., and Tiffney, B. H. 1987. The reciprocal
Paleontol. 19: 83A. interaction of angiosperm evolution and tetrapod
Wilhite, R. 2003. Digitizing large fossil skeletons herbivory. Rev. Paleobot. Palynol. 50: 179–210.
for three-dimensional applications. Paleontol. Woolnough, A. P., and du Toit, J. T. 2001. Vertical
Electronica 5(1): 1–10 zonation of browse quality in tree canopies
Wilson, J. A., and Sereno, P. C. 1998. Early evolution exposed to a size-structured guild of African
and higher-level phylogeny of sauropod browsing ungulates. Oecologia 129: 585–590.
dinosaurs. Soc. Vertebr. Paleontol. Mem. 5: 1–68. Young, T., and Isbell, L. 1991. Sex differences in
Wiman, C. 1929. Die Kriede-dinosaurier aus giraffe feeding ecology: Energetic and social con-
Shantung. Palaeontol. Sinica (Ser. C) 6: 1–67. straints. Ethology 87: 79–89.
SEVEN
Postcranial Skeletal Pneumaticity

in Sauropods and Its Implications
for Mass Estimates
Mathew J. Wedel
standing problems, and outline possible direc-

O ne of the signal features
of sauropods, and one of the corner-
stones of our fascination with them, is their ap-
tions for future studies. To that end, the chapter
is organized around three questions. What cri-
parent efficiency of design. The presacral neural teria do we use to infer pneumaticity in sauro-
spines of all sauropods have a complex of bony pod fossils? What characteristics of pneumatic
ridges or plates known as vertebral laminae (fig. bones have been (or could be) described? and
7.1; abbreviations used in the figures are listed How can we apply data on skeletal pneumatic-
below). In addition, the vertebral centra of most ity to paleobiological problems, such as esti-
sauropods bear deep fossae or have large foram- mating the masses of sauropods? Before
ina that open into internal chambers. The lami- attempting to answer these questions, it will be
nae and cavities of sauropod vertebrae are often useful to review skeletal pneumaticity in living
considered to be adaptations for mass reduction vertebrates.
(Osborn 1899; Hatcher 1901; Gilmore 1925) and
have been important in studies of sauropod evo- Institutional abbreviations: BYU, Earth
lution (McIntosh 1990; Wilson 1999). The pos- Sciences Museum, Brigham Young University,
sibility that these structures were pneumatic— Provo, Utah; CM, Carnegie Museum,
that they contained or partitioned air-filled Pittsburgh, Pennsylvania; DGM, Museo de
diverticula of the lungs or air sacs—has been rec- la Divisão de Geologia y Mineralogia, Rio de
ognized for over a century (Seeley 1870; Janen- Janeiro, Brazil; DMNS, Denver Museum of
sch 1947). However, pneumaticity in sauropods Nature and Science, Denver, Colorado; OMNH,
has received little attention until recently (Britt Oklahoma Museum of Natural History,
1997; Wilson 1999; Wedel 2003a, 2003b). Norman, Oklahoma; USNM, National Museum
My goal here is to review previous work on of Natural History, Smithsonian Institution,
pneumaticity in sauropods, discuss some out- Washington, DC.
201
FIGURE 7.1. Pneumatic features in dorsal vertebrae of Barapasaurus (A–D), Camarasaurus (E–G), Diplodocus (H–J), and
Saltasaurus (K–N). Anterior is to the left; different elements are not to scale. A, A posterior dorsal vertebra of Barapasaurus.
The opening of the neural cavity is under the transverse process. B, A midsagittal section through a middorsal vertebra of
Barapasaurus showing the neural cavity above the neural canal. C, A transverse section through the posterior dorsal shown
in A (position 1). In this vertebra, the neural cavities on either side are separated by a narrow median septum and do not
communicate with the neural canal. The centrum bears large, shallow fossae. D, A transverse section through the middor-
sal shown in B. The neural cavity opens to either side beneath the transverse processes. No bony structures separate the
neural cavity from the neural canal. The fossae on the centrum are smaller and deeper than in the previous example. (A–D
redrawn from Jain et al. 1979:pl. 101, 102.) E, An anterior dorsal vertebra of Camarasaurus. F, A transverse section through
the centrum (E, position 1) showing the large camerae that occupy most of the volume of the centrum. G, a horizontal sec-
tion (E, position 2). (E–G redrawn from Ostrom and McIntosh 1966:pl. 24.) H, A posterior dorsal vertebra of Diplodocus.
(Modified from Gilmore 1932:fig. 2.) I, Transverse sections through the neural spines of other Diplodocus dorsals (similar to
H, position 1). The neural spine has no body or central corpus of bone for most of its length. Instead it is composed of in-
tersecting bony laminae. This form of construction is typical for the presacral neural spines of most sauropods outside the
clade Somphospondyli. (Modified from Osborn 1899:fig. 4.) J, A horizontal section through a generalized Diplodocus dorsal
(similar to H, position 2). This diagram is based on several broken elements and is not intended to represent a specific
specimen. The large camerae in the midcentrum connect to several smaller chambers at either end. K, A transverse section
through the top of the neural spine of an anterior dorsal vertebra of Saltasaurus (L, position 1). Compare the internal pneu-
matic chambers in the neural spine of Saltasaurus with the external fossae in the neural spine of Diplodocus shown in J. L,
An anterior dorsal vertebra of Saltasaurus. M, A transverse section through the centrum (L, position 2). N, A horizontal sec-
tion (L, position 3). In most members of the clade Somphospondyli the neural spines and centra are filled with small
camellae. (K–N modified from Powell 1992:fig. 16.)
Anatomical abbreviations: al, accessory la- lamina; spol, spinopostzygapophyseal lamina;

mina; cit, canalis intertransversarius; cml, ca- sprl, spinoprezygapophyseal lamina; vk, ventral
mella; cmr, camera; dsv, diverticulum superverte- keel.
brale; for, foramen; fos, fossa; lam, lamina; nad,
neural arch diverticulum; naf, neural arch fossa;
SKELETAL PNEUMATICITY IN EXTANT TAXA
ncl, neural canal; ncs, neurocentral suture; ncv,
neural cavity; nsf, neural spine fossa; pcdl, poste- Pneumatization of the postcranial skeleton
rior centrodiapophyseal lamina; podl, postzygodi- in various ornithodiran groups, including
apophyseal lamina; prdl, prezygodiapophyseal sauropods, is just one instance of the more
202 P O S T C R A N I A L S K E L E TA L P N E U M AT I C I T Y I N S AU R O P O D S
general phenomenon of skeletal pneumatiza- enchymal tissue, which degenerates or is
tion. Skeletal pneumatization, which includes resorbed and is in turn replaced by the growing
paranasal, paratympanic, and pulmonary pneu- diverticulum. As the diverticulum bores
matic spaces, is unique to archosaurs and through the cortical bone it produces a pneu-
advanced synapsids (Witmer 1997, 1999). matic foramen, which must remain open for
However, diverticula (epithelium-lined out- pneumatization to proceed normally (Ojala
growths) of the pharynx or trachea are present 1957). Once the bone has been penetrated,
in representative taxa from most major lineages branches of the diverticulum spread through
of tetrapods, including frogs (Duellman and the marrow cavity by replacing bony trabeculae.
Trueb 1986), snakes (Young 1991, 1992), birds The marrow is reduced to small islands of tis-
(King 1966; McClelland 1989a), and primates sue surrounded by the diverticulum. As these
(Janensch 1947). Pharyngeal and tracheal diver- islands of marrow degenerate, the branches of
ticula are often used to inflate specialized struc- the diverticulum anastomose and form a single,
tures used in phonation or visual display. These epithelium-lined air cavity that occupies most
diverticula do not invade any bones except the of the internal volume of the bone. The trabec-
hyoid, which is pneumatized by tracheal diver- ular structure of the bone is greatly reduced,
ticula in the howler monkey Alouatta (Janensch and the inner layers of the cortex are resorbed.
1947; Mycetes of his usage). Diverticula of Witmer (1990) pointed out that a pneumatic
paranasal and paratympanic air spaces extend foramen does not have to be located on the
down the neck in some species of birds, but pneumatic bone in question; the intraosseous
these diverticula are subcutaneous or intermus- diverticulum may have spread across a suture
cular and do not pneumatize the postcranial from an adjacent pneumatic bone. He called
skeleton (King 1966). Extremely rare examples this extramural pneumatization and contrasted
of cervical pneumatization have been reported it with intramural pneumatization, in which a
in humans, but these are pathological cases diverticulum directly invades a bone and pro-
related to occipitoatlantal fusion (Sadler et al. duces a pneumatic foramen. Although Witmer
1996). Among extant taxa, only birds have (1990) was concerned with cranial pneumatiza-
extensive postcranial skeletal pneumaticity tion, extramural pneumatization also occurs in
(PSP). the postcranial skeleton, for example, between
Extant birds have relatively small, inflexible fused vertebrae in the chicken (King 1957;
lungs and an extensive system of air sacs in the Hogg 1984a).
thorax and abdomen. The air sacs are flexible The term air sac has been used by some
and devoid of parenchymal tissue, and their pri- authors for any reservoir of air in an animal that
mary function is to ventilate the lungs (King is lined by epithelium and devoid of parenchy-
1966; Duncker 1971; McClelland 1989b). In mal tissue (e.g., Brattstrom 1959; Cranford et al.
most birds, the air sacs also give rise to a net- 1996). The same term is often used in the
work of diverticula. Diverticula pass into the ornithological literature to refer specifically to
viscera, between muscles, and under the skin in the pulmonary air sacs of birds (e.g., Müller
various taxa (Richardson 1939; King 1966; 1907). In this paper, the term air sac is restricted
Duncker 1971). If a diverticulum comes into to indicate the pulmonary air sacs of birds. All
contact with a bone, the bone may become other epithelium-lined air reservoirs, including
pneumatized. Bremer (1940) described the those that develop from the lungs and air sacs,
pneumatization of the humerus in the chicken are called diverticula. Another important differ-
(Gallus) as follows. The diverticulum enters the ence is between a pneumatic diverticulum,
bone because osteoclasts break down the bony which is a soft-tissue structure, and the bony
tissue ahead of it. The bony tissue immediately recess that it may occupy (Witmer 1999). In
adjacent to the diverticulum is replaced by mes- many cases, the bony recess is produced by the
P O S T C R A N I A L S K E L E TA L P N E U M AT I C I T Y I N S AU R O P O D S 203
diverticulum through the process of pneumati- advanced; as Janensch (1947:10: translated
zation. This causal relationship allows us to from the German by G. Maier) said, “There is
infer the presence of diverticula from certain no basis to consider the pleurocentral cavities
kinds of bony recesses. The study of skeletal in sauropod vertebrae as different from similar
pneumaticity in fossil taxa is founded on such structures in the vertebrae of birds.” In short,
inferences. no soft tissues other than pneumatic diverticula
are known to produce large foramina that lead
to internal chambers, and these chambers con-
WHAT CRITERIA DO WE USE TO INFER
stitute unequivocal evidence of pneumaticity.
PNEUMATICITY IN FOSSILS?
One of the primary differences among the
How do we recognize skeletal pneumaticity? pneumatic vertebrae of different sauropod taxa
More specifically, what are the osteological cor- is the subdivision of the internal chambers.
relates (sensu Witmer 1995, 1997) of pneumatic Some taxa, such as Camarasaurus, have only a
diverticula, such that the presence of the latter few large chambers, whereas others, such as
can be inferred from the former? Several Saltasaurus, have many small chambers (fig.
authors, including Hunter (1774) and Müller 7.1). Vertebrae with many small chambers have
(1907), list differences between pneumatic and been characterized as “complex” (Britt 1993;
apneumatic bones. These authors focused on Wedel 2003b), in contrast to “simple” vertebrae
recognizing pneumaticity in extant birds and with few chambers. The concept of “biological
thus referred to attributes that tend not to fos- complexity” has several potential meanings
silize, such as vascularity, oil content, and color. (McShea 1996). In this paper, complexity refers
Britt (1993, 1997) provided the most compre- only to the level of internal subdivision of
hensive list of pneumatic features identifiable pneumatic bones; complex bones have more
in fossil bones: internal chambers with foram- chambers than simple ones. This is “nonhier-
ina, fossae with crenulate texture, smooth or archical object complexity” in the terminology
crenulate tracks (grooves), bones with thin of McShea (1996).
outer walls, and large foramina.
EXTRAMURAL PNEUMATIZATION
INTERNAL CHAMBERS WITH FORAMINA The only obvious opportunities for extramural
The most obvious osteological correlate of pneumatization in the postcranial skeletons of
pneumaticity is the presence of foramina that sauropods are between fused sacral and caudal
lead to large internal chambers. Large cham- vertebrae and between the sacrum and the
bers, often called “pleurocoels,” are present in ilium. Sacral vertebrae of baby sauropods have
the presacral vertebrae of most sauropods. They deep fossae (Wedel et al. 2000:fig. 14), and at
may also be present in the sacral and caudal least in Apatosaurus, a complex of internal
vertebrae, as in Apatosaurus and Diplodocus (see chambers is present before the sacral vertebrae
Ostrom and McIntosh 1966:pl. 30 and Osborn fuse (Ostrom and McIntosh 1966:pl. 30). The
1899:fig. 13, respectively). In extant birds, such co-ossified blocks of caudal vertebrae in
chambers are invariably associated with pneu- Diplodocus often include centra with large pneu-
matic diverticula (Britt 1993). The presence of matic foramina (Gilmore 1932:fig. 3). It is pos-
similar chambers in the bones of sauropods, sible that co-ossified centra without foramina
theropods, and pterosaurs has been accepted by could be pneumatized by intraosseous diverticula
most authors as prima facie evidence of pneu- of adjacent pneumatic vertebrae, although this
maticity (Seeley 1870; Cope 1877; Marsh 1877; has not been demonstrated.
Janensch 1947; Romer 1966; Britt 1993, 1997; Sanz et al. (1999) reported that “cancellous
O’Connor 2002). As far as I am aware, no sub- tissue” is present in the presacral vertebrae, ribs,
stantive alternative hypotheses have been and ilium of Epachthosaurus and Saltasaurus. The
presacral vertebrae of Saltasaurus are pneumatic communicate with the neural canal (fig. 7.1C).
and have a camellate internal structure (fig. Neural cavities are interpreted as pneumatic for
7.1K–N), and pneumatic ribs are known in sev- the same reason that the more familiar cavities
eral titanosaurs (Wilson and Sereno 1998). in vertebral centra are: they are large internal
Further, spongiosa (sensu Francillon-Vieillot et al. chambers connected to the outside through
1990) are present in apneumatic vertebrae of prominent foramina (Britt 1993).
many—possibly all—sauropods (see “Application
to a Paleobiological Problem: Mass Estimates,” PNEUMATIC RIBS
below), so cancellous bone is not limited to The dorsal ribs of some sauropods have large
titanosaurs. For these reasons, it seems that the foramina that lead to internal chambers. The
“cancellous tissue” of Sanz et al. (1990) is syn- best-known examples of costal pneumaticity in
onymous with camellate pneumatic bone. If so, sauropods are the pneumatic ribs of
then the ilia of some titanosaurs may have been Brachiosaurus (Riggs 1904; Janensch 1950).
pneumatic. Two possible routes for pneumatiza- Pneumatic dorsal ribs are also present in
tion of the ilium are by diverticula of abdominal Euhelopus and some titanosaurs (Wilson and
air sacs and by extramural pneumatization from Sereno 1998). Gilmore (1936) described a fora-
the sacrum. However, the possibility of ilial men that leads to an internal cavity in a dorsal rib
pneumatization must remain speculative until of Apatosaurus, and pneumatic dorsal ribs have
better evidence for it is presented. also been reported in the diplodocid Supersaurus
(Lovelace et al. 2003). Pneumatic dorsal ribs have
NEURAL CAVITIES not been found in Haplocanthosaurus,
In many sauropods, the neural spines of the Camarasaurus, or any basal diplodocoids, so the
dorsal vertebrae contain large chambers. These character evidently evolved independently in
chambers communicate with the outside by way diplodocids and titanosauriforms. Pneumatic
of large foramina beneath the diapophyses. ribs are part of a growing list of pneumatic char-
Upchurch and Martin (2003) called such cham- acters that evolved in parallel in diplodocids and
bers neural cavities and discussed their occur- titanosauriforms, along with complex vertebral
rence in Cetiosaurus, Barapasaurus, and chambers and pneumatic caudal vertebrae (see
Patagosaurus. According to Upchurch and below).
Martin (2003:218), “In Barapasaurus and
FOSSAE AND LAMINAE
Patagosaurus, the neural cavity is linked to the
external surface of the arch by a lateral foramen PNEUMATIC FOSSAE
which lies immediately below the base of the Fossae are ubiquitous in sauropod vertebrae
transverse process, just in front of the posterior and are often the sole evidence of pneumaticity.
centrodiapophyseal lamina [pcdl]” (see fig. 7.1A). For example, basal sauropods such as
In some dorsal vertebrae of Barapasaurus, the Barapasaurus have shallow fossae on the pre-
neural canal is open dorsally and communicates sacral centra and neural spines but lack the
with the neural cavity (Jain et al. 1979). large internal chambers typical of later
Upchurch and Martin (2003) mentioned that sauropods (fig. 7.1). Are these fossae pneu-
similar cavities are present in some matic? The naive assumption that all fossae are
neosauropods, and Bonaparte (1986:fig. 19.7) pneumatic will surely lead to the overestima-
illustrated neural cavities in Camarasaurus and tion of pneumaticity. On the other hand, to
Diplodocus. Jain et al. (1979) and Upchurch and deny that any fossae are pneumatic unless they
Martin (2003) also described a second morphol- contain foramina that lead to large internal
ogy (in Barapasaurus and Cetiosaurus, respec- chambers is equally false. We need criteria to
tively), in which the neural cavity is divided into distinguish pneumatic fossae from nonpneu-
two halves by a median septum and does not matic fossae.
FIGURE 7.2. A cervical vertebra of Brachiosaurus and a hypothetical reconstruction of the pneumatic diverticula. A, BYU
12866, a midcervical vertebra of Brachiosaurus, in left lateral view. The neural spine fossae are bounded on all sides by the
four laminae that connect the pre- and postzygapophyses to the neurapophysis and diapophysis. Some of the neural spine
fossae contain large, sharp-lipped foramina. B, Possible appearance of the pneumatic diverticula, shown in black. We can be
fairly certain that pneumatic diverticula occupied the fossae on the neural arch, neural spine, and centrum, but the connec-
tions between various diverticula and their order of appearance during ontogeny remain speculative. Here the diverticula
have been restored based on those of birds, with the canalis intertransversarius running alongside the centrum and the di-
verticulum supervertebrale occupying the neural spine fossae (see Müller [1907:figs. 3–5, 7, 11, 12] for the appearance of
these diverticula in the pigeon). Any connections between the canalis intertransversarius and the diverticulum superverte-
brale probably passed intermuscularly, because the laminae bounding the neural spine fossae are uninterrupted by tracks
or grooves. C, A transverse section through the midcentrum (A, position 1) traced from a CT image (Wedel et al. 2000:fig.
12C) and corrected for distortion. The volume of air filling the fossae and camellae in the neural arch and spine is un-
known, but it may have equaled or exceeded the volume of air in the centrum. Lamina terminology after Wilson (1999).
Scale bar equals 20 cm.
The best case for a pneumatic fossa is a inference of pneumaticity. “These well defined,
fossa that contains pneumatic foramina within smooth-walled depressions are present in many
its boundaries. The Brachiosaurus vertebra sauropods and seem to be analogous to the
shown in figure 7.2 has large, sharply lipped more pronounced coels [ foramina] that charac-
pneumatic foramina in most of the fossae on terize Brachiosaurus. Like the coels, these
the lateral sides of the centrum and neural depressions may have housed smaller pneu-
spine (see also Janensch 1950; Wilson 1999). matic diverticuli [sic] in life” (Wilson 1999:651).
Similar foramina-within-fossae are present in This hypothesis is supported by the complex
the vertebrae of many other neosauropods, morphology of some pneumatic diverticula in
including Diplodocus (Hatcher 1901:pl. 3, 7), birds. In the ostrich, the large diverticula that lay
Tendaguria (Bonaparte et al. 2000:fig. 17, pl. 8), alongside the cervical vertebrae consist of bun-
and Sauroposeidon (Wedel et al. 2000:fig. 8b). dles of smaller diverticula (Wedel 2003a:fig. 2).
The inference that these fossae are pneumatic It seems reasonable to expect that when such a
relies on the presence of unequivocally pneu- bundle comes into contact with a bone, the
matic features within the fossae. The inferred aggregate would produce a fossa, within which
presence of pneumaticity is less supported in each diverticulum would produce a subfossa.
the case of blind fossae that contain no foram- This hypothesis can and should be tested in
ina, such as the large fossae on the dorsal cen- future computed tomography (CT) studies.
tra of Barapasaurus (fig. 7.1). Gower (2001:121) argued that the “multipartite
Wilson (1999) proposed that “subfossae,” or fossae” and “deep multi-chambered concavi-
fossae-within-fossae, might further support the ties” in the dorsal vertebrae of Erythrosuchus
were more consistent with pneumaticity than cally pneumatic. Their presence demonstrates
with muscular or vascular structures (but see that pneumatic diverticula were in close contact
O’Connor 2002). with all of the preserved cervical vertebrae.
Britt (1993) proposed that crenulate texture of Because we already know that pneumatic diver-
the external bone is evidence that some fossae ticula contacted the cervical vertebrae, it seems
are pneumatic. In Sauroposeidon the difference safe to infer that the neural spine fossae are
in texture between the pneumatic fossae and the pneumatic in origin. At least, the inference of
adjacent bone is striking, and this allows the pneumaticity is better founded than it would be
boundaries of the fossae to be precisely plotted based on the neural spine fossae alone.
(Wedel et al. 2000:fig. 7). However, there is little (As an aside, the nomenclature for vertebral
doubt that the fossae of Sauroposeidon are pneu- laminae has been thoroughly reviewed and
matic, because they contain pneumatic foram- standardized [Wilson 1999], but no standard
ina. The inference that a blind fossa is pneu- nomenclature for vertebral fossae exists. It is
matic based on texture alone is less certain. tempting to propose such a nomenclature, if
Blind fossae can also contain muscles or adipose only to avoid circumlocutions like that used
tissue (O’Connor 2002). It is not known if these above [“the fossae bounded by the podl, prdl,
three kinds of fossae can be reliably distin- spol, and sprl”]. However, a separate nomencla-
guished on the basis of bone texture. Until this is ture for fossae is unnecessary and could be
tested, inferring pneumaticity on the basis of misleading. Hatcher [1901] named several fos-
bone texture alone may not be warranted. sae, such as the “infraprezygapophyseal cavity,”
For the time being, I know of no test that using the same spatial orientation terms that
can definitively determine whether a blind were commonly used for naming laminae [e.g.,
fossa housed a pneumatic diverticulum or Osborn 1899]. Such a position-based nomen-
some other soft tissue. Pneumatic diverticula clature for fossae shares all of the faults of the
often induce bone resorption when they come old orientation-based systems for naming lam-
into contact with the skeleton, and it is possible inae [ for further discussion see Wilson 1999].
that external pneumatic features might be rec- Laminae should be defined by the structures
ognized by some distinctive aspect of cortical they connect [Wilson 1999]. Similarly, I think
bone histology. I do not suggest that this must that fossae should be defined by the laminae
be the case, but it is worth investigating. that bound them. To list all of the bounding
To determine if a fossa is pneumatic or not, laminae when referring to a fossa may be awk-
it is worthwhile to consider other potentially ward, but it is also precise.)
pneumatic features on or in the same bone.
Consider the fossa bounded by the podl, prdl, VERTEBRAL LAMINAE, HOMOLOGY, AND THE
spol, and sprl in Haplocanthosaurus (fig. 7.3). At ORIGINS OF POSTCRANIAL SKELETAL PNEUMATICITY
least in the cervical vertebrae, these fossae do It is tempting to assume that the fossae of basal
not contain any pneumatic foramina or subfos- sauropods are pneumatic because they are
sae, they do not lead to any obvious pneumatic homologous to the unequivocally pneumatic
tracks, and the bone texture is smooth rather features of later sauropods. For example, in
than crenulate (pers. obs.). In other words, noth- Brachiosaurus the fossa bounded by the podl,
ing about the fossae themselves indicates that prdl, spol, and sprl is clearly pneumatic because it
they were pneumatic (as opposed to containing contains pneumatic foramina (fig. 7.2). Does this
adipose deposits or other soft tissues). However, mean that the equivalent fossa in Barapasaurus is
the centra of the same vertebrae contain deep, also pneumatic? After all, phylogenetic analysis
sharp-lipped cavities that penetrate to a narrow indicates that the bounding laminae are homol-
median septum. By the criteria discussed ogous in Barapasaurus and Brachiosaurus
herein, the cavities in the centra are unequivo- (Wilson 1999, 2002). The answer seems to be
FIGURE 7.3. Pneumatic features in a cervical vertebra of Haplocanthosaurus. A, A posterior cervical of
Haplocanthosaurus in right lateral view (CM 879-7; this specimen was erroneously referred to as CM
572 in Upchurch [1998:fig. 8], and as CM 897-7 in Wedel et al. [2000:fig. 2], Wedel [2003a:fig. 3], and
Wedel [2003b:fig. 1]). (Modified from Hatcher (1903:pl. 2.) B–E, Cross sections traced from CT slices.
B, Section at A, position 1. C, Section at A, position 2. The opening of the neural canal and the absence
of the neurocentral suture on one side are due to a break in the specimen. D, Section at A, position 3.
E, Section at A, position 4. The neurocentral sutures are unfused over most of their length, indicating
that this animal was not fully mature. Scale bar equals 5 cm.
that the fossae may be homologous, but that is that bound the fossae would have remained the
no guarantee that they were produced by the same, but the tissue that filled the fossae would
same morphogenetic process. Ontogenetic have changed. The same replacement may also
pathways are themselves subject to evolution- have occurred during ontogeny.
ary change. As Hall (1999:347) stated, “A limb If we order archosaur vertebrae in terms of
built upon one set of rules does not lose its putatively pneumatic features, the resulting
homology with limbs built upon different arrangement has no obvious gaps and is
rules.” Conversely, homology does not neces- roughly congruent with current phylogenies
sarily indicate identical morphogenetic path- (i.e., Sereno 1991; Wilson 2002). At one end of
ways. The shallow fossae of basal sauropods the spectrum are vertebrae that lack laminae,
may have contained deposits of fat such as such as those of extant crocodilians. Very shal-
those identified in birds by O’Connor (2001). It low depressions may be present on the neural
is possible that such adipose deposits were spines or centra, but these depressions are not
replaced by pneumatic diverticula later in bounded by an obvious lip and do not contain
sauropod evolution. In that case, the laminae subfossae or large foramina. The next grade
of vertebral construction is represented by origin of vertebral laminae. The blind fossae of
Marasuchus, which has low ridges below some early saurischians are, at best, equivocal evi-
of the presacral diapophyses (Sereno and Arcucci dence of pneumaticity. However, any explana-
1994); these ridges may represent rudimentary tion that pushes the origin of PSP forward in
laminae (Wilson 1999). At the next level, a series time will accumulate a corresponding number
of diapophyseal and zygapophyseal laminae is of ad hoc hypotheses to explain the early
primitive for Saurischia (Wilson 1999). These appearance of laminae and fossae. For these
laminae are present in Herrerasaurus and reasons, I favor Wilson’s (1999) hypothesis that
prosauropods (Sereno and Novas 1994; laminae are pneumatic in origin and that the
Bonaparte 1986), but the fossae they enclose appearance of laminae marks the appearance of
are blind, lack subfossae, and have no obvious PSP, although as Wilson (1999:651) pointed
textural differences from the adjacent bone out, more work is needed.
(Wedel, pers. obs.). Vertebral centra of these taxa Gower (2001) posited widespread pneu-
lack fossae. Shallow fossae are present on the maticity in Archosauria based on vertebral
centra of early sauropods such as Isanosaurus, fossae. If he is right, PSP originated before
Shunosaurus, and Barapasaurus, and neural the divergence between crocodile- and bird-
chambers may be present in the arch and spine line archosaurs and was present in virtually all
(Jain et al. 1979; Zhang 1988; Buffetaut et al. of the noncrocodilian taxa in the series dis-
2000). In Jobaria and Haplocanthosaurus the cussed above. O’Connor (2002) questioned the
central fossae are bounded by a sharp lip and reliability of blind fossae as indicators of pneu-
penetrate to a median septum (Sereno et al. maticity, but he did not present evidence to fal-
1999; Wedel 2003b, pers. obs.). Finally, most sify Gower’s hypothesis. Indeed, hypotheses of
neosauropods have prominent pneumatic pneumaticity are difficult to falsify; although it
foramina that open into chambers that ramify is often easy to demonstrate that a bone has
within the centrum, and the fossae of the neu- been pneumatized, it is difficult to demonstrate
ral arches and spines contain subfossae or that it has not (Hogg 1980). For now, the possi-
pneumatic foramina. bility that the fossae described by Gower are
It is not clear where pneumaticity first pneumatic cannot be ruled out, but neither can
appears in the preceding series. At one end of less radical alternative hypotheses.
the scale are the vertebrae of crocodiles, which
are known to be apneumatic. At the other end OTHER OSTEOLOGICAL CORRELATES
are the vertebrae of neosauropods, the pneu- OF PNEUMATICITY
matic features of which are virtually identical to Pneumatic tracks, thin outer walls, and large
those of birds (Janensch 1947). In between, the foramina are not likely to be falsely interpreted
inference of pneumaticity receives more sup- as pneumatic features in sauropods. External
port as we approach Neosauropoda, but the tracks are only rarely identified in sauropods.
“break point” between apneumatic and pneu- Wedel et al. (2000:fig. 7) illustrated a pneu-
matic morphologies is debatable. The primitive matic track in Sauroposeidon, but the track
saurischian complex of laminae first appears in was not the basis for the pneumatic interpre-
small dinosaurs and seems to be structural tation; rather, the track was identified as
overkill if pneumatic diverticula were absent pneumatic because it led away from a deep,
(Wilson 1999). An apneumatic interpretation sharply lipped pneumatic fossa. Many sauro-
of these laminae requires that a large number pod vertebrae have thin outer walls, especially
of structures that are clearly related to pneuma- those of the aforementioned Sauroposeidon
tization in later forms be primitively present for (fig. 7.4). However, the thin outer walls of
other reasons, and leaves us (at least for now) sauropod vertebrae invariably bound large
without a satisfying hypothesis to explain the internal chambers that are clearly pneumatic,
FIGURE 7.4. Internal structure
of a cervical vertebra of Sauropo-
seidon, OMNH 53062. A, The
posterior two-thirds of C5 and
the condyle and prezygapoph-
ysis of C6 in right lateral view.
The field crew cut though C6 to
divide the specimen into man-
ageable pieces. B, Cross section
of C6 at the level of the break,
traced from a CT image (A, posi-
tion 1) and photographs of the
broken end. The left side of the
specimen was facing up in the
field and the bone on that side is
badly weathered. Over most of
the broken surface the internal
structure is obscured by plaster
or too damaged to trace, but it is
cleanly exposed in the ramus of
the right prezygapophysis (out-
lined). C, The internal structure
of the prezygapophyseal ramus,
traced from a photograph. The
arrows indicate the thickness of
the bone at several points, as
measured with a pair of digital
calipers. The camellae are filled
with sandstone.
so, again, the inference of pneumaticity does theropods, but most sauropods are simply so
not rest on the equivocal feature. Finally, there large that pneumatic and nutrient foramina
is the question of foramina that are not pneu- are unlikely to be confused. Even juvenile
matic, such as nutrient or nervous foramina. sauropods tend to have large pneumatic fossae
Britt et al. (1998) proposed that pneumatic rather than small foramina (see Wedel et al.
foramina could be distinguished from nutri- 2000:fig. 14).
ent foramina on the basis of relative size, with
pneumatic foramina typically being about an
DESCRIPTION OF PNEUMATIC ELEMENTS
order of magnitude larger, relative to the
length of the centrum. The two kinds of At least four aspects of skeletal pneumaticity
foramina could also be distinguished based on can be described: the external traces of pneu-
the internal structure of the vertebrae. maticity (discussed above), the internal com-
Pneumatic vertebrae typically lack trabecular plexity of an element, the ratio of bone to air-
bone (Bremer 1940; Schepelmann 1990) and space within an element, and the distribution
have compact bone in their outer walls and in of pneumatic features along the vertebral
the septa between pneumatic cavities (Reid column.
1996). The presence of trabecular bone inside
a vertebra is evidence that it is either apneu- INTERNAL COMPLEXITY OF PNEUMATIC
matic or, at least, incompletely pneumatized BONES
(King 1957). Distinguishing pneumatic foram- This variable has received the most attention in
ina from nutrient foramina is a potential prob- previous studies and is only briefly reviewed
lem in studies of birds and other small here. Longman (1933) recognized that sauropod
TABLE 7.1
Classification of Sauropod Vertebrae into Morphologic Categories Based on Pneumatic Characters
Acamerate Pneumatic characters limited to fossae; fossae do not significantly invade the
centrum.
Procamerate Deep fossae penetrate to median septum but are not enclosed by ostial
margins.
Camerate Large, enclosed camerae with regular branching pattern; cameral generations
usually limited to 3.
Polycamerate Large, enclosed camerae with regular branching pattern; cameral generations
usually 3 or more, with increased number of branches at each generation.
Semicamellate Camellae present but limited in extent; large camerae may also be present.
Camellate Internal structure entirely composed of camellae; neural arch laminae not
reduced. Large external fossae may also be present.
Somphospondylous Internal structure entirely composed of camellae; neural arch laminae reduced;
neural spine has an inflated appearance.
NOTE: After Wedel et al. (2000:table 3).
vertebrae with internal chambers fall into two bral camerae. From published descriptions
broad types, those with a few large chambers (Young and Zhao 1972; Russell and Zheng
and those with many small chambers. Longman 1994), the vertebrae of Mamenchisaurus appear
called the first type phanerocamerate and the to be camellate.
second cryptocamerillan (although he did not From the foregoing, it might appear that the
explicitly discuss them as products of skeletal internal structures of sauropod vertebrae, their
pneumatization). Britt (1993, 1997) independ- evolution, and their phylogenetic distribution
ently made the same observation and used the are all well understood. In fact, vertebral inter-
terms camerate and camellate to describe large- nal structure is only known for a small minor-
chambered and small-chambered vertebrae, ity of sauropods. Even in those taxa for which
respectively. Wedel et al. (2000) expanded this the internal structure is known, this knowledge
terminology to include categories for vertebrae is usually limited to a handful of vertebrae or
with fossae only and vertebrae with combina- even a single element, which severely limits
tions of large and small chambers (table 7.1). our ability to assess serial, ontogenetic, and
Wedel et al. (2000) and Wedel (2003b) also dis- population-level variation. Despite these limita-
cussed the phylogenetic distribution of different tions, three broad generalizations can be made.
internal structure types. In general, the verte- First, the vertebrae of very young sauropods
brae of early diverging sauropods such as tend to have a simple I-beam shape in cross sec-
Shunosaurus and Barapasaurus have external tion, with large lateral fossae separated by a
fossae but lack internal chambers. Camerae median septum (Wedel 2003b). This is true
are present in the vertebrae of diplodocids, even for taxa in which the vertebrae of adults
Camarasaurus, and Brachiosaurus. Presacral ver- are highly subdivided, such as Apatosaurus. In
tebrae of Brachiosaurus also have camellae in the these taxa the internal complexity of the verte-
condyles and cotyles, and camellae are variably brae increased during ontogeny. The second
present in the neural spine and apophyses. The generalization is that complex internal struc-
vertebrae of Sauroposeidon and most titanosaurs tures evolved independently in Mamenchisaurus
lack camerae and are entirely filled with camel- and diplodocids and one or more times in
lae, although some titanosaurs may have verte- Titanosauriformes (Wedel 2003b). This suggests
a general evolutionary trend toward increasing Casinos 2000), and the mean for terrestrial
complexity of vertebral internal structure in mammals is 0.53 (calculated from Currey and
sauropods, albeit one that took different forms Alexander 1985:table 1).
in different lineages (i.e., polycamerate verte- The K value is a parameter of tubular bones;
brae in Diplodocidae and somphospondylous it is meaningless when applied to bones with
vertebrae in Somphospondyli) and that may more complex shapes or internal structures,
have been subject to reversals (i.e, camerate ver- such as sauropod vertebrae. I propose the air-
tebrae in some titanosaurs [see Wedel 2003b]). space proportion (ASP), or the proportion of
Finally, the largest and longest- necked the volume of a bone—or the area of a bone
sauropods, such as Mamenchisaurus, the section—that is occupied by air spaces, as a
diplodocines, brachiosaurids, Euhelopus, and variable that can be applied to both tubular and
titanosaurs such as Argentinosaurus and the nontubular bones. One problem is that meas-
unnamed taxon represented by DGM Serie A, uring the volumes of objects is difficult and
all have polycamerate, semicamellate, or fully often imprecise. It is usually easier to measure
camellate internal structures. I have previously the relevant surface areas of a cross section, but
stated that the complex internal structures were any one cross section may not be representative
correlated with increasing size and neck length of the entire bone. For example, the long bones
(Wedel 2003a, 2003b). This may or may not be of birds and mammals are usually tubular at
true; I have not performed any phylogenetic midshaft, but the epiphyses mostly consist of
tests of character correlation. Nevertheless, the marrow-filled trabecular bone or pneumatic
presence of complex internal structures in the camellate bone. Nevertheless, it may be easier
vertebrae of the largest and longest-necked to take the mean of several cross sections as an
sauropods suggests that size, neck length, and approximation of volume than to directly meas-
internal structure are related. ure volume, especially in the case of large, frag-
ile, matrix-filled sauropod vertebrae.
VOLUME OF AIR WITHIN A PNEUMATIC BONE For the avian long bones described above,
The aspect of skeletal pneumaticity that has data were only presented for a single cross sec-
probably received the least attention to date is tion located at midshaft. Therefore, the ASP
the ratio of bone tissue to empty space inside a values I am about to discuss may not be repre-
pneumatic bone. Although many authors have sentative of the entire bones, but they probably
commented on the weight-saving design of approximate the volumes (total and air) of the
sauropod vertebrae (Osborn 1899; Hatcher diaphyses. For tubular bones, ASP may be
1901; Gilmore 1925), no one has quantified just determined by squaring K (if r is the inner
how much mass was saved. The savings in mass diameter and R the outer, then K is r/R, ASP is
could have important paleobiological implica- πr2/πR2 or simply r2/R2, and ASP K2). For
tions, for example, in determining how much the K of pneumatic bones, Currey and
mass to subtract from volumetric mass esti- Alexander (1985) report lower and upper
mates. bounds of 0.69 and 0.86, and I calculate a
Currey and Alexander (1985) and Cubo and mean of 0.80 from the data presented in their
Casinos (2000) reported relevant data on the table 1. Using a larger sample size, Cubo and
long bones of birds, which are tubular and may Casinos (2000) found a slightly lower mean K
be filled with marrow or air. In both studies, the of 0.77. The equivalent values of ASP are 0.48
variable of interest was K, the inner diameter of and 0.74, with a mean of 0.64, or 0.59 for the
the element divided by its outer diameter. Both mean of Cubo and Casinos (2000). This
studies found mean values of K between 0.77 means that, on average, the diaphysis of a
and 0.80 for pneumatic bones. The mean for pneumatic avian long bone is 59%–64% air, by
marrow-filled bird bones is 0.65 (Cubo and volume.
FIGURE 7.5. How to determine the airspace proportion (ASP) of a bone section. (A) A section is traced from a photograph,
CT image, or published illustration; in this case, a transverse section of a Barosaurus africanus cervical vertebra from Janen-
sch (1947:fig. 3). (B) Imaging software is used to fill the bone, airspace, and background with different colors. The number
of pixels of each color can then be counted using Image J (or any program with a pixel count function) and used to compute
the ASP. In this case, bone is black and air is white, so the ASP is (white pixels)/(black pixels + white pixels).
How do these numbers compare with the tained at least 50% air, by volume, and probably
ASPs of sauropod vertebrae? To find out, I somewhat more. This assumes that the cavities
measured the area occupied by bone and the in sauropod vertebrae were entirely filled with
total area for several cross sections of sauropod air and the amount of soft tissue was negligible.
vertebrae (see fig. 7.5 for an example). I obtained Chandra Pal and Bharadwaj (1971) found that
the cross-sectional images from CT scans, pub- the air spaces in pneumatic bird bones are lined
lished cross sections, and photographs of with simple squamous epithelium, so the
broken or cut vertebrae. For image analysis I assumption is probably valid. The ASP values
used Image J, a free program available online presented here for sauropod vertebrae are sim-
from the National Institutes of Health ilar to the range and mean found for pneumatic
(Rasband 2003). Some results are presented in long bones of birds (or at least their diaphyses).
table 7.2 (this research is in progress and I will Second, although only a handful of meas-
present more complete results elsewhere). The urements are available for each taxon, it is
results should be approached with caution: I already clear that ASP can vary widely from
have only analyzed a few vertebrae from a hand- slice to slice within a single vertebra and prob-
ful of taxa, and only one or a few cross sections ably also between vertebrae of different regions
for each bone, so the results may not be repre- of the skeleton and between individuals of the
sentative of either the vertebrae, the regions of same species. As we collect more data we may
the vertebral column, or the taxa to which they find more predictable relationships, for exam-
belong. The sample is strongly biased toward ple, between the ASP values of cervical and
cervical vertebrae simply because cervicals are dorsal vertebrae or between certain taxa. The
roughly cylindrical and fit through CT scanners system may also be so variable that such rela-
better than dorsal or sacral vertebrae. Despite tionships will be impossible to detect, if they
these caveats, some regularities emerge. even exist. Rampant variation seems to be the
First, ASP values range from 0.32 to 0.89, rule for skeletal pneumaticity in general (e.g.,
with a mean of 0.60. Even though the data may King 1957; Cranford et al. 1996; Weiglein
not be truly representative, it seems reasonable 1999), and it would be surprising if ASP were
to conclude that most sauropod vertebrae con- not also highly variable.
TABLE 7.2
The Airspace Proportion (ASP) of Transverse Sections through Vertebrae of Sauropods and Other Saurischians
REGION ASP SOURCE
Apatosaurus Cervical condyle 0.69 Wedel (2003b:fig. 6b)

Cervical midcentrum 0.52 Wedel (2003b:fig. 6c)
Cervical cotyle 0.32 Wedel (2003b:fig. 6d)
Barosaurus Cervical midcentrum 0.56 Janensch (1947:fig. 8)
Cervical, near cotyle 0.77 Janensch (1947:fig. 3)
Caudal midcentrum 0.47 Janensch (1947:fig. 9)
Brachiosaurus Cervical condyle 0.73 Janensch (1950:fig. 70)
Cervical midcentrum 0.67 Wedel et al. (2000:fig. 12c)
Cervical cotyle 0.39 Wedel et al. (2000:fig.12d)
Dorsal midcentrum 0.59 Janensch (1947:fig. 2)
Camarasaurus Cervical condyle 0.49 Wedel (2003b:fig. 9b)
Cervical midcentrum 0.52 Wedel (2003b:fig. 9c)
Cervical, near cotyle 0.50 Wedel (2003b:fig. 9d)
Dorsal midcentrum 0.63 Ostrom and McIntosh
(1966:pl. 23)
(1966:pl. 24)
(1966:pl. 25)
Phuwiangosaurus Cervical midcentrum 0.55 Martin (1994:fig. 2)
Pleurocoelus Cervical midcentrum 0.55 Lull (1911:pl. 15)
Saltasaurus Dorsal midcentrum 0.55 Powell (1992:fig. 16)
Sauroposeidon Cervical prezygapophyseal 0.89 Fig. 7.4
ramus
Cervical midcentrum 0.74 Wedel et al. (2000:fig. 12g)
Cervical postzygapophysis 0.75 Wedel et al. (2000:fig. 12h)
Theropoda Cervical prezygapophysis 0.48 Janensch (1947:fig. 16)
Dorsal midcentrum 0.50 Janensch (1947:fig. 15)
NOTE: Only values for published sections are presented. Much more work will be required to determine norms for different taxa and
different regions of the vertebral column, and the values presented here may not be representative of either. Nevertheless, these values
suggest that pneumatic sauropod vertebrae were often 50%–60% air, by volume. The mean of these 22 measurements is 0.60.
Third, the lowest ASP values—0.32 in Fourth, Sauroposeidon has the highest values
Apatosaurus and 0.39 in Brachiosaurus—are for of ASP, up to a remarkable 0.89. The values for
slices through the cotyle, or bony cup, at the Sauroposeidon are even higher than those for
posterior end of the centrum. Here the cortical the closely related Brachiosaurus, and the
bone is doubled back on itself to form the cup, ranges for the two taxa do not overlap (although
and the wall of the cotyle itself is at an angle to they may come to when a larger sample is con-
the slice and appears wider in cross section. sidered). A very high ASP is probably an
The cotyle is surrounded by pneumatic cham- autapomorphy of Sauroposeidon and may have
bers in both Apatosaurus and Brachiosaurus, but evolved to help lighten its extremely long
these become smaller and eventually disappear (12-m) neck.
toward the end of the vertebra. For these rea- Finally, ASP appears to be independent of the
sons, the cotyle is expected to have a lower ASP internal complexity of the vertebrae. The Salta-
than the rest of the vertebra. saurus vertebra is the most highly subdivided of
the sample. The I-beam-like vertebrae of the In mapping the distribution of pneumaticity
juvenile Pleurocoelus and Phuwiangosaurus are along the vertebral column, it is important to
the least subdivided; the other taxa fall some- consider where on the vertebrae the pneumatic
where in the middle. Nevertheless, most values features are located. In the co-ossified block of
in the table 7.2, including those for Saltasaurus, Diplodocus caudal vertebrae illustrated by
Pleurocoelus, and Phuwiangosaurus, fall between Gilmore (1932:fig. 3), the centra of caudals 15–19
0.50 and 0.60. The means for all taxa other than bear large pneumatic foramina, but the neural
Sauroposeidon also fall within the same range, so spines lack laminae and do not appear to have
there is no apparent trend that relates ASP to been pneumatic. This is in contrast to the pre-
internal complexity. Cast in evolutionary terms, sacral, sacral, and anterior caudal vertebrae,
this indicates that the evolution of complex inter- which have heavily sculpted neural spines with
nal structures from simple ones involved a redis- deep fossae and scattered foramina (see Osborn
tribution rather than a reduction of bony tissue 1899:figs. 7, 13). In the opposite condition, the
within the vertebrae. The ASP values of the juve- neural spines bear laminae and fossae and may
nile Pleurocoelus and Phuwiangosaurus imply that have been pneumatic, but the centra lack pneu-
a similar redistribution was involved in the onto- matic features. Examples include the middle and
genetic derivation of complex chambers from posterior dorsal vertebrae of Jobaria (see Sereno
juvenile fossae. et al. 1999:fig. 3). Sauropod vertebrae can there-
The results presented here are preliminary, fore exist in one of four states: (1) both centrum
and the available data are better suited for sug- and neural spine pneumatic, as in the presacral
gesting hypotheses than for testing them. Much vertebrae of most neosauropods; (2) centrum
work remains to be done, both in gathering pneumatic but neural spine apneumatic, as in
comparative data from extant forms and in the middle caudals of Diplodocus; (3) neural
exploring the implications of pneumaticity for spine pneumatic but centrum apneumatic, as
sauropod biomechanics. in the posterior dorsals of Jobaria (assuming
that the laminate neural spines are pneumatic);
DISTRIBUTION OF PNEUMATICITY ALONG or (4) no signs of pneumaticity in the centrum
THE VERTEBRAL COLUMN or neural spine, as in the distal caudals of most
The two previous sections dealt with the char- sauropods. Pneumatization of the centrum typ-
acteristics of a single pneumatic bone. We must ically results in large internal cavities with
also consider the location of pneumatic features prominent foramina, so the inference of pneu-
in the skeleton, because these features con- maticity is well supported in conditions 1 and 2.
strain the minimum extent of the diverticular In condition 3 the situation may be less clear. In
system. For example, in the USNM 10865 derived neosauropods such as Brachiosaurus
skeleton of Diplodocus, pneumatic foramina are and the diplodocids, the neural spine fossae
present on every vertebra between the axis and often bear small subfossae and foramina,
the nineteenth caudal ([Gilmore 1932; foramina which indicate that these fossae are pneumatic
are only present on caudals 1–18 in the skeleton (see Janensch 1950; Curtice and Stadtman
of Diplodocus described by Osborn [1899] and 2001). In more basal sauropods such as
on caudals 1–16 in the mounted DMNS skele- Haplocanthosaurus, the neural spine fossae are
ton [Wedel, pers. obs.]). This means that in life often blind and lack the heavily sculpted texture
the pneumatic diverticula reached at least as far seen in later forms. The neural spines of these
anteriorly as the axis and as far posteriorly as basal sauropods may have been pneumatic, but
caudal vertebra 19 (fig. 7.6). The diverticular the inference is less well founded.
system may have been more extensive and sim- The earliest sauropodomorph with dis-
ply failed to pneumatize any more bones, but it tinctly emarginated pneumatic fossae is
could not have been any less extensive. Thecodontosaurus caducus (Yates 2003). In
FIGURE 7.6. Hypothetical conformation of the respiratory system of a diplodocid sauropod. The left forelimb, pectoral gir-
dle, and ribs have been removed for clarity. The lung is shown in dark gray, air sacs are light gray, and pneumatic diverticula
are black. Only some of the elements shown in this illustration can be determined with certainty: the minimum length of
the trachea, the presence of at least some air sacs, and the minimum extent of the pneumatic diverticula. The rest of the
respiratory system has been restored based on that of birds, but this remains speculative. The skeleton is modified from
Norman (1985:83).
T. caducus, pneumatic fossae are only present ula of sauropods followed the same pattern of
on the middle cervical vertebrae. This means development as those of birds, then the pres-
that the fossae must have been produced by ence of pneumatic sacral vertebrae in most
diverticula of cervical air sacs similar to those neosauropods indicates the presence of abdom-
of birds (as opposed to diverticula of the lungs inal air sacs (Wedel et al. 2000). There are no
proper). A similar pattern of pneumatization in strong reasons to doubt that neosauropods had
Coelophysis indicates that cervical air sacs were abdominal air sacs. However, the future discov-
present in both sauropodomorphs and ery of a sauropod with a pneumatic hiatus—a gap
theropods by the Norian (Late Triassic), and in the pneumatization of the dorsal vertebrae—
cervical air sacs are probably primitive for would unequivocally demonstrate the presence
saurischians (Wedel 2004). of abdominal air sacs and their diverticula (Wedel
In general, more derived sauropods tended 2003a).
to pneumatize more of the vertebral column.
Except for the atlas, which is always apneu-
APPLICATION TO A PALEOBIOLOGICAL
matic, pneumatic chambers (or prominent fos-
PROBLEM: MASS ESTIMATES
sae) are present in the cervical vertebrae of
Shunosaurus; in the cervical and anterior dorsal The implications of PSP for sauropod paleobi-
vertebrae of Jobaria; in all of the presacral ver- ology are only beginning to be explored. In par-
tebrae of Cetiosaurus; in the presacral and sacral ticular, skeletal pneumaticity may be an impor-
vertebrae of most neosauropods; and in the pre- tant factor in future studies of the biomechan-
sacral, sacral, and caudal vertebrae of diplodocids ics and respiratory physiology of sauropods.
and saltasaurids (Wedel 2003a, 2003b, pers. The most obvious implication of extensive PSP
obs.). This caudad progression of vertebral in sauropods is that they may have weighed
pneumaticity also occurred in the evolution of less than is commonly thought. In this section,
theropods (Britt 1993) and occurs ontogeneti- the problem of estimating the masses of
cally in extant birds (Cover 1953; Hogg 1984b). sauropods is used as an example of how infor-
At a gross level, the system is both homoplastic mation about PSP may be applied to a paleobi-
and recapitulatory. ological question.
In extant birds, diverticula of the cervical Two distinct questions proceed from the
air sacs do not extend farther posteriorly than observation that most sauropod skeletons
the anterior thoracic vertebrae. If the divertic- were highly pneumatic. The first is purely
methodological: (How) Should we take pneu- drilled holes to represent the lungs before esti-
maticity into account in estimating the masses of mating the center of mass of each model and
sauropods? The second question is paleobiologi- the proportion of mass supported by the fore-
cal: If we find that pneumaticity significantly and hindlimbs (see Alexander 1989:figs. 4.6,
lightened sauropods, how does that affect our 5.3). Curiously, he does not seem to have drilled
understanding of sauropods as living animals? If the holes before performing his mass esti-
pneumaticity did not significantly lighten mates; at least, the holes are only mentioned in
sauropods, then the second question is moot, so conjunction with the center of mass and limb
I consider the methodological question first. support studies. Henderson (1999) included
lung spaces in his digital models for mass esti-
METHODS mation purposes and, later, included air sacs
The masses of dinosaurs are generally esti- and diverticula in a buoyancy study (Henderson
mated using allometric equations based on 2004). Paul (1988, 1997) used the alternative
limb bone dimensions (Russell et al. 1980; method of adjusting the density values for the
Anderson et al. 1985) or volumetric measure- mass calculations. He assigned a specific grav-
ments using physical or computer models ity (SG) of 0.9 to the trunk to account for lungs
(Colbert 1962; Paul 1988, 1997; Henderson and air sacs, and an SG of 0.6 to the neck to
1999). If allometric equations are used, then account for pneumatization of the vertebrae.
pneumaticity need not be taken into account; Before attempting to estimate the volume of
the limb bones are assumed to have been as cir- air in a sauropod, it is important to recognize
cumferentially robust as they needed to be to that the air was distributed among four sepa-
support the animal’s mass, regardless of how rate regions: (1) the trachea, (2) the “core” respi-
the body was constituted. If an animal with a ratory system of lungs and, possibly, pulmonary
pneumatic skeleton was lighter than it would air sacs, (3) the extraskeletal (i.e., visceral, inter-
have been otherwise, this should already be muscular, and subcutaneous) diverticula, and
reflected in its limb bone morphology, and no (4) the pneumatic bones. These divisions are
correction is necessary. On the other hand, if important for two reasons. First, the volumes of
volumetric measurements are used, then it is each region are differently constrained by skele-
possible to take skeletal pneumaticity into tal remains. The volume of air in the skeleton
account and failure to do so may result in mass can be estimated with a high degree of confi-
estimates that are too high. dence because the sizes of the airspaces can be
Volumetric mass estimation is performed in measured from fossils. In contrast, the volume
three steps (Alexander 1989). First, the volume of the trachea is not constrained by skeletal
of a scale model of the organism is measured. remains and must be estimated by comparison
Next, the volume of the model is multiplied by to extant taxa. The lung/air sac system and
the scale factor to obtain the volume of the extraskeletal diverticula are only partly con-
organism in life. Finally, the volume of the strained by the skeleton (see below). This leads
organism is multiplied by the estimated density to the second point, which is that estimates of
to obtain its mass. The presence of air in the all four regions can be made independently, so
respiratory system and pneumatic diverticula that skeletal pneumaticity can be taken into
can be accounted for in the first two steps, by account regardless of conformation (birdlike,
reducing the estimated volume of model or the crocodile-like, etc.) and volume of the core res-
organism, or in the third step, by adjusting the piratory system.
density used in the mass calculation. Both
methods have been used in published mass AN EXAMPLE USING DIPLODOCUS
estimates of dinosaurs. Alexander (1989) used Consider the volume of air present inside a liv-
plastic models in his volumetric study, and he ing Diplodocus. Practically all available mass
TABLE 7.3 (1932) and treated the centra as cylinders. The
The Volume of Air in Diplodocus caudal series of CM 84 is incomplete, so I sub-
stituted the measurements for USNM 1065
AIR MASS
from Gilmore (1932); comparison of the meas-
VOLUME (L) SAVINGS (KG)
urements of the elements common to both
Trachea 104 104 skeletons indicates that the two animals were
Lungs and air sacs 1,500 1,500 roughly the same size. I multiplied the volumes
Extraskeletal diverticula ? ? obtained by 0.60, the mean ASP of the sauro-
Pneumatic vertebrae pod vertebrae listed in table 7.2, to obtain the
Centra
total volume of air in the centra.
Cervicals 2–15 136 82
Dorsals 1–10 208 125
The volume of air in the neural spines is
Sacrals 1–5 75 45 harder to calculate. The neural spines are com-
Caudals 1–19 329 198 plex shapes and are not easily approximated
Subtotal for centra 748 450 with simple geometric models. Furthermore,
Neural spines the fossae on the neural arches and spines only
Cervicals 2–15 136 82 partially enclosed the diverticula that occupied
Dorsals 1–10 416 250 them. Did the diverticula completely fill the
Sacrals 1–5 150 90
space between adjacent laminae, did they bulge
Caudals 1–19 165 99
Subtotal for spines 867 520 outward into the surrounding tissues, or did
Subtotal for vertebrae 970 1,455 surrounding tissues bulge inward? In the com-
plete absence of in vivo measurements of diver-
Total 2,574 3,059 ticulum volume in birds, it is impossible to say.
Based on the size of the neural spine relative to
NOTE: See the text for methods of estimation. The total vol- the centrum in most sauropods (see fig. 7.2), it
ume for vertebrae is 1,615.
seems reasonable to assume that in the cervical
vertebrae, at least as much air was present in
estimates for Diplodocus (Colbert 1962; the arch and spine as in the centrum, if not
Alexander 1985; Paul 1997; Henderson 1999) more. In the high-spined dorsal and sacral ver-
are based on CM 84, the nearly complete skele- tebrae (see fig. 7.1), the volume of air in the neu-
ton described by Hatcher (1901). Uncorrected ral arch and spine may have been twice that in
volumetric mass estimates—i.e., those that do the centrum. Finally, proximal caudal vertebrae
not include lungs, air sacs, or diverticula—for have large neural spines but the size of the
this individual range from 11,700 kg (Colbert spines decreases rapidly in successive verte-
1962; as modified by Alexander 1989:table 2.2) brae. On average, the caudal neural spines of
to 18,500 kg (Alexander 1985). Paul (1997) cal- Diplodocus may have contained only half as
culated a mass of 11,400 kg using the corrected much air as their associated centra. These esti-
SGs cited above, and Henderson (1999) estimates are admittedly rough, but they are proba-
mated 14,912 kg, or 13,421 kg after deducting bly conservative and so they will suffice for this
10% to represent the lungs. For the purposes of example.
this example, the volume of the animal is As they developed, the intraosseous divertic-
assumed to have been 15,000 liters. The esti- ula replaced bony tissue, and the density of that
mated volumes of various air reservoirs and tissue must be taken into account in estimating
their effects on body mass are listed in table 7.3. how much mass was saved by pneumatization
Estimating the volume of air in the vertebral of the skeleton. In apneumatic sauropod verte-
centra is the most straightforward. I used pub- brae the internal structure is filled with can-
lished measurements of centrum length and cellous bone and presumably supported red
diameter from Hatcher (1901) and Gilmore (erythropoeitic) bone marrow (fig. 7.7). Distal
in the body. They are assumed to replace soft
tissues (density of 1 g/cm3) in the solid model.
Extraskeletal diverticula include visceral,
intermuscular, and subcutaneous diverticula.
None of these leave traces that are likely to be
fossilized. The bony skeleton places only two
constraints on the extraskeletal diverticula.
First, as previously discussed, the distribution
of pneumatic bones in the skeleton limits the
minimum extent of the diverticular system.
Thus, we can infer that the vertebral diverticula
in Diplodocus must have extended from the axis
FIGURE 7.7. Internal structure of OMNH 27794, a partial to the nineteenth caudal vertebra (at least in
distal caudal vertebra of a titanosauriform. The internal USNM 1065), but the course and diameter of
structure is composed of apneumatic cancellous bone, and the diverticula are unknown. The second con-
no medullary cavity is present. Scale bar equals 1 cm.
straint imposed by the skeleton is that the
canalis intertransversarius, if it existed, could
caudal vertebrae of the theropod Ceratosaurus not have been larger than the transverse foram-
have a large central chamber or centrocoel ina where it passed through them, although it
(Madsen and Welles 2000:fig. 6). This cavity may have been smaller or increased in diameter
lacks large foramina that would connect it to on either side. I am unaware of any studies in
the outside, so it cannot be pneumatic in origin. which the in vivo volume of the avian diverticu-
The medullary cavities of apneumatic avian and lar system is measured. This information vac-
mammalian long bones are filled with adipose uum prevents me from including a volume esti-
tissue that acts as lightweight packing material mate for the diverticular system in table 7.3.
(Currey and Alexander 1985), and the same To estimate the volume of the trachea, I used
may have been true of the centrocoels in the allometric equations presented by Hinds
Ceratosaurus caudals. The presence of a similar and Calder (1971) for birds. The length equation,
marrow cavity in sauropod vertebrae prior to L 16.77M0.394, where L is the length of the tra-
pneumatization cannot be ruled out, but to my chea (cm) and M is the mass of the animal (kg),
knowledge no such cavities have been reported. yielded a predicted tracheal length of 6.8 m for
In birds, the intraosseous diverticula erode the a 12-ton animal. The cervical series of Diplodocus
inner surfaces of the cortical bone in addition to CM 84 is 6.7 m long and the trachea may have
replacing the cancellous bone (Bremer 1940), been somewhat longer, and I judged the corre-
so pneumatic bones tend to have thinner walls spondence between the neck length and the pre-
than apneumatic bones (Currey and Alexander dicted tracheal length to be close enough to jus-
1985; Cubo and Casinos 2000). The tissues that tify using the equations, especially for the coarse
may have been replaced by intraosseous diver- level of detail needed in this example. The vol-
ticula have SGs that range from 0.9 for some ume equation, V 3.724M1.090, yields a volume
fats and oils to 3.2 for apatite (Schmidt-Nielsen of 104 liters.
1983:451, table 11.5). For this example, I estimated Finally, the volume of the lungs and air sacs
that the tissue replaced by the intraosseous diver- must be taken into account. The lungs and air
ticula had an average SG of 1.5 (calculated from sacs are only constrained by the skeleton in that
data presented in Cubo and Casinos 2000), so they must fit inside the ribcage and share space
air cavities that total 970 liters replace 1,455 kg with the viscera. Based on measurements from
of tissue. The extraskeletal diverticula, trachea, caimans and large ungulates, Alexander (1989)
lungs, and air sacs did not replace bony tissue subtracted 8% from the volume of each of his
models to account for lungs. Data presented by 12 metric tons and the SG of the entire body
King (1966:table 3) indicate that the lungs and would have been 0.8. This is lower than the
air sacs of birds may occupy 10%–20% of the SGs of squamates and crocodilians (0.81–0.89)
volume of the body. Hazlehurst and Rayner found by Colbert (1962), higher than the SGs of
(1992) found an average SG of 0.73 in a sample birds (0.73) found by Hazlehurst and Rayner
of 25 birds from 12 unspecified species. On this (1992), and about the same as the SGs
basis, they concluded that the lungs and air sacs (0.79–0.82) used by Henderson (2004) in his
occupy about a quarter of the volume of the study of sauropod buoyancy. Note that the
body in birds. However, some of the air in their amount of mass saved by skeletal pneumatiza-
birds probably resided in extraskeletal divertic- tion is independent of the estimated volume of
ula or pneumatic bones, so the volume of the the body, but the proportion of mass saved is
lungs and air sacs may have been somewhat not. Thus if we start with Alexander’s (1985)
lower. In the interests of erring conservatively, I 18,500-liter estimate for the body volume of
put the volume of the lungs and air sacs at 10% Diplodocus, the mass saved is still 1,455 kg, but
of the body volume. this is only 8% of the solid mass, not 10% as in
The results of these calculations are neces- the previous example.
sarily tentative. The lungs and air sacs were It could be argued that adjusting the esti-
probably not much smaller than estimated mated mass of a sauropod by a mere 8%–10% is
here, but they may have been much larger; the pointless. The mass of the living animal may
trachea could not have been much shorter but have periodically fluctuated by that amount or
may have been much longer, or it may have more, depending on the amount of fat it carried
been of a different or an irregular diameter (see and how much food it held in its gut (Paul
McClelland [1989a] for tracheal convolutions 1997). Further, the proposed correction is tiny
and bulbous expansions in birds); the neural compared to the range of mass estimates pro-
spines may have contained much more or duced by different studies, from 11,700 kg (Paul
somewhat less air; the ASP of Diplodocus verte- 1997) to 18,500 kg (Alexander 1985). However,
brae may be higher or lower; and the tissue there are several reasons for taking into account
replaced by the intraosseous diverticula may the mass saved by skeletal pneumatization. The
have been more or less dense. The extraskeletal first is that estimating the mass of extinct ani-
diverticula have not been accounted for at all, mals is fraught with uncertainty, but we should
although they were certainly extensive in linear account for as many sources of error as possible,
terms and were probably voluminous as well. and PSP is a particularly large source of error if
Uncertainties aside, it seems likely that the ver- it is not considered. Also, the range of mass esti-
tebrae contained a large volume of air, possibly mates for certain taxa may be very wide, but
1,000 liters or more if the very tall neural 8%–10% of the body mass is still a sizable frac-
spines are taken into account. This air mainly tion when applied to any one estimate. The
replaced dense bony tissue, so skeletal pneuma- entire neck and head account for about the same
tization may have lightened the animal by up to percentage of mass in volumetric studies
10%—and that does not include the extraskele- (Alexander 1989; Paul 1997), so failing to
tal diverticula or pulmonary air sacs. In the account for PSP may be as gross an error as
example presented here, the volume of air in omitting the neck and head from the volumetric
the body of Diplodocus is calculated to have model. These are the purely methodological rea-
replaced about 3,000 kg of tissue that would sons for considering the effect of PSP on body
have been present if the animal were solid. If mass. There is also the paleobiological consider-
the total volume of the body was 15,000 liters ation, which is that the living animal was
and the density of the remaining tissue was 8%–10% lighter because of PSP than it would
1 g/cm3, the body mass would have been about have been without. Mass reduction of this
magnitude almost certainly carries a selective SG of the neck may have been as low as 0.6,
advantage (Currey and Alexander 1985), and which would bring the mass of the neck down
this may explain the presence of extensive PSP to about 800 kg (the same result could be
in many sauropods. obtained by applying the air volumes in table
An alternative possibility is that sauropod 7.3 to a more slender neck model than that used
skeletons weighed as much as they would have by Alexander). As the mass of the neck goes
in the absence of PSP but that pneumatization down, so to does the perceived need for a large
allowed the elements to be larger and stronger “nuchal” ligament, the existence of which is
for the same mass. This hypothesis was first controversial (see Wedel et al. 2000; Dodson
articulated by Hunter (1774) to explain skeletal and Harris 2001; Tsuihiji 2004).
pneumatization in birds. It is supported by the Recognition of skeletal pneumaticity in
observation that the skeletons of birds are not sauropods may also affect physiological calcula-
significantly lighter than the skeletons of com- tions. For example, most published studies of
parably sized mammals (Prange et al. 1979). If thermal conductance in dinosaurs (e.g., Spotila
this hypothesis is correct, pneumatic elements et al. 1973, 1991) have modeled dinosaur bodies
should be noticeably larger and more volumi- using solid cylinders. Air is a better insulator
nous than nonpneumatic elements. The transi- than conductor, but moving bodies of air may
tions from pneumatic to apneumatic regions of cool adjacent tissues by convection or evapora-
the vertebral column in Jobaria (Sereno et al. tion. The pneumatic diverticula of birds tend to
1999:fig. 3) and Diplodocus (Osborn 1899:fig. be blind-ended tubes except where they anasto-
13; Gilmore 1932:fig. 3, pl. 6) are not marked by mose (Cover 1953), and most are poorly vascu-
obvious changes in size or form of the verte- larized (Duncker 1971), so there appears to be
brae. This supports the hypotheses that pneu- little potential for evaporative cooling. On the
matic vertebrae were lighter than apneumatic other hand, thermal panting is an important
vertebrae and that PSP really did lighten sauro- homeostatic mechanism for controlling body
pod skeletons. temperature in birds and depends on evapora-
tion from nasal, buccopharyngeal, and upper
PALEOBIOLOGICAL IMPLICATIONS tracheal regions (Lasiewski 1972; Menaum and
The importance of PSP for sauropod paleobiol- Richards 1975). At the very least, the inclusion
ogy is still largely unexplored. To date, of tracheae, lungs and pneumatic diverticula in
Henderson’s (2004) study of sauropod buoy- thermal conductance models would decrease
ancy is the only investigation of the biome- the effective radius of some of the constituent
chanical effects of PSP. Henderson included cylinders. What effect, if any, this would have
pneumatic diverticula in and around the verte- on the results of thermal conductance studies is
brae in his computer models of sauropods, and unknown, which is precisely the point: it has
found that floating sauropods were both highly not been tested.
buoyant and highly unstable. Pneumaticity may
also be important in future studies of neck sup-
PROBLEMS AND PROSPECTS FOR
port in sauropods. Alexander (1985, 1989) cal-
FURTHER RESEARCH
culated that a large elastin ligament would be
better suited than muscles to holding up the Despite a long history of study, research on PSP
neck of Diplodocus. His calculations were based is, in many ways, still in its infancy. Anyone
on a volumetric estimate of 1,340 liters (and, who doubts the accuracy of this statement is
thus, 1,340 kg) for the neck and head. Using the directed to Hunter (1774). In the first published
values in table 7.3, one fifth of that volume, or study of PSP, Hunter developed two of the
268 liters, was occupied by airspaces. If Paul major functional hypotheses entertained today:
(1997) and Henderson (2004) are correct, the pneumaticity may lighten the skeleton, or it
may strengthen the skeleton by allowing bones tical problems. Most medical CT scanners have
of larger diameter for the same mass as marrow- apertures 50 cm or less in diameter, and many
filled bones (see Witmer [1997], for a historical sauropod vertebrae are simply too big to fit
perspective on these and other hypotheses). through the scanners. Furthermore, medical
Although many later authors have documented scanners are not designed to image large, dense
the presence and extent of PSP in certain birds objects like sauropod bones. The relatively low-
(e.g., Crisp 1857; King 1957), most have focused energy X rays employed by medical scanners
on one or a few species (but see O’Connor may fail to penetrate large bones, and this can
2004), some have produced conflicting accounts produce artifacts in the resulting images
(reviewed by King 1957), and few have attempted (Wedel et al. 2000). Industrial CT scanners can
to test functional hypotheses (but see Warncke image denser materials, but the rotating plat-
and Stork 1977; Currey and Alexander 1985; forms used in many industrial scanners are too
Cubo and Casinos 2000; O’Connor 2004). small to accept most sauropod vertebrae. For
Evolutionary patterns of PSP in birds are diffi- the near-future, CT will likely remain a tool of
cult to discern because few species have been great promise but limited application.
studied (King, 1966), usually with little or no Cut sections of bones can yield valuable
phylogenetic context (O’Connor 2002, 2004). information about pneumatic internal struc-
Limits of knowledge of PSP in extant vertebrates tures. The cuts may be made in the field to
necessarily limit what can be inferred from break aggregates of bones into manageable
the fossil record. For example, disagreements pieces, as in the cut Sauroposeidon vertebra
between various published accounts of the shown in fig. 7.4. Less commonly, bones may
development of pneumatization in birds frus- be deliberately cut to expose their cross sections
trate attempts to infer the ontogenetic develop- or internal structures, such as the cut speci-
ment of PSP in sauropods (Wedel 2003a). mens illustrated by Janensch (1947:fig. 5) and
Another problem for studies of PSP in fos- Martill and Naish (2001:pl. 32). Cutting into
sil organisms is small sample sizes. As men- specimens is invasive and potentially danger-
tioned above, few taxa have been intensively ous to both researchers and fossils. Although
studied and the importance of serial, ontoge- cut specimens will continue to appear from time
netic, and intraspecific variation is difficult to to time, they are unlikely to become a major
assess. Sample sizes are mainly limited by the source of data. In contrast, broken bones are
inherent attributes of the fossils: fossilized ubiquitous. The delicate structure of pneumatic
bones are rare, at least compared to the bones bones, even large sauropod vertebrae, may make
of extant vertebrates; they may be crushed or them more prone to breakage than apneumatic
distorted; and they are often too large, too bones. For these reasons broken bones are an
heavy, or too fragile to be easily manipulated. important resource in studies of PSP and could
Even if these difficulties are overcome, most of be exploited more in the future. Published illus-
the pneumatic morphology is still inaccessible, trations of broken sauropod vertebrae are numer-
locked inside the bones. ous; notable examples include Cope (1878:fig. 5),
Hatcher (1901:pl. 7), Longman (1933:pl. 16, fig. 3),
SOURCES OF DATA and Dalla Vecchia (1999:figs. 2, 19). A beautiful
Information on the internal structure of fossil example from outside Sauropoda is the broken
bones comes from three sources: CT studies, transverse process of Tyrannosaurus illustrated by
cut sections of bones, and broken bones. Brochu (2003:fig. 75).
Although CT studies of fossils are becoming
more common, access to scanners is very lim- DIRECTIONS FOR FUTURE RESEARCH
ited and can be prohibitively expensive. Large Four attributes of pneumatic bones are listed
fossils, such as sauropod vertebrae, cause logis- above under “Description of Pneumatic Elements”:
(1) external pneumatic features, (2) internal struc- reasons are unknown, in part because the phys-
ture, (3) ASP, and (4) distribution of pneumatic- iological functions—or exaptive effects (sensu
ity in the skeleton. Only the second attribute has Gould and Vrba 1982)—of diverticula remain
been systematically surveyed in sauropods obscure. Three important physiological ques-
(Wedel 2003b), although aspects of the first are tions that could be answered with existing
treated by Wilson (1999). Knowledge of the methods are: (1) What volume of air is con-
fourth is mainly limited to the observation that tained in the diverticula in life? (2) What is the
diplodocines and saltasaurines have pneumatic rate of diffusion of air into and out of blind-
caudal vertebrae and other sauropods do not ended diverticula? and (3) In cases where diver-
(Wedel 2003b). All existing data on the ASPs of ticula of different air sacs anastomose, is air
sauropod vertebrae are presented in table 7.2. actively circulated through the resulting loops?
Not only do all four attributes need further Finally, more work is needed on the origins
study, but the levels of serial, ontogenetic, and of PSP; if nothing else, Gower’s (2001) uncon-
intraspecific variation should be assessed ventional hypothesis has drawn attention to this
whenever possible. Similar data on PSP in need. Potential projects include histological and
pterosaurs, nonavian theropods, and birds are biomechanical studies to assess the structure
needed to test phylogenetic and functional and functions of vertebral laminae (Wilson
hypotheses. 1999). In addition, criteria for distinguishing
The pneumatic diverticula of birds are mor- the osteological traces of adipose deposits, mus-
phologically and morphogenetically intermedi- cles, vascular structures, and pneumatic diver-
ate between the core respiratory system of ticula are badly needed for the interpretation of
lungs and air sacs and the pneumatic bones. potentially pneumatic features in fossil bones.
Understanding the development, evolution, This problem is the subject of ongoing research
and possible functions of diverticula is thereby O’Connor (1999, 2001, 2002).
fore crucial for interpreting patterns of PSP in
fossil vertebrates. Müller (1907), Richardson
CONCLUSIONS
(1939), Cover (1953), King (1966), Duncker
(1971) and a few others described the form and The best evidence for pneumaticity in a fossil
extent of the diverticular network in the few element is the presence of large foramina that
birds for which they are known, but informa- lead to internal chambers. Based on this crite-
tion on many bird species is lacking or has been rion, pneumatic diverticula were present in the
inadequately documented (King 1966). The vertebrae of most sauropods and in the ribs of
ontogenetic development of the diverticula is some. Vertebral laminae and fossae were clearly
very poorly understood; most of what we think associated with pneumatic diverticula in most
we know is based on inferences derived from eusauropods, but it is not clear whether this was
patterns of skeletal pneumatization (Hogg the case in more basal forms. Measurements of
1984a; McClelland 1989b). Such inferences tell vertebral cross sections indicate that, on aver-
us nothing about the development of the many age, pneumatic sauropod vertebrae were
visceral, intermuscular, and subcutaneous 50%–60% air, by volume. Taking skeletal pneu-
diverticula that do not contact the skeleton or maticity into account may reduce mass esti-
pneumatize any bones. These diverticula could mates of sauropods by up to 10%. Although the
not have evolved to pneumatize the skeleton. functional and physiological implications of
Most diverticula that pneumatize the skeleton pneumaticity in sauropods and other archosaurs
must grow out from the core respiratory system remain largely unexplored, most of the out-
before they reach their “target” bones, so they standing problems appear tractable, and there
probably also evolved for reasons other than is great potential for progress in future studies
skeletal pneumatization (Wedel 2003a). Those of pneumaticity.
ACKNOWLEDGMENTS Senate, and Department of Zoology and the
This work is dedicated to Jack McIntosh, the University of California Museum of Paleontology
dean of sauropod workers, whose generosity and Department of Integrative Biology. This is
and enthusiasm continually inspire me. I am University of California Museum of Paleontology
grateful to my former and current advisors, contribution no. 1841.
Richard Cifelli, Bill Clemens, and Kevin Padian,
for their encouragement and sound guidance LITERATURE CITED
over many years. This work was completed as
part of a doctoral dissertation in the Department Alexander, R. McN. 1985. Mechanics of posture and
gait of some large dinosaurs. Zool. J. Linn. Soc.
of Integrative Biology, University of California,
83: 1–25.
Berkeley. In addition to my advisors, I thank the Alexander, R. McN. 1989. Dynamics of Dinosaurs
other members of my dissertation committee, and Other Extinct Giants. Columbia University
F. Clark Howell, David Wake, and Marvalee Press, New York. 167 pp.
Wake, for sound advice and enlightening dis- Anderson, J. F., Hall-Martin, A., and Russell, D. A.
1985. Long-bone circumference and weight in
cussions. Portions of this work are based on a
mammals, birds and dinosaurs. J. Zool. 207:
CT study conducted in collaboration with R. 53–61.
Kent Sanders, without whose help I would be Bonaparte, J. F. 1986. The early radiation and phylo-
nowhere. I thank the staff of the University of genetic relationships of the Jurassic sauropod
Oklahoma Hospital, Department of Radiology, dinosaurs, based on vertebral anatomy. In:
for their cooperation, expecially B. G. Eaton for Padian, K. (ed.). The Beginning of the Age of
Dinosaurs. Cambridge University Press.
access to CT facilities and Thea Clayborn,
Cambridge. Pp. 247–258.
Kenneth Day, and Susan Gebur for performance Bonaparte, J. F., Heinrich, W. -D., and Wild, R. 2000.
of the scans. I thank David Berman, Michael Review of Janenschia Wild, with the description of
Brett-Surman, Brooks Britt, Sandra Chapman, a new sauropod from the Tendaguru beds of
Jim Diffily, Janet Gillette, Wann Langston, Jr., Tanzania and a discussion on the systematic
value of procoelous caudal vertebrae in the
Paul Sereno, Derek Siveter, Ken Stadtman, Dale
Sauropoda. Palaeontographica A 256: 25–76.
Winkler, and the staff of the Leicester City Brattstrom, B. H. 1959. The functions of the air sac in
Museum for access to specimens in their care. snakes. Herpetologica 15: 103–104.
Translations of critical papers were made by Will Bremer, J. L. 1940. The pneumatization of the
Downs, Nancie Ecker, and Virginia Tidwell and humerus in the common fowl and the associated
obtained courtesy of the Polyglot Paleont- activity of theelin. Anat. Rec. 77: 197–211.
Britt, B. B. 1993. Pneumatic Postcranial Bones in
ologist Web site (http://www.uhmc.sunysb.edu/
Dinosaurs and Other Archosaurs. Ph.D. disserta-
anatomicalsci/paleo). A translation of Janensch tion. University of Calgary, Calgary. 383 pp.
(1947) was made by Gerhard Maier, whose effort Britt, B. B. 1997. Postcranial pneumaticity. In: Currie,
is gratefully acknowledged. I thank Pat P. J., and Padian, K. (eds.). The Encyclopedia
O’Connor for many inspiring conversations of Dinosaurs. Academic Press, San Diego, CA.
Pp. 590–593.
and for gracious access to unpublished data
Britt, B. B. Makovicky, P. J., Gauthier, J., and Bonde, N.
and papers in press. Lauryn Benedict, Rebecca 1998. Postcranial pneumatization in Archaeopteryx.
Doubledee, Andrew Lee, Anne Peattie, and Nature 395: 374–376.
Michael P. Taylor read early drafts of this chap- Brochu, C. A. 2003. Osteology of Tyrannosaurus rex:
ter and made many helpful suggestions. I am insights from a nearly complete skeleton and high-
especially grateful to Pat O’Connor and an resolution computed tomographic analysis of the
skull. Soc. Vertebr. Paleontol. Mem. 7: 1–138.
anonymous reviewer for thoughtful comments
Buffetaut, E., Suteethorn, V., Cuny, G., Tong, H., Le
that greatly improved the manuscript. Funding Loeuff, J., Khansuba, S., and Jongautchariyakal, S.
was provided by grants from the University of 2000. The earliest known sauropod dinosaur.
Oklahoma Graduate College, Graduate Student Nature 407: 72–74.
Chandra Pal and Bharadwaj, M. B. 1971. Histological Gould, S. J., and Vrba, E. S. 1982. Exaptation—A miss-
and certain histochemical studies on the respira- ing term in the science of form. Paleobiology 8: 4–15.
tory system of chicken. II. Trachea, syrinx, brochi Gower, D. J. 2001. Possible postcranial pneumaticity in
and lungs. Indian J. Anim. Sci. 41: 37–45. the last common ancestor of birds and crocodilians:
Colbert, E. H. 1962. The weights of dinosaurs. Am. evidence from Erythrosuchus and other Mesozoic
Mus. Novitates 2076: 1–16. archosaurs. Naturwissenschaften 88: 119–122.
Cope, E. D. 1877. On a gigantic saurian from the Dakota Hall, B. K. 1999. Evolutionary Developmental Biology.
Epoch of Colorado. Palaeontol. Bull. 25: 5–10. 2nd ed. Kluwer Academic, Norwell, MA. 491 pp.
Cope, E. D. 1878. On the saurians recently discovered Hatcher, J. B. 1901. Diplodocus (Marsh): its osteology,
in the Dakota beds of Colorado. Am. Nat. 12: 71–85. taxonomy, and probable habits, with a restoration
Cover, M. S. 1953. Gross and microscopic anatomy of of the skeleton. Mem. Carnegie Mus. 1: 1–63.
the respiratory system of the turkey. III. The air Hatcher, J. B. 1903. Osteology of Haplocanthosaurus,
sacs. Am. J. Vet. Res. 14: 239–245. with a description of a new species, and remarks
Cranford, T. W., Amundin, M., and Norris, K. S., on the probable habits of the Sauropoda, and the
1996. Functional morphology and homology in age and origin of Atlantosaurus beds. Mem.
the odontocete nasal complex: implications for Carnegie Mus. 2: 1–72.
sound generation. J. Morphol. 228: 223–285. Hazlehurst, G. A., and Rayner, J. M. V. 1992. Flight
Crisp, E. 1857. On the presence or absence of air in characteristics of Triassic and Jurassic Pterosauria:
the bones of birds. Proc. Zool. Soc. London 1857: an appraisal based on wing shape. Paleobiology
215–200. 18: 447–463.
Cubo, J., and Casinos, A. 2000. Incidence and mechan- Henderson, D. M. 1999. Estimating the masses and
ical significance of pneumatization in the long centers of mass of extinct animals by 3-D mathe-
bones of birds. Zool. J. Linn. Soc. 130: 499–510. matical slicing. Paleobiology 25: 88–106.
Currey, J. D., and Alexander, R. McN. 1985. The thick- Henderson, D. M. 2004. Tipsy punters: sauropod
ness of the walls of tubular bones. J. Zool. 206: dinosaur pneumaticity, buoyancy and aquatic
453–468. habits. Proc. Biol. Sci. 271 (Suppl.): S180–S183.
Curtice, B., and Stadtman, K. 2001. The demise of Hinds, D. S., and Calder, W. A. 1971. Tracheal dead space
Dystylosaurus edwini and a revision of Supersaurus in the respiration of birds. Evolution 25: 429–440.
vivianae. Mesa Southwest Mus. Bull. 8: 33–40. Hogg, D. A. 1980. A comparative evaluation of meth-
Dalla Vecchia, F. M. 1999. Atlas of the sauropod ods for identification of pneumatization in the
bones from the Upper Hauterivian–Lower avian skeleton. Ibis 122: 359–363.
Barremian of Bale/Valle (SW Istria, Croatia). Hogg, D. A. 1984a. The distribution of pneumatisa-
Natura Nacosta 18: 6–41. tion in the skeleton of the adult domestic fowl. J.
Dodson, P., and Harris, J. D. 2001. Necks of sauro- Anat. 138: 617–629.
pod dinosaurs: Support for a nuchal ligament? Hogg, D. A. 1984b. The development of pneumatisa-
J. Morphol. 248: 224. tion in the postcranial skeleton of the domestic
Duellman, W. E., and Trueb, L. 1986. Biology of fowl. J. Anat. 139: 105–113.
Amphibians. McGraw–Hill, New York. 670 pp. Hunter, J. 1774. An account of certain receptacles of
Duncker, H. -R. 1971. The lung air sac system of air, in birds, which communicate with the lungs,
birds. Adv. Anat., Embryol. Cell Biol. 45: 1–171. and are lodged both among the fleshy parts and
Francillon-Vieillot, H., de Buffrénil, V., Castanet, J., in the hollow bones of those animals. Philos.
Géraudie, J., Meunier, F. J., Sire, Y., Zylberberg, Trans. Roy. Soc. London 64: 205–213.
L., and de Ricqlés, A. 1990. Microstructure and Jain, S. L., Kutty, T. S., Roy-Chowdhury, T. K., and
mineralization of vertebrate skeletal tissues. In: Chatterjee, S. 1979. Some characteristics of
Carter, J. G. (ed.). Skeletal Biomineralization: Barapasaurus tagorei, a sauropod dinosaur from
Patterns, Processes and Evolutionary Trends. the Lower Jurassic of Deccan, India. Proceedings
Vol. 1. Van Nostrand Reinhold, New York. of the IV International Gondwana Symposium,
Pp. 471–548. Calcutta, Vol. 1. Pp. 204–216.
Gilmore, C. W. 1925. A nearly complete articulated Janensch, W. 1947. Pneumatizitat bei Wirbeln von
skeleton of Camarasaurus, a saurischian dinosaur Sauropoden und anderen Saurischien. Palaeonto-
from the Dinosaur National Monument, Utah. graphica (Suppl. 7) 3: 1–25.
Mem. Carnegie Mus. 10: 347–384. Janensch, W. 1950. Die Wirbelsaule von Brachiosaurus
Gilmore, C. W. 1932. On a newly mounted skeleton of brancai. Palaeontographica (Suppl. 7) 3: 27–93.
Diplodocus in the United States National Museum. King, A. S. 1957. The aerated bones of Gallus domesti-
Proc. U. S. Natl. Mus. 81: 1–21. cus. Acta Anat. 31: 220–230.
King, A. S. 1966. Structural and functional aspects of and vascular features. J. Vertebr. Paleontol. 21:
the avian lungs and air sacs. Int. Rev. Gen. Exp. 84A.
Zool. 2: 171–267. O’Connor, P. M. 2002. Pulmonary pneumaticity in
Lasiewski, R. C. 1972. Respiratory function in birds. non-dinosaurian archosaurs with comments on
In: Farmer, D. S., and King, J. R. (eds.). Avian Erythrosuchus and distal forelimb pneumaticity in
Biology. Vol. II. Academic Press, New York. pterosaurs. J. Vertebr. Paleontol. 22: 93A.
Pp. 287–342. O’Connor, P. M. 2004. Pulmonary pneumaticity in
Longman, H. A. 1933. A new dinosaur from the the postcranial skeleton of extant Aves: a case
Queensland Cretaceous. Mem. Queensland Mus. study examining Anseriformes. J. Morphol. 261:
10: 131–144. 141–161.
Lovelace, D., Wahl, W. R., and Hartman, S. A. 2003. Ojala, L. 1957. Pneumatization of the bone and envi-
Evidence for costal pneumaticity in a diplodocid ronmental factors: experimental studies on chick
dinosaur (Supersaurus vivianae). J. Vertebr. humerus. Acta Oto-Laryngol. Suppl. 133: 1–28.
Paleontol. 23: 73A. Osborn, H. F. 1899. A skeleton of Diplodocus. Mem.
Lull, R. S. 1911. Systematic paleontology of the Lower Am. Mus. Nat. Hist. 1: 191–214.
Cretaceous deposits of Maryland: Vertebrata. Ostrom, J. H., and McIntosh, J. S. 1966. Marsh’s
Lower Cretaceous Volume, Maryland Geological Dinosaurs: The Collections from Como Bluff.
Survey. Pp. 183–211. Yale University Press, New Haven, CT. 388 pp.
Madsen, J. H., and Welles, S. P. 2000. Ceratosaurus Paul, G. S. 1988. The brachiosaur giants of the
(Dinosauria, Theropoda): a revised osteology. Morrison and Tendaguru with a description of a
Utah Geol. Surv. Misc. Publ. 00-2: 1–80. new subgenus, Giraffatitan, and a comparison
Marsh, O. C. 1877. Notice of new dinosaurian reptiles of the world’s largest dinosaurs. Hunteria 2(3):
from the Jurassic Formation. Am. J. Sci. 14: 514–516. 1–14.
Martill, D. M., and Naish, D. 2001. Dinosaurs of the Paul, G. S. 1997. Dinosaur models: the good, the bad,
Isle of Wight. Palaeontological Association, and using them to estimate the mass of
London. 433 pp. dinosaurs. Dinofest Int. Proc. 1997: 129–154.
Martin, V. 1994. Baby sauropods from the Sao Khua Powell, J. E. 1992. Osteología de Saltasaurus loricatus
Formation (Lower Cretaceous) in northeastern (Sauropoda–Titanosauridae) del Cretácico
Thailand. GAIA 10: 147–153. Superior del noroeste Argentino. In: Sanz, J. L.,
McClelland, J. 1989a. Larynx and trachea. In King, A. and Buscalioni, A. D. (eds.). Los Dinosaurios y Su
S., and McClelland, J. (eds.). Form and Function in Entorno Biotico: Actas del Segundo Curso de
Birds. Vol. 4. Academic Press, London. Pp. 69–103. Paleontología en Cuenca. Instituto Juan de
McClelland, J. 1989b. Anatomy of the lungs and air Valdes, Cuenca, Argentina Pp. 165–230.
sacs. In King, A. S., and McClelland, J. (eds.). Prange, H. D., Anderson, J. F., and Rahn, H. 1979.
Form and Function in Birds. Vol. 4. Academic Scaling of skeletal mass to body mass in birds
Press, London. Pp. 221–279. and mammals. Am. Nat. 113: 103–122.
McIntosh, J. S. 1990. Sauropoda. In Weishampel, Rasband, W. 2003. Image J. National Institutes of
D. B., Dodson, P., and Osmolska, H. (eds.). The Health, Bethesda, MD (http://rsb.info.nih.gov/ij/).
Dinosauria. University of California Press, Reid, R. E. H. 1996. Bone histology of the Cleveland-
Berkeley, CA. Pp. 345–401. Lloyd dinosaurs and of dinosaurs in general Part 1:
McShea, D. W. 1996. Metazoan complexity: Is there a Introduction: introduction to bony tissues.
trend? Evolution 50: 477–492. Brigham Young Univ. Geol. Stud. 41: 25–72.
Menaum, B., and Richards, S. A. 1975. Observations Richardson, F. 1939. Functional aspects of the pneu-
on the sites of respiratory evaporation in the fowl matic system of the California brown pelican.
during thermal panting. Resp. Physiol. 25: 39–52. Condor 41: 13–17.
Müller, B. 1907. The air-sacs of the pigeon. Riggs, E. S. 1904. Structure and relationships of the
Smithson. Misc. Collect. 50: 365–420. opisthocoelian dinosaurs. Part II: The
Norman, D. 1985. The Illustrated Encyclopedia of Brachiosauridae. Field Columbian Mus. Publ.
Dinosaurs. Crescent Books, New York. 208 Pp. Geol. 2: 229–247.
O’Connor, P. M. 1999. Postcranial pneumatic fea- Romer, A. S. 1966. Vertebrate Paleontology. 3rd ed.
tures and the interpretation of respiratory University of Chicago Press, Chicago, IL. 491 pp.
anatomy from skeletal specimens. J. Vertebr. Russell, D. A., and Zheng, Z. 1994. A large mamen-
Paleontol. 19: 67A. chisaurid from the Junggar Basin, Xinjiang,
O’Connor, P. M. 2001. Soft-tissue influences on People’s Republic of China. Can. J. Earth Sci. 30:
archosaurian vertebrae: interpreting pneumatic 2082–2095.
Russell, D. A., Beland, P., and McIntosh, J. S. 1980. from the Middle Jurassic of England. J. Vertebr.
Paleoecology of the dinosaurs of Tendaguru Paleontol. 23: 208–231.
(Tanzania). Mem. Soc. Geol. France. 59: 169–175. Warncke, G., and Stork, H. -J. 1977. Biostatische und
Sadler, D. J., Doyle, G. J., Hall, K., and Crawford, P. J. thermoregulatorische Funktion der Sandwich-
1996. Craniocervical bone pneumatisation. Strukturen in der Schädeldecke der Vögel. Zool.
Neuroradiology 38: 330–332. Anzeiger 199: 251–257.
Sanz, J. L., Powell, J. E., LeLoeuff, J., Martinez, R., and Wedel, M. J. 2003a. Vertebral pneumaticity, air sacs,
Pereda Superbiola, X. 1999. Sauropod remains and the physiology of sauropod dinosaurs.
from the Upper Cretaceous of Laño (northcentral Paleobiology 29: 243–255.
Spain). Titanosaur phylogenetic relationships. Wedel, M. J. 2003b. The evolution of vertebral pneu-
Estud. Mus. Cie. Nat. Alava 14(Numero Esp. 1): maticity in sauropod dinosaurs. J. Vertebr.
235–255. Paleontol. 23: 344–357.
Schepelmann, K. 1990. Erythropoietic bone marrow Wedel, M. J. 2004. The origin of postcranial skeletal
in the pigeon: development of its distribution and pneumaticity in dinosaurs. Proceedings of the
volume during growth and pneumatization of 19th International Congress of Zoology, Beijing,
bones. J. Morphol. 203: 21–34. China Zoological Society. Pp. 443–445.
Schmidt-Nielsen, K. 1983. Animal Physiology: Wedel, M. J. Cifelli, R. L., and Sanders, R. K. 2000.
Adaptation and Environment. Cambridge Osteology, paleobiology, and relationships of the
University Press, Cambridge. 619 pp. sauropod dinosaur Sauroposeidon. Acta Palaeontol.
Seeley, H. G. 1870. On Ornithopsis, a gigantic animal Polonica 45: 343–388.
of the pterodactyle kind from the Wealden. Ann. Weiglein, A. H. 1999. Development of the paranasal
Mag. Nat. Hist. (Ser. 4) 5: 279–283. sinuses in humans. In: Koppe, T., Nagai, H., and
Sereno, P. C. 1991. Basal archosaurs: phylogenetic Alt, K. W., (eds.). The Paranasal Sinuses of
relationships and functional implications. Soc. Higher Primates. Quintessence, Chicago, IL. Pp.
Vertebr. Paleontol. Mem. 2: 1–53. 35–50.
Sereno, P. C., and Arcucci, A. B. 1994. Dinosaurian Wilson, J. A. 1999. A nomenclature for vertebral lam-
precursors from the Middle Triassic of Argentina: inae in sauropods and other saurischian
Marasuchus lilloensis, gen. nov. J. Vertebr. dinosaurs. J. Vertebr. Paleonotol. 19: 639–653.
Paleontol. 14: 33–73. Wilson, J. A. 2002. Sauropod dinosaur phylogeny:
Sereno, P. C., and Novas, F. E. 1994. The skull and critique and cladistic analysis. Zool. J. Linn. Soc.
neck of the basal theropod Herrerasaurus 136: 217–276.
ischigualastensis. J. Vertebr. Paleontol. 13: 451–476. Wilson, J. A., and Sereno, P. C. 1998. Early evolution
Sereno, P. C., Beck, A. L., Dutheil, D. B., Larsson, H. and higher-level phylogeny of sauropod dinosaurs.
C. E., Lyon, G. H., Moussa, B., Sadleir, R. W., Soc. Vertebr. Paleontol. Mem. 5: 1–68.
Sidor, C. A., Varricchio, D. J., Wilson, G. P., and Witmer, L. M. 1990. The craniofacial air sac system
Wilson, J. A. 1999. Cretaceous sauropods and the of Mesozoic birds (Aves). Zool. J. Linn. Soc. 100:
uneven rate of skeletal evolution among 327–378.
dinosaurs. Science 286: 1342–1347. Witmer, L. M. 1995. The extant phylogenetic bracket
Spotila, J. R., Lommen, P. W., Bakken, G. S., and and the importance of reconstructing soft tissues in
Gates, D. M. 1973. A mathematical model for fossils. In: Thomason, J. J., (ed.). Functional
body temperatures of large reptiles: implications Morphology in Vertebrate Paleontology. Cambridge
for dinosaur ecology. Am. Nat. 107: 391–404. University Press, Cambridge. Pp. 19–33.
Spotila, J. R., O’Connor, P. M., Dodson, P., and Witmer, L. M. 1997. The evolution of the antorbital
Paladino, F. V. 1991. Hot and cold running cavity of archosaurs: a study in soft-tissue recon-
dinosaurs: body size, metabolism and migration. struction in the fossil record with an analysis of
Modern Geol. 16: 203–227. the function of pneumaticity. Soc. Vertebr.
Tsuihiji, T. 2004. The ligament system in the neck of Paleontol. Mem. 3: 1–73.
Rhea americana and its implication for the bifur- Witmer, L. M. 1999. The phylogenetic history of
cated neural spines of sauropod dinosaurs. J. paranasal air sinuses. In: Koppe, T., Koppe,
Vertebr. Paleontol. 24: 165–172. Nagai, H., and Alt, K. W. (eds.). The Paranasal
Upchurch, P. 1998. The phylogenetic relationships Sinuses of Higher Primates. Quintessence,
of sauropod dinosaurs. Zool. J. Linn. Soc. 124: Chicago, IL. Pp. 21–34.
43–103. Yates, A. M. 2003. A new species of the primi-
Upchurch, P., and Martin, J. 2003. The anatomy and tive dinosaur Thecodontosaurus (Saurischia:
taxonomy of Cetiosaurus (Saurischia, Sauropoda) Sauropodomorpha) and its implications for the
systematics of basal dinosaurs. J. Syst. Palaeontol. Young, C. C., and Zhao, X. -J. 1972. [Mamenchi-
1: 1–42. saurus hochuanensis, sp. nov.] Inst. Vertebr.
Young, B. A. 1991. Morphological basis of “growling” Paleontol. Paleoanthropol. Monogr. A 8: 1–30.
in the king cobra, Ophiophagus hannah. J. Exp. (In Chinese.)
Zool. 260: 275–287. Zhang, Y. 1988. [The Middle Jurassic dinosaur fauna
Young, B. A. 1992. Tracheal diverticula in snakes: pos- from Dashanpu, Zigong, Sichuan.] J. Chengdu
sible functions and evolution. J. Zool. 227: 567–583. Coll. Geol. 3: 1–87. (In Chinese.)
eight
The Evolution of Sauropod Locomotion
MORPHOLOGICAL DIVERSITY OF A SECONDARILY QUADRUPEDAL RADIATION
Matthew T. Carrano
fruitful but also have tended to become canal-

S auropod dinosaur locomotion,
like that of many extinct groups, has his-
torically been interpreted in light of potential
ized. In this regard, the words of paleontologist
W. C. Coombs (1975:23) remain particularly apt,
modern analogues. As these analogies—along as much for their still-relevant summary of the
with our understanding of them—have shifted, status quo in sauropod locomotor research as
perspectives on sauropod locomotion have fol- for their warning to future workers:
lowed. Thus early paleontologists focused on the
It is a subtle trap, the ease with which an entire
“whalelike” aspects of these presumably aquatic
reptilian suborder can have its habits and habi-
taxa (e.g., Osborn 1898), reluctantly relinquish- tat preferences deduced by comparison not
ing such ideas as further discoveries began to with all proboscideans, not with the family
characterize sauropod anatomy as more terres- Elephantidae, not with a particular genus or
trial. Although this debate continued for over a even a single species, but by comparison with
century, the essentially terrestrial nature of certain populations of a single subspecies.
Deciding that a particular modern animal is
sauropod limb design was recognized by the
most like sauropods is no guarantee of solving
early 1900s (Hatcher 1903; Riggs 1903). Aside the problem of sauropod behavior.
from a few poorly received attempts (e.g., Hay
1908; Tornier 1909), comparisons have usually Similarly, modern analogues play a limited
been made between sauropods and terrestrial role in illuminating the evolution of sauropod
mammals, rather than reptiles. Particular simi- locomotion. What information may be gleaned
larities were often noted between the limbs of from terrestrial mammals and applied to
sauropods and those of elephants and rhinos sauropods is more likely to inform aspects of
(Holland 1910; Bakker 1971). general limb design, for the simple reason that
With respect to sauropod locomotion, these such information is likely to be rather general
comparisons with proboscideans have been in scope. For example, it has been suggested
229
that sauropods employed a limited locomotor was intrinsic to sauropod biology and was
repertoire relative to other, smaller dinosaurs in established at the outset of sauropod history.”
order to maintain limb safety factors (e.g., Even the most primitive known sauropods were
Wilson and Carrano 1999). This was based on large relative to other dinosaurs, usually reach-
structural limits and behavior changes that had ing at least 5 metric tons; this is often close to
been observed between large- and small-bodied the largest size attained by most other dinosaur
extant mammals (Biewener 1990). clades (e.g., Anderson et al. 1985; Peczkis 1994).
For finer-scale patterns of locomotor evolution Ultimately, some sauropods achieved maxi-
to be understood, data on locomotor morphology mum body sizes exceeding 40 metric tons, with
must first be brought into a phylogenetic context. some estimates suggesting even higher masses
Fortunately, several recent cladistic studies have (e.g., Colbert 1962).
greatly clarified systematic relationships within Not surprisingly, several morphological
Sauropoda (e.g., Upchurch 1995, 1998; Salgado features associated with early sauropod evolu-
et al. 1997; Wilson and Sereno 1998; Curry tion appear to be size-related. This implies that
Rogers and Forster 2001; Wilson 2002), the morphological hallmarks of the earliest
although greater resolution is still needed, and sauropods (and, by extension, the diagnosis of
many taxa have yet to be studied in detail. Sauropoda itself) cannot be entirely separated
Furthermore, the changes apparent within sauro- from the acquisition of large body size.
pod evolution will almost certainly be subtler These features include (1) a columnar,
than those between sauropods and other graviportal limb posture, (2) increased limb
dinosaurs. Therefore interpretations cannot rely bone robusticity, (3) shortened distal limb seg-
solely on general analogies with extant taxa, but ments (Carrano 2001), and (4) increased
must also address the specific differences per- femoral midshaft eccentricity (Wilson and
ceived between different taxa. In this regard, not Carrano 1999; Carrano 2001). In other groups
all aspects of sauropod locomotor morphology of terrestrial amniotes, these features are corre-
may be interpretable, or even explicable, although lated with increased body size (e.g., Alexander
their presence may still be noteworthy. et al. 1979; Scott 1985; Carrano 1998, 1999,
In this chapter, I use a phylogenetic frame- 2001). Most are similarly correlated within
work to analyze sauropod locomotor evolution. other dinosaur clades and across Dinosauria as
This aspect of sauropod biology is approached a whole (Carrano 2001). Additionally, the acqui-
quantitatively with measurements taken directly sition of a graviportal limb posture may be tied
from specimens and is integrated with qualitative to reduction of lower-limb flexion and extension,
morphological observations. I interpret the implying a functional correlation with reduction
resultant patterns in light of the evolution of in the lower-limb extensor attachments sites
body size and quadrupedalism, and note several such as the olecranon and cnemial crest (see
sauropod locomotor specializations. Finally, “Sauropod Locomotor Specializations,” below).
three large ingroups (diplodocoids, basal The simultaneous appearance of these char-
macronarians, and titanosaurians) are described acters at the base of Sauropoda (Upchurch
in greater detail, to illustrate a portion of the 1995, 1998; Wilson and Sereno 1998; Wilson
smaller-scale diversity evident in sauropod loco- 2002) is probably an artifact of an insufficiently
motor morphology. resolved and recorded early sauropod history.
Nonetheless, their close ties to large size in
other vertebrate groups suggest that they may
BODY SIZE AND BODY-SIZE EVOLUTION
be genuinely intercorrelated. In other words,
Body size likely played a central role in sauro- the appearance of more than one of these char-
pod evolution. As Dodson (1990:407) noted, acters within a single clade or large-bodied taxa
“Large size with all its biological implications may be predictable.
230 THE EVOLUTION OF SAUROPOD LOCOMOTION

Limb scaling relationships in dinosaurs dinosaurs with increasing body size (Carrano
were examined by Carrano (2001), although 2001). The apparent disparity between sauropod
sauropods were not highlighted in that study. femora and those of other dinosaurs is an artifact
The results of that analysis demonstrated that of the size discrepancy between them; extraordi-
dinosaur hindlimb bones scale strongly lin- narily large ornithopod and thyreophoran
early regardless of posture, with the only evi- femora show eccentricities comparable to those
dent differences being attributed to changes in of sauropods. Similarly, the proportionally short
body size. Specifically, although hindlimb ele- distal limb segments of sauropods are also mir-
ments of quadrupedal dinosaurs often appear rored among large thyreophorans, ceratopsians,
to be more robust than those of bipedal forms, and ornithopods, as well as large terrestrial
they are similarly robust at any given body size mammals. Furthermore, the length of the distal
(fig. 8.1). Indeed, quadrupedalism is associated limb decreases with increasing body mass in
with an increase in body size over the primitive dinosaurs and terrestrial mammals generally
condition in all dinosaur clades in which it (Carrano 1999, 2001). Thus, sauropod limb scal-
evolves. Dinosaur forelimb elements showed a ing properties are again entirely consistent with
trend opposite to that expected, becoming rela- size-related trends.
tively longer and less robust in quadrupeds. Beyond this, body-size evolution has not
This shift in scaling was interpreted as a received specific attention in sauropods, except
response to the need for a longer forelimb to in the larger context of body-size evolution in
accompany the hindlimb in generating stride dinosaurs (Carrano 2005). Here I used squared-
length. change parsimony to reconstruct ancestral body
Among thyreophorans and ceratopsians, size values using terminal taxon body sizes and
scaling trends are consistent with this overall a composite phylogeny (fig. 8.3). Femoral length
pattern, indicating that hindlimb bones did not and midshaft diameters were used as proxies for
undergo significant proportional changes dur- body size because they scale strongly linearly
ing the transition to quadrupedalism apart with it (see appendix 8.1). Once ancestral values
from those associated with size increases. are reconstructed, ancestor–descendant com-
Forelimb changes are similar to those seen in parisons can be made throughout Sauropoda.
quadrupeds generally (fig. 8.2). “Semibipedal” Specifically, I analyzed evolutionary patterns by
dinosaurs, such as prosauropods and derived comparing each ancestor to its descendants
ornithopods, tend to scale intermediately (including all internal nodes), as well as by com-
between bipeds and quadrupeds in both fore- paring the reconstructed ancestral value for
and hindlimb elements (fig. 8.1). They do not Sauropoda with each terminal taxon (table 8.1).
show significant differences from either but, In addition, I used Spearman rank correlation to
instead, overlap the two postures in forelimb investigate correlations with patristic distance
scaling. The general picture available from from the base of the phylogeny, testing whether
these analyses is one of remarkable uniformity more derived taxa tend to be larger than more
among all dinosaurs regardless of posture, and primitive forms (table 8.2).
of the overwhelming influence of body size Not surprisingly, early sauropod evolution
(versus posture) on limb bone dimensions. is characterized by a steady increase in body
These patterns are apparent whether compar- size from the condition seen in Vulcanodon
isons are made using anteroposterior, mediolat- and larger prosauropods to that in basal
eral, or circumferential femoral dimensions neosauropods, representing at least a dou-
(Carrano 2001). bling in size over some 40 million years (Fig.
Other features of sauropod limbs are also 8.4). This pattern predominates throughout
best viewed as responses to increased body size. basal sauropods and into Neosauropoda, but the
Femoral midshaft eccentricity increases in all pattern within Neosauropoda is more complex.
THE EVOLUTION OF SAUROPOD LOCOMOTION 231

A 2.75
log x = -1.07 + 1.050 * log y; r2 = 0.955 (bipeds)
log femoral anteroposterior diameter

2.50 log x = -0.904 + 1.035 * log y; r2 = 0.902 ("semi-bipeds")
log x = -0.937 + 1.142 * log y; r2 = 0.743 (quadrupeds)
2.25
2.00
1.75
1.50
1.25
1.00
.75
.50
.25
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3
bipeds
log femoral length
quadrupeds
"semi-bipeds"
B 2.50
log humeral anteroposterior diameter
2.25
2.00
1.75
1.50
1.25
1.00
.75
.50
log x = -1.194 + 1.154 * log y; r2 = 0.872 (bipeds)

.25
log x = -0.917 + 1.089 * log y; r2 = 0.818 ("semibipeds")
log x = -0.483 + 0.990 * log y; r2 = 0.735 (quadrupeds)
.00
12 14 16 18 20 22 24 26 28 30 32 3
FIGURE 8.1. Limb bone scaling associated with limb posture in dinosaurs. Reduced major
axis regressions showing bipeds (open circles, solid lines), “semi-bipeds” (filled circles, dotted
lines), and quadrupeds (open squares, dashed lines) from all dinosaur clades. (A) Femoral
anteroposterior diameter versus length; (B) humeral anteroposterior diameter versus length.
There are few scaling differences among the femora of different groups, but increasingly negative
scaling of the humerus accompanies quadrupedalism.
2.2 2.2

A Ceratopsia B Ceratopsia
2.0 2.0
1.8 1.8
1.6 1.6
1.4 1.4
1.2 1.2
1.0 1.0
log x = -0.750 + 0.950 * log y; r2 = 0.919 log x = -1.132 + 1.100 * log y; r2 = 0.941
0.8 0.8
1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0
log femoral length bipeds log humeral length

"semi-bipeds" or
quadrupeds
2.4

2.2 C Ornithopoda 2.2 D Ornithopoda

2.0 2.0
1.8
1.8
1.6
1.6
1.4
1.4
1.2
1.2
1.0
1.0
0.8
0.8
0.6
0.6 0.4
log x = -0.931 + 1.013 * log y; r2 = 0.963 log x = -1.183 + 1.146 * log y; r2 = 0.916
0.4 0.2
1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
log femoral length log humeral length

FIGURE 8.2. Limb bone scaling associated with changes in limb posture in nonsauropod dinosaurs. Reduced major axis re-
gressions showing bipeds (open circles) and either “semibipeds” or quadrupeds (filled circles) in two clades for which the
transition to quadrupedalism is represented. (A) Femoral diameter versus length in Ceratopsia; (B) humeral diameter versus
length in Ceratopsia; (C) femoral diameter versus length in Ornithopoda; (D) humeral diameter versus length in Ornithopoda.
Different postural types scale very similarly within each group.
Diplodocoids and macronarians both show di- body sizes over about 30 million years. This
vergences in body-size evolution, with members trend includes titanosaurians and reaches its
of each group having increased and decreased in nadir among saltasaurines.
size from the primitive condition. The largest This analysis supports previous assertions
members of both clades (Apatosaurus among of body-size increase in sauropods, emphasiz-
diplodocoids, Brachiosaurus, Argyrosaurus, and ing its persistence throughout much of the
Argentinosaurus among macronarians) are simi- clade. However, it also highlights an unappreci-
lar in size, but macronarians reach substantially ated complexity to this pattern, demonstrating
smaller adult body sizes (Saltasaurus, Neuquen- that most body-size increases occur early in
saurus, Magyarosaurus; 1.5–3 metric tons) sauropod evolution and were largely complete
than the smallest diplodocoids (Dicraeosaurus, by the Upper Jurassic neosauropod radiation.
Amargasaurus; 5–10 tons). More notable is the Subsequent decreases are also evident, particu-
steady, consistent decrease in body size among larly among macronarians. These may be a
derived macronarians, from some of the largest response to having reached an upper bound
(e.g., Argentinosaurus; 50 tons) to some of the on body size or represent size-based diversifica-
smallest (e.g., Saltasaurus; 3 tons) sauropod tion in later sauropod lineages. In this study,

THEROPODA
PROSAUROPODA
Isanosaurus
Vulcanodon
Gongxianosaurus
Shunosaurus
Volkheimeria
Datousaurus
Barapasaurus
Kotasaurus
Patagosaurus
Tehuelchesaurus
Omeisaurus
Mamenchisaurus
Lapparentosaurus
Cetiosaurus
Diplodocus
Barosaurus
Apatosaurus
Amphicoelias
Dicraeosaurus
Amargasaurus
Rayososaurus
"Antarctosaurus"
Cetiosauriscus
Haplocanthosaurus
Camarasaurus
Brachiosaurus
"Bothriospondylus"
Euhelopus
Janenschia
Phuwiangosaurus
Andesaurus
Chubutisaurus
Rapetosaurus
Argyrosaurus
Aegyptosaurus
Magyarosaurus
Isisaurus
Opisthocoelicaudia
Ampelosaurus
Laplatasaurus
Rocasaurus
Lirainosaurus
Saltasaurus
Neuquensaurus
FIGURE 8.3. Sauropod phylogeny used for this chapter. Based primarily on Wilson (2002),
but also Salgado et al. (1997), Wilson and Sereno (1998), Upchurch (1998), and Curry
Rogers (pers. comm.). Taxa with dashed lines were not analyzed cladistically but are placed
according to their presumed position in table 13 of Wilson (2002).
TABLE 8.1
Results of Body-Size Analyses for Sauropoda
MEAN SUM SKEW MEDIAN N
A. All ancestor–descendant comparisons

FL 0.014 1.203 0.771 0.006 89 47 42
FAP 0.009 0.641 0.331 0.011 70 38 32
FML 0.013 0.911 0.239 0.008 70 41 29
B. Basal node–terminal taxon comparisons
FL 0.209 10.879 0.588 0.239 52 46 8
NOTE: Results using squared-change parsimony reconstructions based on measurements of femoral length. In A, comparisons are be-
tween each reconstructed ancestral node and each descendant taxon; in B, they are between the basal reconstructed ancestral node for
each clade and all its descendant terminal taxa. These results are drawn from a larger study (Carrano 2005), in which all available dinosaur
taxa were included. Abbreviations: skew, skewness; +, number of positive ancestor–descendant changes; –, number of negative
ancestor–descendant changes; FAP, femoral anteroposterior diameter; FL, femoral length; FML, femoral mediolateral diameter.
TABLE 8.2
Spearman Rank Correlations of Body Size and Patristic Distance for Sauropoda
Z P † z† P†
All Sauropoda
FL 0.104 0.973 0.331 0.106 0.998 0.318
FAP 0.201 1.671 0.095* 0.204 1.697 0.090*
FML 0.219 1.833 0.067* 0.223 1.863 0.063*
Diplodocoidea
FL 0.526 2.409 0.016* 0.519 2.377 0.018*
FAP 0.456 1.766 0.078* 0.449 1.738 0.082*
FML 0.705 2.639 0.008* 0.700 2.618 0.009*
All non-macronarian Sauropoda
FL 0.663 4.499 0.001* 0.662 4.491 0.001*
FAP 0.415 2.490 0.013* 0.413 2.478 0.013*
FML 0.379 2.208 0.027* 0.376 2.193 0.028*
Macronaria
FL 0.718 3.515 0.001* 0.727 3.526 0.001*
FAP 0.205 0.869 0.385 0.211 0.894 0.371
FML 0.552 2.467 0.014* 0.563 2.519 0.012*
NOTE: Abbreviations as in Table 8.1, Note. Daggers (†) indicate values corrected for ties; asterisks (*) indicate significant p values.
Magyarosaurus—which has been described as a weak negative size correlation within

“dwarf” sauropod (Jianu and Weishampel Sauropoda. However, when macronarians are
1999)—is not a unique taxon but instead rep- analyzed separately, they show a significant size
resents the size endpoint for an entire clade of decrease, while other sauropod groups
relatively small taxa. Because smaller (diplodocids and all nonmacronarians) exhibit
sauropods tend to be the end products of their significant size increases (table 8.2).
respective lineages (and therefore have rela- Trend analyses (McShea 1994; Alroy 2000a,
tively high patristic distance values), there is a 2000b; Carrano 2005) suggest that sauropod

3.5
A
3.4
3.3
3.2
log femur length

3.1
3.0
2.9
2.8
2.7
2.6
6 8 10 12 14 16 18 20 22 24 26 28
patristic distance
.25
B
net change (descendant - ancestor)
.20
FIGURE 8.4. Body-size evolu-
tion in Sauropoda, based on
.15
measurements of femur
length. (A) Body size versus pa-
.10
tristic distance, showing in-
creases throughout most of
.05
sauropod evolution, particu-
larly from basal sauropods
.00
(open circles) to diplodocoids
(open squares) and basal
−.05
macronarians (filled circles).
Note the steady size decrease in
−.10
derived macronarians (filled
squares). (B) Net change (de-
−.15
scendant size minus ancestral
size) versus ancestral body size,
−.20
showing that changes during
2.8 2.9 3.0 3.1 3.2 3.3
sauropod evolution are nearly
random (i.e., no correlation). ancestral log femur length
body-size evolution is largely characterized by metric tons, substantially greater than those for
“active” macroevolutionary processes. Two details other dinosaur groups (which tend to be between
of this pattern support such an interpretation: (1) 5 and 10 tons). Their apparent lower size bound,
the range of sauropod body sizes expands approximately 1–3 metric tons, is also consider-
through time (and through the phylogeny), with ably larger than that for other dinosaurs (usually
concomitant loss of smaller taxa as larger taxa between 0.05 and 0.5 tons) (Carrano 2005).
appear; and (2) there is a weak positive correla-
tion between ancestor–descendant changes and
QUADRUPEDALISM
ancestral sizes. The apparent upper size bound
for sauropods (which may well be altered by Quadrupedalism represents a second funda-
subsequent discoveries) is minimally about 50 mental, dominant characteristic of all sauropods,

having been achieved in the earliest known become weight-bearing through consolidation
forms and retained in all known descendants. of the carpus (via fusion and/or loss) and
Much of the appendicular skeleton shows adap- reduction in its mobility, incorporation of the
tations for quadrupedal posture, particularly in metacarpals into a bound unit, and transfor-
the manus and forelimb (McIntosh 1990; mation of the unguals into blunt, hooflike
Upchurch 1995, 1998; Wilson and Sereno structures. Aside from relative lengthening of
1998). These changes include lengthening of the the forelimb, these are the primary features
forelimb elements relative to those of the used to infer quadrupedalism in dinosaur taxa.
hindlimb, as well as morphological modifica- Most of these changes are already present in
tions of the distal forelimb elements associated basal sauropods (e.g., Shunosaurus, Barapa-
with locomotion and weight bearing (see below). saurus), indicating that these animals
Based on such criteria, the primitive sauro- were already fully (and probably obligatorily)
pod Vulcanodon appears to have been fully quadrupedal. The manus of Vulcanodon is
quadrupedal (Cooper 1984). Its more primitive incomplete, but the long humerus and forearm
relatives Blikanasaurus and Isanosaurus entirely suggest that this taxon was also probably fully
lack forelimb materials, so their postural status quadrupedal (Raath 1972; Cooper 1984).
cannot be assessed. The recently described Nevertheless, the transition to permanent
basal sauropod Antetonitrus (Yates and Kitching quadrupedalism likely involved a series of mor-
2003) has a relatively long forelimb but retains phological changes that occurred across a range
prosauropod-like manual elements, suggesting of facultatively quadrupedal taxa. Antetonitrus
that quadrupedalism may have been facultative may represent one such intermediate, combin-
in this taxon. More complete forelimb materials ing a relatively long forelimb with a primitive
are needed to better elucidate its postural and manus (Yates and Kitching 2003).
locomotor capabilities. More derived basal sauropods record the com-
Other outgroup sauropodomorph taxa pletion of this evolutionary process (fig. 8.5).
(prosauropods and stem-sauropodomorphs) all Despite our fragmentary understanding of its
appear to have retained at least facultative early stages, relics of the transition to
bipedalism (e.g., Galton 1990). Thus the transi- quadrupedalism remain in basal sauropods
tion to quadrupedalism in sauropods is only such as Shunosaurus, Barapasaurus, Omeisaurus,
hinted at by the known fossil record. Mamenchisaurus, and “cetiosaurs” (e.g., Lapparen-
Nonetheless, it can be examined indirectly tosaurus, Patagosaurus, Cetiosaurus). Although
through comparison with other dinosaur groups these forms were fully quadrupedal, they docu-
in which the transition is recorded (Ceratopsia, ment the continued acquisition of additional
Thyreophora), along with “semibipedal” taxa characters related to this posture. Their phyloge-
such as prosauropods and derived ornithopods. netic positions help to clarify the original
Among thyreophorans and ceratopsians, sequence of character acquisition in earlier forms
the shift to quadrupedalism is also accompa- by “removing” later-appearing features from the
nied by marked changes in forelimb morphol- early stages of the evolution of sauropod
ogy. Not surprisingly, hindlimb morphology is quadrupedalism.
usually little altered, instead retaining many of The primitive saurischian manus was proba-
the locomotor features already present in the bly digitigrade, as suggested by the posture of
bipedal members of each clade. Most differ- manus prints of otherwise bipedal dinosaurs
ences in the hindlimb can be attributed to (e.g., Lull 1953; Baird 1980). Additionally, other
those associated with concomitant body size dinosaur clades that underwent a transformation
increases (Carrano 2001; see also “Body Size to quadrupedalism also appear to have passed
and Body-Size Evolution,” above). In the fore- through (and sometimes remained at) a phase
limb, however, the manus is modified to involving a digitigrade manus. The major

A
EA
D
PO
ID
IA
us
O
s
AR
O
ru
us
on
C
ur
R
au
N
ur
sa
U
od
O
SA
sa
os
pa
an
LO
R
ei
un
AC
O
ra
lc
IP
m
PR
Vu
Sh
Ba
M
O
D
columnar manus
FIGURE 8.5. The transition to entaxonic pes

quadrupedalism in Sauropoda.
Although basal sauropods were
likely obligate quadrupeds, nu- subunguligrade pes
merous additional modifications
obligatory quadrupedalism?
for this posture were developed
in basal taxa. Phylogeny from
Wilson and Sereno (1998).
FIGURE 8.6. Changes in the metacarpus in Sauropoda. The metacarpals acquire a semicircular, columnar arrangement
within sauropods, with a reduction in length disparities between elements. Proximal (top row) and dorsal (bottom row)
views of the articulated right metacarpus; proximal views are shown with the palm toward the top of the page. (A) Mas-
sospondylus carinatus (modified from Cooper 1981). (B) Shunosaurus lii (modified from Zhang 1988). (C) Brachiosaurus bran-
cai (modified from Janensch 1922; Wilson and Sereno 1998). (D) Janenschia robusta (modified from Janensch 1922; Wilson
and Sereno 1998). Numbers refer to metacarpals I through V. Not to scale.
changes in manus morphology, therefore, Omeisaurus), the intermetacarpal facets are

began from a primitively digitigrade structure already somewhat elongated distally, indicating
in which the individual metacarpals are most that the bones were arranged into a tight,
firmly interconnected proximally, arranged in a amphiaxonic arch while being held nearly verti-
dorsoventral arch, and held at an oblique angle cally relative to the substrate (Osborn 1904;
to the substrate (fig. 8.6). This morphology Gilmore 1946; Christiansen 1997). Coombs
would have served to resist long-axis bending (1975) described this arrangement as mimick-
during flexion and extension at the carpus by ing columns beneath a supported floor (i.e., the
allowing the metacarpals to act together in dis- carpus), an analogy that is probably mechani-
tributing and deflecting these loads. In the cally appropriate. Compressional loading was
most primitive sauropods for which complete likely increased relative to bending in such a
manus remains are known (e.g., Shunosaurus, structure.

A
EA
D
PO
ID
IA
us
O
s
AR
O
ru
us
on
C
ur
R
au
N
ur
sa
U
od
O
SA
sa
os
pa
an
LO
R
ei
un
AC
O
ra
lc
IP
m
PR
Vu
Sh
Ba
M
O
D
laterally flared ilia
FIGURE 8.7. Sauropod locomo-
tor specializations. The acquisi-
ventrally displaced tion of unique morphological
iliac postacetabulum features within sauropods oc-
modified lesser trochanter curred throughout basal taxa,
prior to the divergence of
reduced olecranon and cnemial crest diplodocoids and macronarians.
Phylogeny from Wilson and
Sereno (1998).
Meanwhile, the sauropod pes also under- early in sauropod evolution and were essentially
went significant changes in morphology and in place by the appearance of Neosauropoda
posture. Primitively, the saurischian (and (fig. 8.7).
dinosaurian) pes was a digitigrade structure All sauropods show evidence of reduction
with four primary weight-bearing digits, in the lower-limb muscle attachments. This
an arrangement retained by prosauropods, occurs in both the fore- and the hindlimbs,
primitive theropods (Herrerasaurus, Eoraptor), where the insertion areas for these muscles are
and basal ornithischians (Pisanosaurus, reduced relative to the condition in other
Lesothosaurus). Basal sauropods enlarged digit V saurischians (and other dinosaurs). In the fore-
into an additional support element (Raath 1972; limb, the ulnar olecranon process is reduced to
Cruickshank 1975), representing a reversal to a a flat, rugose surface in primitive sauropods;
more primitive archosaurian condition. These this condition is reversed only in titanosaurians
same taxa also show a shift in metatarsal orien- (Christiansen 1997; Wilson and Sereno 1998).
tation from relatively upright (subvertical) to This process represents the insertion of mus-
more horizontal. Sauropod footprints show evi- culi (mm.) triceps, which are the major forearm
dence that these low-angled metatarsals were extensors (e.g., Christiansen 1997). The
supported by a fleshy pad (Coombs 1975; Gallup hindlimb shows reduction of the cnemial crest
1989), analogous to those of modern pro- on the tibia, representing the insertion for the
boscideans and rhinocerotids, indicating that knee extensors (mm. iliotibiales, musculus [m.]
the sauropod pes was subunguligrade rather ambiens, and mm. femorotibiales). These
than strictly digitigrade. changes suggest a shift in basal sauropods away
from significant use of the lower limb during
locomotion and toward relying primarily on
SAUROPOD LOCOMOTOR
protraction and retraction at the hip and shoul-
SPECIALIZATIONS
der to generate stride length.
Other changes documented in basal sauropod As in other quadrupedal dinosaurs, both the
evolution do not appear to be related to either manus and the pes of sauropods show pha-
quadrupedalism or large body size but, instead, langeal reduction (Osborn 1904; Coombs 1975;
represent unique components of the sauropod Upchurch 1995, 1998; Wilson and Sereno
locomotor apparatus. Like the changes dis- 1998). Usually reduction occurs primarily in
cussed previously, these modifications occurred length, with individual phalanges becoming

TABLE 8.3
Loss of Manual Phalanges in Sauropoda
SPECIMEN PRESERVED RECONSTRUCTED
Theropoda 2-3-4-2*-0
Prosauropoda 2-3-4-3*-2*
Shunosaurus lii ZDM T5402 2-2-2*-2*-2* 2-2-2*-2*-2*
Omeisaurus tianfuensis ZDM T 2-2-?-?-1 2-2-2-2-2*
Jobaria tiguidensis MNN TIG3 2-2-2-2-2 2-2-2-2-2
Apatosaurus louisae CM 3018 2-2*-1-1-1 2-2*-1-1-1
Apatosaurus excelsus CM 563 2-1-1-1-1 2-2*-1-1-1
Camarasaurus sp. FMNH 25120 2-1-1-1-1 2-2*-1-1-1
Camarasaurus sp. AMNH 823 2-1-1-1-1 2-2*-1-1-1
Camarasaurus grandis GMNH-VP 101 2-1-1-1-1 2-2*-1-1-1
Brachiosaurus brancai HMN S II 2-1-1-1-1 2-2*-1-1-1
Janenschia robusta HMN Nr. 5 2-2*-1-1-1 2-2*-1-1-1
Opisthocoelicaudia skarzynskii ZPal MgD-I/48 0-0-0-1*-0 0-0-0-1*-0
Alamosaurus sanjuanensis USNM 15660 0-0-0-0-0 0-0-0-0-0
NOTE: Formulas list numbers of phalanges in order from digit I through digit V. Asterisks (*) indicate vestigial phalanges; question
marks (?) indicate questionable numbers.
compact and often disclike. However, sauropods that this pattern is more complex. For example,
continue this trend to its extreme, eliminating Apatosaurus louisae (CM 3018) retains two pha-
many of the manual and pedal phalanges alto- langes on digit II (Gilmore 1936). Many other
gether. In the manus particularly, this reduction neosauropods (perhaps most nontitanosaurs)
is carried to the point that all the unguals, and retain a distal articular facet on II-1 that implies
nearly all the phalanges, are eventually lost the presence of an additional phalanx, albeit a
(Osborn 1904, 1906; Gilmore 1946; Wilson small one. The vestigial IV-1 in Opisthocoelicaudia
and Sereno 1998; table 8.3). Phalanges typically (Borsuk-Bialynicka 1977) might also be a gen-
serve as points of flexion and extension for the uine relic. Thus, additional stages of reduction
manus, and their loss indicates the loss of may have existed between Neosauropoda and
these functions within this portion of the limb. titanosaurs (table 8.3).
This represents the end point of the transfor- In addition, the primitive dinosaurian pes is a
mation from a propulsive, digitigrade manus mesaxonic structure, with the weight-bearing
to one that was almost entirely dedicated to axis passing down the central digit (III) (fig. 8.8).
columnar support. This condition, in which the central metatarsals
Tracking this set of changes is problematic, are the longest and most robust, is evident in
however, because phalanges are often lost prior prosauropods and in basal sauropods
to discovery of the specimen. This is particularly (Vulcanodon). However, it is altered in more
true of vestigial, nubbinlike phalanges such as derived sauropods (e.g., Shunosaurus, Omei-
those in the manus of Shunosaurus (Zhang saurus) such that the medial metatarsals are
1988). Conservatively, manual phalangeal reduc- larger than the central elements. In particular,
tion appears to occur primarily at two points in metatarsal I becomes the most robust, and the
sauropod evolution. First, most of the phalanges weight-bearing axis shifts to a more medial,
on digits II–V were lost in Neosauropoda, and entaxonic position (Coombs 1975). This type of
then the remainder were lost in derived pedal structure is highly unusual among
titanosaurs. However, some evidence suggests tetrapods, being evident elsewhere among

FIGURE 8.8. Changes in the metatarsus in Sauropoda. The metatarsals change from mesaxonic to entaxonic, as the medial
digits become relatively larger. Proximal (top row) and dorsal (bottom row) views of the articulated left metatarsus; the proxi-
mal views are shown plantar-upward. (A) Massospondylus carinatus (modified from Cooper 1981) . (B) Vulcanodon karibaensis
(modified from Cooper 1984). (C) Shunosaurus lii (modified from Zhang 1988; Wilson and Sereno 1998). (D) Apatosaurus
louisae (modified from Gilmore 1936). Numbers refer to metatarsals I through V. Not to scale.
FIGURE 8.9. Changes in the ilium in Sauropoda. The ilium is expanded anteroposteriorly, and the postacetabulum is de-
pressed ventrally to the level of the ischial peduncle. Right pelves are shown in lateral view. (A) Massospondylus carinatus
(modified from Cooper 1981). (B) Shunosaurus lii (modified from Zhang 1988). (C) Dicraeosaurus hansemanni (modified
from Janensch 1961). (D) Brachiosaurus brancai (modified from Janensch 1961). Abbreviations: ip, ischial peduncle; llf, orig-
ination area for lower limb flexors; pifi, space anterior to ilium for passage of m. puboischiofemoralis internus 2; poa,
postacetabulum; pra, preacetabulum. Not to scale.
megatheriid xenarthrans. This shift may have processes are expanded anteroposteriorly and
accompanied a more general change from nar- dorsoventrally. The preacetabulum is large and
row-bodied, predominantly bipedal taxa in lobate (Raath 1972; Wilson and Sereno 1998),
which the limbs were positioned well under the arching above a large space anteroventral to the
body to wider-bodied quadrupedal taxa in which ilium through which m. puboischiofemoralis
the limbs were more laterally positioned. In internus 2 likely passed (Romer 1923; Carrano
such wider-bodied taxa, the weight-bearing axis and Hutchinson 2002). This space is larger than
would pass through the medial portion of the that seen in neotheropods, implying that m. pub-
pes instead of the central portion, as reflected in oischofemoralis internus 2 did not undergo the
the modified proportions of metatarsals I and II. alterations in size or position seen in that group
The ilium is substantially modified in (Carrano and Hutchinson 2002).
sauropods (fig. 8.9). Primitively (in prosauropods The preacetabular ilium is also flared laterally
and basal theropods), this bone is dorsoventrally in most neosauropods, with the blade curving
narrow and relatively short anteroposteriorly outward from a point just dorsal to the pubic
(Carrano 2000). In all known sauropods, both peduncle. This is carried to an extreme in
the preacetabular and the postacetabular titanosaurians (Powell 1986, 1990, 2003; Wilson

FIGURE 8.10. Changes in the femoral lesser trochanter within Sauropoda. The lesser trochanter is reduced from a distinct
process to a rugose bump and shifts in position from anterolateral to lateral. Right femora are shown in anterior view. (A)
Massospondylus carinatus (modified from Cooper 1981). (B) Vulcanodon karibaensis (modified from Cooper 1984). (C) Ap-
atosaurus excelsus (modified from Riggs 1903). (D) Brachiosaurus brancai (modified from Janensch 1961). Abbreviations: fh,
femoral head; lt, lesser trochanter. Not to scale.
and Sereno 1998), in which the preacetabular and rugosities that mark the origins of several of
ilium reaches laterally nearly as far as the greater these muscles. Therefore, depression of the
trochanter of the femur. A similar condition is postacetabulum would have brought the lower-
seen in ankylosaurs, and in both groups the effect limb flexor origins farther ventrally, reducing
would have been to shift the origins of m. iliotib- their dorsoventral components relative to their
ialis 1 and m. iliofemoralis externus laterally. As a anteroposterior actions.
result, these muscles would have had reduced On the femur, the lesser trochanter was
mediolateral actions relative to their anteroposte- primitively part of the dinosauriform
rior and/or dorsoventral ones. trochanteric shelf, a ridge running around the
The iliac postacetabulum is also large but anterolateral corner of the proximal femur that
extends much farther ventrally than the preac- served as the insertion for m. iliofemoralis
etabulum. In Vulcanodon (Raath 1972; Cooper externus (Hutchinson 2001). In most
1984) and Shunosaurus (Zhang 1988), the dinosaurs, this shelf is differentiated into a dis-
postacetabulum is depressed close to the level of tinct process (the lesser trochanter, for portions
the ischial peduncle, much more so than the of mm. iliotrochanterici) as well as a rugose
condition in prosauropods and theropods. It mound (for m. iliofemoralis externus)
reaches the level of the ischial peduncle in many (Hutchinson 2001; Carrano and Hutchinson
diplodocoids, and actually exceeds it in many 2002). This occurs independently in several
titanosaurians. Several lower-limb flexors origi- dinosaur clades (Carrano 2000), but sauropods
nated on the iliac postacetabulum, including are a persistent exception to this trend (fig.
mm. flexores tibiales internii, m. flexor tibialis 8.10). Instead, although the trochanteric shelf is
externus, and m. iliofibularis (Cooper 1984), modified in sauropods, the lesser trochanter is
although the exact placement and bounds of reduced to a rounded anterolateral ridge in
these muscles are not clear (Carrano and Vulcanodon (Cooper 1984) and a rugose lateral
Hutchinson 2000). In most neosauropods, this bump in other sauropods. The m. iliofemoralis
portion of the ilium often bears complex fossae insertion was not drawn proximally as in other

dinosaurs (Carrano 2000), instead remaining functional) differences among these taxa.
well below the femoral head. Future work in this area is likely to be quite
This transition would have accomplished fruitful.
two things. First, m. iliofemoralis externus
would have been placed laterally, not anteriorly DIPLODOCOIDEA
as in most dinosaurs (Carrano 2000). Thus this Diplodocoids are the sister-group to Macronaria
muscle would have retained a role in femoral within Neosauropoda (fig. 8.3). The group
abduction (and possibly rotation), rather than includes three major clades (Diplodocidae,
femoral protraction. Second, it remained rela- Dicraeosauridae, and Rebbachisauridae) and
tively distal in position and was (apparently) over 25 taxa, of which at least 15 are well known.
never elaborated into separate muscles; no ana- They are generally perceived as smaller and more
logues to mm. iliotrochanterici of theropods slender than other sauropods, based largely on
and birds were present. Therefore sauropods reconstructions of Diplodocus, Amargasaurus,
likely relied on m. puboischiofemoralis inter- and Dicraeosaurus. However, specimens of
nus for femoral protraction, as did most non- Apatosaurus (YPM 1860) and Seismosaurus
theropod dinosaurs. (NMMNH P-3690; possibly synonymous with
Diplodocus [Lucas and Heckert 2000]) rival the
largest macronarians in overall size. (As noted
LOWER-LEVEL PATTERNS IN SAUROPOD
earlier, this gigantism was achieved independ-
LOCOMOTOR EVOLUTION
ently in these two clades.)
Although acquired stepwise in basal sauropods, Diplodocoids have gracile limbs that are
the morphological specializations discussed short relative to the trunk, making them some-
above characterize nearly all known sauropod what low-slung animals (fig. 8.11A). Compared
taxa. Their appearances mark the emplacement to basal macronarians, diplodocoids would have
of the sauropod locomotor apparatus, which had shorter strides relative to their trunk
remained fundamentally unaltered throughout length. A marked discrepancy between the fore-
subsequent neosauropod evolution for the next and the hindlimb lengths exaggerates the con-
100 million years. Yet on this functional and dition typically present in other quadrupedal
morphological foundation was built a diversity dinosaurs. The manus remains relatively short,
of sauropod locomotor habits that has gone similar to that in more primitive sauropods. A
mostly unappreciated. This is largely thanks to few diplodocoids (Diplodocus and particularly
the perceived uniformity of the sauropod Amphicoelias) are notable in having a reduced
appendicular skeleton, which, while genuine at femoral midshaft eccentricity (Osborn and
one scale (and in that sense remarkable), has Mook 1921) compared to other, similar-sized
also helped to obscure smaller-scale patterns in sauropods.
sauropod locomotor evolution (e.g., Hatcher In addition, diplodocoids have a modified
1903:55). Discovering and interpreting lower- metatarsus in which metatarsals III and IV are
level evolutionary patterns require an under- “twinned” in morphology. This contrasts (albeit
standing of smaller-scale differences between subtly) with the condition in other sauropods,
sauropod groups. where these two bones are similar in length but
Here I discuss the differences among distinct from one another in shape. The func-
three major sauropod groups—diplodocoids, tional implications of such a pedal structure are
basal macronarians, and titanosaurians. I do unclear; it may belie some minor locomotor
not seek to present a detailed study of these specialization or simply be a product of
taxa; my goal is rather to demonstrate that increased limb bone slenderness. In other
existing information is sufficient to illustrate respects, however, the diplodocoid appendicu-
marked morphological (and, by inference, lar skeleton is relatively conservative, retaining

FIGURE 8.11. Comparative morphology of three sauropod groups. (A) Diplodocoidea (Apatosaurus louisae); (B) basal
Macronaria (Brachiosaurus brancai); (C) Titanosauria (Opisthocoelicaudia skarzynskii). Skeletal reconstructions modified from
Wilson and Sereno (1998). Approximately to scale.
many sauropod and neosauropod symple- the hindlimbs (again, opposite to the diplodo-
siomorphies. coid condition), a tendency most evident in
Brachiosaurus (FMNH P25107, HMN D) but
BASAL MACRONARIA also seen in other taxa (“Bothriospondylus mada-
“Basal macronarians” are a paraphyletic assem- gascariensis,” MNHN specimen [Lapparent
blage of more than 20 species of taxa including 1943]; Euhelopus, PMU R234 [Wiman 1929];
the Brachiosauridae, Camarasaurus, Euhelopus, Lusotitan atalaiensis, MIGM holotype [Lapparent
Janenschia, and the most basal titanosaurians and Zbyszewski 1957]). In addition, the metacar-
(e.g., Aeolosaurus, Pleurocoelus) (Wilson and pus is elongate relative to that of diplodocids
Sereno 1998; Wilson 2002). Many of these and primitive sauropods, allowing it to con-
forms are large even among sauropods, with tribute increasingly to stride length. Overall,
taxa such as Brachiosaurus, Sauroposeidon, and most of these changes would have allowed the
Argentinosaurus representing some of the forelimb to approach or exceed the hindlimb in
largest known terrestrial vertebrates. its ability to contribute to forward progression.
Basal macronarians are characterized by rel- This represents a reversal of the typical (and
atively gracile limbs that are long relative to the primitive) dinosaurian condition.
trunk (fig. 8.11B). As a result, these taxa would Many of these forms also show a reduction in
have had longer stride lengths compared to the size of the pelvic girdle relative to the pectoral
diplodocoids. The forelimbs are long relative to girdle. This would have reduced the origination

areas for most of the proximal hindlimb mus- interpretation of titanosaurians as the authors
culature, again suggesting a shift in locomotor of “wide-gauge” sauropod trackways (Wilson
dominance toward the forelimb. Some and Carrano 1999).
of the “wide-gauge” features that exemplify Titanosaurians are also characterized by
titanosaurians (see below) are present in these highly flared iliac preacetabular processes
taxa, but to a much lesser degree. For example, (Powell 1986, 1990, 2003; Christiansen 1997),
brachiosaurids show a medial offset of the somewhat analogous to the condition seen in
femoral head, moderate flaring of the preac- ankylosaurs. This flaring would have displaced
etabular ilium, and reduction of the pollex the origins of m. iliofemoralis and m. iliotib-
ungula—all changes eventually carried further ialis 1 more laterally, as noted above (Carrano
by titanosaurians (Wilson and Carrano 1999). 2000). The pelvic elements are relatively small,
Recently, Stevens and Parrish (1999) sug- as in basal macronarians. In addition, the
gested that sauropods generally did not raise ischium is very short relative to the pubis
the neck above the level of the shoulder. The (Wilson and Sereno 1998), and the anterior
resulting clade of long-necked but low-browsing caudals have ball-in-socket articulations, per-
animals stands in marked contrast to conven- haps associated with an increased predomi-
tional reconstructions of sauropods with nance of bipedal rearing in these taxa (Borsuk-
giraffe-like neck postures (e.g., Paul 1988). I do Bialynicka 1977; Wilson and Carrano 1999).
not attempt to evaluate this hypothesis here, Reversing the primitive condition for
but it is worth noting that an additional effect of sauropods, titanosaurian humeri have antero-
forelimb lengthening in basal macronarians posteriorly long distal articular surfaces
would be to raise the height of the shoulder, (Wilson and Sereno, 1998; table 8.4), indicating
thereby elevating the maximum potential a wider range of flexion–extension at the elbow
height of the neck in animals otherwise con- than in other sauropods (Osborn 1900).
strained to low browsing. Thus basal macronar- Although the limb joint surfaces are not well
ians may have been relatively high-browsing modeled and are now missing what were prob-
sauropods thanks to their forelimb adaptations. ably substantial cartilage caps (Hatcher 1901;
Coombs 1975), it is reasonable to assume that
TITANOSAURIA the true articular cartilage would have been at
Titanosaurian sauropods comprise approxi- least as anteroposteriorly extensive as the bony
mately 50 taxa of derived macronarians (Wilson attachment surface. I measured the linear
and Sereno 1998; Curry Rogers and Forster height of the articular surface on the anterior
2001; Wilson 2002) such as Rapetosaurus, face of the distal humerus and compared it
Ampelosaurus, and the Saltasaurinae. Of these, among sauropod groups relative to total
approximately 35 are known from diagnostic humerus length. These differences are signifi-
materials, and many fewer from relatively com- cant when compared using both the unpaired,
plete specimens. As noted above, titanosaurians two-group t-test (t 5.346, df 16; p 0.001)
are unusual in being markedly smaller than and the Mann–Whitney U-test (U 2.000,
most other sauropods, but in addition, they dis- Z 3.184; p 0.015).
play a suite of features that readily distinguish Both fore- and hindlimbs are short relative to
them from other neosauropods (fig. 8.11C). the trunk. The particularly short titanosaurian
These include an outward-canted femur with a forefoot is the most derived among sauropods
correspondingly angled distal articular surface, in having entirely lost the carpus and phalanges
a widened sacrum, and a significantly more (including the manual ungual) or having both
elliptical femoral midshaft relative to body size reduced to cartilage (Gilmore 1946; Salgado et al.
(Wilson and Sereno 1998; Wilson and Carrano 1997; Wilson and Sereno 1998). Whatever func-
1999). Such features are consistent with an tion the enlarged pollex had originally evolved

TABLE 8.4
Humeral Articulation Angles in Sauropods
SPECIMEN HL AHD %
Shunosaurus lii T5401 648 37 5.7

Omeisaurus tianfuensis T5701 790 27 3.4
Mamenchisaurus constructus IVPP V.947 1,080 52 4.8
“Cetiosaurus” mogrebiensis Tamguert n’Tarit 1,370 42 3.1
Barosaurus africanus HMN k 37 970 16 1.6
Dicraeosaurus hansemanni HMN Q 11 740 24 3.2
Camarasaurus grandis YPM 1901 1,288 90 7.0
Brachiosaurus altithorax FMNH P25107 2,040 105 5.1
Brachiosaurus brancai HMN S II 2,130 104 4.9
Euhelopus zdanskyi GSC specimen 1,090 40 3.7
Tehuelchesaurus benitezii MPEF-PV 1125 1,140 41 3.6
Janenschia robusta HMN P 8 890 40 4.5
Gondwanatitan faustoi MN 4111-V 729 71 9.7
Aegyptosaurus baharijensis BSP 1912 VIII 61 1,000 83 8.3
Alamosaurus sanjuanensis USNM 15560 1,360 90 6.6
Opisthocoelicaudia skarzynskii ZPAL MgD-I/48 1,000 58 5.8
Saltasaurus loricatus PVL 4017-67 611 66 10.8
Isisaurus colberti ISIR 335/59 1,480 148 10.0
NOTE: Measurements are in millimeters. Abbreviations: AHD, anterior height of distal articulation; HL, humeral length.
% (AHD/HL) 100.
for in ancestral bipedal saurischians, its conver- conforms to a “passive” rather than an “active”
sion into an entirely locomotor structure was trend (sensu McShea 1994).
complete in titanosaurians. These barrel- Features associated with the acquisition of
chested taxa also exhibit a modified shoulder obligate quadrupedalism—including a relatively
joint (Powell 1986, 2003; Wilson and Sereno long forelimb and supportive manus—are already
1998) that, in conjunction with changes at the present in the primitive sauropod Vulcanodon, but
elbow, suggests increased forelimb mobility and the transition is completed in successively more
flexibility compared to other sauropods. derived basal sauropods. Few scaling changes are
inferred to have accompanied this postural shift
aside from lengthening of the forelimb.
CONCLUSIONS
Other aspects of sauropod limb morphology
The basic sauropod appendicular morphology, are unique to the clade, including the develop-
including its functional relationships to locomo- ment of a columnar, tightly arched manus with
tion, is fundamentally tied to the constraints a reduced phalangeal count. The subun-
imposed by large size on a vertebrate in a terres- guligrade pes is entaxonic, indicating a shift in
trial environment. Many sauropod appendicular primary weight support toward the medial
synapomorphies are also found in other large aspect of the pes. The femoral lesser trochanter
dinosaurs and in large terrestrial mammals, and is fully lateral and reduced to a rugose bump,
their appearance is likely size related. Body size representing the least elaborate such structure
within sauropods increases steadily throughout within Dinosauria. The iliac preacetabulum
the clade, with subsequent size decreases in both (and associated muscle origins) are flared later-
diplodocoids (dicraeosaurids) and macronarians ally, whereas the postacetabulum (and associ-
(saltasaurines). Trend analyses suggest that this ated muscle origins) are depressed ventrally.

These basic aspects of sauropod appendicu- Baird, D. 1980. A prosauropod dinosaur trackway
lar design are never radically altered, but sauro- from the Navajo Sandstone (Lower Jurassic) of
Arizona. In: Jacobs, L. L. (ed.). Aspects of
pod locomotor evolution nevertheless includes
Vertebrate History: Essays in Honor of Edwin
an unappreciated morphological diversity. Harris Colbert. Museum of Northern Arizona
Diplodocoids, basal macronarians, and tita- Press, Flagstaff. Pp. 219–230.
nosaurians each show locomotor specializa- Bakker, R. T. 1971. Ecology of the brontosaurs. Nature
tions that distinguish them from one another. 229: 172–174.
Although many of these differences are subtle Biewener, A. A. 1990. Biomechanics of mammalian
terrestrial locomotion. Science 250: 1097–1103.
relative to those that distinguish sauropods
Borsuk-Bialynicka, M. 1977. A new camarasaurid
from other dinosaurs, they nonetheless sug- sauropod Opisthocoelicaudia skarzynskii gen. n.,
gest that all sauropods did not employ identical sp. n. from the Upper Cretaceous of Mongolia.
locomotor styles or habits. Future studies Palaeontol. Polonica 37: 5–64.
should focus on these smaller-scale morpho- Carrano, M. T. 1998. Locomotion in non-avian
dinosaurs: integrating data from hindlimb kine-
logical differences to fully interpret and appre-
matics, in vivo strains, and bone morphology.
ciate the true locomotor diversity within Paleobiology 24(4): 450–469.
Sauropoda. ———. 1999. What, if anything, is a cursor?
Categories versus continua for determining loco-
ACKNOWLEDGMENTS motor habit in dinosaurs and mammals. J. Zool.
I would like to thank J. A. Wilson, K. A. Curry 247: 29–42.
———. 2000. Homoplasy and the evolution of
Rogers, and D. Chure for inviting me to con-
dinosaur locomotion. Paleobiology 26: 489–512.
tribute to this volume honoring Jack McIntosh. I ———. 2001. Implications of limb bone scaling, cur-
would also like to thank J. A. Wilson and K. A. vature and eccentricity in mammals and non-
Curry Rogers for their many comments and avian dinosaurs. J. Zool. 254: 41–55.
advice on preparing this chapter, particularly ———. 2005. Body-size evolution in the Dinosauria.
In: Carrano, M. T., Gauden, T. J., Blob, R. W. and
regarding many of the more esoteric (and unpub-
Wible, J. R. (eds.). Amnioti Paleobiology:
lished) aspects of sauropod phylogeny. The man- Perspectives on the Evolution of Mammals, Birds
uscript was markedly improved thanks to addi- and Reptiles. University of Chicago Press,
tional comments from J. A. Wilson and G. R. Chicago.
Erickson. The translations of Lapparent (1943) Carrano, M. T., and Hutchinson, J. R. 2002. The pelvic
and Lapparent and Zbyszewski (1957) are avail- and hind limb musculature of Tyrannosaurus rex. J.
Morphol. 253(3): 201–228.
able at The Polyglot Paleontologist Web site
Christiansen, P. 1997. Locomotion in sauropod
(http://ravenel.si.edu/paleo/paleoglot). dinosaurs. GAIA 14: 45–75.
Colbert, E. H. 1962. The weights of dinosaurs. Am.
Mus. Novitates 2076: 1–16.
LITERATURE CITED
Coombs, W. P., Jr. 1975. Sauropod habits and habitats.
Alexander, R. M., Jayes, A. S., Maloiy, G. M. O., and Palaeogeogr. Palaeoclimatol. Palaeoecol. 17: 1–33.
Wathuta, E.M. 1979. Allometry of limb bones of Cooper, M. R. 1981. The prosauropod dinosaur
mammals from shrews (Sorex) to elephant Massospondylus carinatus Owen from Zimbabwe:
(Loxodonta). J. Zool. London 189: 305–314. its biology, mode of life and phylogenetic signifi-
Alroy, J. 2000a. Understanding the dynamics of trends cance. Occas. Papers Natl. Mus. Monum.
within evolving lineages. Paleobiology 26(3): Rhodesia Ser. B Nat. Sci. 6(10): 689–840.
319–329. ———. 1984. A reassessment of Vulcanodon karib-
———. 2000b. New methods for quantifying aensis Raath (Dinosauria: Saurischia) and the ori-
macroevolutionary patterns and processes. gin of the Sauropoda. Palaeontol. Afr. 25: 203–231.
Paleobiology 26(4): 707–733. Cruickshank, A. R. I. 1975. The origin of sauropod
Anderson, J. F., Hall-Martin, A., and Russell, D. A. dinosaurs. So. Afr. J. Sci. 71: 89–90.
1985. Long-bone circumference and weight in Curry Rogers, K. A., and Forster, C. A. 2001. The last of
mammals, birds and dinosaurs. J. Zool. London the dinosaur titans: a new sauropod from
(A) 207: 53–61. Madagascar. Nature 412: 530–534.

Dodson, P. 1990. Sauropod paleoecology. In: Lucas, S. G., and Heckert, A. B. 2000. Jurassic
Weishampel, D. B., Dodson, P., and Osmólska, H. dinosaurs in New Mexico. In: Lucas, S. G., and
(eds.). The Dinosauria. University of California Heckert, A. B. (eds.). Dinosaurs of New Mexico.
Press, Berkeley. Pp. 402–407. New Mexico Museum of Natural History and
Gallup, M. R. 1989. Functional morphology of the Science Bulletin, Albuquerque. Pp. 43–45.
hindfoot of the Texas sauropod Pleurocoelus sp. Lull, R. S. 1953. Triassic life of the Connecticut Valley
indet. In: Farlow, J. O. (ed.). Paleobiology of the (revised). State Conn. State Geol. Nat. Hist. Surv.
Dinosaurs. Geological Society of America Special Bull. 81: 1–336.
Paper 238, Boulder, CO. Pp. 71–74. McIntosh, J. S. 1990. Sauropoda. In: Weishampel, D.
Galton, P. M. 1990. Basal Sauropodomorpha— B., Dodson, P., and Osmólska, H. (eds.). The
Prosauropoda. In: Weishampel, D. B., Dodson, P., Dinosauria. University of California Press,
and Osmólska, H. (eds.). The Dinosauria. University Berkeley. Pp. 345–401.
of California Press, Berkeley. Pp. 320–344. McShea, D. W. 1994. Mechanisms of large-scale evo-
Gilmore, C. W. 1936. Osteology of Apatosaurus, with lutionary trends. Evolution 48(6): 1747–1763.
special reference to specimens in the Carnegie Osborn, H. F. 1898. Additional characters of the great
Museum. Mem. Carnegie Mus. 11(4): 175–278. herbivorous dinosaur Camarasaurus. Bull. Am.
———. 1946. Reptilian fauna of the North Horn Mus. Nat. Hist. 10(12): 219–233.
Formation of central Utah. U. S. Dept. Interior ———. 1900. The angulation of the limbs of
Geol. Surv. Prof. Paper 210-C: 29–53. Proboscidia, Dinocerata, and other quadrupeds,
Hatcher, J. B. 1901. Diplodocus (Marsh): its osteology, in adaptation to weight. Am. Nat. 34: 89–94.
taxonomy, and probable habits, with a restoration ———. 1904. Manus, sacrum, and caudals of
of the skeleton. Mem. Carnegie Mus. 1(1): 1–63. Sauropoda. Bull. Am. Mus. Nat. Hist. 20(14):
———. 1903. Osteology of Haplocanthosaurus, with 181–190.
description of a new genus, and remarks on the ———. 1906. The skeleton of Brontosaurus and skull
probable habits of the Sauropoda and the age and of Morosaurus. Nature 1890(73): 282–284.
origin of the Atlantosaurus beds. Mem. Carnegie Osborn, H. F., and Mook, C. C. 1921. Camarasaurus,
Mus. 2(1): 1–72. Amphicoelias, and other sauropods of Cope. Mem.
Hay, O. P. 1908. On the habits and pose of the Am. Mus. Nat. Hist. New Ser. 3: 247–387.
sauropodous dinosaurs, especially of Diplodocus. Paul, G. S. 1988. The brachiosaur giants of the
Am. Nat. 42: 672–681. Morrison and Tendaguru with a description of a
Holland, W. J. 1910. A review of some recent criticisms new subgenus, Giraffatitan, and a comparison of
of the restorations of sauropod dinosaurs existing the world’s largest dinosaurs. Hunteria 2(3): 1–14.
in the museums of the United States, with special Peczkis, J. 1994. Implications of body-mass estimates
reference to that of Diplodocus carnegiei in the for dinosaurs. J. Vertebr. Paleontol. 14(4): 520–533.
Carnegie Museum. Am. Nat. 44: 259–283. Powell, J. E. 1986. Revision de los titanosauridos de
Hutchinson, J. R. 2001. The evolution of femoral America del Sur. Ph.D. thesis. Universidad
osteology and soft tissues on the line to extant birds Nacional de Tucumán, Facultad de Ciencias
(Neornithes). Zool. J. Linn. Soc. 131: 169–197. Naturales.
Janensch, W. 1922. Das Handskelett von ———. 1990. Epachthosaurus sciuttoi (gen. et sp. nov.),
Gigantosaurus robustus und Brachiosaurus brancai un dinosaurio sauropódo del Cretácico de
aus den Tendaguru-Schichten Deutsch-Ostafrikas. Patagonia (Provincia de Chubut, Argentina). Actas
Centralblatt Mineral. Geol. Paläontol. 1922: del V Congreso Argentino de Paleontologia y
464–480. Bioestratigrafia, Tucumán, Argentina. Pp. 123–128.
———. 1961. Die Gliedmaszen und Gliedmaszengürtel ———. 2003. Revision of South American
der Sauropoden der Tendaguru-Schichten. titanosaurid dinosaurs: palaeobiological, palaeo-
Palaeontogr. Suppl. VII(1) (teil 3, leif 4): 177–235. biogeographical and phylogenetic aspects. Rec.
Jianu, C. -M. and Weishampel, D. B. 1999. The small- Queen Victoria Mus. 111: 1–173.
est of the largest: a new look at possible dwarfing in Raath, M.A. 1972. Fossil vertebrate studies in Rhodesia:
sauropod dinosaurs. Geol. Mijnbouw 78: 335–343. a new dinosaur (Reptilia: Saurischia) from near the
Lapparent, A. F. de. 1943. Les dinosauriens juras- Tria-Jurassic boundary. Arnoldia 5(30): 1–37.
siques de Damparis (Jura). Mem. Soc. Géol. France Riggs, E. S. 1903. Structure and relationships of opistho-
Nouvelle Ser. 47: 1–21. coelian dinosaurs. Part I: Apatosaurus Marsh. Field
Lapparent, A. F. de, and Zybszewski, G. 1957. Les Columbian Mus. Geol. Ser. 2(4): 165–196.
dinosauriens du Portugal. Serv. Geol. Portugal Salgado, L., Coria, R. A., and Calvo, J. O. 1997.
Nova Ser Mem. 2: 1–63. Evolution of titanosaurid sauropods. I: Phylogenetic

analysis based on the postcranial evidence. Wilson, J. A., and Carrano, M. T. 1999. Titanosaurs
Ameghiniana 34(1): 3–32. and the origin of “wide-gauge” trackways: a
Scott, K. M. 1985. Allometric trends and locomotor biomechanical and systematic perspective
adaptations in the Bovidae. Bull. Am. Mus. Nat. on sauropod locomotion. Paleobiol. 25:
Hist. 179(2): 197–288. 252–267.
Stevens, K. A., and Parrish, J. M. 1999. Neck posture Wilson, J. A., and Sereno, P. C. 1998. Early evolution
and feeding habits of two Jurassic sauropods. and higher-level phylogeny of sauropod
Science 284: 798–800. dinosaurs. J. Vertebr. Paleontol. Mem. 5, 18(2;
Tornier, G. 1909. Wie war der Diplodocus carnegii Suppl.): 1–68.
wirklich gebaut? Sitz. Berliner Gesselschaft Wiman, C. 1929. Die Kreide-Dinosaurier aus
Naturf. Fr. Berlin 1909: 193–209. Shantung. Palaeontol. Sinica Ser. C 6(1): 1–67.
Upchurch, P. 1995. The evolutionary history of Yates, A. M., and Kitching, J. W. 2003. The earliest
sauropod dinosaurs. Philos. Trans. Roy. Soc. known sauropod dinosaur and the first steps
London B 349: 365–390. towards sauropod locomotion. Proc. Roy. Soc.
———. 1998. The evolutionary history of sauropod London B270(1525): 1753–1758.
dinosaurs. Philos. Trans. Roy. Soc. London B 349: Zhang, Y. 1988. The Middle Jurassic Dinosaur fauna
365–390. from Dashanpu, Zigong, Sichuan. Vol. I.
Wilson, J. A. 2002. Sauropod dinosaur phylogeny: Sauropod Dinosaurs (1). Shunosaurus. Sichuan
critique and cladistic analysis. Zool. J. Linn. Soc. Publishing House of Science and Technology,
136: 217–276. Chengdu. Pp. 1–89.

APPENDIX 8.1. TAXA AND MEASUREMENTS USED IN THIS CHAPTER
Taxa are listed in a general phylogenetic hierarchy, but see figure. 8.3 for details of their hypothesized interrelationships.
All measurements are in millimeters and were log-transformed prior to analysis. Abbreviations as in table 8.1.
Institutional abbreviations: AMNH, American Museum of Natural History, New York, New York; BMNH, The
Natural History Museum, London, UK; BSP, Bayerische Staatssammlung für Paläontologie, München, Germany;
CH/P.W., Department of Mineral Resources, Bangkok, Thailand; CM, Carnegie Museum of Natural History,
Pittsburgh, Pennsylvania; CMNH, Cleveland Museum of Natural History, Cleveland, Ohio; DINO, DINOLab,
Fruita, Colorado; FMNH, Field Museum of Natural History, Chicago, Illinois; GMNH, Gunma Museum of Natural
History, Gunma, Japan; HMN, Humboldt Museum für Naturkunde, Berlin, Germany; ISI, Indian Statistical
Institute, Kolkata, India; IVPP, Institute of Palaeoanthropology and Palaeontology, Beijing, China; MACN, Museo
Argentino de Ciencias Naturales “Bernardino Rivadavia,” Buenos Aires, Argentina; MCNA, Museo de Ciencias
Naturales de Alava, Vitoria-Gasteiz, Spain; MDE, Musée des Dinosaures, Esperaza, France; MIGM, Museu
Geológico do Instituto Geológico e Mineiro, Lisboa, Portugal; MLP, Museo de La Plata, La Plata; Argentina; MN,
Museu Nacional, Rio de Janeiro, Brazil; MNHN, Muséum National d’Histoire Naturelle, Paris, France; MPCA,
Museo Provincial “Carlos Ameghino,” Cipoletti, Argentina; MPEF, Museo Paleontolólogico “Egidio Feruglio,” Trelew,
Argentina; MUCP, Museo de Ciencias Naturales de la Universidad Nacional del Comahue, Neuquén, Argentina;
OUM, Oxford University Museum, Oxford, UK; PMU, Palaeontological Museum, University of Uppsala, Uppsala,
Sweden; PVL, Instituto Miguel Lillo, Tucumán, Argentina; QG, Queen Victoria Museum, Harare, Zimbabwe; State
University of New York (SUNY), Stony Brook, New York; YPM, Peabody Museum of Natural History, Yale University,
New Haven, Connecticut; ZDM, Zigong Dinosaur Museum, Zigong, China; ZPAL, Muzeum Ziemi Polska
Akademia Nauk, Warsaw, poland.
SPECIMEN FL FAP FML
Sauropoda
Gongxianosaurus shibeiensis Holotype 1,164.0 270.0
Isanosaurus attavipachi CH4-1 760.0 71.5 121.6
Vulcanodon karibaensis QG 24 1,100.0 140.0 174.0
Eusauropoda
Barapasaurus tagorei Holotype 1,365.0 131.0 187.0
Cetiosaurus oxoniensis OUM 1,626.0 305.0
Datousaurus bashanensis IVPP V.7262 1,057.0 147.0
Kotasaurus yamanpalliensis 111/S1Y/76 1,130.0 80.0 160.0
Lapparentosaurus madagascariensis “Individu taille max” 1,590.0 240.0
Mamenchisaurus constructus IVPP V.948 1,280.0 207.0
Mamenchisaurus hochuanensis IVPP holotype 860.0
Omeisaurus junghsiensis IVPP holotype 103.0
Omeisaurus tianfuensis ZDM T5701 1,310.0 206.0
Patagosaurus fariasi PVL 4076 1,542.0 135.5
PVL 4170 255.0
Shunosaurus lii IVPP V.9065 1,250.0 188.0
Tehuelchesaurus benitezii MPEF-PV 1125 1,530.0 243.0
Volkheimeria chubutensis PVL 4077 1,156.0 148.0 75.1
Neosauropoda
Diplodocoidea
Amargasaurus cazaui MACN-N 15 1,050.0 128.8 180.0
Amphicoelias altus AMNH 5764 1,770.0 210.0 216.0
Apatosaurus ajax YPM 1860 2,500.0
Apatosaurus excelsus FMNH 7163 1,830.0 310.0 310.0
Apatosaurus louisae CM 3018 1,785.0 174.0 332.3
Barosaurus africanus HMN NW 4 1,361.0 150.6 204.2
SPECIMEN FL FAP FML
Barosaurus lentus AMNH 6341 1,440.0 120.2 204.3

Cetiosauriscus stewarti BMNH R.3078 1,360.0 190.0 195.0
Dicraeosaurus hansemanni HMN m 1,220.0 142.5
HMN dd 3032 192.3
Diplodocus carnegii CM 84 1,542.0 174.0
CM 94 186.0
Diplodocus longus YPM 1920 1,645.0
AMNH 223 143.0
Haplocanthosaurus priscus CMNH 10380 1,745.0
CM 572 207.0
Rayososaurus tessonei MUCPv-205 1,440.0 220.0
Macronaria
“Bothriospondylus madagascariensis” MNHN uncat. 1,460.0 110.0
Brachiosaurus altithorax FMNH P25107 2,000.0 365.0
Brachiosaurus brancai HMN St 1,913.0 151.7 299.0
Camarasaurus lentus DINO 4514 1,470.0 252.0
CM 11338 86.5
Camarasaurus supremus AMNH 5761a 1,800.0 255.0
GMNH-PV 101 228.0
Euhelopus zdanskyii PMU R234 955.0 100.0 142.0
Titanosauriformes
Aegyptosaurus baharijensis BSP 1912 VIII 61 1,290.0 75.0 223.0
Ampelosaurus atacis MDE uncat. 1 802.0 157.5
MDE uncat. 2 66.0
Andesaurus delgadoi MUCPv-132 1,550.0 226.0
“Antarctosaurus wichmannianus” FMNH 3019 1,770.0
MACN 6904 77.5 217.0
Argyrosaurus superbus PVL 4628 1,910.0 160.0 300.0
Chubutisaurus insignis MACN 18222 1,715.0 265.0
Janenschia robusta HMN IX 1,330.0
HMN P 131.5 188.8
Laplatasaurus araukanicus MLP-Av 1047/1128 1,000.0
Lirainosaurus astibiae MCNA 7468 686.0 97.0
Magyarosaurus dacus BMNH R.3856 488.0 43.7 66.8
Neuquensaurus australis MLP CS 1121/1103 700.0 110.0
Neuquensaurus robustus MLP CS 1094 799.0 134.5
MLP CS 1480 120.0
Opisthocoelicaudia skarzynskii ZPAL MgD-I/48 1,395.0 108.0 280.0
Phuwiangosaurus sirindhornae CH/P.W. 1-1/1-21 1,250.0 85.0 215.0
Rapetosaurus krausei SUNY uncat. 687.0 91.2 63.1
Rocasaurus muniozi MPCA-Pv 56 768.0 117.0
Saltasaurus loricatus PVL 4017-80 875.0 164.6
PVL 4017-79 90.0
Titanosaurus indicus BMNH R.5934 865.0 67.3 80.9
nine
Steps in Understanding Sauropod Biology
THE IMPORTANCE OF SAUROPOD TRACKS
Joanna L. Wright
auropod tracks are spectacular Little progress was made in documenting

S and have captured public imagination
since they were first recognized and described in
sauropod tracks for many years after this, until
the late 1980s, when both the Glen Rose track-
the first half of the twentieth century. Roland T. sites (Farlow et al. 1989) and the Purgatoire site
Bird publicized the bathtub-sized tracks in the (MacClary 1938; Lockley et al. 1986; fig. 9.1)
Glen Rose Formation (Texas) at sites such as were redescribed. These studies are indicative
Paluxy River and Davenport Ranch in Natural of an increased interest in dinosaur tracks in
History (Bird 1939, 1944) soon after he had first general, including sauropod tracks, which has
described them, and dinosaur tracks were as continued until the present day.
popular then as they have ever been. Bird’s re- This study is the first comprehensive review
search on sauropod tracks led him to a number of sauropod tracks since 1994. It differs from pre-
of conclusions that were radical for the time. He vious reviews (Farlow 1992; Lockley et al. 1994b)
noticed that these trackways seldom showed tail not only with the addition of tracksite informa-
drag marks (Bird 1944), and despite some tion published since 1994, but also with the con-
rather sprightly sauropod reconstructions struction of predicted sauropod track morpholo-
around the turn of the twentieth century, gies for comparison with fossilized sauropod
sauropods had always been depicted dragging tracks. In addition, synapomorphies from recent
their long tails on the ground. He came to the phylogenetic studies of sauropods (Salgado et al.
conclusion that sauropods, then thought to be 1997; Upchurch 1998; Wilson and Sereno 1998)
semiaquatic, like long-necked hippopotamuses, are used to determine sauropod track character-
had left these tracks when they were walking istics and to predict differences between the
half-submerged in water and floating their tails trackways of different sauropod clades (see also
behind them. While not the dynamic pose seen Carrano and Wilson 2001). Data from published
in modern reconstructions (Wilson & Sereno sauropod tracksites are used herein to investigate
1998), this could be construed as a step in the patterns in global and temporal distribution,
right direction. lithological and paleoenvironmental associations
252
FIGURE 9.1. The Purgatoire tracksite in southeastern Colorado. (Photo courtesy of USDA Forest Service.)
and to test hypotheses about the paleoenviron- Fifteen of Upchurch’s 204 characters (7%)
mental and paleolatitudinal preferences of might be reflected in sauropod trackways, and
sauropods (Lockley 1991; Farlow 1992; Lockley 24 of Wilson and Sereno’s 109 characters
et al. 1994b). Recently reported Late Triassic (22%): 9 of these characters are used by both
sauropod tracks conform to sauropod track studies. In addition, Salgado et al. (1997) used
synapomorphies. three additional characters in their phylogeny
of the titanosaurs. In total 33 osteological fea-
tures might directly affect trackway morphol-
SAUROPOD FOOT MORPHOLOGY
ogy. However, some of these characters com-
AND PREDICTIVE ICHNOLOGY
bine to produce the same results in trackways
As with many dinosaurs, pedal elements of so that 33 separate trackway changes will not be
sauropods are rare and complete feet and hands apparent (see appendix 9.1).
even rarer. However, patterns do emerge from A simplified sauropod cladogram with spe-
what is known about sauropod feet and certain cial reference to these features is shown in
generalizations can be made. Despite the com- figure 9.2. Omission of clades and taxa not dis-
monly held view that sauropod feet are essen- tinguished by any pedal characters removes
tially conservative in morphology (Farlow most of the differences between Upchurch’s
1992), osteological features of the manus and (1998) and Wilson and Sereno’s (1998) clado-
pes have been used in recent phylogenies grams, yielding a “podial consensus cladogram”
(Upchurch 1998; Wilson and Sereno 1998). showing the clades that can be recognized
S T E P S I N U N D E R S TA N D I N G S AU R O P O D B I O L O G Y 253
FIGURE 9.2. Simplified cladogram of the sauropods (after Salgado et al. 1997; Upchurch 1998; Wilson and Sereno 1998),
showing only taxa that can be discriminated partially based on podial characters of representative sauropods, with pes (above)
and manus (below) in anterior and proximal views: Vulcanodon (after Raath 1972); Shunosaurus (after Zhang et al. 1984);
Omeisaurus (after Wilson and Sereno 1998); Apatosaurus (after Gilmore 1936); Camarasaurus (after Farlow 1992); Brachiosaurus
(after Wilson and Sereno 1998); Diplodocus (after Farlow, 1992); Brachiosaurus (after Janensch 1961); Opistocoelicaudia (after
Wilson and Sereno 1998). Figures not to scale but drawn to similar foot widths for ease of morphological comparisons.
based on synapomorphies. These are num- morphotype has not yet been discovered; or (b)
bered consecutively and are referred to as (#n) the skeletal reconstructions are incorrect.
in the following sections. Sauropods had massive bones and their fos-
Traditionally, fossil vertebrate tracks are silized remains are usually relatively uncrushed
assigned to producers by attempting to fit known so it is possible to reconstruct the feet with a rea-
manus and pes skeletons to tracks. Such an sonable degree of accuracy. It should be possible
approach has allowed identification of several dif- to predict the types of tracks that different clades
ferent types of tracks, including those of of sauropods would produce using the character-
sauropods. However, it is limited by the desire to istics listed above. Hypothetical sauropod tracks
fit skeletons into their purported tracks. The pre- have been constructed using the method outlined
dictive approach (Unwin 1989) uses the structure by Unwin (1989; fig. 9.3). The predicted tracks
of the pedal skeleton of potential trackmakers to are “ideal”; they show the maximum amount of
construct hypothetical tracks that can then be detail possible. Many sauropod tracks are found
compared to fossil tracks. This method con- in sediments that did not preserve the features of
strains the morphological range of tracks made the trackmaker’s foot with such fidelity and so
by the animal in question and has the potential to few tracks will show all the features of the pre-
exclude certain trackmaker candidates from con- dicted tracks.
sideration. However, if none of the tracks fit, then The cladistic approach (Olsen 1995; Carrano
there are two possible conclusions: (a) none of and Wilson 2001) is similar to the predictive
the animals considered is a possible trackmaker, approach (Unwin 1989), but rather than con-
which might mean that the producer of that track structing “ideal” tracks based on the osteological
254 S T E P S I N U N D E R S TA N D I N G S AU R O P O D B I O L O G Y
FIGURE 9.3 Predicted tracks of sauropods in phylogenetic context based on skeletal morphology. The exact relative
positions of manus–pes pairs cannot be predicted. Not to scale. Drawn to same pes print length for ease of morphological
comparison.
structure of known sauropod pedes, it relies prosauropod tracks, except that five digits con-
exclusively on the recognition of synapomor- tacted the ground and digit I is very short, with
phies revealed in tracks (e.g. Carrano and a laterally rather than a medially directed claw
Wilson 2001). The approach outlined here is a impression. No tracks like these are known.
combination of the two methods (predicted Eusauropod tracks would show five short dig-
tracks based on synapomorphies). its subequal in length on the manus, with no dis-
Basal sauropod (e.g., Vulcanodon) tracks crete phalangeal pads (#7,8). Manual digit I
should have been made by an animal progress- bears a large claw. The arrangement of the
ing quadrupedally (podial characteristic #1), manus digits is disputed; it is uncertain whether
with five weight-bearing digits on the manus they should be arranged in a crescent (Wilson
and pes (#2, 3, 6). These are characteristic of all and Sereno 1998) or a semicircle (Upchurch
sauropod tracks. The pes was digitigrade and 1998). Pes impressions should show five short
the long axis would probably have been approx- digit impressions (#12–14), and no separate pha-
imately parallel to the trackway midline. The langeal pads (#13). The first three digits should
first pedal digit impression should show a large show narrow claw impressions (#15) decreasing
sickle-shaped claw inclined at an oblique plane in size from digit I (#16). In most eusauropods
to the ground (#4, 5), with the axis of the claw these claws would lie at an angle to the substrate
directed anterolaterally (fig. 9.3). Digit I is and be directed anterolaterally (#18). The claw
shorter than digits II and III and the lengths of impression of digit IV, if present, would be small
digits IV and V are unknown. The manual dig- (17). The pes prints would toe outward from the
its of Vulcanodon are unknown. The reconstruc- trackway midline (#11,19) and bear a large fleshy
tion looks superficially very like some purported heel pad (#9,10) making them elongate (fig. 9.3).
FIGURE 9.4. Examples of sauropod manus and pes fossilized footprints in phylogenetic context. Scale bars equal 1 m except
for the Triassic tracks, where the scale bar equals 20 cm. Track illustrations (left to right): Triassic tracks from Peacock
Canyon, New Mexico (after Lockley and Hunt 1995); Parabrontopodus (after Lockley et al. 1986, 1994a); Breviparopus (after
Ishigaki 1989); narrow gauge trackways from Ardley Quarry, UK (after Day et al. 2002); trackways from Fatima, Portugal
(after Santos et al. 1994); Brontopodus (after Farlow et al. 1989).
Neosauropod manus impressions would be probably the greatest change in the track mor-
semicircular (#20, 21), but see the discussion phology for this clade (fig. 9.3), but in addition,
below. Both manus and pes tracks would have the manus impressions should show only a
even shorter digits (#22, 23). Neosauropod pes small claw on digit I (#27, 28).
impressions are likely to be very similar to one Titanosaurids should produce even wider
another—the main differences being in the rel- trackways than titanosauriforms (#33), the
ative sizes of the pedal claws and how many anteromedial part of the manus impression
there are: two, three, or four (fig. 9.4). might be even deeper (#30), and no digit or claw
Diplodocid pes impressions should only impressions should be visible in manus tracks
show three claw impressions; pes impressions (#31,32), although pads on the distal ends of the
of Barosaurus would only show two (fig. 9.2). metacarpals might leave impressions (fig. 9.3).
The shape of camarasauromorph manus Comparison of predicted sauropod tracks
impressions would probably be the same as with the sauropod track record (fig. 9.5) yields
those of neosauropods (i.e., semicircular) but some correspondence, indicating that (a) these
they might be slightly deeper relative to the pes tracks were made by sauropods; (b) no basal
impressions (#25) and the anteromedial part of sauropod (e.g., Vulcanodon) tracks have been
the manus impression might be deeper than discovered; and (c) tracks predicted to be made
the posterolateral portion (#26). by titanosaurs can be found from at least the
The lateral deflection of the femoral head in middle Jurassic onwards (Day et al. 2002).
titanosauriforms (#29) may indicate that they The predicted tracks show substantial digit I
would produce a wider trackway than other manus claw impressions for all sauropods
sauropods (Wilson and Carrano 1999). This is except titanosaurs, as well as claw impressions
FIGURE 9.5. Representative sauropod trackways. (A) Peacock Canyon, New Mexico (after Lockley and Hunt 1995). (B)
Trackway from the Lias of Italy (after Avanzini and Leonardi 1994) changes from wide to narrow gauge along the trackway.
(C) Breviparopus taghbloutensis (after Ishigaki 1989). (D) and (E), Parabrontopodus mcintoshi (after Lockley et al. 1986,
1994a). (F) Large manus narrow gauge trackways from Ardley Quarry, UK; note the large manus claw impression (after
Day et al. 2002). (G) Brontopodus birdi (after Farlow et al. 1989). (H) and (I) Trackways from Fatima, Portugal; note the very
large manus and claw impression (after Santos et al. 1994). (J) Rotundichnus muenchehagensis (after Lockley et al. 2004).
(K) Sauropodichnus giganteus (after Calvo 1999). (L) Large manus narrow gauge trackways from China; the wide gauge
trackways this smaller trackway was associated with showed the same morphology, and the narrowness of this trackway
may reflect the youth of the trackmaker (after Lockley et al. 2002b). Scale bars equal 1 m except for (A) and (L) where scale
bars equal 20 cm. All tracks drawn to the same pes length for ease of morphological and trackway pattern comparison.
Note the differences in internal trackway width (gauge), stride lengths, and manus position and orientation.
on all pes prints. However, few sauropod track- distal end. Such claw impressions would be
ways preserve manus claw impressions, and more susceptible to sediment collapse and corre-
many pes impressions show no signs of claw spondingly less likely to be preserved. Pittman
marks. Thus, the absence of manus claw impres- and Gillette (1989) suggested that sauropod
sions cannot by itself indicate a titanosaur origin, pedal “claws were ‘wrapped’ along the lateral
but is more likely to be a preservational artifact. edge of the foot by sauropods walking across
Manus impressions with claw marks indicate a firm or moist substrates, or extended forward for
nontitanosaur origin. The first metacarpal of braking, turning or traction in muddy areas”
many sauropods is shorter than the rest and thus (Pittman and Gillette 1989:331), providing a
the digital pad of digit I might be smaller than functional reason for the absence of these claw
shown and the claw might only impress at its marks in many sauropod tracks.
The ideal tracks would also seem to indicate of claw shape or vertical orientation visible in
that the great majority of tracks were made by trackways. Notable exceptions are trackways of
titanosauriforms or basal eusauropods and that Brontopodus birdi at the type locality (Bird 1944;
nontitanosauriform neosauropods left very few Farlow et al. 1989) (fig. 9.5) and trackways at
tracks. This is because, according to the charac- Ardley Quarry, UK (Day et al. 2002) (fig. 9.5),
ters presented above, all neosauropods, possibly and Galinha, Portugal (Santos et al. 1994). The
all eusauropods, had metacarpals arranged in a absence of manual claw impressions has
semicircular column, which would have pro- incited speculation about the function and posi-
duced horseshoe-shaped tracks. Yet many sauro- tion of the sauropod manus claw and several
pod tracks show crescentic manus impressions hypotheses have been proposed (see Upchurch
so a question arises over the reconstruction of the 1994). Impressions of other manual digits are
sauropod manus. It seems most likely that some even more uncommon but have been reported
neosauropods (diplodocoids), while having (Leonardi and Avanzini 1994; Farlow et al.
vertically oriented columnar ligament-bound 1989; Farlow 1992; Ishigaki 1988; Lockley et al.
metacarpals, did not have them arranged in a 2002b). Manus impressions are always pre-
tight 270° arc but in a wider 210°crescentic arc served in front of the pes impressions but the
(Hand 1999). This is only a slightly tighter con- exact position may vary. The manus may be
figuration than that of Omeisaurus (fig. 9.2). This closer or farther away from the trackway mid-
would then allow neosauropods to be considered line than the pes and it may be very close to the
potential trackmakers of crescentic manus anterior margin of the pes or up to a pes length
impressions (figs. 9.3, 9.4), which are very com- away (fig. 9.5). Manus impressions are angled
mon in the Late Jurassic, when neosauropods are outward from the axis of the trackway at angles
the most common sauropods known from osteo- of 5–75°.
logical remains. Most sauropod tracks are very Pes impressions are triangular to oval, with
difficult to place in a phylogenetic context, either the axis of the print toed outward at an angle of
because they are too poorly preserved to show the 10–30° and the widest part of the footprint
relevant details or because their morphology does directed anterolaterally (figs. 9.4–9.6). The
not fit predicted morphology. back of the footprint narrows to the posterior
The most notable predicted differences in margin; it may be rounded or pointed (fig. 9.6).
sauropod tracks are: (a) basal sauropods to Pedal claw impressions are more often pre-
eusauropods, where the pes should change from served than those of the manus but many track-
digitigrade to semidigitigrade with a heel pad and ways still do not preserve them, for instance,
be toed out from the trackway axis; (b) the transi- most trackways at the Purgatoire locality (Lockley
tion from a crescentic to a semicircular manus, et al. 1986) and pedal claw impressions are not
which may occur in camarasauromorphs; and (c) preserved in many of the trackways at Ardley
the change from narrow to wide gauge trackways, Quarry (Day et al. 2002). Where preserved, pedal
which should occur in titanosauriforms. Most claw impressions are directed laterally or antero-
other transitions would only be evident in very laterally and decrease in size from digit I. Pes
well-preserved trackways. tracks may show two to four claw impressions
(figs. 9.4–9.6). Non-claw-bearing digits may
only impress as a slight bulge on the lateral side
SAUROPOD TRACK MORPHOLOGY AND
of the footprint immediately anterior to the
CLASSIFICATION
“heel” pad, although some pes tracks preserve
Manus tracks are crescentic to horseshoe- separate impressions of rounded toes (Lockley
shaped, with digit III directed anterolaterally and Hunt 1995).
(figs. 9.4, 9.5). Manus claw impressions are sel- Sauropod tracks range in pes length from
dom preserved, and even more rarely are details 20 cm (Lim et al. 1994; Lockley et al. 2002b) to
FIGURE 9.6. Well-preserved sauropod pes tracks showing claw impressions from the Late Jurassic. (A) CU-MWC 194.2,
Lost Springs site, Morrison Formation, Utah; (B) CU-MWC 188.28, Summerville Formation, Arizona.
more than 1 m. Even the smallest sauropod applied to wide gauge trackways (e.g., Lockley
tracks show the characteristic subtriangular et al. 2001).
toed-out pes impressions with anterolaterally Despite the proposed link between wide
directed claw impressions. The relative sizes of gauge trackways and titanosauriforms (Wilson
manus and pes impressions are considered and Carrano 1999), trackway gauge does not
important in sauropod track classification but it seem to reflect the anatomy of the trackmaker;
is difficult to predict this from osteology other characteristics such as locomotor style
because of the soft tissue “heel” pad. and ontogeny contribute, e.g. some change
from one to the other along their length
SAUROPOD TRACKWAY GAUGE (Leonardi and Avanzini 1994; fig. 9.5B) and
Sauropod trackways are termed narrow gauge some sauropods may have made a narrow
when the inside margins of the pes impres- gauge trackway as juveniles and a wide gauge
sions overlap the trackway midline and wide one when older (Lockley et al. 2001, 2002a;
gauge when they do not. When this division fig.9.5L). The pes impressions of some other
was proposed (Farlow 1992; Lockley et al. trackways are very close to the trackway mid-
1994a) an intermediate category, medium line so that they do not qualify as narrow gauge
gauge, was illustrated (Meyer et al. 1994) but in a strict sense, yet to classify them as wide
not described. The illustration seemed to be a gauge seems excessive (figs. 9.5, 9.6). The dis-
difference of scale (the internal width of the tinction is thus not as clear-cut as often implied
trackway relative to pes print size was the but it may reveal useful information on the
same as in the wide gauge example) and this identity of the trackmaker. Wide gauge track-
category has not been used subsequently ways are known from rocks as old as the
in trackway descriptions; occasionally modi- Sinemurian and Pliensbachian (Ishigaki 1988;
fiers such as “moderately” and “slightly” are Farlow 1992; Leonardi and Avanzini 1994)
(fig. 9.5), but the earliest sauropod trackways deflected femoral head (# 29 above), which may
(Lockley et al. 2002b) are narrow gauge. have given them a wider stance and hence a
Brontopodus birdi is the classic ichnotaxon to wider trackway (Wilson and Carrano 1999).
which the description wide gauge was first Titanosauriforms are not known from osteolog-
applied (Farlow et al. 1989; Farlow 1992). The ical remains before the Late Jurassic. However,
ichnogenus Parabrontopodus was erected to pro- the first purported titanosaur tracks are reported
vide a similar standard for narrow gauge track- in the Middle Jurassic (Day et al. 2002), prior to
ways (Lockley et al. 1994a). Sauropod trackways their appearance in the osteological record. It
are most often referred to one of these two has been suggested that wide gauge trackways
ichnogenera based on the internal width of the with large manus claw impressions from the
trackway. These two ichnogenera are also used Middle Jurassic of Portugal (Santos et al. 1994)
to exemplify the other characteristics of these were made by camarasauromorphs, although
tracks, such as the relative size of the manus. they do not have a deflected femoral head
Some Paluxy Brontopodus show narrow gauge (Wilson and Sereno 1998).
locomotion along some trackway segments It has been claimed that wide gauge track-
(Farlow, pers. comm., 2004) so gauge is par- ways are the dominant type of sauropod track-
tially behavioral in even derived sauropods. way in the Cretaceous and that narrow gauge
Thus sauropod tracks should be classified pri- ones are more common before the Late
marily on the basis of footprint morphology Jurassic; the Late Jurassic being a transitional
and only secondarily on the internal width of period (Lockley, Meyer, Hunt and Lucas 1994;
the trackway. Wilson and Carrano 1999). Lockley, Meyer,
The type trackway of Brontopodus has a Hunt and Lucas (1994) plotted sauropod track-
manus-pes area ratio of 1:3; in Parabrontopodus ways, indicating numbers of narrow, wide, and
or Breviparopus this ratio is 1:5 (Lockley et al. unknown gauge, but on a logarithmic scale.
1994a). Other wide gauge trackways have been This was in order to accommodate the large
reported with a ratio as high as 1:2 (Santos et al., range in data (anywhere from one to 100 track-
1994). While an association of a relatively large ways plotted per stage), but it masks differences
manus with a wide gauge is most common, in numbers of known trackways per interval
there are exceptions: both large manus, narrow because stages with one trackway are shown as
gauge (Day et al. 2002; Lockley et al. 2001) and over a quarter of the length of one with 10
small manus, wide gauge (Leonardi and trackways, or an eighth of the length of one
Avanzini 1994; Day et al. 2002; fig. 9.5). In with a hundred. Wilson and Carrano (1999)
addition, a continuum of manus print sizes rel- plotted wide and narrow gauge sauropod track-
ative to the pes impression has been reported ways as percentages of total known sauropod
(from 1:2 to 1:6), which indicates that the divi- trackways which allows easier comparison of
sion between large and small manus print track- relative proportions. However, a plot of track-
ways may not be as clear-cut as often portrayed. sites, rather than trackways, by stage (fig. 9.7)
Insufficient numbers of manus and pes skele- illustrates the inadequacy of the database on
tons are known to be able to determine patterns which this conclusion is based. Most tracksites
of manus and pes relative sizes but even very preserve either narrow or wide gauge trackways
closely related sauropods can show significant so tracksites have been plotted in order that
differences in these ratios; for instance, the large sites like Purgatoire or Paluxy, where many
manus of Diplodocus is relatively small in com- trackways are preserved on a single bedding
parison to that of the pes, whereas that of plane, do not overwhelm data from sites with
Apatosaurus is relatively large (pers. obs.). less areal exposure. Sites which preserve both
Wide gauge trackways have been attributed wide and narrow gauge trackways are plotted as
to titanosauriforms on the basis of their one of each. Figure 9.7 shows the distribution of
FIGURE 9.7 Stratigraphic distribution of sauropod tracksites by stage. () Wide gauge, ()nar-
row gauge, () unknown. Based on 64 sites reported in the literature up to 2003 plus the 102
sites from South Korea, here counted as 10 (Cenomanian). Tracksites from the Morrison Forma-
tion have been split evenly between the Kimmeridgian and the Tithonian. Data in Table 9.1.
tracksites through time. There are insufficient Assuming a crescentic configuration for the
data to determine much about the distribution metacarpals of diplodocoids (Hand 1999), a
of narrow and wide gauge trackmakers through comparison with the sauropod osteological
time. The Late Kimmeridgian has been sug- record (Wilson and Sereno 1998) might sug-
gested as the interval when wide gauge track- gest that Late Jurassic narrow gauge trackways
ways became more numerous than narrow were made by diplodocoids, Late Cretaceous
gauge trackways. However, the picture is not wide gauge trackways by titanosaurs, and Late
that simple. For instance, in Switzerland, seven Jurassic to Early Cretaceous wide gauge track-
early Kimmeridgian sites preserve wide gauge ways by titanosauriforms.
trackways and only in the Late Kimmeridgian
have narrow gauge trackways been reported ICHNOGENERA ATTRIBUTED TO SAUROPODS
(Meyer and Thuring 2003). In most time peri- About a dozen sauropod ichnogenera have
ods (e.g. in the Maastrictian seven of the eight been named, although only a few (three to five)
tracksites are in Bolivia) the data are concen- are considered valid. Sauropod ichnogenera
trated in small geographic areas (see Table 9.1) were reviewed by Farlow et al. (1989) and
and great caution must be exercised in drawing Lockley et al. (1994a); unless otherwise stated,
global conclusions from the small amount of their assessment of the validity of sauropod
data that exist. ichnotaxa is herein considered correct.
FIGURE 9.8. Triassic age tracks attributed to sauropods. (A) Trackway from Furnish
Canyon, Colorado (Lockley et al. 2001). (B) Trackway from Peacock Canyon, New Mexico;
(C) trackway from Cub Creek, Utah (after Lockley and Hunt 1995). (D) Sauropodopus an-
tiquus; (E) Tetrasauropus unguiferous; (F) Pseudotetrasauropus jaquesi (Ellenberger 1972; after
Lockley and Meyer 1999). (G)–(I) CU-MWC 172.22, pes impression showing five digits,
four with claw impressions. Lighting from the right in (H) and the left in (I). Scale bars in
(B)–(I) equal 20 cm. (B)–(F) drawn to same pes print length for ease of comparison. (G)–(I)
drawn to same scale.
EUSAUROPODA diagnostic characters (appendix 9.1) and their

Tetrasauropus (Ellenberger 1972). Sauropod extrapolation to diagnostic features of sauropod
trackways from the Late Triassic were described trackways, these claims can be revisited and
and referred to Tetrasauropus (Lockley et al. evaluated.
2001; fig. 9.8). The recognition of these track- These trackways have four anterolaterally
ways is consistent with the recent discovery of directed claw impressions with a fifth non-claw-
Triassic sauropods (Buffetaut et al. 2000, bearing digit, (fig. 9.8). Digit I is the leading
2002). The Triassic sauropod trackways have digit in the pes but is shorter than the other dig-
been met with a certain amount of skepticism, its, although the claw impression is larger. The
but with the enumeration of sauropod podial deepest part of the track is the inside margin of
the foot, which is oriented anteromedially and foot skeleton and therefore an important differ-
forms the leading edge. The long axis of the pes ence in footprint morphology. Thus, the mor-
impression is angled outward at an angle of phological differences between the tracks from
20–30o to the trackway midline (fig. 9.8A–C). North America and the United Kingdom,
The “heel” pad is substantial and the posterolat- which can be attributed to sauropods, and those
eral part of the track is also well impressed. of the type Tetrasauropus trackway from South
Digits I and V are impressed most deeply, Africa, which is excluded from a sauropod ori-
although all five digits are clear. Well-preserved gin, indicate that these two types of trackways
manus impressions show five digits arranged in can no longer be considered congeneric. When
a gentle crescent; none shows any indications of more well-preserved trackways of this morpho-
claws. The manus impressions are preserved in type are discovered it may be possible to erect a
front of and slightly inside the pes impressions new ichnogenus.
and angled outward at a variable angle of 5–45o Tracks from the Late Triassic attributed to
(fig. 9.8). Manus impressions are 15%–20% of sauropods were first described by Ellenberger
the area of the pes. The trackways are narrow to (1972), although this trackmaker identification
slightly wide gauge (see Lockley et al. 2001:fig. was questioned (Olsen and Galton 1984).
3). These trackways fit the sauropod track diag- These four-toed tracks, Sauropodopus antiquus
nostic criteria and are most likely to have been (fig. 9.8D), do not fit the sauropod track diag-
made by Late Triassic sauropods. Similar tracks nostic criteria outlined above and have been
have been described from other Late Triassic attributed to chirotheres (Lockley and Meyer
sites including Wales (Lockley et al. 1996), 2000). Other tracks from these sites named
Greenland (Jenkins et al. 1994), Italy (Dalla Tetrasauropus unguiferous (fig. 9.8E) were attrib-
Vecchia 1996), and Switzerland (Meyer and uted to prosauropods (Ellenberger 1972;
Thuring 2003). Confirmation of these occur- Lockley and Meyer 1999). These rounded
rences would indicate a Laurasian distribution tracks show the impressions of four large
of sauropods in the Late Triassic. Some tracks anteromedially directed claws on the pes and
from the Pliensbachian of Morocco (Ishigaki three on the manus. The axis of the pes impres-
1988; morphological type 2 of Farlow [1992]) sions is parallel to the trackway midline. This
appear very similar to these tracks and might trackway does not correspond to predicted
also be congeneric. sauropod track morphology. Pseudotetrasauropus
The small size of these tracks indicates a hip includes a number of ichnospecies, some show-
height of 0.8–1.5 m. This is consistent with the ing four or five short toes, with rounded or
fact that some of these Late Triassic sauropods pointed terminations, a large heel pad, and
were small (Buffetaut et al. 2000), although the manus impressions and others showing sepa-
tracks could also have been made by a juvenile rated toes with discrete phalangeal pads made
of a larger Late Triassic sauropod (Buffetaut et by animals progressing bipedally. P. jaquesi is
al. 2002). However, the consistent small size of the only one of these that could be interpreted
these tracks at several widely spaced localities as of sauropod origin; it has a five-fingered
might argue for small adult trackmakers. manus that faces anterolaterally and lies out-
Lockley et al. (2001) referred these track- side the line of the pes impressions and a four-
ways to Tetrasauropus; despite the differences in toed triangular pes oriented parallel to the
foot and claw impression orientations, both trackway midline (fig. 9.8F). The toes are short,
morphotypes were considered sufficiently simi- with blunt terminations; one may show a claw
lar to be referred to the same ichnogenus. impression, which is oriented anteromedially.
However, osteological features (see #5 and 18 The four toes and claw orientation, in addition to
above) indicate that pedal claw orientation in the pes orientation, indicate that the maker of
sauropods is a function of the structure of the this trackway is unlikely to have been a sauropod.
These Triassic sauropod trackways seem to for ease of referral to the footprints at this site
have been made not by basal sauropods such as (Demnat), but they illustrated the trackway with
Vulcanodon, but by eusauropods; they show evi- photographs and diagrams and the name has
dence of a semidigitigrade foot and four later- been used since to refer to tracks from both this
ally compressed sickle-shaped unguals lying at site and others (e.g., Ishigaki 1988, 1989; Calvo
an angle to the substrate (see podial synapo- 1999). Both Farlow et al. (1989) and Lockley et
morphies #5 and 7–15 and the summary of al. (1994a) considered it a valid ichnotaxon.
eusauropod predicted track features; Figs. 9.3, Parabrontopodus mcintoshi (Lockley et al.
9.4A, B, 9.5, 9.6). These tracks provide the first 1994) was erected on the basis of abundant ich-
evidence of eusauropods in the Late Triassic. nofossils from the Purgatoire tracksite of south-
eastern Colorado, although the holotype is a
UNNAMED SAUROPOD TRACK MORPHOTYPE A manus–pes pair (CU-MWC 190.5). All track-
This was recovered from the Early Jurassic of ways from this site were referred to
Morocco (Ishigaki 1985, 1988; Farlow 1992) Parabrontopodus because they are all narrow
and Italy (Leonardi and Avanzini 1994; fig. 9.5). gauge. These tracks are not as well preserved as
The trackway is characterized by long pes those of Brontopodus; pedal claw impressions
impressions showing four clear toes and sug- are rarely preserved and the manus tracks are
gestions of a fifth. crescentic. Parabrontopodus was diagnosed as a
narrow gauge sauropod trackway with out-
NEOSAUROPODA–DIPLODOCIDAE wardly rotated pes impressions, three laterally
Breviparopus taghbaloutensis (Dutuit and directed pedal claw impressions, and small
Ouazzou 1980) was originally described from semicircular manus impressions (Lockley et al.
the Middle Jurassic of Morocco. The trackway 1994a: 140). A diplodocid has been suggested
was narrow gauge, with medium to small as a possible trackmaker for the tracks from the
(manus:pes ratio, 1:3.5–5) crescentic manus type locality, based on the smallish size of the
prints (fig. 9.5C). Well-preserved pes impres- tracks and their preservation in the Morrison
sions show three claws directed anterolaterally Formation.
with two pads, presumably from digits IV and V The original diagnosis did not state how
behind the claw of digit III (fig. 9.5). The pes Parabrontopodus could be differentiated from
impressions are approximately 1 m long, and Breviparopus. Breviparopus is considered a valid
the manus impressions 50 cm wide. Well- ichnotaxon (Farlow et al. 1989; Lockley et al.
preserved manus impressions show indications 1994a) and the type trackway of Breviparopus is
of a single medially directed claw (Ishigaki well preserved (Ishigaki 1985, 1989). It is not
1989). Manus impressions are farther from the clear how Parabrontopodus can be distinguished
trackway midline than pes impressions. The from Breviparopus, and thus Parabrontopodus
manus impressions are very close to the pes may be a junior synonym of Breviparopus. Since
impressions and it has been suggested that the the erection of Parabrontopodus, few trackways
pes impressions truncated those of the manus have been referred to Breviparopus (exceptions
(Dutuit and Ouazzou 1980). However, manus- include Calvo 1999), however, if Parabrontopodus
only trackways from the same area show a sim- is indeed a junior synonym of Breviparopus,
ilar morphology (Farlow et al., 1989), the tracks referred to Parabrontopodus should be
manus area, at one-fourth to one-fifth that of the referred to Breviparopus. The trackway from
pes, is typical for a narrow gauge trackway and which the type specimen was taken (CU-MWC
the presence of raised rims around both types of 190.5; fig. 9.5D) is very similar in manus and
tracks indicates that little overstepping is likely pes morphology to Breviparopus (fig. 9.5C), but
to have taken place. Dutuit and Ouazzou (1980) the topotype trackway (CU-MWC 190.3) shows
suggested the name Breviparopus taghbaloutensis a rather different trackway pattern (fig. 9.5E).
In this trackway the manus impressions not has been attributed to a titanosauriform such as
only are generally farther in front of the pes, but the contemporary Pleurocoelus (Farlow et al.
also tend to be closer to the midline, than those 1989; Gallup 1989; Wilson and Carrano 1999).
of the type trackway, and rather than being ori-
ented at approximately 20–25o to the midline, POSSIBLE TITANOSAURIFORMES
they make an angle of 45–70o to it (fig. 9.7B, C). (WIDE GAUGE TRACKWAYS)
This may indicate an important difference in Rotundichnus muenchehagensis (Hendricks 1981)
locomotion styles between the two types of was erected on the basis of the sauropod tracks
sauropods. CU-MWC 190.3 also has more elon- at Münchehagen (Berriasian), Germany. The
gate pes impressions and a slightly shorter rel- size of these tracks alone indicates that they
ative stride length: 2.2–2.5 times the pes print were made by sauropods, but the arrangement
length (although stride length is related to and orientation of manus and pes impressions
speed rather than taxonomy), in comparison to are also consistent with a sauropod origin (fig.
the stride length 3 times the pes length in the 9.5J). No claw impressions or other morpholog-
type trackway and Breviparopus. Farlow et al. ical details are preserved and the trackways have
(1989) previously suggested that there might be a high manus:pes ratio (ca. 1:3) and are wide
two sauropod track morphotypes at Purgatoire. gauge. The lack of morphological detail means
that these tracks cannot be diagnostic of any par-
TITANOSAURIFORMES ticular type of sauropod or trackway and should
Brontopodus birdi (Farlow et al. 1989) is based not be applied to tracks from other sites. The
on well-preserved and substantial material stride length is 3.2 times the pes print length. In
from Texas (fig. 9.5G). The length and width of contrast to Brontopodus (see above), the manus
Brontopodus manus tracks are approximately is semicircular in shape; the manus length is
equal, clawless, and U-shaped, with digit half the dimension of its width, and there is no
impressions I and V slightly separated from indication of the separation of the lateral digits
conjoined digit impressions II–IV. Manus track from the central digits typical of Brontopodus.
centers are closer to the trackway midline than Sauropodichnus giganteus (Calvo 1991) was
pes track centers. Well-preserved pes tracks erected on the basis of a trackway from
show three large, laterally directed claw marks, Argentina, consisting of large (90-cm maxi-
decreasing in size from digit I to digit III, a mum dimension) rounded impressions inter-
small claw on digit IV, and a pad on digit V. preted as having been made by sauropod pedes.
Manus and pes impressions are well away from The manus impressions were presumably over-
the trackway midline (more than 0.5 pes print stepped completely by those of the pes
widths)—the classic wide gauge pattern. Manus (Mazzetta and Blanco 2001). This ichnogenus
impressions are about one-third the area of the was considered a nomen dubium by Lockley
pes impressions. Brontopodus stride length at et al. (1994a). Calvo (1999) revised the diagno-
the type locality varies from approximately two sis based on undoubted sauropod trackways
to five times the pes print length, and the pace with manus prints (fig. 9.5K) from the localities
angle (related to stride length) is 100–120o; of Cerrito del Bote and El Chocón (Albian–
stride length should increase with increased Cenomanian). These trackways are wide gauge,
speed and so should not be expected to be a with crescentic manus impressions almost one
constant within trackways or ichnogenera. quarter the area of the pes impressions (range,
Brontopodus differs from all other sauropod ich- 1:3–1:5; average, 1:3.7). This spans the range of
notaxa in that the manus length and width are manus heteropody from large to small manus.
approximately equal; this indicates that the No claw impressions are preserved but the cres-
trackmaker possessed metacarpals arranged in centic manus morphology may mean that these
a tight arc, rather than a crescent. Brontopodus trackways do not fall into Brontopodus, although
less well-preserved manus impressions of associated with limy substrates in tropical and
Brontopodus from the type locality are crescentic subtropical latitudes.” Farlow (1992) tested this
in shape (Farlow et al. 1989). Sauropodichnus hypothesis by examining the distribution of
is more similar to Rotundichnus than to sauropod tracks by environment and paleolati-
Brontopodus. The main differences are that tude and comparing this to the distribution of
Sauropodichnus has a relatively smaller manus other dinosaur tracks and skeletal remains. He
and slightly narrower trackway, although it may concluded that current knowledge of sauropod
eventually turn out to be a junior synonym of tracksites was “inadequate for determining lati-
Rotundichnus. Trackways referred to Breviparopus tudinal and habitat preferences of sauropods”
were also reported from these localities (Calvo (Farlow 1992:90). Lockley et al. (1994b) per-
1999). Titanosaurs have been suggested as formed an analysis similar to that of Farlow but
trackmakers for both types of trackways (Calvo added recently discovered tracksites, raising the
1999), despite discrepancies between the number of known tracksites from 22 to 190, of
known titanosaur skeletal morphology and the which 102 were in a single formation in one
trackways (crescentic manus impressions and small area of South Korea. Lockley et al.
narrow gauge trackways). (1994b) defined a tracksite as a specific geo-
Elephantopoides barkhausensis (Kaever and graphic location on a particular geographic area
Lapparent 1974; Freise and Klassen 1979) was and thus split up several of Farlow’s (1992)
established on the basis of sauropod tracks sites; for example, the Purgatoire site has four
from Barkhausen (Late Jurassic), Germany. track-bearing levels. Lockley et al. (1994b) only
These tracks are wide gauge, with a crescentic considered sauropod tracks and concurred with
manus one-third the area of the pes and the Lockley’s (1991) original hypothesis. The
trackway width half the width of the pes. No method of Lockley et al. (1994b) is followed in
morphological details such as claw impressions this report.
are preserved. The lack of morphological detail It is possible to test the contention that
makes it difficult to refer tracks from other sites sauropods preferred paleolatitudes of 0–30oN
to this ichnotaxon. (Lockley et al. 1994b). However, most sauropod
tracks are found at paleolatitudes of 30–45oN
(figs. 9.9, 9.10). The second-hightest number
SAUROPOD TRACK DISTRIBUTION
of tracksites is found at paleolatitudes of
Sauropod tracks are now known from the Late 15–30oN, which happens to be the paleolatitudi-
Triassic to the latest Cretaceous (fig. 9.7) and nal range of much of North America and
from every continent except Antarctica (fig. 9.9), Europe during the Mesozoic where prospecting
although some intervals and areas are better doc- for tracksites has been most concentrated.
umented than others. Table 9.1 is a summary of Tracks from other areas are usually discovered
tracksite data used in figures. 9.7 and 9.9–9.11. by local residents, with subsequent documenta-
tion by scientists from developed countries
SPATIOTEMPORAL DISTRIBUTION (Lockley 2002b), and are generally concentrated
The patchy distribution of sauropod tracksites near population centers (Lockley 2002a).
(fig. 9.9) places limitations on interpretation of Sauropod tracksite distribution is more inform-
the data. However, bearing this in mind, and ative about the state of modern countries than
using the phylogenetic correspondences from about the habitat preferences of sauropods.
figure 9.5, the track record is generally consis- To test if sauropod tracks are found at dif-
tent with the osteological record. ferent paleolatitudes than other types of
Lockley (1991:123, 124) suggests that “from dinosaur tracks, especially ornithopods, as sug-
their first appearance . . . right through to gested by Lockley (1991; Lockley et al. 1994b), it
Cretaceous times brontosaur tracksites are would be necessary to plot other dinosaur
TABLE 9.1
Summary of Sauropod Tracksite Publications Around the World
AGE COUNTRY NO. SITES LITHOLOGY PALEOENVIRONMENT GAUGE REFERENCE(S)
Late Triassic
Norian-Rhaetian Western USA 5 Clastic n Lockley et al. 2001
Norian-Rhaetian Wales 1 Clastic, redbeds Fluvial n Lockley et al. 1996
Switzerland 1 Carbonate Coastal Meyer & Thuring 2003
Greenland 1 Jenkins et al. 1994
Early Jurassic
Pliensbachian Northeastern Italy 1 Carbonate Coastal w Avanzini et al. 2001
Pliensbachian Morocco 3 Carbonate b Ishigaki 1988
Early Sinemurian Northern Italy 1 Carbonate Coastal b Dalla Vecchia 1994; Avanzini
et al. 1997
Hettangian Poland 2 Sandstone Fluvial n Gierlinski & Sawicki 1998
Middle Jurassic
Aalenian–Bathonian Yorkshire, UK 9 Terrigenous clastics Fluvial b Romano & Whyte 2003
Baj–Bath Portugal 1 Limestone Coastal w Antunes & Mateus 2003;
Santos et al. 1994
Baj–Bath Morocco 4 Ishigaki 1989; Dutuit &
Ouazzou 1980
Bathonian Oxford, UK 1 Coastal b Day et al. 2002
Callovian UT, USA 1 Sandstone Coastal w Foster et al. 2000
Southeastern Mexico 1 Arenitic siltstone Fluvial Ferrusquía-Villafranca et al. 1996
Portugal 1 Lockley & Meyer 1999; Antunes
& Mateus 2003
Late Jurassic
Oxfordian Chile 1 Sandstone Marine platform Leonardi 1989
Oxfordian Tadjikistan 1 Novikov et al. 1993
Early Kimmeridgian Switzerland 4 Mudstone Coastal w Meyer & Thuring 2003
Early Kimmeridgian Switzerland 3 w Meyer & Thuring 2003
Kimmeridgian Portugal 1 Limestone b Lockley & Meyer 1999
Kimmeridgian Portugal 1 Limestone Antunes & Mateus 2003
TABLE 9.1 (continued)
AGE COUNTRY NO. SITES LITHOLOGY PALEOENVIRONMENT GAUGE REFERENCE(S)
Kimmeridgian CO, USA 1 Sandstone Floodplain

Kimmeridgian 1 Sandstone w Barnes & Lockley 1994
Late Kimmeridgian Switzerland 4 n Meyer & Thuring 2003
Late Kimmeridgian Switzerland 1 w Meyer & Thuring 2003
Tithonian–Kimmeridgian CO, USA 6 n
Tithonian–Kimmeridgian CO, USA 3 Limestone Lacustrine n Prince & Lockley 1989
Tithonian–Kimmeridgian WY, USA 2 Limestone Coastal w
Tithonian UT, USA 5
Tithonian UT, USA 1 Muddy sandstone
Tithonian Switzerland 3 w Meyer & Thuring 2003
Tithonian France 1 Allain & Pereda-Suberbiola 2003
Tithonian Portugal 8 Limestone Coastal w Lockley & Meyer 2000;
Lockley et al. 1994c
Portugal 6 Antunes & Mateus 2003
Niger 1 Weishampel et al. 1990
Early Cretaceous
Berriasian UK 1 Limestone Coastal Wright et al. 1997
Berriasian Germany 1 Sandstone Coastal n Fischer 1998
Berriasian Spain 1 Calcareous n Lockley & Meyer 2000
Berriasian Chile 1 Coastal swamp Leonardi 1989
Berriasian Thailand 1 Sandstone Coastal Le Loeff et al. 2002
Hauterivian Portugal 2 Antunes & Mateus 2003
Haut–Barr Italy 1 Limestone Coastal Dal Sasso 2003
Neocomian Brazil 2 Lacustrine Leonardi 1989
Neocomian Australia 1 Sandstone Coastal n Thulborn et al. 1994
Neocomian Australia 1 Sandstone Swamp/ forest n Thulborn et al. 1994
Barremian–Aptian UK 1 Terrigenous clastics Fluvial Radley 1994; Insole
& Hutt 1994
Barremian–Aptian Spain 1 w Moratalla et al. 1994
Barremian–Aptian Spain 1 Silty sandstone w Moratalla et al. 1994
Barremian–Aptian China 4 Sand/mudstone, Fluvial b Lockley et al. 2002b
clastics
Aptian Spain 10 Carbonates Coastal b Casanovas et al. 1997
Albian Spain 2 Conglomerate, Fluvial Moratalla et al. 1994
Sandstone
Albian Argentina 3 Mudstone Fluvial w Calvo 1999
Albian TX, USA 8 Limestone Coastal w Pittman 1989
Albian AR, USA 1 b Pittman & Gillette 1989
Late Albian Croatia 1 Carbonate Marine platform Dalla Vecchia 1994
Brazil 1 Sandstone Fluvial Leonardi 1989
Tadjikistan 2 w
Late Cretaceous
Cenomanian Croatia 1 w Lockley & Meyer 2000
Cenomanian Portugal 1 Antunes & Mateus 2003
Cenomanian South Korea 102 Shales, siltstone Lacustrine w Lim et al. 1994
Cenomanian TX, USA 1 Limestone Coastal w Farlow et al. 1989
late Cenomanian Croatia 1 Carbonate Marine platform, n Dalla Vecchia 1994
humid
Turonian–Coniacian China 1 Terrigenous clastics Subtropical w Lockley et al. 2002b
Maastrichtian Spain 1 Silty limestone Coastal w Schulp & Brokx 1999
Maastrichtian Bolivia 1 Sandstone Fluvial w Lockley et al. 2002a
Maastrichtian Bolivia 5 Limestone Lacustrine w Lockley et al. 2002a
Maastrichtian Bolivia 1 Marine platform w Lockley et al. 2002
(Leonardi & Avanzini 1994)
Note: Abbreviations: n, narrow gauge; w, wide gauge; b, both wide and narrow gauge.
FIGURE 9.9. The global distribution of sauropod tracksites. (A) Late Triassic to Middle Jurassic tracksites
shown on an Early Jurassic paleogeographic map; (B) Late Jurassic tracksites; (C) Early and Late Cretaceous
tracksites plotted on a Maastrichtian paleogeographic map. Data in Table 9.1.
FIGURE 9.10. Paleolatitudinal distributions of sauropod tracksites. Tracksite occur-
rences plotted as conservatively as possible, given the available data (i.e., where
tracksites fell between two categories, they were placed in the one nearer the equator).
tracksite distributions against paleolatitude. and carbonate-evaporite lake basins.” Their dia-
Farlow (1992) did this, not only for tracks but grams illustrating the relationship between
also for skeletal remains, and found few differ- sauropod tracksites and sedimentary facies show
ences; this has not been done since. In South that the majority of sauropod tracks are found in
Korea, where more than a hundred tracksites carbonate environments. However, their dia-
have been reported, there is obviously no lati- grams show that they conflated lithology and
tudinal separation, even if sauropod and paleoenvironments, counting all environments
ornithopod tracks are seldom found on the as carbonates apart from “fluviolacustrine.”
same bedding plane. Some sauropod tracks from lacustrine paleoen-
vironments are preserved in siliciclastic sedi-
PALEOENVIRONMENTAL DISTRIBUTION ments (e.g., Paik et al. 2001). Figure 9.11 shows
As biogenic sedimentary structures produced sauropod tracksites plotted by lithology and by
by the behavior of animals, tracks reveal where paleoenvironment, with the South Korean local-
living animals roamed. Tracks may have the ities counted as 10 (rather than 102). If the total
potential to reveal habitat preferences of their number of Korean tracksites were inserted,
producers if favored habitats are as likely to be lacustrine environments would be boosted to
preserved as any other and if these habitats 52%, marine and coastal environments would
have track preservation potential. However, sed- decrease to 28%, and other inland environ-
iments and tracks in lowland and coastal envi- ments, mainly fluvial, would decrease to 20%.
ronments are much more likely to be preserved This demonstrates how this one small geo-
than those in more upland or inland environ- graphic area, during one time period, can
ments, and certain other environments, such as overwhelm the whole data set. The Korean lacus-
forests, are much less likely to preserve tracks. trine deposits are not evaporitic and therefore do
Lockley et al. (1994b:242) suggested that not necessarily indicate a semiarid climate.
“sauropod tracks were most commonly associ- These data show no support for the idea that
ated with marine carbonate platform sequences sauropods preferred carbonate substrates. The
FIGURE 9.11. Paleoenvironmental and lithological associations of sauropod tracksites. South Ko-
rean localities counted as 10 (rather than 102), but because they are siliciclastic (Paik et al. 2001),
the addition of the total number would shift the balance away from carbonate localities, to only
25% carbonates. Number of sites in each plot is not the same because published reports do not al-
ways give both sets of data.
65% of tracksites from coastal or lacustrine envi- common in the Cretaceous, perhaps only
ronments may simply indicate that tracks in Cretaceous sites should be considered in deter-
such localities are more likely to be preserved. mining track association preferences (fig. 9.12).
There is no major shift in environmental pref- No attempt has been made herein to assess
erences through time (table 9.2), nor do the mak- paleoclimate of the tracksites, so claims that
ers of narrow or wide gauge trackways seem to many sauropod tracksites are found in semiarid
have had different habitat preferences (table 9.3). to seasonal climates cannot be assessed, except
Overall, most sauropod tracksites from the Late to note that this covers a wide range of climates
Jurassic and Early Cretaceous are preserved in and could be interpreted as describing the
carbonates, and more sauropod tracksites from majority of climates around the world.
other times are preserved in siliclastic sedi-
ments. This may reflect the predominant types DIFFERENCE IN DISTRIBUTION OF SAUROPOD
of terrestrial depositional environments in TRACKS AND BODY FOSSILS
prospected areas during these time periods, One of the most potentially interesting areas of
rather than trackmaker habitat preferences. A sauropod track research is the contrasting infor-
predominance of coastal tracksites may corre- mation that is revealed by tracks and skeletal
spond with an increase in coastal plain environ- remains. Most sauropod bones represent indi-
ments related to the breakup of Pangaea. viduals at least 80% of adult body size, while
Associations of sauropod tracks with other many tracks are those of very small individuals,
types of fossils are not reexamined herein so probably juveniles, as no known sauropods are
the assertion that sauropod tracks are preferen- this small as adults (Lockley et al. 1994c). This is
tially preserved with theropod tracks is not especially true of sauropod tracks from Asia;
tested. However, sauropod tracks have been recent discoveries from China have continued
found in association with ornithopod tracks this trend (Lockley et al. 2002b). The relatively
(Thulborn et al. 1994; Calvo 1999; Paik et al. high proportion of juvenile sauropod trackways
2001), although an association with no tracks may be interpreted in several ways. Juveniles
or theropod tracks seems to be more common. may have inhabited areas where footprints have
However, as ornithopod tracks only become a higher preservation potential, for instance,
TABLE 9.2
Lithology and Paleoenvironments by Trackway Gauge Through Time
AGE GAUGE LITHOLOGY PALEOENVIRONMENT
Late Triassic n Terrigenous clastics Fluvial

Early Jurassic 2n Terrigenous clastics River floodplain
5n&w Carbonates Marginal marine
Middle Jurassic 2w 1 sandstone Coastal
1 limestone Coastal
5n&w 1 limestone Coastal
4 clastics Fluvial
Late Jurassic w Clastics or limestone Mainly coastal
n Limestone or clastics Coastal or inland
Early Cretaceous w&n Ubiquitous Ubiquitous
Late Cretaceous w Siliciclastic Lacustrine
w Siliciclastic Marginal marine
NOTE: In the Late Cretaceous two geographic areas preserved the majority of these sites. Abbreviations: n, narrow; w, wide.
lakeshores, riverbanks, and lagoonal shorelines, indicate the sequence of weight distribution in
whereas the adults spent more time in areas the sauropod step cycle (Wilson and Carrano
where bones are more likely to be preserved, 1999). Where footprint depth variation has been
such as low-lying alluvial plains. Sauropod described the patterns are consistent. The
clutch sizes are thought to have been large, and anteromedial side of the pes print is the deepest;
many more sauropods would hatch out than sur- this forms the leading edge of the track and
vive to adulthood. Their greater numbers might would have been the push-off point for the recov-
account for the high proportion of small sauro- ery stroke. The next deepest part of the pes
pod trackways. impression is the lateral side of the track (mainly
the “heel”); this would have been the first part of
LOCOMOTION AND BEHAVIOR the foot to contact the substrate for the begin-
Trackways preserve direct evidence of behavior; ning of the power stroke. Brontopodus tracks
they are the nearest we are likely to get to a from the type locality, the topotype trackway of
“movie” of a dinosaur. Trackways can be used to Parabrontopodus (CU-MWC 190.3 [Lockley et al.
constrain locomotor hypotheses of their pro- 1994a]), and the Late Triassic sauropod track
ducers and have the potential to reveal details of CU-MWC 172.22 (fig. 9.8) all show this depth
the animal’s locomotion such as weight distri- distribution pattern.
bution, step cycle, or gait. Pittman and Gillette (1989) suggested from
Sauropod trackways unequivocally show that footprint evidence that sauropod pedal claws
the trackmakers had an upright stance and a might be wrapped along the lateral edge of the
generally parasagittal gait. All known sauropod foot when walking on firm substrates and
trackways show slow walking speeds, in keeping extended for added traction when turning, brak-
with their graviportal anatomy. Sauropod pes ing, or walking on more slippery substrates.
impressions range in length from 20 cm to Wrapping of the claws along the edge of the
more than 1 m, which gives estimated hip foot might help explain the lack of pedal claw
heights of 1–5 m (Thulborn 1990). impressions in many sauropod trackways.
Well-preserved sauropod trackways show Manus impressions seem to be deepest at
depth variations within footprints, which may the anterior side (digits I–III) and shallowest at
TABLE 9.3
Sauropod Track Morphotypes with Possible Trackmakers
CLAW/DIGIT GEOGRAPHIC POSSIBLE

MORPHOTYPE MANUS:PES GAUGE IMPRESSIONS AGE LOCATION TRACKMAKER REFERENCE:(S)
1 Small manus Narrow Distinct digit L Tr–E Jr USA, Wales, Eusauropod Lockley et al. 1996, 2001
“Tetrasauropus” impressions (Greenland, (Meyer & Thuring 2003;
& 4 pedal claw Italy, Dalla Vecchia 1996;
impressions Switzerland, Jenkins et al. 1994;
Morocco) Ishigaki 1988) (fig. 9.5A)
2 Small manus Wide or Indications of E Jr Italy, Morocco Eusauropod Ishigaki 1988; Leonardi &
variable manus digits; Avanzini 1994 (fig. 9.5B)
up to 4 pedal
claw impressions
3 Small manus Narrow May have 1 manus M–L Jr USA, N. Africa Neosauropod Lockley et al. 1994.
Breviparopus (1:5) claw impression; ?Diplodocid (fig. 9.5C–E)
(Parabrontopodus) up to 3 pedal claw
impressions
4 Large manus Narrow Large digit M Jr UK ??? Day et al. 2002 (fig. 9.5F)
I claw impression
5 Small manus Wide No manual claw M Jr UK Titanosaurid Day et al. 2002
impressions
6 Large manus Wide Very small manus EK USA, Germany, Titanosauriform Farlow et al. 1989
Brontopodus, (1:3) claw impression if S. America (fig. 9.5G)
(?Rotundichnus, present; 4 pedal
?Sauropodichnus) claw impressions
7 Very large Wide Large claw on L Jr Portugal ?Camarasauromorph Santos et al. 1994
manus (1:2) manual digit I; (fig. 9.5H, I)
4 large pedal
claw impressions
NOTE: These do not necessarily correspond to ichnogenera; several valid ichnogenera could be contained within one of these broad morphotypes. Abbreviations: E, Early; M, Middle; L, Late; J, Jurassic; K, Creta-
ceous; Tr, Triassic..
FIGURE 9.12. Sauropod tracksites preserving narrow, wide or both gauges of trackways plotted by
(A) lithology (n 90) and (B) environment (n 83). Numbers of sites are not the same because
not all publications give both sets of data.
the back, regardless of the orientation of the their large size, although large hadrosaur and
manus impressions. This may indicate that iguanodontid tracks with skin impressions seem
most of the weight on the manus was borne by to be more common, or it may simply be that the
the middle three digits. This might also partly substrates in which sauropod tracks have so far
explain the lack of manus claw impressions in been preserved do not provide sufficient resolu-
most sauropod tracks. tion to reveal such details.
Occasionally extra traces are preserved in Manus-only sauropod trackways have been
sauropod trackways, which provide extra evi- interpreted as sauropod swimming tracks by
dence of limb motion. Drag marks posterior to Ishigaki (1989), although because of the weight
manus impressions have been described for the distribution of the sauropod body, it is consid-
New York slab of Brontopodus from the Paluxy ered more likely that these are a preservational
River site (Farlow et al. 1989) and for tracks from artifact and the manus prints are undertracks
the Purbeck Limestone Group (Wright et al. punched through from a level above (Lockley
1997). These drag marks are slightly curved and and Rice 1990).
show that the forelimb described a slight lateral Several sauropod tracksites preserve a num-
arc during the recovery stroke. ber of parallel trackways (Bird 1944; Lockley
Finer morphological details of sauropod et al. 1986; Farlow et al. 1989). At some sites,
tracks, such as skin impressions, are not often such as Davenport Ranch, Texas, in the Glen
preserved, although they have been reported Rose Formation, gregarious behavior has been
(Lockley and Hunt 1995). These tracks showed demonstrated (Bird 1944; Lockley 1991). At
small polygonal scales in the sole of the foot; this others such as the Purgatoire tracksite, which is
is similar to the pattern reported for the soles of thought to have been an ancient lake shore, a
the feet of other dinosaurs such as hadrosaurs, geographic control is strongly indicated
iguanodontids, and theropods. The rarity of (Lockley et al. 1986). Many sites with multiple
sauropod skin impressions may be because of sauropod trackways either do not preserve a
large enough area to assess gregarious behavior up to 30 from the trackway midline and the dor-
(e.g., Lockley et al. 2002b) or are not preserved sal surface of the manus directed anterolaterally
in sufficient detail. at angles of up to 75 (figs. 9.5, 9.8). The gauge of
Multiple sauropod trackways heading in the sauropod trackways is mainly a product of pelvic
same direction on a single bedding plane are and hindlimb anatomy (Wilson and Carrano
preserved in numbers up to 25. This may be an 1999) of some sauropod groups (e.g., titanosauri-
indication of “herd” size, or we may not be find- forms), although in some instances it seems to be
ing larger “herd” sizes because bedding planes related to ontogeny or locomotory style. Sauropod
extensive enough to show larger groups are so tracksites indicate that sauropods sometimes
seldom exposed. Sauropod trackways at the traveled in groups, although no indications of
Davenport Ranch site in Texas (Glen Rose specific herd structure have been identified
Formation) were interpreted as showing that (Lockley 1991).
the larger animals were protecting the smaller Sauropod tracks have been found on every
ones, although a subsequent examination dis- continent except Antarctica. Their apparent dis-
proved this (Lockley 1991). No sauropod track- tribution solely in North America and Europe
sites have shown evidence of complex social before the Late Jurassic may well be the result
behavior such as a consistent or strategic herd of more intensive prospecting there, because
structure. Really large herds may have trampled sauropod skeletal remains have been found in
the ground so extensively that no discrete track- Asia in the Late Triassic (Buffetaut et al. 2001,
ways are preserved. There is no indication that 2002) and in southern Africa in the Early
sauropods traveled in herds of hundreds or Jurassic (Raath 1972). The tracks and bone fos-
thousands as has been inferred from bonebeds sil records appear to sample different assem-
of ceratopsians or hadrosaurs, even though blages and reveal complementary information.
tracks of these ornithischians do not give any This synthesis of currently known sauropod
indication of the huge size of their herds. tracksites indicates that sauropod tracksites are
preserved in approximately equal numbers in
siliciclastic and carbonate lithologies. Marginal
SUMMARY AND PROSPECTUS
marine and inland environments are also rep-
Despite the purported conservative nature of resented in approximately equal numbers. The
the sauropod foot, characteristic features of patterns of tracksite preservation in these
sauropod tracks can be identified and used to lithologies and environments through time
determine if footprints were indeed produced may reflect predominant depositional environ-
by sauropods. These characteristics allow us to ments with a high preservation potential at any
identify certain Late Triassic tracks as being of given time. If sauropods preferred mountain-
sauropod origin and allow us to distinguish a sides or forests as a living environment, their
number of other sauropod footprint morpho- fossils (tracks or bones) would be unlikely to be
types. In addition to predicted morphotypes, a preserved, so only statements about differing fre-
couple of other sauropod ichnomorphotypes quencies of occurrence in paleoenvironments
can be distinguished. These show sauropod with a high preservation potential can be made,
footprint characteristics but do not correspond and the data show no such difference. There are
to the podial anatomy of any known sauropod. currently insufficient data in both the skeletal
However, the feet of many sauropods are and the ichnological records to make any general
unknown. statements about sauropod habitat preferences.
Sauropod trackways show that sauropods pro- Sauropods were a globally distributed, long-lived
gressed quadrupedally with a parasagittal gait. group probably capable of exploiting a wide
Slow walking speeds are preserved at all track- range of habitats, lithologies, and climates
sites. Sauropods walked with their feet toed-out at (Farlow 1992, Thulborn et al. 1994).
Increasingly sophisticated computer pro- Bird, R. T. 1944. Did Brontosaurus ever walk on land?
grams now allow us to begin to investigate Nat. Hist. 53: 61–67
Buffetaut, E., Suteethorn, V., Cuny, G., Tong, H., Le
sauropod tracks in new ways. Computer ani-
Loeff, J., Khansubha, S., and Jongautcharlyaku, S.
mation programs allow the use of trackways as 2000. The earliest known sauropod dinosaur.
a control on models of dinosaur locomotion Nature 407: 72–74.
(Henderson 2002). Statistical modeling allows Buffetaut, E., Suteethorn, V., Le Loeff, J., Cuny, G.,
more detailed analysis of trackway patterns Tong, H., and Khansubha, S. 2002. The first
independent of variables related to the speed of giant dinosaurs: a large sauropod from the Late
Triassic of Thailand. C. R. Palevol 1: 103–109.
the trackmaker, and photogrammetric tech-
Calvo, J. 1991. Huellas de dinosaurios en la
niques can produce contour maps of footprints, Formacion Río Limay (Albiano-Cenomaniano?),
which can be used in more detailed locomotor Picun Leufu, Provincia de Neuquen, República
studies than have previously been possible. Argentina. Ameghiniana 28: 241–258.
Laser surveying can be used to make detailed ———. 1999. Dinosaurs and other vertebrates of the
Lake Ezequiel Ramos Mexia area, Neuquen–
maps accurate to a few millimeters (Wright et
Patagonia, Argentina. In: Tomida,Y., Rich, T., and
al. 1997). Such technology is likely to become Vickers-Rich, P. (eds.). Proceedings of the Second
ubiquitous and less expensive and will allow Gondwanan Dinosaur Symposium. National
tracks to be used in increasingly more sophis- Science Museum Monographs No. 15, Tokyo.
ticated ways to investigate sauropod biology Pp. 13–45.
further. Carrano, M. T., and Wilson, J. A. 2001. Taxon distri-
butions and the tetrapod track record.
Paleobiology 27: 564–582.
ACKNOWLEDGMENTS Casanovas, M., Fernandez, A., Pérez-Lorente, F., and
I thank Jeff Wilson, Kristi Curry Rogers, and Dan Santafé, J. V. 1997. Sauropod trackways from site
Chure for asking me to participate in this sympo- el Sobaquilla (Muniulla, La Rioja, Spain) indicate
amble walking. Ichnos 5: 101–107.
sium and to contribute a chapter to this book.
Dalla-Vechia, F. M. 1994. Jurassic and Cretaceous
Thanks are also due to Liz Cook, Jordon Hand, sauropod evidence in the Mesozoic carbonate
and Deborah Thomas for reading drafts of the platforms of the southern Alps and Dinards.
manuscript. This chapter was greatly improved GAIA 10: 65–74
by the comments of an anonymous referee. ———. 1996. Archosaurian trackways in the Upper
Carnian of Dogna Valley (Udine, Fruili, NE Italy).
Natura Nascosta 12: 5–17.
LITERATURE CITED Dal Sasso, C. 2003. Dinosaurs of Italy. C. R. Palevol.
2: 45–66.
Allain, R., and Superbiola, X. P. 2003. Dinosaurs of Day, J. J., Upchurch, P., Norman, D. B., Gale, A. S.,
France. C. R. Palevol. 2: 27–44. and Powell, H. P. 2002. Sauropod trackways, evo-
Antunes, M. T., and Mateus, O. 2003. Dinosaurs of lution and behavior. Science 296: 1659.
Portugal. C. R. Palevol. 2: 77–95. Dutuit, J.-M., and Ouazzou, A. 1980. Découverte
Avanzini, M., Frisia, S., van den Driessche, K., and d’une piste de Dinosaure sauropode sue le site
Keppens, E. 1997. A dinosaur tracksite in an early d’empreintes de Demnat (Haut-Atlas marocain).
Liassic tidal flat in northern Italy: paleoenviron- Mem. Soc. Geol. France 139: 95–102.
mental reconstruction from the sedimentology Ellenberger, P. 1972. Contribution à la classification
and geochemistry. Palaios 12: 538–551. des Pistes de Vertébrés du Trias: Les types du
Avanzini, M., Leonardi, G., Tomasoni, R., and Stormberg d’Afrique du Sud Palaeovertebrata,
Campolongo, M. 2001. Enigmatic dinosaur track- Mémoire Extraordinaire. Laboratoire de
ways from the Lower Jurassic (Pliensbachian) of Paléontologie de Vertébrés, Montpellier. 104 pp.
the Sarca Valley, northeast Italy. Ichnos 8: 235–242. Farlow, J. O. 1992. Sauropod tracks and trackmakers:
Barnes, F. A., and Lockley, M. G. 1994. Trackway evi- integrating the ichnological and skeletal records.
dence for social sauropods from the Morrison Zubia 10: 89–138.
Formation, eastern Utah (USA). GAIA 10: 37–42. Farlow, J. O., Pittman, J. G., and Hawthorne, J. M.
Bird, R. T. 1939. Thunder in his footsteps. Nat. His. 1989. Brontopodus birdi, Lower Cretaceous sauro-
43: 254–261 pod footprints from the US Gulf coastal plain. In:
Gillette, D. D., and Lockley, M. G. (eds.). Dinosaur Janensch, W. 1922. Das Handskelett von Gigantosaurus
Tracks and Traces. Cambridge University Press, und Brachiosaurus aus den Tendaguru-Schichten
Cambridge. Pp. 371–394. Deutsch-Ostafrikas: Centralblatt für Mineralogie,
Ferrusquía-Villafranca, I., Jiménez-Hidlago, E., and Geologie, und Paläontologie. Pp. 464–480.
Bravo-Cuevas, V. M. 1996. Footprints of small Janensch, W. 1961. Die Gliedmaszen und
sauropods from the Middle Jurassic of Oaxaca, Gliedmaszengürtel der Sauropoden der
southeastern Mexico. In: Morales, M. (ed.). The Tendaguru-Schiechten. Palaeontographica (Suppl.
Continental Jurassic. Mus. North. Ariz. Bull. 60: 7[1]) 3(4): 177–235.
119–126. Jenkins, F. A., Shubin, N. H., Amaral, W. W., Gatesey,
Fischer, R. 1998. Die Saurierfährten im S. M., Schaff, C. R., Clemensen, L. B., Downs,
Naturdenkmal Münchehagen. Mitteilung. Inst. W. R., Davidson, N. R., Bonde, B., and Osbaeck, F.
Geol. Paläontol. Univ. Hannover 37: 3–59. 1994. Late Triassic continental vertebrates and
Foster, J., Hamblin, A., and Lockley, M. G. 2000. The depositional environments of the Fleming Fjord
oldest evidence of a sauropod dinosaur in the Formation, Jameson Land East Greenland.
western United States and other important verte- Meddelelser Grønland Geosci. 32: 3–25.
brate trackways from Grand Staircase– Escalante Kaever, M., and Lapparent, A. F. de. 1974. Les traces
National Monument, Utah. Ichnos 7: 169–181. des pas de dinosaures du Jurasique de
Freise, H., and Klassen, H. 1979. Die Dinosaurier- Barkhausen (Basse Saxe, Allemagne). Bull. Soc.
fahrten von Barkhausen im Wiehengebirge. Geol. France 16: 516–525.
Veroffentl. Landkr. Osnabruck 1: 1–36. Le Loeff, J., Khansubha, S., Buffetaut, E.,
Gallup, M. R. 1989. Functional morphology of the Suteethorn, V., Tong, H., and Souillat, C. 2002.
hindfoot of the Texas sauropod Pleurocoelus sp. Dinosaur footprints from the Phra Wihan
Indet. In: Farlow, J. O. (ed.). Paleobiology of the Formation (Early Cretaceous of Thailand). C. R.
Dinosaurs. Geol. Soc. Am. Spec. Paper 238: 71–74. Palevol. 1: 287–292.
Gierlinski, G., and Sawicki, G. 1998. New sauropod Leonardi, G., 1989. Inventory and statistics of the
tracks from the Lower Jurassic of Poland. Geol. Q. South American dinosaurian ichnofauna and its
42: 477–480. paleobiological interpretation. In Gillette, D. D.
Gilmore, C. W. 1936. Osteology of Apatosaurus, with and M. G. Lockley, (eds) Dinosaur Tracks and
special reference to specimens in the Carnegie Traces, Cambridge University Press, Pp. 371–394.
Museum. Mem. Carnegie Mus. 9(4): 175–272. Leonardi, G. and Avanzini, M. 1994. Dinosauri in
Hand, J. D. 1999. A fully articulated Apatosaurus Italia. Le Scienze. Quaderni 76: 69–81, Milan.
manus from the Morrison Formation of Wyoming. Lim, S. K., Lockley, M. G., Yang, S.-Y., Fleming, R. F.,
Geological Society of America, Abstracts with and Houck, K. A. 1994. Preliminary report on
Programs. Vol. 30. sauropod tracksite from the Cretaceous of Korea.
Henderson, D. M. 2002. Wide and narrow gauge GAIA 10: 109–117.
sauropod trackways as a consequence of body Lockley, M. G. 1991. Tracking Dinosaurs. Cambridge
mass distribution and the requirement for stabil- University Press, Cambridge. 238 pp.
ity. J. Vertebr. Paleontol. 22: 64A. Lockley, M. G., and Hunt, A. P. 1995. Dinosaur
Hendricks, A. 1981. Die Saurierfährten von Tracks and Other Fossil Footprints of the Western
Münchehagen bei Rehburg-Loccum (NW-Deuts- United States. Columbia University Press, New
chland). Abhandlung. Landesmus. Naturkunde York. 338 pp.
Münster 43: 1–22. Lockley, M. G., and Meyer, C. A. 1999. Dinosaur
Insole, A. N., and Hutt, S. 1994. The palaeoecology of Tracks and Other Fossil Footprints of Europe.
the dinosaurs of the Wessex Formation (Wealden Columbia University Press, New York. 323 pp.
Group, Early Cretaceous), Isle of Wight, southern Lockley, M. G., and Rice, A. 1990. Did Brontosaurus
England. Zool. J. Linn. Soc. 112: 197–215. ever swim out to sea? Ichnos 1: 81–90.
Ishigaki, S. 1985. Nature Study 31: 5–8. (In Japanese.) Lockley, M. G., Houck, K., and N. K. Prince. 1986.
———. 1988. Les empreintes des dinosaures du North America’s largest dinosaur tracksite: impli-
Jurasique inferieur du Haut Atlas central maro- cations for Morrison Formation paleoecology.
cain. Notes Serv. Geol. Maroc 44: 79–86. Geol. Soc. Am. Bull. 57: 1163–1176.
———. 1989. Footprints of swimming sauropods Lockley, M. G., Farlow, J. O., and Meyer, C. A.
from Morocco. In: Gillette, D. D., and Lockley, 1994a. Brontopodus and Parabrontopodus ichno-
M. G. (eds.). Dinosaur Tracks and Traces. gen nov. and the significance of wide and nar-
Cambridge University Press, Cambridge. row gauge sauropod trackways. GAIA 10:
Pp 83–86. 126–34.
Lockley, M. G., Meyer, C. A., Hunt, A. P., and Lucas, Olsen, P. E., and Galton, P. 1984. A review of the rep-
S. G. 1994b. The distribution of sauropod tracks tile and amphibian assemblages from the
and trackmakers. GAIA 10: 233–248. Stormberg from Southern Africa with special
Lockley, M. G., Meyer, C. A., and Santos, V. F. 1994c. emphasis on the footprints and the Stormberg.
Trackway evidence for a herd of juvenile Paleontograp. Afr. 25: 87–110.
sauropods from the Late Jurassic of Portugal. Paik, I. S., Kim, H. J., and Lee, Y. I. 2001. Dinosaur
GAIA 10: 43–48. track-bearing deposits in the Cretaceous Jindong
Lockley, M. G., King, M., Howe, S., and Sharp, T. Formation, Korea: occurrence, palaeoenvironments
1996. Dinosaur tracks and other archosaur foot- and preservation. Cretaceous Res. 22: 79–92.
prints from the Triassic of South Wales. Ichnos 5: Pittman, J. G. 1989. Stratigraphy, lithology, deposi-
23–41. tional environment and track type of dinosaur
Lockley, M. G., Wright, J. L., Hunt, A. P., and Lucas, trackbearing beds of the Gulf Coastal Plain. In:
S. G. 2001. The Late Triassic sauropod track Gillette, D. D., and Lockley, M. G. (eds.). Dinosaur
record comes into focus: old legacies and new Tracks and Traces. Cambridge University Press,
paradigms. New Mexico Geological Society Cambridge. Pp. 135–154.
Guidebook, 52nd Field Conference, Geology of Pittman, J. G., and Gillette, D. D. 1989. The Briar
the Llano Estacado. Site: a new sauropod dinosaur tracksite in Lower
Lockley, M. G., Schulp, A. S., Meyer, C. A., Leonardi, G., Cretaceous beds of Arkansas, USA. In: Gillette,
and Mamani, D. K. 2002a. Titanosaur trackways D. D., and Lockley, M. G. (eds.). Dinosaur Tracks
from the Upper Cretaceous of Bolivia: evidence for and Traces. Cambridge University Press,
large manus, wide gauge locomotion and gregari- Cambridge. Pp. 313–332.
ous behaviour. Cretaceous Res. 23: 383–400. Prince, N. K., and Lockley, M. G. 1989. The sedimen-
Lockley, M. G., Wright, J. L., White, D., Li., J., Feng, tology of the Purgatoire Tracksite region,
L., Li, H., and Matsukawa, M. 2002b. The first Morrison Formation of southestern Colorado. In:
sauropod trackways from China. Cretaceous Res. Gillette, D. D., and Lockley, M. G. (eds.). Dinosaur
23: 363–381. Tracks and Traces. Cambridge University Press,
Lockley, M. G., Wright, J. L., and Thies, D. 2004. Cambridge. Pp. 155–164.
Some observations on the dinosaur tracks at Raath, M. A. 1972. Fossil vertebrate studies in
Münchehagen (Lower Cretaceous), Germany. Rhodesia: a new dinosaur (Reptilia: Saurischia)
Ichnos 11: 261–274. from near the Trias-Jurassic boundary. Arnoldia 5:
MacClary, J. S. 1938. Dinosaur trails of Purgatory. Sci. 1–37.
Am. 158: 72. Radley, J. D. 1994. Field meeting, 24–5 April 1993:
Mazzetta, G. V., and Blanco, R. E. 2001. Speeds of the Lower Cretaceous of the Isle of Wight. Proc.
dinosaurs from the Albian–Cenomanian of Geol. Assoc. 105: 145–152.
Patagonia and sauropod stance and gait. Acta Romano, M., and Whyte, M. A. 2003. Jurassic
Palaeontol. Polonica 46: 235–246. dinosaur tracks and trackways of the Cleveland
Meyer, C.A., Lockley, M.G., Robinson, J.W. and V. F. Basin, Yorkshire: preservation, diversity and
Santos, V.F. 1994. A comparison of well- distribution. Proc. Yorkshire Geol. Soc. 54:
preserved sauropod tracks from the Late Jurassic of 185–215.
Portugal and the Western United States: evidence Salgado, L., Coria, R.A., and Calvo, J.O. 1997. Evolution
and implications. Gaia: Revista de Geociencias, of titanosaurid sauropods. I: Phylogenetic analysis
Museu Nacional de Historia Natural, Lisbon, based on the postcranial evidence. Ameghiniana
Portugal. 10: 57–64. 34: 3–32.
Meyer, C.A. and Thuring, B., 2003. Dinosaurs in Santos, V. F., Lockley, M. G., Meyer, C. A., Carvalho,
Switzerland, Comptes Rendu Palevol 2: 103–117. J., Galopim de Carvalho, A. M., and Moratalla, J. J.
Moratalla, J. J., Garcia-Mondejar, J., Lockley, M. G., 1994. A new sauropod tracksite from the Middle
Sanz, J. L. and Jimenez, S. 1994. Sauropod track- Jurassic of Portugal. GAIA 10: 5–14.
ways from the Lower Cretaceous of Spain. GAIA Schulp, A. S., and Brokx, W. A. 1999. Maastrichtian
10: 75–83. sauropod tracks from the Fumanya site,
Novikov, V. P., Suprichev, V. V., and Salikhov, F.S. Berguedà, Spain. Ichnos 6: 239–250.
1993. Dinosaur tracksites in the Obikhingou Thulborn, R. A. 1990. Dinosaur Tracks. Cambridge
River Basin (Peter the Great range). Rep. Acad. University Press, Cambridge. 410 pp.
Sci. Tadjikistan Repub. 1(36): 55–58. Thulborn, T., T., Hamley, and Foulkes, P. 1994.
Olsen, P. E. 1995. A new approach for recognizing Preliminary report on sauropod dinosaur tracks
track makers. Geol. Soc. Am. Abstr. Progr. 27: 86. in the Broome Sandstone. GAIA 10: 85–94.
Unwin, D. M. 1989. A predictive method for the Wilson, J. A., and Sereno, P. C. 1998. Early evolution
identification of vertebrate ichnites and its appli- and higher-level phylogeny of sauropod dinosaurs.
cation to pterosaur tracks. In: Gillette, D. D., and Society of Vertebrate Paleontology Memoir 5.
Lockley, M. G. (eds.). Dinosaur Tracks and Traces. Wilson, J.A. and Upchurch, P., 2003. A revision of
Cambridge University Press, Cambridge. Pp. Titanosaurus Lydekker (Dinosauria–Sauropoda),
259–274. the first dinosaur genus with a ‘gondwanan’ dis-
Upchurch, P. 1994. Manus claw function in sauro- tribution. Journal of Systematic Palaeontology
pod dinosaurs. GAIA 10: 161–72. 1: 125–160.
——. 1998. The phylogenetic relationships of sauro- Wright, J. L., Radley, J. D., Upchurch, P., and
pod dinosaurs. Zool. J. Linn. Soc. 124: 43–103. Wimbledon, W. A. 1997. Keates’ Quarry dinosaur
Weishampel, D., Dodson, P., and Osmolska, H. footprint site, Intermarine Member, Purbeck
1990. The Dinosauria. University of California Limestone Group (Berriasian). UK National Trust
Press, Berkeley. 733 pp. Site Report.
Wilson, J. A., and Carrano, M. T. 1999. Titanosaurs Zhang, Y., Yang, D., and Peng, G. 1984. [New materi-
and the origin of “wide gauge” trackways: a bio- als of Shunosaurus from the Middle Jurassic of
mechanical and systematic perspective on sauro- Dashanpu, Zigong, Sichuan.] J. Chengdu Coll.
pod locomotion. Paleobiology 25: 252–267. Geol. Suppl. 2: 1–12. (In Chinese.)
APPENDIX 9.1. PODIAL SYNAPOMORPHIES OF SAUROPODOMORPHS AND THEIR
POTENTIAL EXPRESSION IN TRACKS
Abbreviations: C, Upchurch (1998) numbered characters; WS, Wilson and Sereno (1998) numbered characters; Sa,
Salgado et al. (1997) numbered characters. See also Carrano and Wilson (2001).
CHARACTER PREVIOUS NO. SYNAPOMORPHY ICHNOLOGICAL EXPRESSION

NO.
Sauropoda: sauropodomorphs more closely related to Saltasaurus than Plateosaurus (Wilson & Sereno 1998)
Posture 1 C 186, C 158, vQuadrupedal posture Trackway should show manus
WS 1 and pes impressions
Pes 2 WS 14 Proximal ends of metatarsals I Tracks showing that fewer
& V subequal in area to those than five digits contacted the
of metatarsals II & IV: metatarsal ground cannot have been
I may even be larger in area made by sauropods
than metatarsal II (fig. 9.3)
3 WS 15 Length of metatarsal V 70%
metatarsal IV (fig. 9.3)
4 WS 16 Entaxonic pes structure Strong decreasing gradient
in ungual size from digit I
to digit V should be visible in
footprints
5 WS 17 Ungual of pedal digit I Digit I pedal claw impressions
deep and narrow are predicted to be visible,
and sickle-shaped and oblique
to the ground rather than
triangular
Manus 6 C 168 Metacarpal V robust and 90% Digital impressions,
of the length of the longest if present, should show
metacarpal; all five manual five digits.
digits participated in weight
bearing
Eusauropoda (Upchurch 1995): sauropods more closely related to Saltasaurus than to Vulcanodon
Manus 7 C 170, Phalanges on manual digits II Digital impressions should be
WS 43 & III reduced; all digits subequal of a similar depth. Large claw
in length apart from digit I, impression should be apparent
which bears a very large claw on the medial side of the
manus print
8 Manual phalanges (other than Short, wide digit impressions,
unguals) broader than long, with no discrete phalangeal
without well-formed collateral pads
ligament pits (fig. 9.3).
9 WS 50 Metatarsal III length 25% that Eusauropod pes impressions
of tibia; overall reduction in would be relatively larger, in
relative length of metatarsals length and width, than those
of more basal sauropods
10 WS 52 Metatarsals with spreading Presence of fleshy pad
configuration; indicates a (Janensch 1922)—pes
sloping orientation for the impressions would be
metatarsals; semidigitigrade widest anteriorly,
pedal structure narrowing to the rear.
NO.
11 WS 51, Metatarsal I thicker than all Sauropod pes tracks are

C 198 other metatarsals, short, and turned outward at up to
very robust; asymmetric pes 30 from the trackway
midline; digit I is the leading
digit in the foot and bears
the most force during push-off.
Leading margin of sauropod
pes impressions (anteromedial
side of the pes) often deepest
part of print (Wilson & Sereno
1998)
12 WS 53 Nonterminal phalanges of Reduced digit lengths; pedal
pedal digits broader than long digit impressions will be very
short and show no discrete
phalangeal pads
13 C 202 Loss/extreme reduction of
collateral ligament pits on
pedal phalanges
14 WS 55, Penultimate phalanges of pedal
C 200 digits II–IV rudimentary or
absent
15 WS 56 Pedal digits I–III have Deep narrow claws for all three
sickle-shaped unguals digits should be evident
16 WS 54 Ungual of pedal digit I longer Ungual I very large and
than metatarsal I should be evident
17 WS 57 Ungual of pedal digit IV Tracks should show three
rudimentary or absent large claws, decreasing in
size from I to III (see
no. 4 above) and either no
or a small claw impression
for digit IV; digit V should
not bear a claw, although it
might leave a short blunt
digit impression. Any track
with five pedal claw
impressions could not
have been made by a
eusauropod
18 WS 64 Pedal unguals asymmetrical. Footprints should show pedal
All pedal unguals lay with part claw impressions pointing
of their medial surface on the anterolaterally with respect
ground and with the distal end to the pes axis. Individual
of the ungual directed laterally claw impressions oblique to
the ground, flattened, shallow,
and curved
19 WS 73 Metatarsals III & IV with May indicate that even
minimum transverse shaft more weight is being
diameters 65% of placed on the inner side
NO.
metatarsals I & II of the pes, but such a

subtle difference is unlikely
to be discernible in footprints
Eusauropoda (Upchurch 1995): sauropods more closely related to Saltasaurus than to Vulcanodon.
Manus 20 WS 80 Long intermetacarpal Shapes and size of manus
articulations; metacarpals impressions within trackways
bound by ligaments into a should be very consistent;
rigid column; digitigrade more consistent morphology
manus than those of basal sauropods
21 WS 81 Metacarpal proximal ends Horseshoe-shaped rather than
subtriangular; arc of ca. 270a crescentic manus impressions
22 C 171 Manual phalangeal formula Unlikely to be evident
reduced to 2-2-1-1; further in footprints
reduction of manus digit
length
Pes 23 C 201 Pedal phalangeal formula Unlikely to be evident
reduced to 2-3-4-2-1; further in footprints.
reduction of pes digit length
24 Proximal pedal phalanges
narrow toward their lateral
and palmar margins.
Camarasauromorpha (Salgado et al. 1997): the most recent common ancestor of Camarasauridae and
Titanosauriformes
Manus 25 Sa 12, Lengthened metacarpals May have deeper manus
C167 relative to radius; forelimb than pes impressions
WS 93 relatively longer; Shift in
weight distribution?
26 WS 94 Metacarpal I subequal in length As above
to metacarpal IV; relative
increase in length; shift from
metacarpals II–IV being the
longest to I–III; Related to
Titanosauriformes: Brachiosaurus titanosaurs
Manus 27 WS 98 Distal condyle of metacarpal I Even deeper manus
undivided, phalangeal articular impressions?—unlikely
surface reduced; Shift in body to be evident in footprints
28 Sa 16(1) Claw on manual digit I reduced Claw impressions in
titanosauriform trackways
should be small
Pes 29 C 187, Femoral head deflected Linked to production of
WS 100 medially wide gauge trackways
(Wilson & Carrano 1999)
NO.
Titanosauria: (Wilson & Upchurch 2003)

Manus 30 C 166 Metacarpal I longer than Even deeper manus
metacarpal III and subequal impressions?—unlikely to
to or longer than metacarpal II; be evident in footprints
completion of the transition
of the main weight- bearing
digits of the manus from II–IV
to I–III
31 Sa 16(2) Claw on manual digit I reduced No claw impressions should be
visible in the manus tracks
32 Sa 27 Manual phalanges absent, No phalangeal impressions
related to no. 31 above; there should be visible in manus
can be no claws if there are no tracks of titanosaurids,
phalanges although distal ends of
the metacarpals may have
borne pads that left
impressions visible in tracks
Pes 33 Deflected femoral axis and Even wider gauge trackways
beveled knee joint (Wilson &
Carrano 1999); development of
no. 29 above
TEN
Nesting Titanosaurs from Auca Mahuevo and

Adjacent Sites
UNDERSTANDING SAUROPOD REPRODUCTIVE BEHAVIOR AND EMBRYONIC DEVELOPMENT
Luis M. Chiappe, Frankie Jackson, Rodolfo A. Coria, and Lowell Dingus
housands of sauropod egg clutch spatial distribution and egg-bed strati-

T clutches, some containing eggs with ex-
quisitely preserved embryonic bone and integu-
graphic position, together with studies of their
sedimentary context, have provided the basis for
ment, have been discovered in the Late Creta- inferring aspects of the nesting behavior of
ceous nesting site of Auca Mahuevo (Chiappe et these dinosaurs (Chiappe et al. 2000; Chiappe
al. 1998, 2000, 2001, 2004; Dingus et al. 2000; and Dingus 2001). Combining taxonomic con-
Chiappe and Dingus, 2001; Coria et al. 2002) and straint with extensive sampling, research at
adjacent localities in northwestern Patagonia, Auca Mahuevo and its adjacent localities offers
Argentina (fig. 10.1). Five expeditions to this ex- the clearest picture to date of sauropod repro-
traordinary area (1997, 1999, 2000, 2001, and duction and embryonic development. In this
2002) have yielded a wealth of information for chapter, we summarize the major develop-
understanding the prehatching development, ments of this research program and discuss
the nesting structure, the egg morphology and their significance in light of previous interpreta-
malformation, and the reproductive behavior of tions of the reproductive biology of these colos-
these dinosaurs. Cranial characters of the in ovo sal dinosaurs.
embryos allowed the identification of the eggs as
those of titanosaurs (Chiappe et al. 1998,
GEOLOGICAL SETTING AND ASSOCIATED
2001; Salgado et al. 2005). Textural differences
DINOSAUR FAUNA
in the sediments containing some clutches
have illuminated aspects of the nest structure of Auca Mahuevo lies approximately 120 km north-
these animals (Garrido et al. 2001; Chiappe et al. west of the city of Neuquén in the homonymous
2004). Microstructural studies have expanded Argentine province (fig. 10.1). Two adjacent
our understanding of the eggshell variation nesting sites, Barreales Norte and Barreales Es-
(Grellet-Tinner et al. 2004) and incidence of egg condido, are 15 and 22 km south of Auca
malformation (Jackson et al. 2001, 2004) within Mahuevo, respectively. These three sites occur
a titanosaur population. Detailed mapping of within an 85-m-thick sequence of sandstone,
285
FIGURE 10.1. Map of the
province of Neuquén (Argentina)
indicating the location of Auca
Mahuevo. Barriales Norte and
Barreales Escondido are 15 and
22 km south of Auca Mahuevo,
respectively.
siltstone, and mudstone of the Anacleto Forma- splay deposits occur within the Auca Mahuevo
tion (fig. 10.3), one of the lithostratigraphic units section, and in egg-bed 4 they occasionally pre-
of the fossiliferous Cenomanian–Campanian serve nesting trace fossils (Chiappe et al. 2004).
Neuquén Group (Ramos 1981; Legarretta and Less than 1 m above the nesting structures
Gulisano 1989; Ardolino and Franchi 1996; in egg-bed 4 is a sandy red mudstone that con-
Leanza 1999; Dingus et al. 2000). Recent paleo- tains a great number of sauropod tracks (Loope
magnetic analysis of rocks from the lower por- et al. 2000). The sauropod tracks are recogniz-
tion of the Auca Mahuevo section containing able as thin (1-cm-thick), laterally discontinuous
egg-beds 1–3 established the presence of a Re- limy deposits that measure up to 80 cm in
versed magnetozone in the Anacleto Formation diameter and are oval to circular in shape.
(Dingus et al. 2000). In conjunction with earlier These platter-shaped features are interpreted to
biochronologic correlations, this magnetozone contain precipitates of calcium carbonate
was tentatively correlated with C33R, in the within the track depression, possibly from evap-
early–middle Campanian, between 83.5 and 79.5 oration of standing water (Loope et al. 2000).
million years ago (Dingus et al. 2000). Traceable over several kilometers, the track
Exposures at Auca Mahuevo include at least horizon provides an index layer useful for veri-
four distinct egg-bearing layers (figs. 10.2, 10.3; fying that the underlying nesting traces occur
egg-beds 1–4), which occur in uniform mud- on the upper surface of a single sandstone stra-
stones representing overbank deposits on a flu- tum. Additional layers of calcium carbonate
vial plain (Chiappe et al. 2000). Two of these lay- precipitates, also interpreted as footprints,
ers (egg-beds 2 and 3) can be subdivided into two occur elsewhere in the Auca Mahuevo section.
horizons of eggs, separated by a few centime- Clutches from egg-bed 3 occur in paleoverti-
ters of sediments in the case of egg-bed 3 and sols (Chiappe and Dingus 2001), recognizable
about 1 m in the case of egg-bed 2. Egg-beds 3 by the abundance and widely varying orienta-
and 4 are laterally continuous for at least sev- tion of slickensides—striated surfaces pro-
eral kilometers (Chiappe et al. 2000; fig. 10.2). duced by soil movement within the nesting
Most egg clutches exhibit no discernible evidence ground. Vertisols today are associated with
of nest structure. However, thin sandstones clay-rich parent materials and are widespread
representing abandoned channel and crevasse in regions that experience wet–dry climatic
286 N E S T I N G T I TA N O S AU R S F R O M AU C A M A H U E V O A N D A D J A C E N T S I T E S
FIGURE 10.2. Aerial photograph showing egg-beds 3 and 4 at Auca Mahuevo. These egg-beds can be traced later-
ally for several kilometers.
cycles under semiarid to subhumid environmen- tion above an egg-bed at Barreales Norte. A
tal conditions. Similar depositional conditions footprint layer similar to those of Auca
have been inferred for the Late Jurassic–Late Mahuevo also occurs at this locality at compara-
Cretaceous of the northern third of Patagonia, ble stratigraphic positions. In addition, compa-
where the climate regime has been recon- rable thicknesses separate egg-beds at these
structed as warm, arid to semiarid, and with a three localities.
distinct dry season (Andreis 2001). Several fossils of adult sauropod and thero-
Several egg-beds also occur at Barreales pod dinosaurs have been found at Auca
Norte and Barreales Escondido. Although Mahuevo. Remains of titanosaurs have been
stratigraphic correlations between these locali- collected from egg-bed 4 and from strata
ties and Auca Mahuevo are still preliminary, the between this egg layer and egg-bed 3 (fig. 10.3).
available data suggest that these egg-beds can These remains are yet to be studied in detail.
be traced across the vast distances that separate Among the theropods found at this site is the
the three sites. Abundant geodes of pale blue nearly complete skeleton of Aucasaurus garridoi
celestite crystals that occur in a discrete horizon (Coria et al. 2002), an abelisaurid collected
above Auca Mahuevo’s egg-bed 3 are also found from a laminated mudstone unit some 25 m
at approximately the same stratigraphic posi- above egg-bed 4. Isolated teeth comparable in
N E S T I N G T I TA N O S AU R S F R O M AU C A M A H U E V O A N D A D J A C E N T S I T E S 287
FIGURE 10.3. Composite strati-
graphic section at Auca
Mahuevo. Note the presence of
four stratigraphically distinct
beds of essentially identical ti-
tanosaur eggs. Egg-beds 2 and 3
can each be subdivided into two
layers.
morphology with those of dromeosaurids and titanosaur remains. The eggs are spherical to
the fragmentary remains of a large indetermi- subspherical and approximately 13–15 cm in
nate theropod the size of a charcharodon- diameter, with a tubercular surface ornamenta-
tosaurid were also collected from Auca tion consisting of single, rounded nodes (fig.
Mahuevo (Coria and Arcucci 2005). 10.4). The eggshell consists of a single structural
layer of calcite—approximately 1.3 mm thick in
well-preserved samples—pierced by a pore net-
EGGS, CLUTCHES, AND NESTS
work of vertical and horizontal canals that inter-
Eggs from all these localities exhibit similar size, sect one another at the bases of the eggshell units
shape, microstructure, and surface ornamenta- (Grellet-Tinner et al. 2004). The overall
tion as the eggs that contain diagnostic microstructure of the eggshell is similar to that
FIGURE 10.4. (A) Auca Mahuevo clutch (Museo Carmen Funes, Plaza Huincul, Argentina; MCF-PVPH-258; quarry
of egg-bed 3 illustrated in fig. 10.5) containing nearly 40 eggs. (B) and (C) Scanning electron micrograph and thin
section of the Auca Mahuevo eggshell. Scale bars in B and C equal 10 micron and 1 mm, respectively.
described as the ootaxon Megaloolithus patagoni- named from various localities around the world.
cus from the Anacleto Formation at Neuquén Vianey-Liaud and others (2003) took an initial
City (Calvo et al. 1997; fig. 10.4). Megaloolithus step in this direction and greatly reduced the
patagonicus has recently been proposed as a pos- number of valid megaloolithid oospecies from
sible junior synonym of Megaloolithus jabalpuren- India.
sis (Vianey-Liaud et al. 2003), a Late Cretaceous Although it is difficult to ascertain the com-
(Maastrichtian) oospecies from India. pleteness of a fossil egg-clutch, Auca Mahuevo’s
Megaloolithus patagonicus is also very similar to clutches are composed of numerous eggs, most
Late Cretaceous eggshells from Perú identified typically from about 20 to nearly 40 eggs
by Vianey-Liaud et al. (1997) as Megaloolithus (Chiappe et al. 2004; Jackson et al. 2004; fig.
pseudomamillare (Grellet-Tinner et al. 2004). 10.4). Eggs are stacked one on top of the other
Future studies are likely to synonymize many of without any internal spatial arrangement (up to
the megaloolithid oospecies that have been three stacked layers of eggs have been described
FIGURE 10.5. Map in plan view of eggs exposed at a quarry in Auca Mahuevo’s egg-bed 3. The total surface area excavated in this quarry is
approximately 65 m2.
FIGURE 10.6. Clutch maps of two sites contained within erosional surfaces of Auca Mahuevo’s egg-
bed 3. Seventy-four and 31 randomly distributed egg clutches were mapped within 1,701 m2 (A) and
486 m2 (B). Even if recognition of egg-clutch boundaries is sometimes difficult (see text), these
numbers provide a minimal estimate of the clutches laid on these surfaces.
for some clutches). More than 500 whole eggs to edaphic processes—a substantial displace-
were quarried over a 65-m2 surface of egg-bed ment along the friction planes of slickensides
3 during the 1999, 2002, and 2002 field sea- has been observed on several instances. It is also
sons (fig. 10.5). Spatial analysis of the eggs possible that the eggs were somewhat displaced
exposed during 1999, in conjunction with a by flotation prior to their final burial. Detailed
three-dimensional map constructed from field stratigraphy and mapping of egg depths within
data, revealed an unexpectedly high egg density this quarry also revealed two distinct levels of
(11 eggs/m2) in this quarry. At this quarry, the eggs, separated by several centimeters of sedi-
boundaries of individual egg clutches are some- ment, which were interpreted as different egg-
times difficult to determine. In some cases, eggs laying events (Chiappe et al. 2000).
occur as large accumulations that clearly repre- Mapping of egg clutches exposed on ero-
sent more than one clutch (fig. 10.5). To a certain sional surfaces of egg-bed 3 produced a concen-
degree, such a pattern can be explained as the tration of 74 and 31 randomly distributed egg
consequence of postburial egg displacement due clutches within 1,701 and 486 m2, respectively
FIGURE 10.7. Egg-clutch map of a site exposed on an erosional surface of Auca Mahuevo’s egg-bed 4. This map shows
a density comparable to that recorded for egg-bed 3 (fig. 10.6).
(Chiappe et al. 2000; fig. 10.6). Egg clutches on of clutches provides opportunities for studying
erosional surfaces are often recognized as accu- aspects of sauropod reproductive biology that
mulations of partially weathered eggs whose are otherwise difficult to assess. For example,
periphery (eggshell) is still vertically oriented the discovery of abnormal, multilayered eggs in
within the substrate. Heavily weathered clutches 6 of nearly 400 in situ clutches surveyed at Auca
are recognized as large accumulations of broken Mahuevo’s egg-beds 2 and 3 (Jackson et al. 2001,
eggshells that fan out from a core and that are 2004) provided the first assessment of the inci-
separated from other egg clutches by areas with dence of egg malformation in a sauropod popu-
minimal eggshell. If any, maps constructed lation. This study detected three different types
with these criteria for recognition underesti- of abnormal eggshells and the fact that patho-
mate the number of egg clutches that were laid logic eggs were laid in clutches containing a
within these paleosurfaces because of the above- majority of normal eggs.
mentioned causes of egg accumulation and dis- Several egg clutches from egg-bed 4 pre-
placement. The maximum stratigraphic thick- serve evidence of nest architecture. The eggs of
ness of strata containing egg clutches within these clutches are similar in size, shape, and
these areas was less than 70 cm. A similar map microstructure to other Auca Mahuevo eggs
of an erosional surface of egg-bed 4 showed a containing embryonic remains of titanosaur
comparable high density of randomly distributed sauropods (Chiappe et al. 2001). The clutches
egg clutches (fig. 10.7). This high concentration are contained in large, subcircular to subellipti-
FIGURE 10.8. Four titanosaur nests from Auca Mahuevo’s egg-bed 4. These nests consist of surface
depressions whose periphery is defined by an elevated rim of structureless sand. Textural differences
between the depression fill and the substrate indicate that titanosaurs laid their eggs on the surface.
cal to kidney-shaped depressions in sandstone, We have interpreted the structureless sand of

although the depression and interstitial spaces the depression’s rim as the piled debris produced
between the eggs are filled with mudstone during the construction of the nest and the
(fig. 10.8). All of the depressions truncate pri- mudstone surrounding the eggs as the result of
mary stratification of the host substrate and are a flooding event that is similar to most over-
encircled by a rim of structureless sandstone. bank deposition of the Anacleto Formation in
The maximum axis of the depressions ranges the Auca Mahuevo section. The fact that the
in length between 100 and 140 cm and their eggs in the recognized nests are not entombed
depth is about 10 to 18 cm. These egg-filled by sandstone but by mudstone resulting from
depressions are interpreted as excavated nests the flooding event indicates that the nesting
on the basis of lithologic criteria (Garrido et al. sauropod did not bury the eggs after they were
2001; Chiappe et al. 2004): (1) the depressions laid (Chiappe et al. 2004). All other clutches
truncate the primary stratification of the host from Auca Mahuevo are likely to have been laid
sandstone, (2) a massive rim surrounds the in similarly constructed surface nests, which
perimeter of the depression, and (3) textural are not recognizable due to the lack of textural
differences exist between the host sandstone differences between the host mudstone and the
and the in-filling sediment. in-filling sediment. These nests could have
FIGURE 10.9. Titanosaur embryonic remains from Barreales Norte, 15 km south of Auca Mahuevo. Note the
remains of limb bones covered by the eggshell of tuberculate ornamentation.
been lined and covered with vegetation (Grellet- erosional surfaces and (2) compressed against
Tinner et al. 2004), a suggestion that several the bottom inner shell of in situ eggs. Despite the
researchers have made for other megaloolithid- abundance of embryos, no remains of hatchlings
type eggs from the Late Cretaceous of France or early juveniles have been discovered. So far,
(Erben 1970; Kérourio 1981; Cousin 1997). most embryonic remains have been collected
from Auca Mahuevo’s egg-bed 3 (Chiappe et al.
1998, 2001), although some have also been
EMBRYONIC MORPHOLOGY
found at Barreales Norte (fig. 10.9) and Barreales
AND SYSTEMATICS
Escondido.
In ovo embryonic remains have been found in Patches of integument, preserved as calcitic
two situations: (1) encased in highly cemented impressions (negative and positive), are com-
fragments of eggs that occur as “float” on mon in highly cemented egg fragments that
FIGURE 10.10. Photos and interpretive drawings of embryonic titanosaur skulls from Auca Mahuevo’s egg-
bed 3 in left lateral view. (A) MCF-PVPH-272; (B) MCF-PVPH-263. Arrows point to the approximate location
of external nares. Abbreviations: af, antorbital fenestra; an, angular; d, dentary; f, frontal; itf, infratemporal
fenestra; j, jugal; la, lacrimal; m, maxilla; mf, mandibular fenestra; orb, orbit; p, parietal; pmx, premaxilla; po,
postorbital; prf, prefrontal; pt, pterygoid; q, quadrate; qj, quadratojugal; scp, scleral plates; sq, squamosal; stf,
supratemporal fenestra.
occur on erosional surfaces. These skin impres- forming part of the ventral margin of the skull,
sions display a range of nonoverlapping tuber- minimally retracted nares).
cular patterns including rosettes, flowerlike The anatomical information available in the
arrangements, and rows of larger tubercles handful of embryos initially prepared sup-
(Chiappe et al. 1998). ported the identification of these embryos as
Osteological remains are most typically neosauropod dinosaurs (Chiappe et al. 1998),
found flattened against the bottom inner shell the clade originating from the common ances-
of in situ eggs. Cranial material is better ossi- tor of the Late Jurassic Diplodocus longus and the
fied than limb material, which typically lacks Late Cretaceous titanosaur Saltasaurus loricatus
ends (fig. 10.9). Some embryos are clearly (Wilson and Sereno 1998). Originally, although
larger (by as much as 25%) than others. synapomorphies of Sauropoda were identified
Although it is difficult to estimate the degree of among the disarticulated skulls (e.g., jugal
development of the embryos in comparison to process of the postorbital much longer than the
the embryonic stages of extant reptiles, the rostrocaudal extension of the dorsal end of this
basic sauropod morphogenetic plan is readily bone [Chiappe et al. 1998]), the dental mor-
visible in the available skulls. A general phology played a key role in the taxonomic
“Haeckelian pattern” is also evident in these identification of the embryos: the smooth
skulls, which in some respects resemble condi- enamel (devoid of denticles) of the crowns and
tions found outside Sauropoda (e.g., jugal the pencil shape (straight margins and tapering
FIGURE 10.11. Simplified cladogram of sauropod relationships. Derived dental characters present in
the Auca Mahuevo embryos support their inclusion in a Subgroup of Titanosauria that excludes the
African Malawisaurus dixeyi. (After Salgado et al. 1997.) Icons from Wilson and Sereno (1998).
crowns) of the teeth, derived characters of postorbital bar that is broader transversely than
neosauropods and diplodocoids titanosaurs, rostrocaudally (Chiappe et al. 2001). Most
respectively (Wilson and Sereno 1998; Wilson importantly, comparisons between these more
2002), supported the placement of the embryos recently discovered embryonic skulls and
within Neosauropoda. However, because pencil- titanosaur cranial remains revealed several
shaped teeth are commonly interpreted as inde- apparent synapomorphies of this sauropod
pendently evolved in both diplodocoids and clade. For example, the embryos exhibit the
titanosaurs (e.g., Salgado and Calvo 1997; same ventral notch of the dentigerous margin,
Wilson and Sereno 1998; Curry Rogers and between the maxilla, jugal, and quadratoju-
Forster 2001; Wilson 2002), the presence of gal, as the titanosaurs Rapetosaurus krausei
this condition alone was insufficient for placing (Curry Rogers and Forster 2001, 2004) and
the Auca Mahuevo embryos within either of Nemegtosaurus mongoliensis (Nowinski 1971),
these two neosauropod clades. Subsequent dis- from the Late Cretaceous of Madagascar and
coveries of embryos yielding more complete Mongolia, respectively. They also share with
and articulated skulls (Chiappe et al. 2001; this taxon and with other titanosaurs from the
Salgado et al. 2005) (fig. 10.10) revealed addi- Late Cretaceous of Argentina (i.e., Antarctosaurus
tional synapomorphies of neosauropods and of wichmannianus and an undescribed titanosaur
the more inclusive taxon Eusauropoda (fig. from northwestern Patagonia) the low rostral
10.11)—taxa more closely related to Saltasaurus portion of the dentary (Chiappe et al. 2001).
loricatus than to Vulcanodon karibaensis (Wilson The remarkable width of the skull roof, as
and Sereno 1998). For example, the new inferred from the size of the frontals and pari-
embryos display the eusauropod condition of a etals, and the presence of a large mandibular
stepped snout, an absence of a fossa surround- fenestra are other derived features shared by
ing the antorbital fenestra, a rostral expansion the embryos and the above-mentioned unde-
of the quadratojugal, and a lack of contact scribed titanosaur from northwestern Patagonia
between this bone and the squamosal (Chiappe (Coria and Salgado 1999). These embryonic
et al. 2001; fig. 10.10). Likewise, the new skulls represent the most complete titanosaur
embryos exhibit the neosauropod condition of a crania. Furthermore, the embryonic material
suggests that previous skull reconstructions of inverted tear-shaped appearance. The rostrum
these dinosaurs as “camarasauroid” (Salgado became substantially enlarged, probably as a
and Calvo 1997) are incorrect. Neither a vertical consequence of maxillary expansion, and the
orientation of the quadrate nor the presence of maxilla developed a connection with the
broad nares separated by an elevated premaxil- quadratojugal, thus excluding the jugal from
lary–nasal arch, conditions typical of “cama- the ventral margin of the skull. In addition, the
rasauroid” skulls, is present in the well-pre- external nares expanded in size and migrated
served skulls of the Auca Mahuevo embryos backward, to be relocated on top of the orbits.
(Chiappe et al. 2001). In addition, the Auca Mahuevo embryos
Although the recently discovered embryos have provided evidence that may potentially clar-
have provided support for identification of the ify the sequence of transformations that
Auca Mahuevo eggs as those of titanosaurs, occurred during the long evolution of sauropods
systematic placement of the embryos beyond (Chiappe et al. 2001). An example of how these
Titanosauria (i.e., all sauropods more closely new developmental data could potentially eluci-
related to Saltasaurus loricatus than to either date aspects of sauropod evolution is provided
Brachiosaurus brancai or Euhelopus zdanskyi by examining two salient features of their cra-
[Wilson and Sereno 1998]) remains problematic. nial architecture: the narial retraction and the
This is primarily because of the paucity of cra- forward rotation of the braincase. Salgado and
nial anatomical information available for adult Calvo (1997) suggested that the partial to
titanosaurs (Salgado and Calvo 1997; Curry extreme retraction of the eusauropod nares
Rogers and Forster 2001). The presence of could have been evolutionarily coupled to the
pencil-like teeth, so far known only for forward rotation of the braincase, best exempli-
Nemegtosaurus mongoliensis (Nowinski 1971), fied in diplodocids (McIntosh 1997) and some
Alamosaurus sanjuanensis (Kues et al. 1980), titanosaurs (Salgado and Calvo 1997; Curry
and saltasaurines (most recent common ances- Rogers and Forster 2001, 2004). Because the
tor of Neuquensaurus australis and Saltasaurus best-preserved embryos are exposed in lateral
loricatus plus all its descendants [Salgado and view, crushed against the inner shell (fig.
Calvo 1997]) among titanosaurs, provides sup- 10.10), the exact location of the external nares
port for the placement of the embryos within a cannot be directly observed. However, the loca-
subgroup of titanosaurs that excludes the most tion of the external nares can be inferred from
primitively toothed Malawisaurus dixeyi (Jacobs the orientation of the lacrimals, which in
et al. 1993; fig. 10.11). Nonetheless, this inter- eusauropods mark the approximate caudal end
pretation becomes more complex in light of of the nares. The rostrodorsal orientation of the
recent phylogenetic inferences indicating that lacrimal suggests that the nares of the embryos
this specialized dental condition could have opened in front of the orbit, dorsorostral to the
evolved more than once within titanosaurs antorbital fenestra. This position is also sup-
(Curry Rogers and Forster 2001). ported by the rostral extension of the frontals,
Despite these reservations, comparisons which in the Auca Mahuevo embryos nearly
between the embryos and the best-preserved reach the rostral margin of the orbit (fig. 10.10).
skulls of adult titanosaurs, those of Nemegtosaurus The paucity of cranial material of adult
mongoliensis and Rapetosaurus krausei, suggest titanosaurs has prevented determination of the
that dramatic transformations must have location of the nares in this group of sauropods—
occurred during the ontogeny of these dinosaurs. however, Rapetosaurus krausei (Curry Rogers and
The frontals and parietals became greatly Forster 2004) and Nemegtosaurus mongoliensis
reduced in size and they migrated to the dorso- appear to have fully retracted nares (this condi-
caudal and caudal region of the orbit. The latter tion is best observed in a yet undescribed speci-
became ventrally constricted and adopted an men of Nemegtosaurus mongoliensis housed at
the Mongolian Natural History Museum in that the conclusion of gregariousness seems
Ulaanbataar). Although it is likely that the min- unavoidable. It is highly implausible that soli-
imally retracted nares of the embryos migrated tary females laid the thousands of egg clutches
backward during postnatal allometric develop- contained in these localities. Based on clutches
ment, the rostroventral orientation of the that have been quarried in situ, many eggs were
quadrate and squamosal suggests that at that preserved whole, suggesting that they were
particular stage of development, the embryonic buried quickly and did not sit out on the paleo-
braincase was partially rotated (see Salgado and surface for very long, rendering them vulnera-
Calvo [1997] for correlations between braincase ble to natural processes of disintegration and
rotation and quadrate orientation). This evi- trampling during subsequent breeding seasons.
dence is contrary to Salgado and Calvo’s (1997) The six stratigraphically distinct egg layers
hypothesis of a concerted evolution of the nar- (egg-beds 1–4, with egg-beds 2 and 3 each con-
ial retraction and braincase rostral rotation of sisting of two egg levels) containing eggs of
eusauropods (Chiappe et al. 2001). These con- similar morphology suggest that one sauropod
ditions are likely to have evolved independently, species nested at this site at least six separate
although confirmation of this awaits the discov- times. A minimum of two beds of morphologi-
ery of adult skulls with conditions resembling cally similar eggs also occurs at Barreales Norte
those of the Auca Mahuevo embryos—unre- and Barreales Escondido, supporting site
tracted nares and already rotated braincase. fidelity for these localities as well. The most
parsimonious assumption, therefore, is that all
eggs were laid by the same titanosaur species.
BEHAVIORAL INFERENCES
Discovery of additional embryonic remains in
Although the behavior of extinct organisms eggs from these egg-bearing layers will provide
cannot be directly observed, it can be inferred a means for testing this hypothesis.
when the product of an organism’s activity is The discovery at Auca Mahuevo of well-
preserved in the fossil record (Clark et al. preserved nest traces provided indisputable
1999). The in situ eggs of Auca Mahuevo and evidence of nest construction and architecture
adjacent localities are the preserved physical (Chiappe et al. 2004). Sedimentological evi-
evidence of the sauropods’ egg-laying behavior. dence showed that, contrary to most modern
The mapping and collection of eggs, clutches, reptiles, titanosaurs laid eggs in excavated
and nests at these localities have led us to infer depressions without burying them. Although
several aspects of the reproductive behavior of nest attendance by titanosaurs may be
titanosaur sauropods (Chiappe et al. 2000; inferred by phylogenetic bracketing (all living
Chiappe and Dingus 2001). archosaurs attend their nests), adult size and
The high concentration of egg clutches dis- proximity between clutches (fig. 10.5) suggest
tributed in a relatively narrow stratigraphic hori- little or no parental care of their clutches, a
zon (e.g., 70 cm in Auca Mahuevo’s egg-bed 3) conclusion again supported by the lack of evi-
suggests a gregarious nesting behavior. Even if dence of trampling in our quarry of egg-bed 3,
each egg-bed could preserve eggs laid during where most eggs show minimal crushing.
more than one closely occurring nesting season,
and taking into consideration that the specifics
COMPARISONS TO OTHER
of the gregarious behavior we have envisioned
NESTING SITES
(the number of females nesting at approxi-
mately the same time and in a given season, the A large number of dinosaur egg localities contain
frequency of reproductive seasons, and other clutches of eggs similar to those from Auca
similar questions) remained unanswered, the Mahuevo (e.g., subspherical eggs with a relatively
density of eggs contained in these layers is such thick eggshell comprised of a single structural
layer of calcite, with shell units well separated radius of the egg-laying female (Cousin et al.
from each other and a tuberculate surface orna- 1990). By comparing published limb dimen-
mentation) (Sahni et al. 1990; Powell 1992; sions of the European Late Cretaceous titanosaur
Vianey-Liaud et al. 1990; Mohabey 1996; Calvo Hypselosaurus to radii of the arcs, Cousin et al.
et al. 1997). Traditionally, these eggs have been (1990) proposed a crouching position for the
classified within the Megaloolithidae category adult sauropod during egg laying. However,
of eggshell parataxonomy and considered to none of the eggs assumed to be of Hypselosaurus
have been laid by sauropod dinosaurs (Zhao contain embryonic remains, thus rendering
1979; Mikhailov 1991, 1997). However, Auca their identification as indeterminate. Such ad
Mahuevo and its adjacent localities are the only hoc assumptions regarding the taxonomy of the
sites in the world where diagnosable remains of egg-laying dinosaur can potentially bias inter-
sauropod dinosaurs have been found inside pretations of data and obscure the true tapho-
eggs, and although these embryos support the nomic picture.
identification of some megaloolithid-type eggs Colonial nesting and/or site fidelity have
(e.g., Megaloolithus patagonicus) as sauropod eggs, also been hypothesized for sauropod dinosaurs
it would be risky to extrapolate such a conclu- (Sanz et al., 1995; Figueroa and Powell 2000;
sion to all other eggs of similar morphology. López-Martínez 2000; Mohabey 2000). One
Despite this paucity of eggs containing iden- study reported abundant fragmented eggshell
tifiable embryonic remains at other sites, several and 24 megaloolithid nests arranged in three
assumptions have been made regarding the nest clusters in a 6,000-m2 area. Extrapolation of
construction, egg-laying behavior, and physiol- the data, however, extended the number of eggs
ogy of sauropod dinosaurs. Inference of nest to 300,000, purportedly laid by sauropods nest-
architecture typically is based on clutch geome- ing on a seashore (Sanz et al. 1995; López-
try rather than primary lithologic attributes of Martínez 2000). Inferences made from these
the surrounding sediment (Dughi and Sirugue calculations included territorial behavior, high
1966; Kérourio 1981; Williams et al. 1984; population density, site fidelity, and site prefer-
Faccio 1990; Sahni et al. 1990; Powell 1992; ence (Sanz et al. 1995). These interpretations
Sanz et al. 1995; Mohabey 1996; Calvo et al. were challenged on the basis of time-averaging
1997). For example, a single layer of mega- of the deposit, nonsynchronous deposition of
loolithid eggs from India that occurred in a egg horizons, pedogenesis, and other sedimen-
1-m2 area was used to infer a saucer-shaped tological/taphonomic evidence (Sander et al.
sauropod nest (Mohabey 1996, 2000). However, 1998). Without detailed taphonomic analysis,
the margins of the “nest” are described as homo- such paleobiological inferences appear unwar-
geneous with the host rock, with no observable ranted (Sander et al. 1998).
lithological differences. Fossil eggs such as these Taxonomically unidentified eggs have also
provide no evidence of nest structure that results been used in studies of sauropod physiology
from excavation by adult dinosaurs and are, (Case 1978; Erben et al. 1979; Bakker 1986;
therefore, more appropriately called clutches. To Paul 1990). For example, egg size data and esti-
the best of our knowledge, the Auca Mahuevo mated sauropod hatchling weight were used for
sauropod nests provide the only documentation determining lifetime reproductive potential as a
of sauropod nest architecture based on lithologic function of sauropod body mass (Paul 1990)
attributes (Chiappe et al. 2004). and to estimate the age of sauropods at sexual
A number of egg-laying behaviors have been maturity (Case 1978). In light of the taxonomic
attributed to sauropod dinosaurs (Moratalla and uncertainty of the eggs used for these studies,
Powell 1990). For example, it has been sug- their conclusions are unwarranted.
gested that arcs comprised of 15 to 20 eggs Many megaloolithid eggs found worldwide
(radii, 1.3–1.7m) resulted from the turning share a similar structural morphology with
specimens from Auca Mahuevo and most likely Provincia del Neuquén, Rep. Argentina. Publi-
represent eggs laid by sauropod dinosaurs. cación Conjunta de la Dirección Provincial de
Minería de la Provincia del Neuquén y Dirección
However, taxonomic identification of eggs
Nacional del Servicio Geológico. Anaies (Geol.) 25:
based on bones within the same stratigraphic 9–106.
unit carries a high potential for error, as shown Bakker, R. T. 1986. The Dinosaur Heresies. William
with the misidentification of both oviraptorid Morrow, New York. 482 pp.
and troodontid eggs (Norell et al. 1994; Horner Calvo, J. O., Engelland, S., Heredia, S. E., and Salgado,
and Weishampel 1996). In the case of mega- L. 1997. First record of dinosaur eggshells
(?Sauropoda–Megaloolithidae) from Neuquén,
loolithid eggs, this situation is further compli-
Patagonia, Argentina. GAIA 14: 23–32.
cated by the discovery of neonate remains of the Case, T. J. 1978. Speculations on the growth rate and
hadrosaurid Telmatosaurus transylvanicus in the reproduction of some dinosaurs. Paleobiology 4:
proximity of megaloolithid egg-clutches from 320–328.
the late Cretaceous of Romania (Grigorescu et Chiappe, L. M., and Dingus, L. 2001. Walking on
Eggs. Scribner, New York. 219 pp.
al. 1994, 2003), an association that, if con-
Chiappe, L. M., Coria, R. A., Dingus, L., Jackson, F.,
firmed, would highlight the paraphyletic nature Chinsamy, A., and Fox, M. 1998. Sauropod
of this egg category. Egg studies that depend on dinosaur embryos from the Late Cretaceous of
taxonomic or ontogenetic comparisons (e.g., Patagonia. Nature 396: 258–261.
embryo to adult size), therefore, should await Chiappe, L. M., Dingus, L., Jackson, F., Grellet-
definitive identification based on embryonic Tinner, G., Aspinall, R., Clarke, J., Coria, R. A.,
Garrido, A., and Loope, D. 2000. Sauropod eggs
remains within the egg. Thus far, only the in
and embryos from the Upper Cretaceous of
ovo sauropod remains discovered at Auca Patagonia. 1st Symposium of Dinosaur Eggs and
Mahuevo provide this crucial evidence. Embryos, Isona, Spain. Pp. 23–29.
Chiappe, L. M., Salgado, L., and Coria, R. A. 2001.
ACKNOWLEDGMENTS Embryonic skulls of titanosaur sauropod dinosaurs.
Science 293: 2444–2446.
We thank Michelle Schwengle for rendering the
Chiappe, L. M., Schmitt, J. G., Jackson, F., Garrido, A.,
illustrations and Richard Aspinall for assisting Dingus, L., and Grellet-Tinner, G. 2004. Nest struc-
with the statistical analyses of egg distribution of ture for sauropods: Sedimentary criteria for recog-
Auca Mahuevo’s egg-bed 3. We are also grateful nition of dinosaur nesting traces. Palaios 19: 89–95.
to Kristi Curry Rogers and Jeff Wilson for invit- Clark, J. M., Norell, M. A., and Chiappe, L. M. 1999.
ing us to contribute to this book and to David An oviraptorid skeleton from the Late Cretaceous
of Ukhaa Tolgod, Mongolia, preserved in an
Varricchio for his thorough review of the original
avian-like brooding position over an oviraptorid
manuscript. Field and postfield research for this nest. Am. Mus. Novitates 3265: 1–36.
research was supported by the Ann and Gordon Coria, R. A., and Arcucci, A. 2005. Nuevos
Getty Foundation, the Charlotte and Walter dinosaurios terópodos de Auca Mahuevo, Provincia
Kohler Charitable Trust, the Dirección General del Neuquén (Cretácico tardío, Argentina).
Ameghiniana (in press).
de Cultura de Neuquén, the Fundación
Coria, R. A., and Salgado, L. 1999. Nuevos aportes a
Antorchas, the Infoquest Foundation, and the la anatomia craneana de los sauropodos
Municipalidad de Plaza Huincul. titanosauridos. Ameghiniana 36: 98.
Coria, R. A., Chiappe, L. M., and Dingus, L. 2002. A
LITERATURE CITED new close relative of Carnotaurus sastrei Bonaparte
1985 (Theropoda: Abelisauridae) from the Late
Andreis, R. R. 2001. Paleoecology and environments Cretaceous of Patagonia. J. Vertebr. Paleontol. 22:
of the Cretaceous sedimentary basins of 460–465.
Patagonia (southern Argentina). VII International Cousin, R., 1997. Les gisements d’oeufs de
Symposium on Mesozoic Terrestrial Ecosystems. dinosauriens des Hautes Corbiéres et des
Publ. Espec. Asoc. Paleontol. Argentina 7: 7–14. Corbiéres Orientales (Aude): Ponte, nidification,
Ardolino, A. A., and Franchi, M. R. 1996. Geología y microstructre des coquilles. Bull. Soc. Etudes Sci.
Recursos Minerales del Departamento Añelo. Aude 97: 29–46.
Cousin, R., Bréton, G., Fournier, R., and Watté, J. ———. 2003. The puzzle of Tustea incubation site
1990. Dinosaur egglaying and nesting in France. (Upper Maastrichtian, Hateg Basin, Romania):
In: Carpenter, K., Hirsch, K. F., and Horner, J. R. Hadrosaur hatchlings close to Megaloolithidae
(eds.). Dinosaur Eggs and Babies. Cambridge type of eggshell. 2nd International Symposium
University Press, New York. Pp. 56–74. on Dinosaur Eggs and Babies, Abstracts, 16.
Curry Rogers, K., and Forster., C. 2001. The last of Horner, J. R., and Weishampel, D. B. 1996. A com-
the dinosaur titans: a new sauropod from parative embryological study of two ornithis-
Madagascar. Nature 412: 530–534. chian dinosaurs—A correction. Nature 332:
———. 2004. The skull of Rapetosaurus krausei 256–257.
(Sauropoda: Titanosauria) from the Late Jackson, F. D., Garrido, A. C., and Loope, D. B., 2001.
Cretaceous of Madagascar. J. Vertebr. Paleontol. 24: Sauropod egg clutches containing abnormal eggs
121–144. from the ate Cretaceous of Patagonia: clues to
Dingus, L., Clarke, J., Scott, G. R., Swisher, C. C., III, reproductive biology. J. Vertebr. Paleontol. 21: 65A.
Chiappe, L. M., and Coria, R. A. 2000. First mag- Jackson, F. D., Garrido, A. C., Schmitt, J., Chiappe,
netostratigraphic/faunal constraints for the age L. M., Dingus, L., and Loope, D. B. 2004.
of sauropod embryo-bearing rocks in the Abnormal, multilayered titanosaurid (Dinosauria:
Neuquén Group (Late Cretaceous, Neuquén Sauropoda) eggs from in situ clutches at the Auca
Province, Argentina). Am. Mus. Novitates 3290: Mahuevo locality, Neuquén Province, Argentina.
1–11. J. Vertebr. Paleontol. 24: 913–922.
Dughi, R., and Sirugue, F. 1966. Sur la fossilization Jacobs, L. L., Winkler, D. A., Downs, W. R., and
des oeufs de dinosaurs. C. R. Sea. Acad. Sci. 262: Gomani, E. 1993. New material of an Early
2330–2332. Cretaceous titanosaurid sauropod dinosaur from
Erben, H. 1970. Ultrastrukturen und Mineralisation Malawi. Palaeontology 26: 523–534.
rezenter und fossiler Eischalen bei Vögeln und Kérourio, P. 1981. Nouvelles observations sur le mode
Reptilien. Biomineral. Forschungsber. 1: 1–66. de nidification et de ponte chez les dinosauries du
Erben, H., Hoefs, J., and Wedepohl, K. H. 1979. Cretace terminal du Midi de la France. C. R.
Paleobiological and isotopic studies of eggshells Sommaire Soc. Geol. France 1: 25–28.
from a declining dinosaur species. Paleobiology Kues, B. S., Lehman, T., and Rigby, J. K. 1980. Teeth
5: 380–414. of Alamosaurus sanjuanensis, a Late Cretaceous
Faccio, G. 1990. Dinosaurian eggs from the Upper sauropod. J. Paleontol. 54: 864–869.
Cretaceous of Uruguay. In: Carpenter, K., Hirsch, Leanza, H. A. 1999. The Jurassic and Cretaceous ter-
K. F., and Horner, J. R. (eds.). Dinosaur Eggs and restrial beds from southern Neuquén Basin,
Babies. Cambridge University Press, New York. Argentina. Inst. Super. Correl. Geol. Ser. Misc. 4:
Pp. 47–55. 1–30.
Figueroa, C., and Powell, J. E. 2000. Structure and Legarretta, L., and Gulisano, C. 1989. Analisis estrati-
permeability properties of thick dinosaur grafico secuencial de la Cuenca Neuquina (Triasico
eggshells from the Upper Cretaceous of Superior–Terciario Inferior). In: Chebli, G., and
Patagonia, Argentina. Results of a computer sim- Spalletti, L. (eds.). Cuencas Sedimentarias
ulation analysis. In: Bravo, A. M., and Reyes, T. Argentinas, Serie de Correlacion Geologica Nro. 6.
(eds.). 1st International Symposium of Dinosaur Universidad Nacional de Tucumán, Tucumán,
Eggs and Embryos, Isona, Spain. Pp. 51–60. Argentina. Pp. 221–243.
Garrido, A. C., Chiappe, L. M., Jackson, F., Schmitt, Loope, D., Schmitt, J., and Jackson, F. 2000.
J., and Dingus, L. 2001. First sauropod nest struc- Thunder platters: thin discontinuous limestones
tures. J. Vertebr. Paleontol. 21: 53A. in Late Cretaceous continental mudstones of
Grellet-Tinner, G., Chiappe, L. M., and Coria, R. A. Argentina may be chemically infilled sauropod
2004. Eggs of titanosaurid sauropods from the tracks. Geological Society of America, Abstracts
Upper Cretaceous of Auca Mahuevo (Argentina). with Programs: A450.
Can. J. Earth Sci. 41: 949–960. López-Martínez, N. 2000. Eggshell sites from the
Grigorescu, D., Weishampel, D., Norman, D., Creatceous-Tertiary transition in south–central
Seclamen, M., Rusu, M., Baltres, A., and Pyrenees (Spain). In: Bravo, A. M., and Reyes, T.
Teodorescu, V. 1994. Late Maastrichtian dinosaur (eds.). 1st International Symposium of Dinosaur
eggs from the Hateg Basin (Romania). In: Eggs and Embryos, Isona, Spain. Pp. 95–115.
Carpenter, K., Hirsh, K. F., and Horner, J. R. McIntosh, J. 1997. Sauropoda. In: Currie, P. J., and
(eds.). Dinosaur Eggs and Babies, Cambridge Padian, K. (eds.). Encyclopedia of Dinosaurs,
University Press, Cambridge. Pp. 75–87. Academic Press, New York. Pp. 654–658.
Mikhailov, K. E. 1991. Classification of fossil dinosaur eggs and nesting sites from the Deccan
eggshells of amniotic vertebrates. Acta Paleontol. volcano-sedimentary province of peninsular India.
Polonica 36: 193–238. In: Carpenter, K., Hirsch, K. F., and Horner, J. R.
———. 1997. Fossil and recent eggshell in amniotic (eds.). Dinosaur Eggs and Babies. Cambridge
vertebrates: fine structure, comparative morphol- University Press, Cambridge. Pp. 204–226.
ogy and classification. Special Papers in Salgado, L., and Calvo, J. O. 1997. Evolution of
Paleontology, No. 56. Paleontological Association, titanosaurid sauropods. II: The cranial evidence.
London. 80 pp. Ameghiniana 34: 33–48.
Mohabey, D. M. 1996. A new oospecies, Megaloolithus Salgado, L., Coria, R. A., and Chiappe, L. M., 2005.
matleyi, from the Lameta Formation (Upper Osteology of the sauropod embryos from the
Cretaceous) of Chandrapur district, Maharashtra, Upper Cretaceous of Argentina. Acta Paleontol
India, and general remarks on the palaeoenviron- Polonica 50: 79–92.
ment and nesting behavior of dinosaurs. Sander, P. M., Peitz, C., Gallemi, J., and Cousin, R.
Cretaceous Res. 17: 183–196. 1998. Dinosaurs nesting on a red beach? Earth
———. 2000. Indian Upper Cretaceous Planet. Sci. 327: 67–74.
(Maestrichtian) dinosaur eggs: their parataxon- Sanz, J. L., Moratalla, J. J., Díaz-Molina, M., López-
omy and implication in understanding the nest- Martínez, N., Kälin, O., and Vianey-Liaud, M.
ing behavior. In: Bravo, A. M., and Reyes, T. 1995. Dinosaur nests at the sea shore. Nature
(eds.). 1st International Symposium of Dinosaur 376: 731–732.
Eggs and Embryos, Isona, Spain. Pp. 95–115. Vianey-Liaud, M., Mallan, P., Buscail, O., and C.
Moratalla, J. J., and Powell, J. E. 1990. Dinosaur nest- Montgelard. 1990. Review of French dinosaur
ing patterns. In: Carpenter, K., Hirsch, K. F., and eggshells: morphology, structure, mineral, and
Horner, J. R. (eds.). Dinosaur Eggs and Babies. organic composition. In: Carpenter, K., Hirsch,
Cambridge University Press, Cambridge. Pp. K. F., and Horner, J. R. (eds.). Dinosaur Eggs and
37–46. Babies. Cambridge University Press, Cambridge.
Norell, M. A., Clark, J. M., Dashzeveg, D., Barsbold, Pp. 151–183.
R., Chiappe, L. M., Davidson, A., McKenna, M. C., Vianey-Liaud, M., Hirsch, K. F., Sahni, A., and Sigé, B.
Altangerel, P., and Novacek, M. J. 1994. A thero- 1997. Late Cretaceous Peruvian eggshells and
pod dinosaur embryo and the affinities of the their relationships with Laurasian and Eastern
Flaming Cliffs dinosaur eggs. Science 266: Gondwanian material. Geobios 30: 75–90.
779–782. Vianey-Liaud, M., Khosla, A., and Garcia, G. 2003.
Nowinski, A. 1971. Nemegtosaurus mongoliensis n. Relationships between European and Indian
gen., n. sp. (Sauropoda) from the Uppermost dinosaur eggs and eggshells of the oofamily
Cretaceous of Mongolia. Palaeontol. Polonica 25: Megaloolithidae. J. Vertebr. Paleontol. 23:
57–81. 575–585.
Paul, G. 1990. Dinosaur reproduction in the fast Williams, D. L. G., Seymour, R. S., and Kérourio, P.
lane: implications for size, success, and exten- 1984. Structure of fossil dinosaur eggshell from
sion. In: Carpenter, K., Hirsch, K. F., and Horner, the Aix Basin, France. Palaeogeogr. Palaeoclimatol.
J. R. (eds.). Dinosaur Eggs and Babies. Cambridge Palaeoecol. 45: 23–37.
University Press, Cambridge. Pp. 244–255. Wilson, J. A. 2002. Sauropod dinosaur phylogeny:
Powell, J. 1992. Hallazgo de huevos asignables a critique and cladistic analysis. Zool. J. Linn. Soc.
dinosaurios titanosáuridos (Saurischia, Sauropoda) 136: 217–276.
de la Provincia de Río Negro, Argentina. Acta Zool. Wilson, J. A., and Sereno, P. C. 1998. Early evolution
Lilloana 41: 381–389. and higher-level phylogeny of sauropod
Ramos, V. A. 1981. Descripción geológica de la hoja dinosaurs. J. Vertebr Paleontol 18 (Suppl. 2):
33c, Los Chihuidos Norte, Provincia del Neuquén. 1–68.
Servicio Nacional de Minería y Geología. Boletín Zhao, Z. 1979. The advancement of researches on
182. the dinosaurian eggs in China. South China
Sahni, A. S. K., Tandon, A., Jolly, S., Bajpai, S., Sood, Mesozoic and Cenozoic “Red Formation” Science,
A., and Srinivasan, S. 1990. Upper Cretaceous Beijing. Pp. 329–340.
ELEVEN
Sauropod Histology
MICROSCOPIC VIEWS ON THE LIVES OF GIANTS
Kristina Curry Rogers and Gregory M. Erickson
auropod hatchlings may have adults, these giants are estimated to have
S been only a meter long from head to tail
(Chiappe et al. 1998, 2001) and weighed in at
reached lengths of more than 30 m and
weighed between 30 and 100 tons.
less than 10 kg (Breton et al. 1985; Weishampel The growth strategies that permitted
and Horner 1994), while many adult sauropods to attain such massive proportions
sauropods attained sizes rivaling those of ex- have historically been a topic of great interest,
tant whales (Appenzeller 1994; Seebacher but until recently they remained among the
2001; Erickson et al. 2001). This range of sizes greatest of biological mysteries. In early investi-
is greater than for any other dinosaurian line- gations, workers extrapolated from reptilian
age, and includes the largest terrestrial verte- growth rates and hypothesized that giant
brates ever to inhabit the planet. In addition to sauropods would have taken decades to reach
this dramatic range of ontogenetic size vari- sexual maturity and well over a century to attain
ability, sauropod sizes vary interspecifically their enormous adult sizes (e.g., Case 1978b;
(fig. 11.1). The Romanian titanosaur Mag- Calder 1984). More recent workers have
yarosaurus has been dubbed the “smallest of focused specifically on evidence gleaned from
the largest” because of its minute adult stature bones and have revealed astounding new data
(5 m [Jianu and Weishampel 1999]). Simi- that indicate that sauropod growth rates soared
larly, the saltasaurids Saltasaurus and Neuquen- through most of ontogeny (e.g., Ricqlès
saurus are on the “small” end of sauropod size, 1968a,b; Reid 1981, 1990, 1997; Rimblot-Baly
reaching adult lengths of only 7 m and weights et al. 1995; Curry 1998; Sander 2000, 2003;
of 10,000 kg (Powell 1986, 1990, 2003). Erickson et al. 2001; Curry Rogers et al. 2003;
Conversely, Argentinosaurus, Paralititan, and Sander and Tückmantel 2003).
Seismosaurus push the envelope of size for On the outside, fossilized bones provide the
land-dwelling vertebrates (Anderson et al. raw data on which we base our interpretation of
1985; Coe et al. 1987; Gillette 1991; Appenzeller sauropod anatomy, relationships, function, and
1994; Seebacher 2001; Smith et al. 2001). As biomechanics, and our confidence in the
303
FIGURE 11.1. Interspecific and ontogenetic variation in sauropod size. At hatching, sauropods were likely 1 m long and
under 10 kg. As adults, sauropod sizes ranged from 7 m long and 10,000 kg (e.g., Saltasaurus) to 30 m long and approach-
ing 100 tons (e.g., Argentinosaurus). Most sauropod adults (e.g., Diplodocus, Camarasaurus ) fall in the middle of this range.
Scale bar equals 1 m.
restored version of long-extinct sauropods rests data for choosing among extant analogues, and
almost exclusively on evidence obtained from allow for sauropod growth rate quantification
fossil bones. The sizes and body proportions on microstructural and organismal levels. Bone
outlined above are readily reconstructed from histology breathes new life into fossil bones and
the skeleton, even when remains are relatively offers a rigorous, testable means of addressing
fragmentary. Muscle scars on bones help deter- hypotheses of sauropod growth rates and life
mine where muscles originate and insert, and history. Did sauropods grow indeterminately?
the resultant pattern of musculature dictates Did sauropods grow at constant rates through-
reconstructed body shape, as well as our inter- out ontogeny, or did they experience regular
pretations of bone and muscle synergy and bio- cycles of relative growth rate variation? When
mechanical function (chapters 6 and 8). did sauropods attain maturity? How long did it
Similarly, osteological correlates of soft tissues take sauropods to reach their massive adult
preserved in fossil bones (e.g., pneumatic fos- sizes? Did all sauropods exhibit the same basic
sae; chapter 7) make it possible to reconstruct growth strategy? Along the same lines, is sauro-
unpreserved soft parts that may never be found pod histology relevant to sauropod phylogeny?
in the fossil record. For example, Witmer The answers to these questions lie under
(2001) reconstructed Diplodocus with terminally microscopes and deep within sauropod bone
positioned external nares on the basis of soft tissue.
tissue correlates in the skull. The gross mor- In this chapter we consider sauropod
phology of bones thus enriches the story of growth throughout ontogeny from both histo-
sauropod evolution. Through osteology we doc- logical and whole-body perspectives. We first
ument the debut and demise of major lineages provide a primer on relevant bone histological
and the evolutionary tale told by the trail of patterns, followed by a brief overview of the evi-
characters observed in fossil skeletons. dence from living and fossil vertebrate taxa rel-
The internal architecture of fossil bones doc- evant to qualitative interpretations of sauropod
uments patterns of bone deposition and remod- bone tissue. We outline a method for garnering
eling and can sometimes allow access to the quantified tissue growth rates directly from
portions of the bone growth record throughout sauropod bones and then pull away from
ontogeny. The clues preserved in fossil bone tis- microstructural analysis to a broader focus on
sue thus provide the necessary comparative sauropod whole-body growth rates through
304 SAUROPOD HISTOLOGY

developmental mass extrapolation. We present ORGANIZATION OF PRIMARY BONE TISSUE
a case study for Apatosaurus that incorporates Primary bone organization is among the most
all methods discussed in the text and conclude important aspects for determining relative bone
the chapter with a prospectus for future work depositional rates. As a bone begins an active
on sauropod growth rates. phase of growth, osteoblasts secrete osteoid,
which contains the elastic protein, collagen.
The mineral component of bone is composed
BONE HISTOLOGY: STUDYING
of hydroxyapatite crystals that align in parallel
SAUROPOD GROWTH FROM
with the deposited collagen fibers. In living ver-
THE INSIDE OUT
tebrates, the degree of primary bone organiza-
In extant vertebrates, distinctive bone tissue tion thus reflects the rate of collagen/hydroxya-
types are associated with different growth rates patite deposition. Though collagen is only
and trajectories. Fortunately, patterns of verte- rarely preserved in the fossil record (Wykoff
brate bone tissue organization are readily 1975; Currey 1987; Schwietzer et al. 1999,
retrieved from fossils, in which the mineral 2005; Carter and Beaupré 2001), hydroxyap-
constituents of bone persist, as do vacuities left atite crystal alignment persists and, thus, indi-
by bone’s vascular and nervous supply. Thus, cates relative rates of primary bone deposition.
when analogous tissue patterns occur in extinct There are three general categories of bone
and extant taxa, bone growth rates and overall fibrillar organization, each thought to be the
growth strategies in extinct taxa may be inferred outcome of its formative rate. When the bone
(Enlow and Brown 1956, 1957, 1958; Ricqlès depositional rates are relatively slow, collagen
1977, 1980; Ricqlès et al. 1991; Chinsamy 1990, fibers and hydroxyapatite crystals show a tight
1993; Curry 1998, 1999; Sander 2000, 2003a, parallel alignment and there are endogenous
2003b; Padian et al. 2001; Chinsamy and pulses to the mineralization front. The latter
Elzanowski 2001). Sauropod bone tissue gives the bone a striated texture under polar-
exhibits a range of typologies comparable to the ized light and is thus commonly called lamellar
range of tissue types found in extant vertebrate bone (fig. 11.2A); each increment in bone
lineages. Several different levels of bone tissue growth with parallel fiber alignment is a single
organization are relevant to discussions of lamella. Stronger, endogenous biorhythms
growth rates in sauropods, including the depo- (usually annual) often cause physiological dis-
sitional pattern of primary bone protein and ruptions to bone formation and lead to the for-
mineral, as well as the pattern and extent of mation of lines of arrested growth (LAG; see
bone vasculature (also termed porosity, because below). The second type of primary bone tissue
of the inclusion of lymphatic and nervous tis- is known as parallel fibered and is deposited at
sue in bone canals Margerie et al. 2002; Starck slightly faster rates (fig. 11.2B). In the third
and Chinsamy 2002) and the degree and nature case, bone is formed at more rapid rates, and
of remodeling (Francillon-Vieillot et al. 1990). collagen fibers and adhering hydroxyapatite
The combination of these characteristics in dif- crystals align haphazardly. This type of rapidly
ferent patterns allows us to draw conclusions growing bone tissue is specified as woven, or
regarding sauropod bone formation rate, age, fibrous, bone (fig. 11.2C). Its genesis is rarely
and longevity, and can even be used to develop disrupted by endogenous biorhythms in living
sigmoidal growth curves that allow comparison taxa, although LAG are not uncommon in
of whole-body growth rates among extinct dinosaurs (Reid 1981, 1984, 1990; Chinsamy
dinosaurs and living vertebrates. The bone clas- 1990; Varricchio, 1993).
sificatory scheme for descriptive tissue histol- In actively growing taxa that generally
ogy developed by Francillon-Vieillot et al. exhibit rapid growth rates (e.g., eutherian mam-
(1990) is employed in the following discussion. mals and birds), the woven and lamellar types
SAUROPOD HISTOLOGY 305

FIGURE 11.2. Bone fiber organization. (A) Endosteal, lamellar zonal bone presents a parallel alignment of hydroxyapatite
crystals and collagen fibers, and results in a striated texture under polarized light. (B) Periosteal, parallel-fibered bone and
flattened, circumferential vascular canals. (C) Periosteal, woven-fibered bone, with collagen fibers and hydroxyapatite crys-
tals arranged haphazardly. This example includes several scattered secondary osteons. Scale bars equal 100 m.
of bone tissue occur in combination. The gross vascular canals oriented circularly and longitu-
size and shape of the bone are dictated by appo- dinally in superimposed laminae (fig. 11.5B).
sitional woven bone growth at the outer sur- Reticular vascularization is an obliquely and
faces of elements. Later, vascular canals trapped irregularly oriented arrangement of the vascu-
within the woven bone are partially infilled by lar canals (fig. 11.5C).
layers of slower growing lamellar bone. The Haversian bone tissue results from the cen-
resultant, complexly organized tissue is termed trifugal obliteration of primary vascular canals
the “fibrolamellar complex” (fig. 11.3). and cortices and subsequent centripetal redeposi-
tion of bone to form secondary osteons. Multiple
ORGANIZATION OF BONE VASCULATURE cycles of remodeling result in “dense Haversian
Like the organization of primary bone tissue, bone,” in which little primary bone remains.
patterns of primary bone vascular supply are Primary bone remodeling and the formation of
thought to vary with bone depositional speeds. secondary osteons are a time-dependent process.
As vascular canals are trapped by newly It is most frequently linked to age, localized
deposited bone, they are termed primary mechanical stress (e.g., muscle attachment),
osteons. Primary osteons in a single direction mineral homeostasis, and fatigue damage repair
may be oriented longitudinally, circularly, radi- (Amprino 1948; Lacroix 1971; Martin and Burr
ally, or obliquely relative to the long axis of the 1989). Haversian bone is most common in the
bone under examination (fig. 11.4). In more bone tissue of mammals (Ricqlès et al. 1991), but
rapidly growing bone tissue, primary vascular also occurs in other vertebrates including birds,
canals run in multiple directions and commonly “reptiles,” and dinosaurs (Mantell 1850; Queckett
intercalate in three major patterns (fig. 11.5). 1855; Owen 1859; Foote 1913; Enlow and Brown
Plexiform vascularization (Enlow 1963) occurs 1956, 1957, 1958).
when longitudinal, circular, and radial vascular Recent experimental work on extant birds
canals interweave to form a three-dimensional (Castanet et al. 1996, 2000; Margerie et al.
vascular plexus (fig. 11.5A). Laminar vascular- 2002; Starck and Chinsamy 2002) indicates
ization (Foote 1913) is characterized by primary that bones characterized by laminar and retic-

FIGURE 11.3. The fibrolamellar complex with lamellar–zonal bone deposited centrifugally around primary vascular canals
in a woven bone matrix to form primary osteons. (A) Apatosaurus radius in polarized light. Lamellar–zonal bone is bright;
woven matrix is darker. (B) Apatosaurus radius in normal light. Scale bar equals 100 m.
ular vascular organization overlap in their and normally consist of woven and/or parallel-
growth rate ranges and are the fastest-growing fibered matrices. In the fibrolamellar complex,
tissue types (centripetal deposition in bony calli zones are comprised of superimposed laminae.
can actually be even faster, at 1 mm/day Annuli are narrower than zones and reflect
[Lanyon and Rubin 1984]). Plexiform vascular- periods of relatively slow growth rates. They are
ity occurs at significantly lower rates (Castanet normally composed of parallel and/or lamellar
et al. 2000). Bones exhibiting longitudinal vas- fibered matrices and may occur as thin rings in
cular organization are typically the slowest- localized regions of cortical bone. LAG (also
growing type of vascularized primary bone. known as “rest lines”) represent periods of tem-
These findings are particularly significant for porary, complete, or near-complete cessation of
studies of dinosaurian bone growth rates, appositional growth and represent the external
because they provide a means of constraining surface of the bone at the time the LAG were
the estimates for local tissue formation rates in deposited. The length of time represented by an
extinct taxa (table 11.1). individual LAG varies, and may span from a
few days to as long as six months (Castanet and
THE PERIODICITY OF BONE GROWTH Smirina 1990).
Growth marks in bone tissue represent a perio- Bones characterized by LAG, annuli, zones,
dicity in bone depositional rate and can be dif- or any combination of these is typically termed
ferentiated on the basis of general morphology zonal bone (fig; 11.6). Bone tissue occurring in
and biological significance. Zones correspond zones may be lamellar and poorly vascularized
to periods of active growth and osteogenesis. if an individual deposits bone slowly through-
They are thicker than other growth marks out its active growth period (e.g., wild crocodil-
(except perhaps “polish marks” [Sander 2000]) ians; fig. 11.6A). Alternatively, bone within

FIGURE 11.4. Organization of primary osteons. (A) Longi- FIGURE 11.5. Intercalating primary osteons. (A) Laminar
tudinal vascular canals parallel the long axis of the bone; (B) vasculature, with a predominance of circular and radial in-
circular vascular canals ring the circumference of the bone; terwoven canals. (B) Plexiform bone, with circular, radial,
(C) radial vascular canals radiate from the endosteal regions and longitudinal interwoven canals. (C) Reticular bone,
of the bone. (Modified from Francillon-Viellot et al. 1990.) with an oblique and irregular orientation of primary
osteons. Depositional rates of reticular and laminar bone
generally overlap and reflect the most rapid rates of
primary bone growth.
TABLE 11.1
Relationship Between Bone Growth Rates and Bone Tissue Types
AVERAGE GROWTH RATE (M/DAY)
SUBRETICULAR/
LAMINAR BONE RETICULAR BONE LONGITUDINAL CANALS
Ostrich 24.7 33.2 4.9

Emu 27.1 31.6 5.4
Mink 13.7 — —
Duck 10.0 — —
NOTE: Rates outlined below are average rates as determined by Castanet et al. (1993, 1996, 2000) and Ricqlès et al. (1991).
FIGURE 11.6. The periodicity of bone growth. (A) Cortical bone with lamellar–zonal, poorly vascularized zones of active
growth punctuated by numerous lines of arrested growth (LAG) indicating periods of dramatic decrease in bone deposition.
(B) Cortical bone of highly vascularized, fibrolamellar bone punctuated by one LAG (indicated by arrow). (C) Zonal bone not
punctuated by LAG but demarcated by a regular decrease in the size and density of primary vascular canals. Arrows indicate
the onset of each new cycle. Scale bars in (A) and (C) equal 500 m. Scale bar in (B) equals 100 m.
zones may be fibrolamellar, with abundant vas- small amounts of poorly vascularized lamellar
culature, and thus tell the tale of relatively rapid bone tissue (the “external fundamental system”
growth during an ontogenetic stage (fig. 11.6B). [Francillon-Vieillot et al. 1990; Ricqlès et al.
This pattern is typical among diverse 1991]). A low porosity and lamellated structure
dinosaurian taxa, including Maiasaura (Horner predispose accretionary bone to resist loads in
et al. 2000), Hypacrosaurus (Cooper and vivo, and may serve as a significant structural
Horner 1999), Syntarsus (Chinsamy 1990), and reinforcement in maturing animals (Francillon-
Troodon (Varricchio 1993). In some taxa, includ- Viellot 1990). This avascular, accretionary tis-
ing humans, LAG are deposited at the external sue is frequently marked with LAG and resorp-
margins of a bone once the majority of skeletal tion lines and has been associated with growth
growth has ceased. In these taxa, minor thick- rate plateaus in extant taxa (e.g., Curry 1998;
ening of the bone may still occur by accretion of Sander 2000, 2003a; fig. 11.6C).

An additional type of primary bone stratifica- age estimate for an individual, particularly
tion has been observed in extant cetaceans when the biological meaning of the growth
(Altman and Ditmer 1972), sirenians (Fawcett mark is considered, and an effort is made to
1942), leatherback turtles (Rhodin 1985), account for growth cycles effaced by medullar
ichthyosaurs (Buffrénil and Mazin 1990), and expansion (e.g., Chinsamy 1990; Varricchio
some sauropods (Ricqlès 1968b, 1983; Reid 1981; 1993; Rimblot-Baly et al. 1995; Curry 1999;
Rimblot-Baly et al. 1995; Curry 1999; Sander Sander 2000; Erickson et al. 2001; Sander and
2000, 2003; Sander and Tückmantel 2003). In Tückmantel 2003). Phylogenetic parsimony
these taxa, continuous fibrolamellar bone depo- has demonstrated that annual cycle/LAG depo-
sition occurs, but a regular variation in this sition is plesiomorphic for vertebrates (e.g.,
continual depositional pattern is indicated by Castanet and Smirina 1990). Thus, in most
variable vascular patterns arranged as superim- studies of dinosaurian growth and longevity,
posed cycles of growth (Ricqlès 1968a,b, 1983; growth marks are regarded as annual increments
Rimblot-Baly et al. 1995; Curry 1999; Sander (Ricqlès 1983; Chinsamy 1990; Varricchio 1993;
2000, 2003; Sander and Tückmantel 2003). Erickson and Brochu 1999; Curry 1999;
Regions of relatively slower growth exhibit longi- Erickson and Tumanova 2000; Erickson et al.
tudinal vascularity, while areas of faster growth 2001).
exhibit more extensive canal networks. Once
again, cycles are not marked by LAG or clear POTENTIAL PITFALLS OF EQUATING BONE
annuli. Rhodin (1985) concluded that such struc- MICROSTRUCTURE AND GROWTH RATES
tures reflect cyclical growth of primary cortices, IN EXTINCT TAXA
with faster and slower growth periods. Such Overall, we have had much success in docu-
cycles as observed in sauropods record mild mod- menting general patterns of tissue morphology
ulations of bone deposition in a fast (to very fast) that can be used for qualifying bone growth
context. They are not equivalent to cycles result- rates. However, most dinosaur bones are
ing from a steep drop in bone formation (where uniquely composed of tissues with slow-growing
annuli or LAG are common), as regularly reptilian and rapid-growing avian/mammalian
observed in extant poikilotherms (Castanet and attributes, leaving results somewhat inconclu-
Naulleau 1985; Castanet and Smirina 1990) and sive. Several characteristics of bone biology must
some other dinosaurs (e.g., Syntarsus [Chinsamy be considered in any analysis of growth strategy
1990], Troodon [Varricchio 1993], Massospondylus in fossil vertebrates.
[Chinsamy 1993]). Sander (2000) provided an The destructive nature of histological analy-
alternative explanation for the presence of corti- sis, and difficulty in processing extremely
cal stratification in Janenschia robusta. Rather robust sauropod elements, has traditionally
than a specific change in growth rate, he postu- restricted investigations of sauropod bone tis-
lated that cycles observed in low magnification sue to single elements. Single-element analyses
resulted from an abruptly lower mineralized are problematic when the goals are determina-
bone tissue, which causes a decrease in reflectiv- tion of generalized growth patterns and quan-
ity in bright-field illumination. tification of overall body growth rates because
Cortical stratification of all sorts, including bone microstructure is influenced by ontogeny,
LAG, annuli, and general cycles, may reflect environment, mechanics, and phylogeny (e.g.,
cyclical annual growth, seasonal stress, and/or Padian et al. 2001; Starck and Chinsamy 2002).
other endogenous rhythms (Castanet and Different bones in the same skeleton may have
Smirina 1990; Erickson and Tumanova 2000). differing numbers of growth lines due to vari-
Thus, it is most commonly utilized in skele- able remodeling rates (Chinsamy 1993; Horner
tochronology. These cyclical patterns of bone et al. 1999, 2000; Padian et al. 2001). Similarly,
deposition can essentially provide a minimum bone tissue architecture varies with the position

of the thin section, with bone at middiaphyses emu (Dromaius novaehollandiae [Castanet et al.
typically composed of compacta and spongiosa, 2000]), and Japanese quail (Coturnix japonica
bone at metaphyses comprised of hypertro- [Starck and Chinsamy 2002]) allowed for the
phied calcified cartilage, and bone at epiphyses correlation of bone growth rates with observed
a mosaic of tissue types, all of which may have patterns of fibrillar and vascular organization.
variable growth rates. The diversity of possible The first quantitative data on growth dynamics
formative influences coupled with the fact that of the primary cortical bone of juvenile ratites
tissue structure varies among elements and demonstrated that high growth rates are linked
within single elements prohibits attempts to to highly vascularized bone tissue but that rates
assess generalized life history attributes (e.g., for the same tissue type could be as diverse as
longevity, growth rates, ages) from single 10–80 m/day (Castanet et al. 2000). Detailed
bones. Thus, studies that cut across some of work on Anas indicated that growth rates for the
these constraints and allow for a comparison of same tissue type vary from 20 to 110 m/day
intraskeletal variation are the best approach for (Margerie et al. 2002). The growth rates of fibro-
determining general histological and growth lamellar bone in living mammals have only
patterns (Horner et al. 1999). been documented in one study (Ricqlès et al.
Remodeling of primary bone tissue over the 1991), at a rate of 2.5–4.5 m/day. Despite this
course of ontogeny is also problematic for doc- growth rate overlap among vascular patterns,
umenting longevity, particularly in extinct ver- transitions between bone fibrillar types do cor-
tebrates. The record of the earliest stages of respond to specific changes in formative rates
ontogenetic growth can be lost either through (table 11.1). Margerie et al. (2002) and Starck
medullar expansion or through remodeling of and Chinsamy (2002) independently cautioned
primary bone tissue through the formation of that the presence and proportion of primary
secondary tissue. Thus, arguments for overall porosity may be more relevant in the interpreta-
growth strategy in extinct taxa must be framed tion of bone growth rates than the primary vas-
in the context of an ontogenetic series. Younger cular pattern.
bones capture growth marks and bone tissue In any case, under simple structurofunc-
types that may be remodeled/obliterated in tional assumptions and using the principle of
later life. Ontogenetic series thus afford a parsimony, when a particular tissue type is pre-
unique opportunity to document a more com- served in a fossil taxon and quantitative growth
plete record of osteogenic variation during the data on similar tissues exist for living animals,
life of a particular taxon (Chinsamy 1990, 1993; application of bone growth rates in living verte-
Curry 1999; Horner et al. 1999; Erickson and brates to similar bone tissues in extinct taxa is
Tumanova 2000; Sander 2000; Erickson et al. possible and potentially informative. Investiga-
2001; Sander and Tückmantel 2003). tion of tissue growth rates in this manner can
With regard to bone growth rates and their help to restrict estimates of age at various
realistic expression of overall skeletal growth, growth stages, especially when skeletochrono-
the following must be considered. The same logic indicators are absent (e.g., Curry 1999;
bone tissue can form at a range of different rates Sander 2000; Sander and Tückmantel 2003).
(Ricqlès et al. 1991; Castanet et al. 2000; The relationship between localized measures
Margerie et al. 2002; Starck and Chinsamy of tissue formation rates and overall body
2002) and depositional rates for different tissue growth is untested for virtually all skeletal ele-
types (as in the case of laminar and reticular ments, even in extant taxa (Erickson et al. 2001).
bone) actually overlap. Recent work on bone tis- This lack of data can lead to problems in draw-
sue growth rates in extant duck (Anas platyryn- ing conclusions from single elements in fossil
chos [Castanet et al. 1996; Margerie et al. 2002]), taxa. These results emphasize the importance of
ostrich (Struthio camelus [Castanet et al. 2000]), examining homologous bones, standardizing

thin sections, and comparing a multielement sizes in sauropods (Ricqlès 1968a, 1980, 1983;
sample in an ontogenetic context. Rimblot-Baly et al. 1995; Curry 1998, 1999;
Sander 2000; Erickson et al. 2001). Counts of
WHAT CAN BONE HISTOLOGY cortical stratification supports this hypothesis,
REALLY TELL US? indicating that sauropods might have grown to
Qualification of the mechanisms and speed of half-size by the age of 5 years and to their adult
deposition recorded by primary bone and vas- size by between 10 and 30 years of age (Curry
culature in extant taxa assists us in (1) register- 1999; Sander 2000; Sander and Tückmantel,
ing general growth patterns in extinct taxa and 2003).
determining appropriate modern analogues; (2)
identifying variation of bone growth patterns HISTOLOGICAL VARIABILITY AND
among individual skeletal elements to gain a SAUROPOD BIOLOGY
more detailed understanding of how function, Significantly, not all sauropod bones demon-
phylogeny, ontogeny, biorhythms, and environ- strate the same depositional patterns, indicat-
ment influence bone development; and (3) ing that formation of different tissue patterns is
allowing for quantification of bone tissue depo- influenced by functional environment or
sitional rates. Each of these questions can be interelement allometry (Curry 1999; Sander
addressed through the analysis of the broad- 2000). In the case of sauropods and other
scale histological patterns and trends observed dinosaurs, the slightly different histological sto-
in an ontogenetic series of extinct taxa from the ries revealed by different elements can provide
same taxon and locality. the key data for addressing more specific life
history, quantitative, and whole-body growth
GENERAL SAUROPOD GROWTH PATTERN rate questions (Varricchio 1993; Castanet et al.
Most studies agree that sauropods commonly 2000; Curry 1998, 1999; Padian et al. 2001).
deposited abundant, fibrolamellar bone. For For example, Sander (2000) documented
most of ontogeny, this bone was highly vascular- dimorphism in his histological sample of
ized in laminar, reticular, and/or plexiform pat- Barosaurus and hypothesized that the difference
terns. Late in ontogeny, bone vasculature drops was due to sexual dimorphism and not related
off and bone matrices are most frequently com- to taxonomy. In one morph, bone growth was
prised of parallel- and lamellar-fibered tissue, rapid and continuous, with relatively little
thus indicating that in late ontogeny growth sub- haversian remodeling. In the second morph,
stantially slowed. LAG are not common among remodeling was extensive and occurred in the
sauropods until late in ontogeny, although some context of slower, discontinuous bone deposi-
taxa exhibit cortical stratification in the form of tion. Sander hypothesized that the highly
cycles or “polish lines” (e.g., Sauropoda indet. remodeled bone was likely from a female, with
[Reid 1981], Lapparentosaurus [Ricqlès 1983; the discontinuities the result of reproduction.
Rimblot-Baly et al. 1995], Apatosaurus [Curry Sander ruled out local tissue variability in his
1999], Janenschia [Sander 2000; Sander and sample because of the lack of intermediate tis-
Tückmantel 2003]). All sauropod bone tissue, sue types deep in the cortical bone. This
including that where “polish lines” or growth methodology highlights the potential for histol-
cycles are common, is inconsistent with the use ogy to reveal details of dinosaur lives, when lit-
of reptilian growth analogues. Instead, histologi- tle other evidence is available.
cal analyses among sauropods reveal a fibrillar
and vascular organization most similar to that of QUANTITATIVE SAUROPOD BONE GROWTH RATES
rapidly growing mammals and birds. These Bone histology has been the basis for interpret-
results suggest that prolonged ontogeny was not ing growth strategies and life histories in
required for the attainment of gigantic body dinosaurs since their earliest discoveries.

Amprino (1947) was the first to recognize the bone, as long as vascular and fibrillar constitu-
functional relationship between primary bone tion is consistent. This growth rate calculation
organization and the speed of its deposition. only relies on the assumption that cycles repre-
“Amprino’s Rule” thus predicts that given tissue sent a year of bone deposition and does not
typologies are deposited within a specific set of require the application of growth rates from the
growth rate parameters, even if that typology poorly sampled modern fauna. To date, these
occurs in different bones, individuals, or taxa. methods have been applied in a handful of
Determination of the quantity of primary tissue sauropod taxa (Curry 1998; Sander 2000;
types in fossil bones thus allows a unique Sander and Tückmantel 2003) and indicate that
insight into quantified bone growth rates in sauropods might have grown at rates on the
extinct or long-dead taxa. conservative end of the laminar, fibrolamellar
When skeletochronologic data are not avail- bone spectrum (table 11.1).
able, we can bound our assessments of the
maximum and minimum number of days that
BONE MICROSTRUCTURE AND OVERALL
an individual grew through the application of
BODY GROWTH RATES
“Amprino’s Rule”. For example, if we measure
the quantity of laminar bone tissue present in a As outlined above, bone histology provides an
sauropod femur, and we know the range of pos- exceptional tool for studying bone growth in
sible daily depositional rates for laminar tissue dinosaurs at the microscopic level. But is it pos-
in a variety of living vertebrates, we can use the sible to extrapolate from the microstructural
highest and lowest laminar tissue growth rates level to a more inclusive view of overall sauro-
in living vertebrates to provide a quick look at pod body growth? Even more significantly, is it
the sauropod lifespan. The robustness of these possible to compare somatic growth rates at the
methods has been poorly tested among living macroscopic level we calculate in sauropods to
vertebrates (Castanet et al. 2000; Chinsamy the rates among extant vertebrates? The most
and Starck 2002; Margerie et al. 2002), and common and physiologically informative meas-
more work is necessary in large-bodied mam- ures of developmental growth rates in living
mals, reptiles, and birds before clear answers taxa are not derived directly from bone
may be assessed, especially in light of the fact microstructure, but from some more easily
that fibrolamellar bone can be deposited at a 30- obtained unit (e.g., mass increase). In most ver-
fold range of rates. tebrates, mass changes with respect to age can
When skeletochronologic data exist in the be represented by sigmoidal curves (Case 1978a;
form of LAG or growth cycles, we can deter- Sussman 1964; Purves et al. 1992) with three
mine the average daily growth rate for highly characteristic growth stages (fig. 11.7). An initial
vascularized primary tissue forming zones by slow growth lag phase is followed by the rapid
calculating the average bone apposition rate per growth exponential stage when the majority of
growth cycle. To compare this number to values mass is typically accrued. The ontogenetic tra-
obtained for living vertebrates, we convert our jectory culminates in a reduction or cessation of
value for the dinosaur to micrometers per day, development during the stationary phase.
using the number of days in the time period of Mass standardized comparisons of maximum
interest (i.e., 377 for the Jurassic [Wells 1963]). overall body growth rates among phylogenetically
Of course, these rates must be considered in and morphologically diverse taxa can be made
light of the fact that some days of nondeposi- using values from the exponential stage of devel-
tion occur during LAGs, and it is impossible to opment. These comparisons indicate that rates
account for this time when calculating rates. absolutely increase with respect to body mass and
Depositional rates obtained for incremental tis- that individual clades have characteristic whole-
sue deposits can then be applied to noncyclic body growth rates (Case 1978a, 1978b; Calder

FIGURE 11.7. Growth curve for Spermophilus richardsonii (“Richardson’s ground squirrel”
[modified from Sussman 1964]). Early ontogeny is characterized by a slow-growth, lag
stage during infancy. This is followed by an exponential growth stage midway through
development, when maximum growth rates are obtained and most mass is accrued. De-
velopment culminates with a stationary stage, when growth slows or comes to a standstill.
1984; Erickson et al. 2001). Retrieving these data diaphyseal circumferences and the interspecific
from fossil bones is impossible without the quan- allometric equation established by Anderson
titative record of age at various ontogenetic stages et al. (1985). Because these equations are not
provided by histological analysis. The monumen- applicable throughout ontogeny, Erickson and
tal scale of most sauropods begs the question, Tumanova (2000) utilized DME, an allometric
What is the typical sauropod body growth trajec- scaling principle. The principle is based on the
tory? Erickson and Tumanova (2000) uniquely fact that if a linear measurement from an
combined skeletochronologic data with a unique anatomical structure scales with body mass
scaling principle they termed developmental throughout ontogeny, and the body masses of
mass extrapolation (DME) and quantified whole- more than one individual are known, then com-
body growth rates for dinosaurs. DME requires a parable linear measurements taken throughout
reliable age determination for bones and mass ontogeny can be used to extrapolate individual
quantifications at different ontogenetic stages. body masses. Femur length is one of the best
possible measures for this method because it
AGE ASSESSMENT scales isometrically during ontogeny in extant
Age is most commonly assessed from growth archosaurs (Dodson 1975; Carrier and Leon
marks in histologically prepared specimens, as 1990; Erickson and Tumanova 2000).
described above. Growth marks/bone tissue lost The first step in applying DME to dinosaurs
due to medullar expansion and bone remodel- is to take the cube of femur length (l3) for each
ing with increased age must be accounted for by specimen in an ontogenetic series. Each value
sequentially superimposing subadult speci- is then taken as a percentage of the adult value
mens on those from larger individuals. (the value for the largest known bone of inter-
est). The percentages are then converted to frac-
BODY MASS ASSESSMENT: DEVELOPMENTAL tional values and multiplied by the adult mass
MASS EXTRAPOLATION estimate (see above) to obtain body masses for
Body mass for adult animals in a sample may each of the included growth stages (Erickson
be estimated using the minimal long-bone and Tumanova 2000; Erickson et al. 2001;

Curry Rogers et al. 2003). The accuracy of this is still predicted to have lower growth rates
method in predicting exponential stage growth (55,638 g/day) than B. musculus (Case 1978a).
rates has been tested in humans and wild alli-
gators and is accurate within 5% in each case
OVERVIEW OF SAUROPOD
(Erickson et al. 2001). Following application of
GROWTH STRATEGIES
DME, growth curves can be reconstructed for
the extinct taxon of interest. Though sauropods were traditionally regarded
In our recent analyses of dinosaurian as long-lived animals with extremely prolonged
growth trajectories, we have demonstrated that ontogenies (Case 1978b; Reid 1981), recent
sigmoidal equations accurately describe growth years have witnessed a shift in our understand-
data for all dinosaurs (Erickson et al. 2001; ing of sauropod growth patterns and mecha-
Curry Rogers et al. 2003). All members of nisms. These new results have been based pri-
Dinosauria grow at rates 2–56 times faster than marily on bone histological data. A handful of
those of any living reptile. As a clade, dinosaurs diverse sauropods have been studied. Ricqlès
did not grow at rates intermediate to those of (1968a) was the first to conduct an explicit
reptiles, birds, and mammals, or at rates equiv- analysis of sauropod bone histological patterns
alent to those of living altricial birds. Instead, in “Bothriospondylus” (Lapparentosaurus) mada-
our results indicate that Dinosauria exhibited gascariensis. In two papers (1968a, 1983),
its own, unique growth trajectory and was Ricqlès documented the qualitative histology of
unique among vertebrate clades in having humeri and noted the presence of vascular
members with growth rates below, equal to, or cycles deposited in the context of highly vascu-
above typical mammalian/avian rates. Inter- larized fibrolamellar bone. Reid (1981) observed
estingly, the growth rates of dinosaurs numerous LAG in the pubis of an unidentified
depended on the size of the taxon of interest: sauropod and challenged Ricqlès’ interpretation
the larger the dinosaur, the more rapid the of sauropods as rapidly growing organisms.
overall body growth rate. The regression equa- Rimblot-Baly et al. (1995) extended the
tion we established for Dinosauria included six Ricqlès qualifications through the inclusion of a
taxa spanning the phylogenetic, size, and tem- Lapparentosaurus ontogenetic series. Rimblot-
poral range for the clade. It allowed us to make Baly et al. (1995) confirmed Ricqlès’ (1968a,
quantified predictions of dinosaurian growth 1983) interpretations of cortical stratification
rates beyond the range of included taxa and also noted lamellar, accretionary tissue at
(Erickson et al. 2001). Sauropod growth rates the cortical periphery in their largest samples.
vary from taxon to taxon and approach those The work by Rimblot-Baly et al. (1995) is also
known for extant whales (absolutely and rela- significant because it constitutes the first
tively some of the fastest-growing eutherians). attempt at quantifying bone growth rates
For example, a fully grown Apatosaurus grew at directly from sauropod bone tissue: they pro-
about 14,460 g/day, compared with 20,700 posed a bone appositional rate of 7 m/day.
g/day for a 30,000-kg gray whale (Eschrichtius This rate is similar to rates observed in living
robustus [Case 1978a]). Though this rapid cows and other large mammals and birds,
growth rate may seem astounding, it is upheld which typically deposit laminar, fibrolamellar
by bone histological data and still does not hold bone.
a candle to the growth rates observed in the Sander (2000, 2003) and Sander and
extant blue whale (Balaenoptera musculus, the Tückmantel (2003) provided histological descrip-
fastest-growing animal ever, at 66,000 g/day, tions for the Jurassic sauropods Brachiosaurus,
[Case 1978a]). In fact, even the enormous Barosaurus, Dicraeosaurus, Janenschia, and
sauropod Argentinosaurus (unconservatively Apatosaurus. In each taxon, highly vascularized
estimated to weigh 100,000 kg at adult size) fibrolamellar bone dominates throughout

ontogeny. In addition, Sander (2000) distin- TABLE 11.2
guished taxa on the basis of the relative degree of Apatosaurus Relative Age Classes, Derived from Maximum
bone remodeling and spatial characteristic. He Length, Muscle Scar Development, and Bone Surface Finish
concluded that Brachiosaurus attained sexual MAXIMUM PERCENTAGE

maturity at 40% maximum size, while ELEMENT LENGTH (CM) OF ADULT SIZE
Barosaurus may not have reached this plateau
until 70% of body size had been obtained. Polish I Radius R1 45.5 61
lines present in the single Janenschia sample Ulna U1 47.9 61
yielded age estimates of sexual maturity at 11 Scapula S1 70.7 34
years, maximum size at 26 years, and death at 38 II Radius R2 54.5 73
years of age. In addition, Sander utilized differing Ulna U2 52.2 67
histological patterns in Barosaurus individuals to Scapula S2 117.4 56
develop a hypothesis of sexual dimorphism III Radius R3 67.5 91
related to reproduction. Ulna U3 59.4 76
Analyses such as those performed by IV Radius R4 74.2 100
Rimblot-Baly et al. (1995) and Sander (2000, Ulna U4 78.5 100
2003) have provided abundant data and the req- Scapula S3 210.9 100
uisite raw material for furthering our under-
standing of sauropod growth trajectories. Both
consider taxa in an ontogenetic series, and tion, and muscle scar development (table 11.2).
Sander (2000) considers multiple skeletal ele- To rule out the effects of environment on the
ments. The building of our base of knowledge interpretation of histological patterns, all but the
about sauropod growth dynamics relies on such largest adult specimen were from a single local-
detailed analytical studies, and we now have the ity. Similarly, multiple elements from different
tools to take this work even further. functional environments were included to
negate the influence of biomechanics on the
overall histological signal. Transverse thin sec-
APPLICATION AND EXAMPLE:
tions were made at standardized locations in
APATOSAURUS AS A CASE STUDY
each element (fig. 11.8). Comparable histologi-
To date, Apatosaurus is the only sauropod for cal organization of bones of similar sizes and
which all the methods outlined above have been inferred relative ages served as an independent
utilized to document patterns in bone and life test of a priori age class assessments. Age classes
history strategy. Here we apply each method to I and II represented juveniles up to 75% of
an ontogenetic sample of Apatosaurus to illus- adult size, age class III represented subadults
trate (1) the kinds of data that can be generated, between 76% and 91% of adult size, and age
(2) the utility and shortcomings of each class IV represented fully grown adults.
method, and (3) how different methods out- Qualitative data on interelement histological
lined in this chapter can be used to build on and variation and ontogenetic bone depositional pat-
test answers derived from their alternatives. terns served as the basis for the first quantifica-
tions of Apatosaurus skeletal growth strategy.
THE SAMPLE
Curry (1999) histologically analyzed five ulnae, HISTOLOGICAL CHARACTERIZATIONS,
four radii, and three scapulae from at least four APPLICATIONS, AND INFERENCES
Apatosaurus individuals at various stages of Laminar and plexiform fibrolamellar bone pre-
ontogenetic development. The available sample dominates during 90% of ontogenetic growth in
was divided into four relative age classes on the all Apatosaurus elements studied. Laminar–
basis of overall bone length, degree of ossifica- plexiform deposits are present in each element

FIGURE 11.8. Histological sam-
ple of Apatosaurus excelsus, from
left to right: Scapula, radius, and
ulna. Scapulae were sectioned at
the proximal end of the scapular
blade. Radii and ulnae were
sectioned at middiaphysis in
areas devoid of muscle scars.
(Modified from Curry 1999:
fig. 1.)
but have different gross morphologies in radii, plateau and the attainment of adult size in
ulnae, and scapulae of similar relative ages Apatosaurus. These histological results parallel
(fig. 11.9). Radii and ulnae exhibit noncyclic, con- the sigmoidal, determinate growth curve for
tinuous deposition of laminar bone throughout mass increase in extant mammals and birds
most of ontogeny (fig. 11.9A). In contrast, cyclic- (Case 1978a; Calder 1984) and contradict the
ity of laminar bone deposition is evident at low traditional view of indeterminate growth in
magnifications in Apatosaurus scapulae (fig. sauropods (Ricqlès 1968b, 1983; Rimblot-Baly
11.9B), and cycles are not demarcated by annuli et al. 1995). The predominance of laminar and
or LAG. Instead, each scapular cycle is demar- plexiform bone throughout most of Apatosaurus
cated only by a regular variation in vascular canal ontogeny is most similar to that observed in
density and width, always in the context of con- actively growing extant mammals and birds (e.g.
tinuous laminar bone deposition. Despite this Ricqlès et al. 1991; Castanet et al. 1996, 2000).
histovariability between forelimb and girdle ele- Amprino’s Rule suggests that laminar tissue in
ments, all three elements maintained rapid rates sauropods grew at the same basic range of rates
of bone growth throughout ontogeny, slowing that occurs in extant mammals and birds. Bone
only after subadult sizes were reached. tissue patterns thus suggest that birds and
After 90% of adult size was attained, the mammals are the most appropriate analogues
highly vascularized fibrolamellar primary bone for investigations of Apatosaurus growth.
was replaced by the first deposits of slower- These qualitative data are a useful starting
growing avascular, lamellar-zonal bone with point for understanding ages and longevity in
numerous annuli and LAG (fig. 11.10). Lamellar– Apatosaurus and serve as the basis for skele-
zonal bone and peripheral LAG indicate a growth tochronologic estimates of age. Cycles similar

FIGURE 11.9. Apatosaurus tissue histology. Radii, ulnae, and scapulae each exhibit continuous, rapid deposition of highly
vascularized, fibrolamellar bone for most of ontogeny. (A) Laminar bone maintains a consistently high rate of growth in
radii and ulnae but is punctuated by areas of reduced vascular density and canal size in scapulae. Scale bar equals 100 m.
(B) Annuli and LAG do not mark vascular cycles observed in scapulae, thus attesting to the continuous, rapid deposition of
primary bone in each element until late ontogeny. Scale bar equals 500 m.
to those in Apatosaurus scapulae also occur in cycles in scapula 2 yielded conservative age esti-
extant sirenians (Fawcett 1942), sea turtles mates of 5 and 10 years, respectively. Because
(Rhodin 1985), some other sauropods (Ricqlès radii and ulnae did not have growth cycles or
1968b, 1983; Reid 1981; Rimblot-Baly et al. LAG, scapular cycle counts were applied to
1995; Sander 2000), and ichthyosaurs radius and ulna in the same relative age class to
(Buffrénil and Mazin 1990). Just as in these provide a testable hypothesis of absolute age for
taxa, Apatosaurus vascular cycles are assumed to all bones in each age class (I–III). Age class I
represent annually periodic bone depositional radius, ulna, and scapula were estimated at 5
events on the grounds of phylogenetic parsi- years. Similarly, class II and III elements were
mony and tissue formation rates consistent estimated at between 8 and 10 years of age
with those in living vertebrates (Castanet and (table 11.3).
Smirina 1990; Erickson and Tumanova 2000).
Endosteal reconstruction characterizes scapu- AMPRINO’S RULE, INFERENCES, AND BONE
lae included in this analysis, but isolated GROWTH RATE QUANTIFICATION
patches of primary bone are retained in the Laminar/reticular–plexiform bone tissue in a
deep cortex of the youngest scapula and allow fibrolamellar context predominates for most of
gaps in the cortical bone record to be recon- Apatosaurus ontogeny. Amprino’s Rule provides
structed. Lost growth rings can also be a means for more precisely defining “rapid”
accounted for by sequentially superimposing bone growth for Apatosaurus. In this case, tissue
subadult specimens on those from larger indi- growth rates for laminar–reticular bone are
viduals. Data from the younger, less remodeled known in several modern taxa: ostrich (Castanet
bones “fill in” missing data in older elements et al. 2000; Sander and Tückmantel 2003), emu
and provide a more complete history of growth. (Castanet et al. 2000), mink (Ricqlès et al. 1991),
Maximum counts of 5 cycles in scapula 1 and 10 duck (Castanet et al. 1996; Margerie et al. 2002),

FIGURE 11.10. Apatosaurus exponential and stationary stages of growth. (A) Periosteal
surface in juvenile Apatosaurus radius (50% of adult size) demonstrating fast-growth
laminar, fibrolamellar bone in the external cortex. This continuous, rapid growth
pattern is consistent with the exponential stage of growth. (B) Periosteal surface in
adult Apatosaurus radius (90% of adult size). The first peripheral lines of arrested
growth (LAG) appear in the external regions of Apatosaurus radius cortical bone. These
LAG document the onset of the stationary phase of growth. Scale bars equal 100 m.
and cow (Ricqlès et al. 1991). These experimental forms at rates of between 10 and 80 m/day.
data indicate that the apposition rates of variably Though tissue growth rates for a given typology
vascularized fibrolamellar bone can overlap and vary among these taxa and can overlap and, at
include a wide range of possible values. For least in Anas (Margerie et al. 2002) the growth
example, in the duck Anas platyrhynchos rates observed may not be directly linked to
(Margerie et al. 2002), the rates vary from less bone vascularity, application of any of these
than 20 to 110 m/day. Similarly, in ostriches growth rates to the laminar tissue observed
and emus laminar–reticular fibrolamellar bone in Apatosaurus does have implications: namely,

TABLE 11.3
Apatosaurus Ages Derived from Application of Apatosaurus Scapular Laminar Bone Growth Rates to Radius and Ulna
AGE (YEARS)
ELEMENT SCAPULA 1 RATE (10.1 M/DAY) SCAPULA 2 RATE (11.4 M/DAY)
Radius 1 5.8 5.1

Radius 2 7.6 6.8
Radius 3 10.0 8.8
Ulna 1 5.5 5.1
Ulna 2 7.1 7.8
Ulna 3 10.5 9.3
NOTE: Scapula 1 and 2 laminar bone depositional rates are calculated by dividing the average amount of laminar bone deposited in
each cycle by 378 (Wells, 1963). This rate is then applied to radius or ulna by:
m radial or ulnar laminar bone 10.1/11.4 m scapular laminar bone

days of total growth 1 day of growth
Converting days to years gives an answer for age for a particular relative age class.
providing a testable array of possible primary higher, at 11.4 m/day. We then apply this
bone depositional rates for Apatosaurus. However, observed rate of bone deposition to other skeletal
it is important to stress that this application of elements that lack cycles (radius and ulna) and
bone growth rates observed in living animals is arrive at an age for each Apatosaurus element
merely a means of bounding the possibilities of under study. Measuring the minimum thick-
growth rates in this dinosaur. ness of continuous laminar bone for each
The second method for aging Apatosaurus radius and ulna provides a conservative esti-
outlined above provides just the data needed to mate of total bone growth. Conversion of days
determine how the speed of sauropod bone to years provides an age estimate for each
growth compares to the speed of bone growth Apatosaurus forelimb element. Age class I
in living vertebrates. Rather than relying on the radius and ulna were between 4 and 6 years of
application of bone appositional rates in living age, class II radius and ulnae were between 6
taxa, we turn to the bone of Apatosaurus itself. and 9 years, and age class III elements were
In the first step, we measure the thickness of 8–11 years of age (table 11.3). Similar work con-
laminar, fibrolamellar bone in scapular cycles ducted by Sander and Tückmantel (2003) yields
and take an average of the growth rate around comparable results for bone apposition rates in a
the cross section (thus negating the influences variety of sauropods (between 1 and 20 m/day,
of the shape of the scapula). We divide this aver- yielding comparable ages for the attainment of
age amount of bone per cycle by the number of adult size). While significant, these quantified
days in the Jurassic year (377 [Wells 1963]) and growth rates must be viewed with some cau-
thus develop an estimate of average laminar tion because they document growth at very
bone growth rates directly from Apatosaurus specific bone locations and only in a single
bone. Average cycle thicknesses are consistent direction (centripetal growth). Extending these
in the two included scapulae, and laminar tis- rates to the entire organism is impossible with-
sue was deposited at comparable rates in each: out a complete understanding of the allometry
for scapula 1 (50% adult size), laminar bone of the skeleton. Because this type of data does
accreted at a rate of 10.1 m/day. In scapula not exist for any living taxon, it is unlikely that
2 (60% adult size), the rate was slightly it will exist for a dinosaur in the near-future,

and we must view these results with this in body growth rates (Erickson et al. 2001) and pro-
mind. vided more detailed evidence on the relation-
ship between bone tissue growth and mass
LONGEVITY ESTIMATES AND DEVELOPMENTAL increase during ontogeny in this taxon. Our
MASS EXTRAPOLATION: APPLICATION analysis revealed that a sigmoidal equation accu-
AND INFERENCES rately described the growth data for Apatosaurus
How do we move from the microscopic to the (fig. 11.11). The exponential stage growth was
somatic level of growth rate interpretation in approximately 6 years long, and the onset of
extinct taxa? In most extant animals, mass somatic maturity occurred at about age 13. In
changes with respect to age show sigmoidal pat- terms of mass increase, Apatosaurus had growth
terns (Sussman 1964). Standardized compar- rates similar to those of extant whales, and may
isons of maximum growth rates among phylo- have added about 14,460 g/day to their body
genetically and morphologically diverse taxa can mass. Interestingly, although Apatosaurus bone
be made using data from the exponential stage tissue is morphologically indistinguishable from
of development (Case 1978a,b; Calder 1984; that of altricial birds (i.e., both deposit abundant
Erickson and Tumanova 2000). To compare laminar, fibrolamellar bone), Apatosaurus accu-
whole-body growth rates between dinosaurs and mulated body mass at rates about half that of a
living vertebrates, similar quantified data are scaled-up altricial bird (123,025 g/day [Erickson
needed. This requires growth series that span et al. 2001]). Our regression analysis of dinosaur
the full range of dinosaur size, shape, and phy- growth also allowed us to extrapolate to larger
logeny, and accurate age and mass assessments dinosaurs, and it is clear that the largest known
at all ontogenetic stages. The merging of bone dinosaur (Argentinosaurus; estimated adult body
histological methods with scaling principles size of 100,000 kg [Appenzeller 1994]) would
provided the requisite tools for the analysis of still have grown at 55,638 g/day, a rate absolutely
how diverse dinosaurs really grew (Erickson and slower than that estimated for a scaled-up altri-
Tumanova 2000; Erickson et al. 2001). Body cial bird.
mass estimates are obtained using long-bone
circumferences and regression equations GROWTH RATE SYNTHESIS
(Anderson et al. 1985). Because these equations All of the methods outlined above reveal that
do not work for all ontogenetic stages, the scal- Apatosaurus grew rapidly throughout most of
ing principle called DME (Thompson 1943; ontogeny. Histological typing identifies highly
Erickson and Tumanova 2000) was employed. vascularized fibrolamellar bone tissue with no
The data for a diverse group of dinosaurs indi- LAG or growth marks until late in ontogeny. The
cate that sigmoidal equations accurately fit the first evidence of a bone growth rate reduction
growth trajectories for Dinosauria and that occurs in subadults, and the largest adult
dinosaurs as a whole did not show growth rates Apatosaurus bones demonstrate continued
intermediate between those of birds/mammals diminished growth with attainment of adult body
and reptiles. Instead, our results indicate that size. Categorization of bone tissue typologies is
Dinosauria was unique among major vertebrate useful in that it provides a “first pass” on growth
clades in having members that show rates rates during ontogeny, as well as longevity esti-
below, equivalent to, or above typical mam- mates for the individuals under study.
malian/avian rates. In addition, the distribution Amprino’s Rule allows direct comparison of
of these rates is related to the size of the organ- quantified growth rates we might obtain directly
ism under consideration: the larger the dinosaur, from sauropod bones to those experimentally
the faster it reaches a somatic maximum obtained in living vertebrates. In summary,
(Erickson et al. 2001). We included Apatosaurus Amprino’s Rule tests the real possibility of our
ontohistogenetic data in our analysis of overall hypothesis: if rates we obtain are observed in

FIGURE 11.11. Apatosaurus growth rates. (A) Growth curve for Apatosaurus. The largest animal in the sample is among the
largest specimens known. The growth curve for Apatosaurus is similar to that known in living vertebrates and includes an
early-ontogeny lag stage, followed by an exponential stage, and, finally, a stationary stage. (B) Comparison of exponential
stage growth rates in Apatosaurus (A), Maiasaura (M), Massospondylus (M), Syntarsus, (S), Psittacosaurus (P), and Shuvuuia (S)
with typical values for extant vertebrates. Standardized comparisons are made using contrasts between animals of compara-
ble adult mass to diminish signal from differences in shape and negate the effects of size. Growth rates for included di-
nosaurs are as follows: Shuvuuia, 3.4 g/day; Psittacosaurus, 12.5 g/day; Syntarsus, 23.9 g/day; Massospondylus, 90.3 g/day;
Maiasaura, 2,793 g/day; and Apatosaurus, 14,460 g/day. (Modified from Erickson et al. 2001.)
living taxa for the same tissue typology, our history studies. Similarly, comparisons among
argument for Apatosaurus is not disproven. sauropods (e.g., Sander, 2000, 2003; Sander
Application of DME and longevity estimates and Tückmantel 2003; Sander et al. 2004) high-
provide quantified measures of growth rates for light the potential of histology to inform about
all stages of whole-body growth, including the seemingly intractable questions such as sauro-
exponential stage. Because growth comparisons pod sexual dimorphism.
can be made with living taxa when the exponen- As our understanding of histology, skele-
tial stage whole-body growth rate is known and tochronology, and DME continues to grow, the
mass is standardized, DME provides one of the phylogenetic significance of sauropod growth
most groundbreaking advances in our ability to strategies can finally be addressed. Useful phy-
decipher whole-organism life history patterns in logenetic characters may be derived from
sauropods and other extinct taxa. growth curves, and it will be possible to com-
pare overall body growth strategies employed
by divergent sauropod taxa. For example,
FUTURE DIRECTIONS
Saltasaurus and Argentinosaurus are both
In this chapter we have demonstrated the utility titanosaurs, but the end points of their growth
of bone histological analysis for interpreting a trajectories vary dramatically. With a combined
wide range of paleobiological questions sur- histological/DME approach, we can specifically
rounding sauropods, which cannot be answered evaluate the onset and length of exponential-
using traditional gross morphological studies. stage growth rates to address the mechanisms
That said, we have only touched the surface on of sauropods gigantism. With a deeper under-
our way to a more comprehensive understand- standing of sauropod life history strategies, our
ing of sauropod growth strategy on micro- and knowledge will extend beyond the realm of
macroscopic levels. Bones hold the keys to sauropod biology to questions of ecology,
addressing even more detailed questions of resource allocation, and evolution.
sauropod life history, overall growth strategy,
evolution, and behavior.
ACKNOWLEDGMENTS
As we have outlined in this chapter, the qual-
ity of available data varies with each step of an We wish to thank E. Schexnayder and C. Wong
histological analysis, and all steps, from qualifi- for assistance with illustrations. We are also
cation of bone depositional rates to the whole- grateful to Martin Sander and Anusuya
body approach of DME, are informative in some Chinsamy, who provided helpful reviews of the
way. At the descriptive stage of an histological manuscript. Of course, thanks would be incom-
research project, we establish hypotheses of rel- plete without acknowledgment of all that Jack
ative growth rates and age estimates based on McIntosh has brought to the study of sauropods.
histological characteristics and skeletochronol- This research is supported by the National
ogy. These hypotheses are based on the compar- Science Foundation, the Paleontology Depart-
isons drawn between the tissue of sauropods ment at the Science Museum of Minnesota, the
and that of living vertebrates, within a single Geology Department at Macalester College, and
ontogenetic series, and within the skeleton of a the Department of Biological Sciences at Florida
single individual. Thus, collecting as many data State University.
as possible, in terms of number of bones, num-
ber of individuals, and variety of comparative
specimens, will significantly bolster our ability LITERATURE CITED
to interpret fossil tissues. Luckily, many sauro- Altman, P. L., and Ditmer, D. S. 1972. Biology Data
pod taxa from around the world include growth Book, Vol. I. Federation of the American Society
series and will prove most valuable for life of Experimental Biology. Bethesda, MD.

Amprino, R. 1947. La structure du tissu osseux envis- Castanet, J., Curry Rogers, K., Cubo, J., and Boisard,
agée comme expression de différences dans la J. J. 2000. Quantification of periosteal osteogene-
vitesse de l’accroissement. Arch. Biol. 58: 315–330. sis in ostrich and emu: implications for assessing
———. 1948. A contribution to the functional mean- growth in dinosaurs. C. R. Acad. Sci. (Paris) III
ing of the substitution of primary by secondary 323: 543–550.
bone tissue. Acta Anat. 5(III): 291–300. Chiappe, L. M., Coria, R. A., Dingus, L., Jackson, F.,
Anderson, J. F., Hall-Martin, A., and Russell, D. A. 1985. Chinsamy, A., and Fox, M. 1998. Sauropod
Long-bone circumference and weight in mammals, dinosaur embryos from the Late Cretaceous of
birds, and dinosaurs. J. Zool. A207: 53–61. Patagonia. Nature 396: 258–261.
Appenzeller, T. 1994. Argentine dinos vie for heavy- Chiappe, L. M., Salgado, L., and Coria, R. A. 2001.
weight title. Science 266: 1805. Embryonic skulls of titanosaur sauropod
Breton, G., Fournier, R., and Watté, J.-P. 1985. Le lieu dinosaurs. science 293: 2444–2446.
de ponte de dinosaures de Rennes-le-Chateau Chinsamy, A. 1990. Physiological implications of the
(Aude): premiers résultats de la campagne de bone histology of Syntarsus Rhodesiensis (Saurischia:
fouilles 1984. Actes Coll. Dinosaures Chine Theropoda). Palaeontol. Afr. 27: 77–82.
France 1985: 127–140. ———. 1993. Bone histology and growth trajectory
Buffrénil, V. de, and Mazin, J. -M. 1990. Bone histol- of the prosauropod dinosaur Massospondylus cari-
ogy of the ichthyosaurs: comparative data and natus Owen. Modern Geology 18: 319–329.
functional interpretation. Paleobiology 16: Chinsamy, A., and Elzanowski, A. 2001. Evolution
435–447. of growth patterns in birds. Nature 412:
Calder, W. A.,III. 1984. Size, Function, and Life History. 402–403.
Harvard University Press, Cambridge, MA. Coe, M. J., Dilcher, D. L., Farlow, J. O., Jarzen, D. M.,
Carrier, D., and Leon, L. R. 1990. Skeletal growth of and Russell, D. A. 1987. Dinosaurs and land
the California gull (Larus Californicus). J. Zool. plants. In: Friis, E. M., Chaloner, W. G., and Crane,
London 222: 375–389. P. R. (eds.). The Origins of Angiosperms and
Carter, D. C., and Beaupré, G. S. 2001. Skeletal Their Biological Consequences. Cambridge
Function and Form, Mechanobiology of Skeletal University Press, Cambridge. Pp. 225–228.
Development,Aging,andRegeneration. Cambridge Cooper, L. N., and Horner, J. R. 1999. Growth rate of
University Press, Cambridge. 318 pp. Hypacrosaurus stebingeri as hypothesized from
Case, T. J. 1978a. On the evolution and adaptive sig- lines of arrested growth and whole femur cir-
nificance of postnatal growth rates in the terres- cumference. J. Vertebr. Paleontol. 19: 39A.
trial vertebrates. Q. Rev. Biol. 53: 243–282. Currey, J. D. 1987. The evolution of the mechanical
———. 1978b. Speculations on the growth rate and properties of amniote bone. J. Biomech. 20:
reproduction of some dinosaurs. Paleobiology 4: 1035–1044.
320–328. Curry, K. A. 1998. Histological quantification of
Castanet, J., and Naulleau, G. 1985. La squelet- growth rates in Apatosaurus. J. Vertebr. Paleontol.
tochronologie chez les Reptiles. II. Résultats 18: 36A.
expérimentaux sur la signification des marques de ———. 1999. Ontogenetic histology of Apatosaurus
croissance squelettiques utilisées comme critère (Dinosauria: Sauropoda): New insights on growth
d’âge chez les serpents. Remarques sur la crois- rates and longevity. J. Vertebr. Paleontol. 19:
sance et la longévité de la Vipère Aspic. Ann. Sci. 654–665.
Nat. Zool. (Paris) 7: 41–62. Curry Rogers, K. A., Erickson, G., and Norell, M. A.
Castanet, J., and Smirina, E. 1990. Introduction to 2003. Dinosaurian life history strategies, growth
the skeltochronological method in amphibians rates, and character evolution: New insights gar-
and reptiles. Ann. Sci. Nat. Zool. (Paris) 11: nered from bone histology and developmental
191–197. mass extrapolation. J. Vertebr. Paleontol. 23(3):
Castanet, J., Francillon-Vieillot, H., Meunier, F. J. and 90A.
Ricqlès, A. de 1993. Bone and individual aging. Dodson, P. 1975. Functional and ecological signifi-
In: Hall, B. K. (ed.). Bone. Bone Growth—B. Vol. cance of relative growth of Alligator. J. Zool.
7. CRC Press, Boca Raton, FL. Pp. 245–283. London 175: 315–355.
Castanet, J., Grandin, A., Arbourachid, A., and Enlow, D. H. 1963. Principles of Bone Remodeling.
Ricqlès, A. de 1996. Expression de l’dynamique Charles C Thomas, Springfield, IL. 123 pp.
de croissance dans la structure de los périostique Enlow, D. H., and Brown, S. O. 1956. A comparative
chez Anas platyrhynchous. C. R. Acad. Sci. (Paris) histological study of fossil and recent bone tissue
Sci. Vie 319: 301–308. (Part 1). Tex. J. Sci. 8: 405–443.

———. 1957. A comparative histological study of fos- Mantell, G. A. 1850. On the Pelorosaurus, an unde-
sil and recent bone tissue (Part 2). Tex. J. Sci. 9: scribed gigantic terrestrial reptile, whose remains
186–214. are associated with those of the Iguanodon and the
———. 1958. A comparative histological study of fos- other saurians in the strata of the Tilgate Forest, in
sil and recent bone tissue (Part 3). Tex. J. Sci. 10: Sussex. Philos. Trans. Roy. Soc. London 140:
187–230. 379–390.
Erickson, G. M., and Brochu, C. 1999. How the “ter- Margerie, E. de, Cubo, J., and Castanet, J. 2002. Bone
ror crocodile” grew so big. Nature 398: 205–206. typology and growth rate: testing ‘Amprino’s
Erickson, G. M., and Tumanova, T. A. 2000. Growth Rule’ in the mallard (Anas platyrhynchos). C. R.
curve of Psittacosaurus mongoliensis Osborn Biol. 325: 221–230.
(Ceratopsia: Psittacosauridae) inferred from Martin, R. B., and Burr, D. B. 1989. Structure,
long bone histology. Zool. J. Linn. Soc. 130: Function and Adaptation of Compact Bone.
551–556. Raven Press, New York. 275 pp.
Erickson, G. M., Curry Rogers, K., and Yerby, S. 2001. Owen, R. 1859. Monograph on the fossil Reptilia of
Dinosaurian growth patterns and rapid avian the Wealden and Purbeck Formations.
growth rates. Nature 412: 429–433. Supplement II. Crocodilia (Streptospondylus etc.).
Fawcett, D. W. 1942. Amedullary bones of the Florida Palaeontogr. Soc. Monogr. 11: 20–44.
manatee (Trichechus Latirostris). Am. J. Anat. 71: Padian, K., Ricqlès, A. J. de and Horner, J. R. 2001.
271–303. Dinosaurian growth rates and bird origins.
Foote, J. S. 1913. Th comparative histology of the Nature 412: 405–408.
femur. Smithsonian Miscellaneous Collection Powell, J. E. 1986. Revision de los titanosauridos de
61(2232). America del Sur. unpublished Ph.D. thesis,
Francillon-Vieillot, H., de Buffrénil, V., Castanet, J., Universidad Nacional de Tucuman, Tucuman.
Géraudie, J., Meunier, F. J., Sire, J. Y., Zylberberg, 340 pp.
L., and Ricqlès, A. de 1990. Microstructure and ———. 1992. Osteologia de Saltasaurus loricatus
mineralization of vertebrate skeletal tissues. In: (Sauropoda–Titanosauridae). In: Sanz, J. L., and
Skeletal Biomineralization: Patterns, Processes, Buscalioni, A. D. (eds.). Los Dinosaurios y su
and Evolutionary Trends. Vol. 1. Van Nostrand Entorno Biotico. Instituto “Juan de Valdes,”
Reinhold, New York. Pp. 473–530. Cuenca.
Gillette, D. D. 1991. Seismosaurus halli (n. gen., n. sp.) ———. 2003. Revision of South American
a new sauropod dinosaur from the Morrison Titanosaurid dinosaurs: palaeobiological, palaeo-
Formation (Upper Jurassic–Lower Cretaceous) of biogeographical, and phylogenetic aspects. Rec.
New Mexico, U.S.A. J. Vertebr. Paleontol. 11: Queen Victoria Mus. Launceston 111: 1–173.
417–433. Purves, W. K., Orians, G. H., and Heller, H. C. 1992.
Horner, J. R., Ricqlès, A., de and Padian, K. 1999. Life: The Science of Biology. 3rd ed. Freeman, Salt
Variation in skeletochronological indicators of the Lake City, UT.
hadrosaurid dinosaur Hypacrosaurus: implications Queckett, J. T. 1855. Descriptive and Illustrated
for age assessment in dinosaurs. Paleobiology 25: Catalogue of the Histological Series Contained in
295–304. the Museum of the Royal College of Surgeons. II.
———. 2000. Long bone histology of the Structure of the Skeleton of Vertebrate Animals.
hadrosaurid dinosaur Maiasaura peeblesorum; Royal College of Surgeons, London.
growth dynamics and physiology based on an Reid, R. E. H. 1981. Lamellar-zonal bone with zones
ontogenetic series of skeletal elements. J. Vertebr. and annuli in the pelvis of a sauropod dinosaur.
Paleontol. 20: 115–129. Nature 292: 49–51.
Jianu, C. -M., and Weishampel, D. B. 1999. The ———. 1984. The histology of dinosaurian bone,
smallest of the largest, a new look at possible and its possible bearing on dinosaurian physiol-
dwarfing in sauropod dinosaurs. Geol. Mijnbouw ogy; pp. 629-663 in M. W. J. Ferguson (ed.),
78: 335–343. The Structure, Development and Evolution of
Lacroix, P. 1971. The internal remodeling of bones. Reptiles: Zoological Journal of London
In: Bourne, G. H. (ed.). The Biochemistry and Symposia 52. Academic Press, London.
Physiology of Bone. Vol. 3. Academic Press, New Pp.629–663.
York. Pp. 119–144. ———. 1990. Zonal “growth rings” in dinosaurs.
Lanyon, L. E., and Rubin, C. T. 1984. Static vs. dynamic Modern Geol. 15: 19–48.
loads as an influence on bone remodelling. Journal ———. 1997. How dinosaurs grew. In: Farlow, J. O.,
of Biomechanics 17: 897–905. and Brett-Surman, M. K. (eds.). The Complete

Dinosaur. Indiana University Press, Bloomington in sauropod humeri and femora. Palaontol. Z.
and Indianapolis. Pp. 403–413. 77(1): 139–150.
Rhodin, A. G. J. 1985. Comparative chondro-osseus Sander, P. M., Laven, T., Mateus, O., and Knoetschke,
development and growth of marine turtles. N. 2004. Insular dwarfism in a brachiosaurid
Copeia 1985: 752–771. sauropod from the Upper Jurassic of Germany.
Ricqlès, A. de. 1968a. Quelques observations paléo- J. Vertebr. Paleontol. 243, (Suppl.): 108A.
histologiques sur les dinosaurien sauropode Schweitzer, M. H., Watt, J. A., Avci, R. Forster, C. A.,
Bothriospondylus. Ann. Univ. Madagascar. 6: Krause, D. W., Knapp, L., Rogers, R. R., Beech, I.,
157–209. and Marshall, M. 1999. Keratin immunoreactivity
———. 1968b. Recherches paléohistologiques sur in the Late Cretaceous bird Rahonavis ostromi. J.
les os longs des tétrapodes. I—Origine du tissu Vertebrate Paleontology 19: 712–722.
osseux plexiforme des dinosauriens sauropodes. Schweitzer, M.H., Wittmeyer, J.L., Horner, J.R., and
Ann. Paleontol. 54: 133–145. Janktoporski. 2005. Soft-tissue vessels and cellu-
———. 1977. Recherches paléohistologiques sur les lar preservation in Tyranosaurus Rex. Science
os longs de Tétrapodes. VII (deuxième partie: 307: 1952–1955.
fonctions) (suite). Ann. Paléontol. (Vertebr.) 62: Seebacher, F. 2001. An new method to calculate allo-
71–126. metric length-mass relationships of dinosaurs.
———. 1980. Tissue structure of dinosaur bone. J. Vertebr. Paleontol. 21: 51–60.
Functional significance and possible relation to Smith, J. B., Lamanna, M. C., Lacovara, K. J., Dodson,
dinosaur physiology. In: Thomas, R. D. K., and P., Smith, J. R., Poole, J. C.,Giegengack, R., and
Olson, E. C. (eds.). A Cold Look at the Warm- Attia, Y. 2001. A giant sauropod dinosaur from an
Blooded Dinosaurs. AAAS Selected Symposium. Upper Cretaceous mangrove deposit in Egypt.
Vol. 28. Westview Press, Boulder, CO. Pp. Science 292: 1704–1706.
103–139. Starck, J. M., and Chinsamy, A. 2002. Bone
———. 1983. Cyclical growth in the long limb bones microstructure and developmental plasticity in
of a sauropod dinosaur. Acta Paleontologica 28: birds and other dinosaurs. J. Morphol. 254:
225–232. 232–246.
Ricqlès, A. de., Meunier, F. J., Castanet, J., and Sussman, M. 1964. Growth and Development.
Francillon-Vieillot, H. 1991. Comparative micro- Prentice–Hall, Upper Saddle River, NJ.
structure of bone. In: Hall, B. K. (ed.). Bone, Bone Thompson, D. 1943. On Growth Form. Cambridge
Matrix, and Bone Specific Products. Vol. 3. CRC University Press. 1116 pp.
Press, Boca Raton, FL. Pp. 1–78. Varricchio, D. J. 1993. Bone microstructure of the
Rimblot-Baly, F., Ricqlès, A. de, and Zylberberg, l. upper Cretaceous theropod dinosaur Troodon for-
1995. Analyse paléohistologique d’une série de mosus. J. Vertebr. Paleontol. 13: 99–104.
croissance partielle chez Lapparentosaurus mada- Weishampel, D. B., and Horner, J. R. 1994. Life his-
gascariensis (Jurassique Moyen): Essai sur la tory syndromes, heterochrony, and the evolution
dynamique de croissance d’un dinosaure of Dinosauria. In: Carpenter, K., Hirsch, K. F.,
sauropode. Ann. Paleontol. 81: 49–86. and Horner, J. R. (eds.). Dinosaur Eggs and
Sander, P. M. 2000. Longbone histology of the Babies. Cambridge University Press, Cambridge.
Tendaguru sauropods: implications for growth Pp. 229–243.
and biology. Paleobiology 26: 466–488. Wells, J. W. 1963. Coral growth and geochronometry.
———. 2003. Long and girdle bone histology in Nature 197: 948–950.
sauropod dinosaurs: methods of study and impli- Witmer, L. M. 2001. Nostril positions in dinosaurs
cations for growth, life history, taxonomy, and and other vertebrates and its significance for
evolution. J. Vertebr. Paleontol. 233 (Suppl.): 93A. nasal function. science 293: 850–853.
Sander, P. M., and Tückmantel, C. 2003. Bone lamina Wykoff, R. W. G. 1975. Pleistocene and dinosaur
thickness, bone apposition rates, and age estimates gelatins. Comp. Biochem. Physiol. 55B: 95–97.

a conversation with jack mcintosh
John S. (“Jack”) McIntosh has been a leading student the entirety of sauropod dinosaurs, a monumental
of sauropod dinosaurs for well over half a century. effort that serves as a standard reference for any inves-
During the course of his long career, Jack has been tigation into sauropods and also provides raw infor-
influenced by legendary paleontologists such as mation for phylogenetic analyses of sauropod relation-
Barnum Brown, Richard Lull, Friedrich von Huene, ships. Incredibly, Jack managed this body of work
and Alfred Romer, and he continues to influence while winning his bread as a theoretical physicist at
young dinosaur paleontologists. Yale, Princeton, and Wesleyan. The conversation
Jack’s two main interests, sauropod dinosaurs below recounts Jack’s formative years and early
and the history of North American paleontology, encounters as a precocious student, his service in the
intersect in the badlands of the western United States Second World War, and paleontology in the badlands
during the last quarter of the nineteenth century, of the western United States. We trace Jack’s interest
when O. C. Marsh and E. D. Cope discovered and in sauropods and delve into the some of the life expe-
described Camarasaurus, Diplodocus, Allosaurus, riences that solidified his position as one of our most
Stegosaurus, and many other dinosaurs. During long celebrated colleagues.
hours at museums studying bones and examining The conversation was recorded and transcribed
field notes, maps, and journals, Jack has recon- from discussions with Jack McIntosh on 3 April 2004
structed the events of many field seasons over (Middletown, Connecticut) and 8 November 2004
decades of dinosaur collecting. His sleuthing has (Denver, Colorado). Material from the two interviews
recovered lost details about the provenance and asso- has been woven together for continuity, and technical
ciations of many dinosaur skeletons and led to recon- terms and colloquial phrasings have been edited to
sideration of many of our old impressions of render them accessible to a broader readership. The
sauropods. The most famous recovered “detail” is recording was transcribed by Carole Goodyear and
Jack’s revision of the skull of Apatosaurus. For the bet- Amanda Kealey.
ter part of a century, the famous Yale mount of an
Apatosaurus skeleton bore a Camarasaurus-like skull KCR: I love to hear the stories about how people get
because of a dotted-in sketch made in 1883 by Marsh. into paleontology. What sparked your interest in
Through examination of the original quarry maps sauropods?
and shipping manifests, Jack discovered a “account- Mcintosh: Well, I think it was the same way as most
ing error” and eventually recapitated Apatosaurus people. I think I was about six. My father took me to
with the correct, Diplodocus-like skull. the Carnegie Museum. I saw the Diplodocus, flipped.
In addition to his historical pursuits, Jack remains From there on, it continued to be part of my life.
one of the foremost experts on sauropod dinosaurs.
Jack was the first to summarize and synthesize skele- JAW: So Jack, you wrote a famous letter at age thirteen
tal, stratigraphic, taphonomic, and taxonomic data on or fourteen to [Richard] Lull. It seems you had already
327
developed a very academic interest in dinosaurs. How “We have found a copy and you can have it for $100.”
did that emerge? Well, of course back in the Depression days, I think I
McIntosh: When I was a kid, I was sick most of the had seen a $20 bill, but the idea of $100 was totally
time. In the sixth grade I missed about one third of out of the question. So I had to write them back
the year. When I was sick my uncle, who had gone sorrowfully saying I couldn’t buy it. But when I got to
to Yale and had taken a course under Lull, had the the Yale Library and looked up von Huene, there was
book Organic Evolution. He gave me that book to a copy of Die fossil Reptil-Ordnung Saurischia, so I
read, and that of course was very exciting. immediately got it down. At the Carnegie Library they
would put me in a room, where they would bring a
JAW: Your letter to Lull is pretty advanced for a 13 or 14 book in. You could sit there and work on it but you
year old. You asked whether Cardiodon and could barely touch it. But here [Yale Library], I said,
Cetiosaurus should be the same thing,” and “What “Can I look at this?” and they said “Oh, you can take
about Apatosaurus?” it out.” Not only could you take it out, but you could
take it out and renew it over and over again. So I took
KCR: Had you already met Lull when you wrote to this thing out, brought it back to my room, and
him, or did you write the letter without knowing him in began copying large parts of it. I wasn’t the least bit
advance? homesick. This was just absolutely wonderful.
McIntosh: When you’re that age, you’re brash, you So the second day I was there, I went up to the
think you can do anything—you’re crazy! Peabody Museum. I went upstairs into the “secret”
tower where Ed Lewis, who was the paleontologist
KCR: Did Lull write back to you? there at the time, had his domain. I went in and
talked to him, and told him that I wanted to work
McIntosh: Well, he wrote me back and said, “I really
on dinosaurs, so he set me to work in the laboratory
don’t know that much about saurischians. Write to
with the big “Atlantosaurus” (Apatosaurus) cervical
Gilmore.”
vertebrae that were all broken. [Othneil C.] Marsh
mounted two of these cervicals that were very good
JAW: Do you still have the copy of Lull’s letter?
ones, but they had been mishandled, so Lewis put
McIntosh: Oh of course, sure. And of course I have me to work getting those things back together. They
another one, which I cherish, that I got many, many were not too badly damaged; there were a number
years later from [Werner] Janensch. Well, it’s in of pieces but you could put them back together. So
German of course. the first thing I did was to get those big cervicals of
“Atlantosaurus” (Apatosaurus) back together.
KCR: Did you get to meet these paleontologists as a Freshman year I was working in one of the din-
young kid? ing halls for a job, but from the second year I had a
McIntosh: I would go in every Saturday when I was regular job putting dinosaur bones back together.
in high school to Pittsburgh, to the Carnegie Library What happened was I was working with a little
and Carnegie Museum, so I got to know all the peo- Coelurus, which was very interesting. I had this
ple there of course. But I’d never even visited New femur and tibia and so forth, was studying them,
York City until my father took me on to college. and I felt somebody’s breath on the back of my
After I graduated from high school in 1941, I neck. It was Barnum Brown. He got all excited. “I
started at Yale. I had never been to Connecticut have an animal,” he said. “It has a nice tibia and
before. So when my Father drove me to Yale, he took fibula and the tibia looks just like yours except it’s
me up there and dumped me and went back home. one and a half times as long but we don’t have the
They thought I was going to be terribly homesick and femur.” He said, “What I would like to ask you to
all that sort of thing. Well, I immediately went to the do is to come down to the American Museum and
Sterling Library and was overwhelmed. I looked up take some measurements of our animal, then we
von Huene because I had been looking for copies of will take a cast of your femur and we’ll expand it.”
the 1932 Die fossil Reptil-Ordnung Saurischia, and they Well, up to then, I had not cut a single class. I was
didn’t have it at the Carnegie Library. In fact I had good about going to classes. But I’ll tell you, the
written to the “Seven Book Hunters,” who were next Monday I was on the train for New York.
advertised in The New York Times, and I made an At the American Museum I rode on the elevator
absolutely silly statement. I said, “Can you possibly to the fifth floor, and I go in to see Rachel Nichols,
get me a copy of Die fossil Reptil-Ordnung Saurischia? who is running the place. She was the secretary; she
I’ll pay any price for it.” Well, that was a mistake knew everything. She sat me down and called
because I got back a letter a month or so later saying, Barnum Brown. He brought in a couple of trays,
328 A C O N V E R S AT I O N W I T H J A C K M C I N T O S H
great big trays, and one of them had an animal that immediately. I mean you were on call. So I was a
he called “Daptosaurus,” but he never did describe it. month into my junior year when I finally got called.
John Ostrom described another specimen of this I finally was sent over to Guam as a flight weather
animal as Deinonychus many years later. The other officer on a B-29. We flew 21 or 22 missions over
tray contained another very small theropod, which Japan. We were in weather planes, so we would fly at
Brown called “Macrodontosaurus.” Years later, high altitudes—32,000 feet, very high altitudes in
Ostrom described this specimen under the name those days—and we would fly over the different tar-
Microvenator since he determined that the big teeth gets and then radio back to our group as to where the
found with the specimen belonged to a different weather was the best and then they would pick them
animal. So Barnum Brown took me and he put me out. These missions would sometimes take 20 hours
in a room with a great big long table. He put this there and back, very long missions. Our group would
box down and said, “Now why don’t you study these fly a couple of hours behind us. We would pick out a
things and you can measure them and draw them, target, and then they would go and bomb that target.
and when you go back to Yale, we can decide what We were flying way up. The Japanese would
to do about this.” When he brought these things in shoot at us, and the stuff would rattle on the plane,
to show me, Rachel Nichols whispered, “He doesn’t but we were never badly hit. We had engines go out
show those to anybody.” So I had the afternoon to a couple of times, and we had to land on Iwo [Jima]
study Deinonychus—brand new from the Cloverly. I on the way back or we would have gone into the
mean this was just absolutely out of this world. And ocean. Iwo [Jima] by that time looked like a huge air-
with Barnum Brown of all people being the one that craft carrier that had been completely macadamized
set me onto this thing. So that was great. over. It didn’t look the way it did when the Marines
were in there. But it was interesting seeing it, and it
JAW: A brush with greatness. Did anything come of it? was very nice to have it there, because as I say, we
McIntosh: I never went up to see Barnum Brown would have been in the ocean. We lost two engines
again. He retired right about that time anyway. on the same side. I’ve forgotten whether this was
So then after two years at Yale I was in the Army due to ack-ack or to engine failure. But anyway, we
Air Corps and I was sent to various colleges to learn did lose two engines. If you lose one on each side,
meteorology and radar. I went to Brown and MIT that’s all right, you can get back. But if you lose two
and Harvard, and I went into the MCZ [Museum of on the same side, that’s not good. And so Iwo [Jima]
Comparative Zoology, Harvard] to see what was in was there and we were able to get back to Guam.
there. Here I was, this little insignificant shave-tail. I About a week after they dropped the atom
went up to the great door, knocked, and introduced bombs, a Twentieth Air Force Colonel decided that
myself to [Alfred S.] Romer. He was just absolutely he wanted to see what was going on and he came
delightful. He invited me in, and he began talking up and commandeered our plane. So we flew over
with me and pulled out these papers by [Chinese both Hiroshima and Nagasaki in the same day, and
paleontologist] C. C. Young that had just come over that was a bit worrisome, too, because you’re a sin-
from Chungking [as it was then spelled]. Chungking, gle plane. But I don’t think they even fired on us.
of course, was the temporary capital of China at that We flew another mission over Japan where we went
time. The Japanese had taken over much of China, out to supply materials for prisoners of war. We were
and Chungking was way back in Sichuan Province. assigned a particular camp to drop our stuff and we
Young had gone there, and he published some of flew over in all directions. We couldn’t find it, and
these early papers on the Lufengosaurus, etc. The we were very sad because this was great, you know, to
paper on which they were printed was of very poor be able to provision these guys that had been POWs for
quality. Somehow Young had gotten copies over to years. We turned around, and as we flew back home we
Romer, and Romer had papers on all of these differ- saw one. I don’t know if it was the one we were sup-
ent prosauropods. He brought them out and showed posed to supply or not, but all these people were down
them to me and said, “Oh, why don’t you borrow there yelling and everything. We flew in at very low alti-
them over the weekend?” That was Romer. He prob- tude and dropped all of this stuff, and they were just
ably had the only copies of those things in the going mad. I mean that was exciting, to say the least.
United States, and he let me borrow them. The last mission over Japan was on V-J Day,
when MacArthur had everybody flying over.
JAW: Were you in the war? However, we wanted to see what was going on. We
McIntosh: Oh, yes, sure. I had just started my junior were supposed to fly formation, but our particular
year. What happened was I volunteered for this group went in over the U.S.S. Missouri at low alti-
business [Army Air Corps], but they didn’t take you tude and really saw what was going on down there.
A C O N V E R S AT I O N W I T H J A C K M C I N T O S H 329
JAW: This is when they were signing the surrender? an avocation, in my spare time. Sometimes I had
McIntosh: Yes. MacArthur was furious. He had us spare time; sometimes I didn’t. But I could do it,
flying training missions for a week after that! There particularly during the seven years that I taught at
were a whole bunch of us that did this kind of Yale. There are now things coming out all the time.
thing. I mean—How many new genera have been
After the war, I was at Tinker Field near described in the last ten years?
Oklahoma City for a while. They were letting people
out and they couldn’t let everybody out at once. And KCR: Have you ever found a good intersection between
so I was just there with absolutely nothing to do your interest in dinosaurs and your interest in physics?
for about a month or so. I got a pass and went McIntosh: No, the things that interest me in paleon-
down to Norman to see Stovall and met Wann tology don’t overlap with my interests in physics.
Langston. Wann showed me all around and we had
an absolutely great time. Of course, there were lots JAW: You had another brush with greatness in your
of sauropod bones there, which Wann was responsi- physics career too—hearing Einstein speak at Princeton.
ble for putting together. Wann had been looking for Did he have a real presence when he showed up?
Die fosil Reptil-Ordnung Saurischia too. I told him McIntosh: Oh wow, when he walked into the room,
that I had access to it, and he was very envious. So you could have heard a pin drop anywhere. He shuf-
when I got back to New Haven, I immediately fled into the room with the crazy old sweater that he
copied all kinds of stuff and sent it to him. A few always wore, and they pulled up a chair at the front
years ago he told me he still has it. of the room, and he sat down to lecture. He talked
I got back about three years and one day after I in a little voice that people in the back row could
went into the Air Corps—I didn’t even go home. easily have heard. I was up near the front row. The
I went directly from where I was discharged at room was so quiet that everybody heard every word
Indiantown Gap, Pennsylvania, out to New Haven, he said. Another time, I was sitting in the theater to
because the next term was starting on about the first hear Linus Pauling, the famous chemist who won
of March. And I signed up for these courses, essen- the Nobel Prize. He was going to give a lecture, and
tially the same ones that I had been taking when I everybody was buzzing; it was a huge noise in the
left: same teachers, in the same rooms, and most of room. And all of a sudden every sound stopped. I
the same students with whom I had been in these mean it was just as if somebody had turned off a
courses three years earlier. The same thing had hap- booming television or something. And I turned
pened to them. They had started, and they had been around and looked and Einstein was walking down
called and so forth. It was incredible. It was as if the aisle.
three years had vanished. I can’t even describe it. Can you imagine? Every time you went anywhere
It was amazing, but anyway, then I got back into to have the whole world stop. It must have been
working in the laboratory. the most embarrassing thing for him. Of course, he
I was majoring in physics, but I wasn’t going into didn’t come out that often. That was the only other
paleontology because you weren’t allowed to “do time that I saw him.
dinosaurs” in those days. You did mammal teeth, and
KCR: It really is hard to imagine that kind of awe. I
I was not interested in mammal teeth at all. But I
wonder if there’s anybody like that in paleontology. [All
could work on these dinosaurs in the laboratory, and
laugh.] What prompted your interest in sauropods?
that was my job. I worked on these things and I
What about them was so exciting?
finished, got my degree, and then went on to get my
Ph.D. in physics and go on to teach in various places: McIntosh: Oh, the Carnegie Museum! The
Princeton, Yale, and finally ended up at Wesleyan. Diplodocus and the Camarasaurus at the Carnegie
Museum were the things that started it off. I’m five
JAW: The years following the war were some lean years years old, and I’m taken into the Carnegie Museum
for vertebrate paleontology. You must have been very by my father. It just never left. It’s just that I wanted
lonely for a while there—the only other person you could to find out as much about sauropods as I possibly
talk to about sauropods was José Bonaparte. Did you could. Most people get over it in a couple of years,
ever think for a minute about abandoning sauropods but I didn’t. See, I’m not really that interested in this
and working on a different group? bird business. The whole thing has been proved to
McIntosh: No. Oh, no. I continued, obviously, to me. I know birds are dinosaurs, and they’re going to
have my interest in sauropods. Joe Gregory made find lots of them, and they’re going to have feathers,
me a research associate at Yale, so I would go down and they’re going to be wonderful. John Ostrom
and work there any time I wanted to. And I did, as tells me that this [Yixian] formation is a mile high,
with layer after layer, where they go in and find one thought was a single animal, and he called it
after another, fossil birds. It’s overwhelming. And it’s Camarasaurus II. A few of his letters have survived,
all little stuff, and you can’t really look at those verte- and it turned out to be two animals, too. The first
brae because [the animal] is so small and has little one of those animals wasn’t exactly articulated, but it
feathers coming out. I just don’t have much interest was partially articulated and it was all together.
in them. But if you collect a sauropod, and you can What those mysterious numbers referred to are box
look at each vertebra, and see each lamina, and so on. numbers. But they [Osborn and Mook] didn’t pub-
I got to the Carnegie Museum as a teenager and lish the box numbers at all because they didn’t real-
got to know “Pop” Kay [J. Leroy Kay], who was the ize they were important. I was able to decode these
curator at the time there. He was very nice and things from the few surviving letters, which listed
showed me all sorts of things. Much later, around each lettered bone and the box it was shipped in.
1960 or so, when I was getting older and when I When [the American Museum’s W. D.] Matthew
was teaching, I wanted to check something. So I went to Philadelphia [to retrieve Cope’s
went down and met Craig Black, and he arranged Camarasaurus bones], he put new numbers on all of
for me to come in and do anything I wanted, these boxes for shipping purposes. So when the
which was extremely nice. And that’s when I people in the American Museum prepared these
started going through the entire sauropod collec- bones, they put Matthew’s box number on them,
tion. I eventually went through all of the dinosaur but they also put Lucas’s box number and the letter
collections. But with the sauropod collection I assigned to each bone on them. These things were
tried to identify everything. numbered in a certain order in the shipping mani-
JAW: And that’s what led to that [1981] Bulletin of the fest, and what I have figured out is that the order of
Carnegie Museum? the boxes corresponds to the way they were shipped.
I now know what was in each box and the order in
McIntosh: Eventually. It didn’t come out until the
which those boxes came. There were 11 shipments
1980s. When I started doing all that cataloging, I
in all, and there might be one or two more of a
didn’t really envision the catalog that came out later.
tooth or something like that, but there are essen-
But when I got through with it, I mean when I was
tially 11 shipments that were sent in by O. W. Lucas.
well along with it, it seemed to be the obvious thing
Then there are about six or eight more that were
to do.
sent in by Ira Lucas. So in the first set of boxes,
JAW: What initially got you interested in working with shipment number 11 was sent in by O. W. Lucas and
the discoveries of Cope and Marsh? Why did you start then the next one, number 12, is the one that’s the
investigating the quarries and maps and all that? key to everything—it tells me how to connect the
Matthew numbers with the Lucas numbers. So I
McIntosh: Of course Marsh and Cope described
now know the order. So it turns out that shipment
many of the original sauropods of North America
numbers 11, 12, and 13 are all one individual. They
and their types are very important to separate out.
started collecting from the tail and they go forward
The Cope bones of Camarasaurus in the American
as you get up to the cervicals and there is no overlap
Museum—as you probably know—have mysterious
at all until you get to the very end. At the very end,
numbers and letters on them. Osborn and Mook
you begin to get a second individual. I have pic-
[who described the Cope collection] had no idea
tures of the quarries that were taken at that time,
what these meant, and they don’t even mention this
and I’m trying to identify exactly which bone is
in their monograph. The trouble was, Cope just didn’t
which. I mean there’s no duplication, and you get
keep records, and he threw all his stuff away. Almost
the right numbers of vertebrae and the right num-
all the letters from his collectors have been lost, or
bers of everything. When you go forward in the first
somebody threw them away after he died, before the
individual of Camarasaurus supremus, there’s a sec-
stuff was taken to the American Museum. But
ond individual that lies beyond that; and unfortu-
[Cope’s collector] Lucas wrote all this stuff out in
nately it’s the front end of that thing that’s begin-
great detail—where these things were found and
ning to come in here, and so I can’t quite separate
how. He drew pictures of a large number of these
them all exactly. I’m pretty sure how it goes.
bones and lettered them, and a few of the drawings
have survived—but only a few of them. JAW: So you’re able to identify individuals?
Cope described Camarasaurus and he had this McIntosh: Exactly. I hope they never do it, but if they
drawing made of the animal, which is rather weird actually wanted to mount a skeleton of Camarasaurus
but good for a starter. He thought it all was one supremus, they can, because I can tell them exactly
individual. It actually is more than one individual, at which individual each bone belongs to. This is the
least two. Lucas found the second animal, which he kind of thing that you can do if you have records.
One of the Carnegie Museum quarries where the diately this was Jim’s site. So I got to Dinosaur
quarry maps may be missing is the one on the Red National Monument, and instead of going up to the
Forks of Powder River, and there are two quarries quarry area to look at things, I went to the telephone
there. We’re not even sure if a map was ever made! booth and called Delta, but I couldn’t get a hold of
The lower quarry, down by the river by the Red Fork, the Joneses [Eddie & Vivian; friends of Jim’s]. But the
is the Diplodocus hayi that’s out in Houston, and telephone operator said, “Oh, yes, I know all about
then there’s another quarry that’s up a bit higher this,” and she told me about the Dry Mesa quarry
where there are huge numbers of sauropod bones, and she told me exactly how to get there, and getting
interesting bones, too. Fortunately Diplodocus hayi is there was no simple problem. There were no num-
a unique individual; there’s no mixture there. We bers on those roads. She told me to go so many
know exactly which bones belong to that, and it miles here and turn on this road and so forth. I hur-
came from a single quarry and then this other stuff. ried right down to Dry Mesa and I got up to this
But this other stuff has all of these bones, and there road that Jim had made to the quarry about a mile or
is no indication at all as to what’s what. I mean there two down. I was in my regular car—it wasn’t four-
are numbers on the bones, but there’s no quarry wheel drive or anything. I decided, I don’t want to try
map to tell you what those numbers correspond to, to go down this thing. So I put my pack on my back
and so that is completely lost. Unless that quarry and walked down. When I got down there, there
map should turn up, and I’m quite sure it won’t now were cars and even trailers from some 24 states that
because Betty Hill [the former secretary at Carnegie had piled up. I mean the word had gotten out. Jim
Museum] and I have gone through absolutely every- came rushing out and handed me a trowel and said,
thing. The trouble is, here you have this huge num- “Dig out that interesting Diplodocus caudal,” and so
ber of bones at the Carnegie Museum that are very that is the kind of thing that happened. Jim was just
interesting, and yet they’re all mixed up. It may have absolutely wonderful. It was just a beehive of activity,
been that even if you had the quarry map, you with everybody around there. It was extremely excit-
wouldn’t be able to do much with it. But it may have ing, and they had already uncovered the big scapula.
been that at least you would know perhaps clusters A couple of years later I went down to see Jim
of bones, you’d know something about it. But that and he decided that I needed to know more about
quarry map is missing and so it’s absolutely too bad the geology of Utah. So he got his truck out and he
that we don’t know anything about that. got Brooks Britt, who was about 16 at the time, and
I have made a collection of quarry maps that I’m we got into the car and he drove us from one end of
keeping—not to work on myself necessarily, but for the San Rafael Swell to the other, showing us every
the future. I mean because these things do get lost, single quarry. We even went to Gilmore’s quarry
and I want copies available, and I don’t want what where he found the Alamosaurus, and Jim showed
happened at the Carnegie Museum to happen again. me everything. I learned more Utah geology that
Gilmore made huge quarry maps at Quarry C at the day than I’ve ever learned [about anything] in that
Sheep Creek, and those things have vanished too. short a period of time in my life.
JAW: Have you spent a lot of time out west ground- KCR: I think one of the big questions that kids have
truthing some of these localities? about dinosaurs is why “Brontosaurus” has a different
McIntosh: Well, up until two or three years ago, I name, and why the skulls were changed and what the
would make a trip out west every summer and I whole situation is with that. Your work on debunking
would go to various museums and so forth. In ‘73 some of those old dinosaur legends is one of the most
and ‘74, I was in the field all summer with Bob familiar examples of how our science works, for people
Bakker, Peter Dodson, and Kay Behrensmeyer. That all over the world. Can you tell us a little about how you
was absolutely marvelous—we went to most of the got started on that, and how you dissected all these prob-
Jurassic dinosaur quarries, particularly the classic lems of sauropod heads and names?
ones, that had ever been found, and I learned an McIntosh: Oh yes, at that time nobody knew what
awful lot there. the skull of an Apatosaurus was. I was working
At that time, I had only known Jim Jensen for a through all of [Earl] Douglass’s records and so forth.
couple of years. I stopped at Dinosaur National I have notebooks full of hundreds of letters like this.
Monument every year, and one year I picked up a I don’t have the letters themselves and I don’t even
couple of hitchhikers on the way. They said that have photocopies of them. But I copied out by hand
they had just heard on the radio on their last ride everything that had to do with dinosaurs in those
that an extraordinary dinosaur quarry had been letters. One of the reasons I did that is because so
found down in western Colorado, and I knew imme- many things have vanished over the years and I
decided that everything that ought to be preserved [of Torvosaurus], and when they knew Jim was coming
should be preserved. And I have an enormous col- Vivian put this claw on the front table in the living
lection that people have allowed me to make copies room, and didn’t say anything. Jim came in and sat
of their records, so they’re now preserved. A num- down on the couch. They talked for about 15 minutes.
ber of quarry maps have been lost or destroyed, and All of a sudden Jim’s eye landed on that claw. . . .
without maps, collections lose an enormous amount “What is that!?” Vivian was just waiting, of course,
of their value, with no chance of determining in a for him to see it. She took him up to the Dry Mesa
multigenera or multi-individual quarry which thing Quarry, and then all bets were off as far as opening
might go with which. the Garden Park quarry. He canceled everything back
in Canyon City, and that was that. But somebody
KCR: You’ve been in the field of sauropod research for a should open that again. There’s still lots more there.
long time, and you’ve seen a lot of students like us enter-
ing and gaining real interest in sauropods, and con- JAW: So what about places. . . . Where would you like to
tributing to our changing view of the group. What do see more sauropods come from?
you think the future holds for sauropods? What ques- McIntosh: Well, of course I want the American
tions are you most excited about finding answers to? Museum to collect that stuff from Mongolia. I’d also
McIntosh: The main thing that I would like to have like to see all that stuff down in South America pre-
happen, which may never happen now, is to have pared and figured properly. You know, some of
them reopen the Marsh Quarry. There definitely is those skeletons hiding around there, and they even
other Diplodocus material in that quarry [Felch have some of the skulls!
Quarry, Garden Park, Colorado]. That quarry is not
finished by any manner or means, and it’s got all JAW: What are you planning next?
kinds of good stuff in it. I probably told you that I McIntosh: Well, I’m 81 years old. I want to get all of
had Jensen ready to open it. The trouble is that the my collections of pictures sorted out, and all the
hill goes up, so you have to dig further and further identifications written on them. I know what all the
into it. When Hatcher and Utterback collected identifications are but some of those notes are very
Haplocanthosaurus, they had to dig back in to get rough: I can read them but nobody else could.
those beautiful skeletons, some of the best stuff they
took out of there aside from the Ceratosaurus. Jensen JAW: Can you hire someone to help you with that?
had already made arrangements to rent a bulldozer to McIntosh: Ah, I don’t want anyone to help me with
get back in there, but on his way from Canyon City it. I want to do it myself. It keeps me busy. Lately I
back to his home in Provo, he drove through Delta, seem to be finding myself falling behind because I
Colorado, where friends of his, the Joneses, Vivian learn an awful lot of things at each SVP! There is
and Eddie, lived. They had found a huge claw just so much to learn about sauropods.
list of contributors
PAUL M . BARRETT ,Department of Palaeontology, J . MICHAEL PARRISH , Department of Biological

The Natural History Museum, Cromwell Road, Sciences, Northern Illinois University, DeKalb, IL
London SW7 5BD, UK 60115
[email protected] [email protected]
MATTHEW T. CARRANO , Department of Paleo- PAUL C . SERENO , Department of Organismal

biology, Smithsonian Institution, P.O. Box 37012, Biology and Anatomy, University of Chicago, 1027
MRC-121, Washington, DC 20013-7012 East 57th Street, Chicago, IL 60637
LUIS M . CHIAPPE , Department of Vertebrate KENT A . STEVENS ,Department of Computer and

Paleontology, Natural History Museum of Los Information Science, 301 Deschutes Hall,
Angeles County, 900 Exposition Boulevard, Los University of Oregon, Eugene, OR 97403
Angeles, CA 90007 [email protected]
[email protected]
PAUL UPCHURCH , Department of Earth Sciences,
RODOLFO A . CORIA , Museo Municipal Carmen University College London, Gower Street, London
Funes, Avenida Córdoba 55, Plaza Huincul 8318, WCIE 6BT, UK
Neuquén 8300, Argentina [email protected]
[email protected]
MATHEW J . WEDEL , University of California
KRISTINA CURRY ROGERS , Paleontology Depart- Museum of Paleontology and Department of
ment, Science Museum of Minnesota, 120 West Integrative Biology, 1101 Valley Life Sciences
Kellogg Boulevard, St. Paul, MN 55102 Building, Berkeley, CA 94720-4780
LOWELL DINGUS , Infoquest Foundation, 160 Cabrini JEFFREY A . WILSON , Museum of Paleontology and
Boulevard, No. 48, New York, NY 10033 Department of Geological Sciences, University of
[email protected] Michigan, 1109 Geddes Road, Ann Arbor, MI
48109-1079
GREGORY M . ERICKSON , Department of Biological [email protected]
Sciences, Florida State University, Conradi
Building, Tallahassee, FL 32306-1100 JOANNA L . WRIGHT, Department of Geology,
[email protected] University of Colorado at Denver, Campus Box
172, P.O. Box 173364, Denver, CO 80217-3364
FRANKIE JACKSON , Department of Earth Sciences, [email protected]
Montana State University, Bozeman, MT 59717
[email protected]
335
INDEX
Aalenian, 105, 115–18, 120, 121, 122, 130–32 angular, 59–60, 158–59, 160, 295
abelisaurids, 287 ankle, 21, 41
abundance, 137, 143, 146, 196 ankylosaurs, 242, 245
acetabulum, 180, 192 Antarctica, 10
adaptation, 144, 148, 197, 201, 237 Antarctosaurus: A. giganteus, 39; phylogeny of, 53, 55,
adaptive zone, 126 68–69, 71–72, 234; A. septentrionalis, 65;
adductor fossa, 21, 36–37, 158 A. wichmannianus, 170
adipose tissue, 207, 208, 219, 223 Antenonitrus, 2, 24, 27, 29, 31, 237
Aegyptosaurus, 53, 68–69, 72, 73, 234, 246 Apatosaurus, 17, 24, 115, 243; bones of, 180–81, 194, 242,
Aeolosaurus, 53, 68–69, 72 244, 307, 315–21; A. excelsus, 17, 240, 242, 317;
Africa, 5, 10–11, 15–16, 274; dicraeosaurids in, 35, 172; feeding mechanism of, 128; feet of, 29, 31, 33, 240–41;
Jurassic of, 35, 142, 145, 276; prosauropods in, 129, growth of, 10, 315; limbs of, 24; A. louisae, 186, 190,
140, 146; rebbachisaurids in, 143, 157; titanosaurs 240–41, 244; neck of, 185–86, 195; phylogeny of, 234;
in, 20, 38–39, 143. See also names of specific countries and pneumaticity, 204, 205, 211, 214; skull of, 61, 327;
age and growth, 311, 313–14, 320; adult, 304, 312, specializations of, 37, 38; and tracks, 254–55, 260
316–17, 319, 321–22; juvenile, 312, 316; subadult, apomorphies, 163
316–17, 321. See also longevity appendicular characters, 36
air sacs: cervical, 216; pulmonary, 203, 217–20, 223 Aptian: and diversity, 105, 115–16, 119, 130, 135, 143;
air volume in bones, 212–15, 217–21, 223 feeding mechanisms in, 132–33, 134; titanosaurs in,
Alamosaurus: locomotion of, 240, 246; phylogeny of, 66–67, 136
53–55, 56, 68–71, 72, 115, 296; A. sanjuanensis, 240, archosaurs (Archosauria), 167, 203, 208, 209, 239, 314
246, 297 Argentina, 5, 119, 139, 287; eggs from, 10, 50, 285–86,
Albian: diplodocoids in, 136–37, 143; and diversity, 105, 289; titanosaurs in, 51, 66–67, 296; tracks in, 265,
112, 115–16, 119–20, 121, 130; feeding mechanisms 269
in, 132–33, 134; teeth in, 135; titanosaurs in, 66, 136 Argentinosaurus, 212, 244; growth of, 315, 323;
alligators, 315 phylogeny of, 53, 68–69, 71–72, 73; size of, 303–4
allometry, 312, 320 Argyrosaurus, 53, 68–69, 72, 234
Allosaurus, 8 Arizona, 259
alveolar trough, 158, 159, 161–63, 165–66, 167–68 Arkansas, 269
alveolus, 164 Asia, 5, 18, 142, 272, 276; titanosaurs in, 19, 39. See also
Amargasaurus, 37, 115, 172, 234, 243 names of specific countries
Ampelosaurus, 53, 68–71, 72, 234, 245 ASP (air-space proportion), 212–15, 220, 223
Amphicoelias, 234 astralagus, 21, 33, 41
Amprino’s rule, 313, 317–18, 321 Atlantosauridae, 16, 51
Andesauridae, 51, 54, 71 Atlasaurus, 115, 117
Andesaurus: A. delgadoi, 38, 52; phylogeny of, 51, 56, 67, Auca Mahuevo eggs. See eggs, Auca Mahuevo
71–72, 73; in trees and cladograms, 53–55, 68–69, Aucasaurus garridoi, 287
234, 296 Augustinia, 53, 68–69, 72
angiosperms, 147, 196 Australia, 10, 268
337
Austrosaurus, 53, 115 312, 317; whole-body, 310–12, 313–15, 321, 323. See
autapomorphies, 37, 60, 108, 214 also growth curves
bones, 174, 273, 300; apneumatic, 204, 205, 219;
Bajocian, 105, 115–16, 120, 122, 130–32, 135 cancellous, 204, 218–19; cortical, 307, 309–10, 311,
Barapasaurus, 2, 111, 115, 139, 234; feet of, 31–32; and 312, 315, 318–19; cut sections of, 222, 311, 316;
pneumaticity, 202, 205–6, 207, 209, 211; embryonic, 285, 294; growth marks in, 307, 311,
quadrupedalism of, 237–39 314, 321; internal complexity of, 210–12, 214–15;
Barosaurus, 8, 115, 172, 234, 256; B. africanus, 213, 246; limb, 217, 230; marrow-filled, 212, 222; morphology
bones of, 30, 213–14, 312, 315–16 of, 304, 307; remodeling of, 306, 310–11, 312, 314,
Barremian: diplodocoids in, 136, 143; and diversity, 105, 316; texture of, 207, 209; with thin outer walls,
115–16, 119, 120, 130, 136; feeding mechanisms in, 204, 209; volume of air in, 212–15, 217–21, 223. See
132–33, 134; teeth in, 135, 170, 174 also histology, bone; osteology; pneumaticity; recon-
Bathonian, 105, 115–16, 117–18, 130, 132, 134–36 structions; skeletons; skulls; vertebrae; names of spe-
behavior, 323. See also feeding; locomotion; rearing; cific bones; under names of specific genera
reproduction bone tissue: fibrolamellar, 309–10, 311–12, 313, 315,
Bellusaurus, 115, 117 316–20, 321; Haversian, 306, 312; lamellar, 305–6,
Berriasian: diplodocoids in, 136, 143; and diversity, 105, 307, 309; parallel-fibered, 305–6, 307, 312; secondary,
115–16, 118–21, 130, 132, 135–36 311; woven, 305–6, 307; zonal, 307, 309, 313, 317
bioenergetics, 194 bone vascularization: degree of, 309, 311–12, 313, 315,
biology, 109, 310, 312. See also paleobiology 317–18, 321; laminar, 306–9, 312, 313, 315, 316–17,
biomass, 125, 140, 197. See also body size/mass 318–20; plexiform, 306–8, 312, 316–17, 318; reticular,
biomechanics, 215, 303–4, 310, 316 306–9, 312, 318–19; variable, 310, 319
biorhythms, 305, 310, 312 “Bothriospondylus,” 234, 315
biotic mixing, 121, 137 Bothrosauropodidae, 16
bipedalism, 23, 231–33, 237, 241, 245–46, 263 brachiosaurids (Brachiosauridae), 212; feeding mecha-
birds, 306, 330–31; altricial, 321–22; growth of, 307, 312, nism of, 128, 194, 196; locomotion of, 244–45;
313, 315, 317, 321–22; necks of, 182, 189–90, 191–92. phylogeny of, 32, 38, 54
See also names of specific birds Brachiosaurus, 28, 115, 316; B. altithorax, 246; bones of,
birds and pneumaticity, 203–4, 209–10, 212–13, 194, 315; B. brancai, 38, 181, 182, 184–85, 186–89,
221–22, 223; adipose tissue in, 208, 219; air sacs in, 238, 240–42, 244, 246; feeding of, 196, 197;
203, 220; diverticula of, 206, 216, 223 locomotion of, 238, 240–42, 244, 246; phylogeny
bison, 179 of, 18, 38–39, 52, 68–71, 234, 296; and pneumatic-
Blikanasaurus, 2, 27, 30–31, 115, 117, 140; limbs of, 23, ity, 205–6, 207, 211, 214, 215; and tracks, 254–55
24, 29, 237 braincase, 297–98
body plan, 1–2, 3, 20, 178 brain size, 7
body size/mass, 6–7, 11, 144, 146, 174, 299; adult, 304, Brazil, 66, 168, 170, 268–69
312, 316–17, 319, 321–22; and bone growth rates, Brazil Series B, 68–72
313–15; decrease in, 233, 235–36, 246; differences in, Breviparopus, 256, 260, 264–65, 266, 274;
126, 138; evolution of, 230–36; increase in, 23, 173, B. taghbloutensis, 257, 264
212, 231, 233, 235–37; and pneumaticity, 201, 212, Brontopodus, 256, 265–66, 273–75; B. birdi, 257, 258,
216–21; similarities in, 140, 144; and specializations, 260, 265
23, 41–42; subadult, 312, 316–17; of titanosaurs, 39, brontosaurs, 266
233, 244, 245, 303, 323 Brontosaurus, 16
Bolivia, 269 Brown, Barnum, 328–29
bone growth, 303–23; annual, 310, 318; and annuli, 307, browsing, 146, 194; by dorsiflexion (BD), 178, 179, 197;
310, 317; cyclicity/periodicity of, 307–10, 312, 313, high, 128, 187, 196, 245; low, 127–28, 145, 179, 186,
317, 320; and medullar expansion, 310, 311, 314; 188–89, 196–97, 245; maximum height of, 144, 195;
noncyclic, 313, 317; and polish lines, 312, 316; quali- medium-level, 127–28, 197; neutral (BN), 178–83,
tative, 304, 312, 323; stages of, 313–15, 319, 321–23. 197; by ventriflexion (BV), 178–79, 195, 197. See also
See also LAG feeding height
bone growth rates, 303–23; and Amprino’s rule, 313, buoyancy, 221
317, 321; of Apatosaurus, 316–23; average, 309, 313;
maximum, 321–22; plateaus in, 309, 317; quantifica- caimans, 219
tion of, 304–5, 312–13, 318–21; rapid, 305, 308, 311, Callovian, 139; diplodocoids in, 134, 136; and diversity,
317–19; remodeling, 310; slow, 306, 307, 309–10, 105, 115–16, 117, 130, 132, 135–36
338 INDEX
camarasaurids (Camarasauridae), 32, 138, 139, 296; classification, 5, 16–18. See also ingroups; nodes;
feeding of, 194, 197 outgroups; synonyms; trees, phylogenetic; taxonomy
camarasauromorphs (Camarasauromorpha), 254–56, claws: manual, 255–58, 260, 263, 274; pedal, 255–58,
258, 260, 274 262–65, 273–74
Camarasaurus, 16, 115, 169; bones of, 159, 181, 244, climate, 286–87. See also paleoclimate
246, 331; feeding of, 128, 196, 197; feet of, 30–32, 33, cnemial crest, 230, 239, 244
240; C. grandis, 26, 240, 246; C. lentus, 160; limb Coelophysis, 216
bones of, 24, 26; phylogeny of, 38, 65, 68–71, 139, 234; Coelurus, 328
and pneumaticity, 202, 204, 205; C. supremus, 17, 331; coevolution, 10, 147, 197
and tracks, 254–55; vertebrae of, 37–38, 186, 189, 211, collagen, 305–6
214 Colorado, 253, 262, 264, 268, 332
camels, 181, 182, 189–92 competition, 140–43, 148
Campanian, 286; and diversity, 105, 115–17, 120, 130, competitive exclusion, 140
132–33, 135–36; feeding mechanisms in, 133, 134; competitive replacement, 138–39, 142, 143
titanosaurs in, 66–67, 131, 136–37 condyle, 40, 41; and feeding, 181–84, 187, 193; and
Canada, 144 pneumaticity, 210–11, 214
Carnegie Museum, 327, 330–31, 332 Coniacian, 66, 105, 115–16, 119–21, 132
Carnian, 2, 23, 139; and diversity, 105, 115–16, 117, conifers, 146, 196, 197
130–32; prosauropods in, 129–30 convergence, 143
carnivores, 125–26 Coombs, W. C., 229
carpal, 21, 27, 33 coracoid, 22, 40, 64
carpus, 22, 41, 237, 245 coronoid process, 158, 173, 174–75
cartilage, 26, 180, 245, 311 cotyles, 181–84, 211, 214
CCRs (candidate competitive replacements), 138–39, cows, 315, 319
141, 144–45, 147 cranial remains, 19, 27, 36; embryonic, 285, 295–98.
Cedarosaurus, 115, 174 See also skulls
Cenomanian, 286; and diversity, 105, 115–16, 119, 130, Cretaceous, 5, 7, 11, 32, 38, 51; diplodocoids in, 171;
135–36; feeding mechanism diversity in, 132–33; diversity in, 105, 115–16, 157; plants in, 195–96;
titanosaurs in, 66–67, 131, 136, 143 radiation in, 104; tracks in, 260–61, 272. See also
ceratopsians (Ceratopsia), 231, 233, 237, 276 names of specific periods
ceratopsids, 168, 173, 174 Cretaceous, Early/Lower, 18, 106, 119, 145, 174;
Ceratosaurus, 219, 333 dicraeosaurids in, 35, 172; diplodocoids in, 137,
cetaceans, 309 142–43; diversity in, 105, 119, 136, 138, 147; feeding
cetiosaurids (Cetiosauridae), 118, 127, 138, 139 mechanism diversity in, 133, 134, 142;
Cetiosauriscus, 234 rebbachisaurids in, 10, 157; titanosaurs in, 10, 137,
cetiosaurs, 237 142–43; tracks in, 261, 268–73
Cetiosaurus: bones of, 16, 26; C. mogrebiensis, 246; and Cretaceous, Late/Upper, 19, 58, 109, 145, 287;
pneumaticity, 205, 216; range of, 113, 115; taxonomy diplodocoids in, 143; diversity in, 105–6, 133, 134;
of, 15, 112, 139, 234 eggs in, 285, 289, 294, 300; feeding mechanisms in,
chambers, vertebral, 201, 204–5, 209, 211, 214, 216 134; NOOs for, 119–20; T. madagascariensis in, 56, 57,
characters, 36, 52, 54, 65, 114, 304 59; teeth in, 170, 171, 173; titanosaurs in, 50, 137, 147,
carcharodontosaurids, 288 261, 296, 299; tracks in, 261, 266, 269–71, 273
cheek embayment, 172 Croatia, 269
chickens, 203 Crocodilians/crocodylians, 15, 39, 182, 208–9, 220
Chile, 267–68 crown: broad (BC), 18, 21, 42, 127–29, 134–35, 141–42;
China: Jurassic of, 5, 18, 19, 145; Lower Lufeng embryonic, 295–96; narrow (NC), 16, 18, 42,
Formation of, 140, 141, 144; prosauropods in, 129, 128–29, 134–35, 141–42, 170–72; of Nigersaurus
140; tracks in, 257, 269, 272 taqueti, 159, 162–63, 167, 168–69; shape of, 21,
chirotheres, 263 173–74. See also wear facets
Chubutisaurus, 234 CT (computed tomography), 159, 162, 206, 213, 222
clades, 23, 65, 67–68, 70–71, 141–42 cycadophytes, 146, 196, 197
cladistic analyses, 104, 108–9
cladistic hypotheses, 18–20 data: on diversity, 113–14, 116; missing, 36, 72;
cladograms, 70, 106, 107, 113–15, 118, 128; podial problematic, 116, 139; qualitative, 317; quantitative, 321;
consensus, 253–54; simplified, 296 skeletochronologic, 313, 314, 317. See also sampling
INDEX 339
dating, radiometric, 111, 129 stratigraphic, 41–42, 144, 146; temporal, 252; of
Datousaurus, 234 tracks and body fossils, 272–73
Deinonychus, 5 divergence, 27, 117, 172
dentary, 21, 59–60, 160, 172, 173; embryonic, 295–96; of diversity, 9, 16, 104–24; absolute, 116, 134; and absolute
Nigersaurus taqueti, 158, 162, 163, 165–67, 168, 174–75 age, 111–12; of amniotes, 125; and artifacts, 117–20;
dentine, 166, 169, 170, 174 comparison of results on, 116–17; in Cretaceous,
Depéret, Charles, 56–58 115–16, 157; data on, 113–14, 116; decrease in, 109,
DGM Serie A, 212 116–17, 118, 119, 120–21, 134, 138, 142; distortion in
diaphyses, 57, 212, 311, 314, 317 estimation of, 111, 113; in Early Cretaceous, 105, 119,
dicraeosaurids (Dicraeosauridae), 133, 134, 158, 171; 136, 138, 147; in Early Jurassic, 117, 129; and feeding
feeding mechanism of, 128, 174; phylogeny of, 32, mechanisms, 132, 137–38, 148; increase in, 104, 105,
35–36; teeth of, 36–37, 159, 167, 172 110, 116–18, 120–21, 134; in Jurassic, 115–17; in Late
Dicraeosaurus, 115, 170, 172, 243, 315; D. hansemanni, 35, Cretaceous, 105–6, 133; in Late Jurassic, 104, 105,
186–87, 241, 246; locomotion of, 241, 246; neck of, 118, 121, 147; of locomotor morphology, 230, 247;
185–86, 187; phylogeny of, 69, 234; specializations low, 117, 143, 145; in Middle Jurassic, 105, 107,
of, 37, 38 116–18, 147; of ornithischians, 144; overestimation
diet, 138, 195. See also food; plants of, 112, 113, 117; patterns of, 117–20, 122; peak, 146;
digitigrady, 30, 34, 237–38, 239, 240, 258 of plants, 147; and preservation biases, 109–11;
digit: manual, 255–58, 265, 274–75; pedal, 255–56, 258, previous work on, 105–7; processes affecting,
262–63 120–21, 122; of prosauropods, 140; relative, 134, 140,
DinoMorph, 185, 188, 190, 194 142; results on, 114–20; and sampling biases, 109–10;
Dinosauria, The (Weishampel et al.), 2, 3 of sauropodomorphs, 125–48; and stratigraphic
Dinosauriforms, 242 ranges, 111–12; taxic approach to, 105; and taxonomic
Dinosauromorpha, 114 level, 112–13; and TDEs vs. PDEs, 107–9, 118, 119, 141,
dinosaurs (Dinosauria), 3, 4, 114, 243; body size of, 230–33; 145; of titanosaurs, 19; underestimation of, 116. See
growth of, 309, 315, 322; locomotion of, 233, 243 also NOOs; PDEs; radiation; tabefaction; TDEs
diphyly, 141 diversity curves, 109, 111, 114–16, 127
diplodocids (Diplodocidae), 134, 297; compared to diverticula, 203; abdominal, 205; extraskeletal, 217–20,
Nigersaurus taqueti, 159, 167, 169–72; and feeding, 223; intraosseous, 204, 220; vertebral, 202, 216, 221
174, 185–86, 194, 196–97; phylogeny of, 32, 35, 296; DME (developmental mass extrapolation), 314–15, 321, 323
and pneumaticity, 205, 216; and tracks, 254–55, Dodson, P., 230
264–65, 274; vertebrae of, 28, 192, 211. See also down-weighting, 111–12
names of specific genera dromaeosaurids, 288
diplodocines, 212, 223 duck, 309, 311, 318–19
diplodocoids (Diplodocoidea), 32, 119, 133; evolution of,
34–38, 243–44; feeding mechanisms of, 21, 134; edaphic processes, 291
locomotion of, 238–39, 242; phylogeny of, 5, 18, 19, egg-beds, 285–87, 289–93, 298. See also nests
32; radiation of, 134–35; size of, 233, 235–36, 243, eggs, Auca Mahuevo: abnormal, 285, 292; age of,
246; specializations of, 8, 21–22, 36–38; TDEs of, 285–86, 291; clutches of, 273, 285, 286, 289–93,
136–37; teeth of, 134, 171–72, 296; and titanosaurs, 298, 299, 300; laying of, 298, 299; megaloolithid,
55, 56, 67, 68, 70, 73, 142–43; and tracks, 258, 261; 299–300; morphology of, 285, 288, 292, 298–300;
vertebrae in, 40–41, 51. See also dicraeosaurids; number of, 289, 291, 299; shells of, 285, 288–89,
diplodocids; rebbachisaurids 292, 294, 298–99; titanosaur, 50, 297. See also
Diplodocus, 7, 115, 128, 243; bones of, 37, 159, 180–81, embryos
194, 304, 332; D. carnegii, 186, 193; head of, 28, 61, Egypt, 66–67
160; D. longus, 1, 2, 32, 34–35, 38–39, 160; neck of, Einstein, Albert, 330
185–86, 190, 195; phylogeny of, 69, 234; and pneu- elbow, 181, 245–46
maticity, 202, 204, 205, 206, 215, 217–21; teeth of, Elephantopoides barkharusensis, 266
18, 166, 172, 196; and tracks, 254, 260 elephants, 229
discoveries, 5–6, 7, 73, 106, 111, 266. See also fossils; embryos, 10, 19, 50, 294–300; development of, 285,
trackways/ichnofossils 295, 297–98; titanosaur, 292, 294. See also eggs
distribution: of clutches, 285; and gauge, 261; geographic, emu, 309, 311, 318–19
104, 110, 144; global, 252, 270; paleoenvironmental, enamel, 21, 170–71, 172, 173–74, 295; of Nigersaurus
271–72; paleolatitudinal, 271; of plants, 147, 196; of taqueti, 159, 162, 166–67, 168, 169
pneumaticity, 215–16, 223; spatiotemporal, 266–71; England, 15, 106, 112, 168
340 INDEX
environment, 287, 310, 312. See also paleoenvironment feet. See forefeet; manus; metacarpals; metatarsals; pes;
Eoraptor, 24, 31–32, 33, 37 pes and tracks; phalanges; tarsals; trackways/
Epachthosaurus, 52–55, 68–69, 71–72, 73, 204 ichnofossils
epithelium, 213 females, 298, 299, 312
erosional surfaces, 294–95 femur: and body size, 231–33, 235–36; and classification,
Erythrosuchus, 206–7 23–24, 26; eccentricity of, 21–22, 230, 231, 243–44;
Euhelopodidae, 18–19 and growth, 313, 314; and locomotion, 22, 40, 242–45;
Euhelopus, 65, 115; locomotion of, 244, 246; neck of, of titanosaurs, 38–39, 61, 63–64, 242, 256, 260
186, 187–89; phylogeny of, 18, 38, 55, 67, 68–71, fenestra, 59–60, 160; embryonic, 295–96, 297; of
139, 234; and pneumaticity, 205, 212; E. zdanskyi, Nigersaurus taqueti, 158–59, 163, 164, 169, 175
187–89, 246 ferns, 146, 196–97
euornithopods, 9 fibrolamellar complex, 306–7
Europe, 18, 111, 129, 142, 145, 146; rebbachisaurids in, fibula, 41, 64
36, 157, 172; titanosaurs in, 20, 38, 51, 299; tracks fish, 322
in, 266. See also names of specific countries Flagellicaudata, 22, 35, 37
eusauropods (Eusauropoda), 3, 19; basal, 118, 258; flooding, 293
evolution of, 27–31, 297; feeding mechanism of, 21, fodder. See plants
127–28; and pneumaticity, 223; posture of, 21; food, 143; processing of, 126, 144, 174, 194, 196. See
synamorphies of, 296; teeth of, 36–37; and tracks, also diet; mastication; plants
254–56, 258, 262–64, 274. See also names of specific foot morphology, 253–58, 262, 264
genera footprints. See trackways/ichnofossils
eutherians, 322 foramina, 60, 160; of Nigersaurus taqueti, 158, 159, 161,
Eutitanosauria, 55 162–63, 164–66, 167–68; replacement, 162–63,
evolution, 8, 20–42, 118; of body size, 230–36; cra- 164–66, 167–68, 174; vertebral, 201–2, 203–5, 207,
nial, 121, 297; of Diplodocoidea, 35–38, 243–44; of 209–10, 215, 219
Eusauropoda, 27–31, 297; lower-level patterns in, forearm, 237
243–46; macro-, 137, 138, 236; of Macronaria, forefeet, 21, 33–34, 245. See also manus; metacarpals
38–41; of Neosauropoda, 32–34, 243; of plants, forelimbs: and feeding, 180, 181; length of, 21, 179, 188,
195; of posture, 23–27; of quadrupedalism, 230, 231, 237, 243–46; and locomotion, 239, 245; of
236–39; and synapomorphies, 20–22; theropod, Nigersaurus taqueti, 61, 64, 158; and posture, 23–25,
216; of tooth batteries, 171–75; of vertebrae, 211–12, 40–41; of titanosaurs, 245–46
215. See also locomotion, evolution of; coevolution; forests, 195–96, 276
specializations fossa, vertebral, 201–2, 204, 211, 215–16, 218; blind,
extinction, 110, 111; of broad-crowned sauropods, 134; of 207, 209, 215; pneumatic vs. nonpneumatic, 205–9;
cetiosauridae, 139; of dicraeosaurids, 133; of subfossae of, 206, 209, 215
diplodocids, 142; at Jurassic-Cretaceous boundary, fossil-bearing formations and sites, 109; Allen, 288;
105, 106, 145; mass, 109, 118–19, 120; of Anacleto, 286, 288–89, 293; Ardley Quarry, 256–58;
Nemegtosauridae, 143; of prosauropods, 129, 140, Auca Mahuevo, 285–300; Barreales Escondido,
144; of rebbachisaurids, 143; regional, 146; and sea 285–86, 287, 294, 298; Barreales Norte, 285–86,
level, 121 287, 294, 298; Clarens, 144; Dry Mesa Quarry, 332,
333; Fatima, 256–57; Gadoufaoua beds, 167; Glen
Farlow, J. O, 266 Rose, 252, 275–76; Kayenta, 144; Lower Elliot, 140;
feeding, 41, 121; head-down, 187, 197; underwater, 190, Lower Lufeng, 140, 141, 144; Maevarano, 59;
197. See also herbivory; jaw; teeth; vegetation Marsh Quarry, 333; McCoy Brook, 144; Morrison,
feeding height, 180, 185, 192, 194–95. See also browsing 142, 195–96, 259, 261, 264; number of, 116, 119,
feeding mechanisms, 142, 144; Apatosaurus type, 128, 121, 267–69; Paluxy, 260, 275; Portland, 144;
132–33; brachiosaurid type, 128, 132–33, 134; prosauropod, 129; Purbeck Limestone, 275;
Camarasaurus type, 128, 132–33; Cetiosaurid type, Purgatoire, 252–53, 258, 260, 264–65, 266, 275;
127, 132–33, 134; dicraeosaurid type, 128, 132–33, 134; Summerville, 259; Tendaguru Beds, 142, 196;
diplodocoid, 171; Diplodocus type, 128, 132–33, 134; Upper Elliot, 144
and diversity, 137–38, 148; higher eusauropod type, Fossil Reptil-Ordnung Saurischia, Die (Heune), 328
127–28, 132–33, 134; nemegtosaurid type, 128, 132, fossils, 2–4, 195, 287; distribution of, 272–73; earliest,
133; Nigersaurus type, 128, 132, 134; prosauropod, 117; gaps in, 107–8; and growth, 303–4; quality of,
126, 141; sauropod, 127–29; Shunosaurus type, 127, 5, 10, 105, 110–11, 118, 121. See also preservation;
132–33, 134; titanosaur type, 128, 131–33, 134 trackways/ichnofossils
INDEX 341
France, 66, 268, 294 histology, bone, 304–19, 321, 323; of Apatosaurus,
frogs, 203 316–19; and growth, 307–15; qualitative, 315; and
frontal, 59–60, 295–96, 297 tissue organization, 305–6; variability in, 312; and
vasculature, 306–7
gastric mill, 126 holotypes, 158
gastrolith, 174 Homalosauropodidae, 16
gazelles, 179 homology, 207–8
genera, 53; and diversity estimation, 112–13, 114, 116, homoplasy, 143
127, 130; ichno-, 261–66, 274; number of, 125, horses, 179, 181, 182, 190
135–36. See also names of specific genera human factors, 109
geological age, absolute, 111–12, 114, 116 humans, 309, 315
geological factors, 109, 120 humeri, 203, 315; and locomotion, 22, 244–46; and
Germany, 16, 265–66, 268, 274 posture, 26, 40–41, 181; size of, 232–33, 237; of
ghost lineages, 2, 116–17, 133, 137 titanosaurs, 56–57, 63, 64, 245–46
ginkgophytes, 196 hydroxyapatite crystals, 305–6
giraffes, necks of, 179, 181, 183–84, 186, 193; flexibility Hypselosaurus, 299
of, 189, 191
glenoid, 180, 192 ichnofossils. See trackways/ichnofossils
Gondwana, 146 ichnogenera, 261–66, 274
Gondwanatitan, 53, 115, 246 ichnology, predictive, 252–58
Gongxianosaurus, 141; feet of, 27, 29, 30–31; phylogeny ichthyosaurs, 6, 309, 318
of, 23, 234; G. shibeiensis, 24 iguanodontids, 275
grazing, 179, 196 ilium, 63, 64, 204–5; and locomotion, 41, 239, 241–42, 245
Greenland, 267, 274 India, 5, 10, 111, 139; eggs in, 289, 299; titanosaurs in,
gregariousness, 298 20, 38, 51, 56–57, 65, 66
growth. See bone growth; bone growth rates ingroup, 52, 65, 68, 114
growth curves, 305, 314–15, 317, 322, 323 innovation, key, 126, 141
gymnosperms, 196, 197 integument, 55, 285, 294. See also skin
Isanosaurus, 117, 140, 237; bones of, 26, 209; phylogeny
habitats, 106, 138, 140, 146, 189; preferences of, 266, of, 23, 234
271–72 ischium, 63, 64, 244–45
hadrosaurids/hadrosaurs, 168, 173, 275, 276, 300 Isisaurus: colberti, 65, 69; phylogeny of, 20, 53, 56, 68,
Haplocanthosaurus, 35, 55, 115, 139, 234, 333; and 70–72, 234. See also ‘Titanosaurus,’ colberti
pneumaticity, 207–9, 215 Isle of Wight, 56, 170, 174
hatchlings, 303–4 Italy, 34–35, 257, 263, 264, 267–68, 274
Hauterivian, 105, 115–16, 130, 132, 133, 135–36
head height, 178–79, 180, 183, 188, 194; low, 188, 195 Jabalpur Titanosaur indet., 65, 68–69, 72
herbivores, extant, 179, 183, 195 Jainosaurus, 53, 67, 68–69, 72; J. septentrionalis, 65
herbivory, 146; facultative, 125, 140; obligate, 125, 126, Janensch, W., 16, 18
140; and ornithischians, 157, 172; of Janenschia, 72, 137, 316; bones of, 310, 315; locomotion
sauropodomorphs, 125–26; and specializations, 21, of, 238, 240, 244, 246; phylogeny of, 38, 53, 68–69,
27–28, 36–37, 42. See also browsing 234; J. robusta, 238, 240, 246, 310
herds, 10, 276 Japan, 329
Herrerasaurus, 28; bones of, 24, 26, 31–32, 33, 37, jaw, 21, 36–37, 126, 127, 173; of Nigersaurus taqueti, 158,
209 159, 163, 167, 170, 172
heterodontosaurids, 144 Jensen, Jim, 332, 333
heteropody, 265 Jianu, C.-M., 107, 113, 116–17
Hettangian, 138; and diversity, 105, 111, 115–16, 118, Jingshanosaurus, 24, 31–32, 38
130–32, 135; prosauropods in, 129–30 Jobaria, 115, 117; bones of, 21, 24, 32–34, 35, 169;
hindfeet, 21 J. tiguidensis, 240; vertebrae of, 209, 215, 216, 221
hindlimbs, 27, 29–30, 158; and locomotion, 40, 195, joints, 23–24, 34, 181, 246
239, 245, 276; and posture, 23–25; size of, 179, 180, jugal, 59–60, 295–96, 297
231, 243–44, 245; of titanosaurs, 61, 64, 245 Jurassic, 4; diversity in, 105, 115–17; NOOs for, 120;
hips, 239, 263, 273 plants in, 195–96; teeth in, 173; tracks in, 261. See
histograms, 130–33, 135–36, 271, 275 also names of specific periods
342 INDEX
Jurassic, Early/Lower, 2, 32, 125, 139, 276, 287; and quadrupedalism, 236–39, 246; and specializa-
competition with prosauropods in, 140–41; diversity tions, 230, 239–43, 247
in, 117, 129, 133; feeding mechanisms in, 133, 134; longevity, 305, 310, 311, 321, 323
prosauropods in, 129, 143–44, 146; tracks in, 34, Losillasaurus, 143
264, 267, 270–71, 273–74 Lourinhasaurus, 115
Jurassic, Late/Upper, 11, 19, 32, 35, 139, 145; Lufengosaurus, 24–25, 31–32, 33, 37, 139
abundance in, 106; competition in, 142; Lull, Richard, 327–28
diplodocids in, 36, 172; diversity in, 104, 105, 118, lung, 203, 216, 217–20, 221, 223. See also air sacs,
121, 133, 147; feeding mechanisms in, 133, 134, 148; pulmonary
neosauropods in, 39, 233; radiation in, 137–38, 233; lycopods, 196–97
sympatry in, 194; titanosauriforms in, 260–61;
titanosaurs in, 50, 137; tracks in, 258–59, 261, Maastrichtian, 270; and diversity, 105, 115–17, 120, 130,
266–68, 270–74 132–33, 135–36; feeding mechanisms in, 132–33, 134;
Jurassic, Middle, 2, 27, 104, 112, 139, 145; in China, 5, titanosaurs in, 66–67, 136–37
18; diversity in, 105, 107, 116–18, 147; feeding macronarians (Macronaria), 174; basal, 33, 134, 139,
mechanisms in, 134, 142, 148; neosauropods in, 236, 243, 244–45; derived, 233, 236; evolution of,
10, 33; radiation in, 122; and titanosaurs, 38–39, 38–42; locomotion of, 238–39, 244–45; phylogeny
50, 106, 256, 260; tracks in, 256, 264, 267, of, 5, 8, 18, 32–33, 36, 38–39; posture of, 39–41; size
270–71, 273–74 of, 233, 235–36, 246, 265. See also brachiosaurids;
Jurassic-Cretaceous boundary, 105, 106, 118–19, 145 Opisthocoelicaudiinae; saltasaurids; saltasaurines;
juveniles, 63, 314, 316, 319. See also hatchlings Somphospondyli; Titanosauriformes; titanosaurs;
names of specific genera
Kazakhstan, 119 Madagascar, 5, 145; titanosaurs in, 38, 51, 56–59, 67, 296
Kimmeridgian: diplodocoids in, 134–36; and diversity, 105, Magyarosaurus, 53, 68–71, 72, 234–35, 303
112, 115–18, 120, 129–30, 135; feeding mechanisms in, Maiasaura, 309, 322
132, 134; titanosaurs in, 66, 136–37, 138–39 Malagasy Taxon B, 53, 68–69, 72
knee, 40, 180, 239 Malawi, 67
Kotasaurus, 115, 234 Malawisaurus, 72, 115; M. dixeyi, 52, 296; phylogeny of,
52–55, 56, 68–71, 73, 296
lacrimal, 59–60, 295, 297 Mamenchisaurus, 19, 30, 115, 211–12; M. constructus, 246;
LAG (lines of arrested growth), 305, 307, 309–10, 313; locomotion of, 237, 246; phylogeny of, 19, 115, 234
in sauropods, 312, 315, 317, 319, 321 mammals, extant, 168, 182, 189, 306; growth of, 311,
laminae, vertebral, 201–2, 205–9, 211, 223 312–13, 315, 317; and locomotion, 229, 230, 231, 246;
Laplatasaurus, 57, 234 and pneumaticity, 212, 219, 221. See also names of
Lapparentosaurus, 234, 315 specific fauna
life history, 312, 323 mandibular symphysis, 158, 163, 175
ligament, 37, 179, 181, 187, 221 manus, 35, 41, 64; and locomotion, 237–38, 239–40,
Limaysaurus, 20, 38 243
limbs, 9, 15, 239, 241; bones of, 217, 230; length of, manus and tracks, 35, 39, 253–54; and claws, 255–57,
230, 243–44; proportions of, 126, 127–28, 299; 260, 263, 274–75; crescentic, 258, 263, 264,
scaling of, 231–33. See also forelimbs; hindlimbs; 265–66; depth of, 273–75; and digits, 255–57, 265,
locomotion; posture 275; semicircular/horeshoe-shaped, 258, 264, 265;
Lirainosaurus, 53, 65, 68–71, 72, 115, 234 size of, 259, 260, 265
lithology: carbonate, 267, 271–73, 275, 276; clastic, manus-pes ratio, 260, 264, 265–66, 274
267–69, 273, 275; limestone, 267–69, 272–73, 275; Marasuchus, 209
mudstone, 267, 269, 272, 286, 287–88, 292–93; marginocephalians, 2, 3
sandstone, 267–69, 272–73, 285–86, 288, 293; marsupials, 322
siliciclastic, 271–73, 276; siltstone, 268–69, 286, 288 mass. See body size/mass; DME
Lithostrotia, 52, 56, 67–71, 73 Massospondylidae, 138–39
Lockley, M. G., 266, 271 Massospondylus, 139, 140, 322; bones of, 26, 33–34, 144;
locomotion, 39, 144; and gait, 39, 273; and tracks, 265, M. carinatus, 238, 241–42; locomotion of, 238,
273–76, 277; and weight distribution, 273, 275. See 241–42
also bipedalism; posture; quadrupedalism; rearing mastication, 144, 169
locomotion, evolution of, 229–47; and body size, maxilla, 59, 60, 160, 172, 173; embryonic, 295–96, 297;
230–36, 246; lower-level patterns in, 243–46, 247; of Nigersaurus taqueti, 158, 162–64, 168, 169
INDEX 343
McIntosh, John “Jack,” 327–33 201–2, 205–6, 209, 211, 215, 218, 220; presacral, 22,
Meckel’s canal, 158, 165 37, 201–2
Megaloolithidae, 299–300 New Mexico, 256–57, 262
Megaloolithus, 289 New York, 275
Melanorosauridae, 138–39 niche partitioning, 138, 142, 148, 194, 195
Mesozoic, 10, 125, 145, 146, 266; plants in, 195, 196–97 Niger, 157, 158, 168, 268
metacarpal, 33–34, 41, 61, 64; evolution of, 33, 238; and Nigersaurus, 9, 37, 38, 119, 128, 172
locomotion, 237, 244; and tracks, 256–58, 261, 265 Nigersaurus taqueti, 157–72; skull of, 157–66, 169–70;
metacarpus, 21 teeth of, 158–60, 162–69, 170–75
metapodial, 34 nodes, 3, 20, 22, 52, 68–71
metatarsal, 24, 27, 28, 29–30, 64; and locomotion, 21, NOOs (number of opportunities to observe), 110,
239, 240–41, 243 117–18, 119–20, 121
Mexico, 267 Norian, 2, 23, 146, 216; and diversity, 105, 115–16, 117,
mink, 309, 318 118, 130–32, 135; prosauropods in, 129, 140
Mongolia, 19, 56, 67, 73, 296 North America, 23, 39, 111, 139, 146, 196; coexistence
monophyly, 19, 125, 139; of prosauropods, 20, 22, 125; in, 142; diplodocids in, 35, 172; Late Jurassic of, 11,
of Titanosauriformes, 65, 67, 70, 73; of titanosaurs, 35, 145; prosauropods in, 129; skeletons in, 15–16;
52, 54–56, 65, 68, 69, 71 tracks in, 34, 263, 266. See also names of specific
monospecificity, 112, 127 countries
Montana, 196
Morocco, 34, 263, 264, 267, 274 occlusion, 126, 174; development of, 141, 144; interdigi-
MPTs (most parsimonious trees), 52, 55, 69, 73, 109, 114 tating, 127–28; and wear facets, 167, 169
muscles, 203, 207, 223; and growth, 304, 316; and olecranon, 21–22, 230, 239, 244
locomotion, 239, 242–43, 245 olecranon process, 41, 239
Omeisaurus, 115; bones of, 24–25, 28, 37; feet of, 30,
nares, 59, 60–61, 171, 172; embryonic, 295, 297–98; of 31–32, 33, 34, 258; locomotion of, 237–40, 246;
Nigersaurus taqueti, 158, 159, 161, 163, 169 phylogeny of, 18–19, 234; O. tianfuensis, 240, 246;
narial fossa, 158, 159, 160, 161 and tracks, 254, 258
neck, 221; curvature of, 172, 182, 185–89; dorsiflexion of, ontogeny, 16, 276, 297, 300; and growth, 303–4, 310–11,
185–86, 187–89, 190–92, 195; elongation of, 21, 28, 38; 312, 314; and pneumaticity, 206, 208, 216, 223
evolution of, 121; flexibility of, 180, 189–92; length of, operational taxonomic units (OTUs), 109
126, 127–28, 179, 195, 212, 245; ventriflexion of, 185–86, Opisthocoelicaudia, 115; bones of, 24, 31–32, 40, 41, 254;
189. See also vertebrae, cervical; vertebral column locomotion of, 240, 244, 246; phylogeny of, 53–55,
nemegtosaurids (Nemegtosauridae), 55, 128, 133, 137, 68–72, 139, 234, 254; O. skarzynskii, 38, 52, 240,
143, 196 244, 246; and titanosaurs, 56, 73
Nemegtosaurus, 72, 115; N. mongoliensis, 296, 297–98; Opisthocoelicaudiinae, 38; 52, 67, 70–71, 73
phylogeny of, 19, 53, 55, 65, 68–71, 73; teeth of, 171, 297 opisthocoely, 181
neoceratopsians, 9 orbit, 59, 159, 295, 297
neosauropods (Neosauropoda), 19, 36, 231; bones of, origination times, 109, 118
21, 38, 160; embryonic, 295–96; evolution of, 18, ornithischians (Ornithischia), 4, 157, 239, 276; basal, 3,
32–35, 243; feet of, 30–31; locomotion of, 239, 240, 34, 239; diversity of, 107, 121, 144; and
242, 243, 244, 245; and pneumaticity, 205, 206, sauropodomorphs, 143–46; teeth of, 172–74
209, 216; radiation of, 107, 118, 233; and ornithopods (Ornithopoda), 2, 3, 237; body size of, 231,
titanosaurs, 55, 62; and tracks, 254–56, 258, 233; teeth of, 173, 174; tracks of, 40, 266, 271, 272
264–65, 274. See also diplodocoids; macronarians; ossification, 41, 57, 316; reduced, 21, 26, 33
names of specific genera osteoblasts, 305
neotheropods, 241 osteoclasts, 203
nests, 285, 292, 293, 298–300. See also egg-beds osteoderm, 56–57
Neuquensaurus, 72, 303; N. australis, 39; phylogeny of, osteology, 253; and feeding, 179, 181, 186, 189, 190–91, 193;
53–55, 56, 68–71, 234, 296 and pneumaticity, 209–10. See also bone growth; bone
neural arches, 63–64, 202, 206, 209, 211, 218 growth rates; bones; bone tissue; bone vascularization
neural canal, 202, 208 osteons, 306–7, 308
neural cavity, 202, 205 ostrich, 189, 206, 309, 311, 318–19
neural spine: caudal, 215, 218; cervical, 171, 173, 206, 218; outgroups, 22, 29–30, 33, 36–37, 114; and titanosaurs,
dorsal, 63–64, 202, 205, 218; and pneumaticity, 68, 71
344 INDEX
Owen, Richard, 15 ingroups; lineages, ghost; monophyly; outgroups;
Oxfordian, 138; diplodocoids in, 134, 136; and diversity, PDEs; trees, phylogenetic; synapomorphies; under
105, 115–18, 120, 130, 132–33, 135–36 specific clades and genera
phylogeny of titanosaurs, 50–73; background on, 51–56;
pachycephalosaurs, 37 and interrelationships, 65–70; and Rapetosaurus
pads, 263; heel/pedal, 255, 258–59, 263, 264, 265 krausei, 56–65
paleobiology, 6, 185, 230; and pneumaticity, 201, 212, physiology, 221, 223, 299, 305
216–17, 220–21; and reproduction, 10, 285, 292. See plants, 141, 145, 167, 195, 294; aquatic, 190, 196; crop-
also biology ping of, 127–28, 129, 174; evolution of, 107, 195;
paleobotany, 195 height of, 179, 194; and interactions with
paleoclimate, 272, 276 sauropodomorphs, 146–47; nipping of, 128, 129. See
paleoecology, 121, 146, 178, 194–96 also names of specific flora
paleoenvironment, 137, 252–53, 316; coastal, 267–69, Plateosaurus, 24–25, 28, 30–31, 37; P. engelhardti, 22–23
271–73, 275, 299; floodplain, 268, 273; fluvial, plesiomorphy, 310
267–69, 271–73, 275, 286; forest, 268, 271; inland, plesiosaurs, 6, 15
271, 273, 276; lacustrine, 268–69, 271–73; marine, Pleurocoelus, 214–15, 265
271–73, 276; plains, 272, 273, 286; subtropical, 266, Pliensbachian: and diversity, 105, 111, 115–16, 130–32;
269; swamp, 268, 272 prosauropods in, 129, 140; tracks in, 259, 263
paleogeography, 270 pneumaticity, 201–23; and air volume, 212–15; correlates
paleolatitude, 253, 266 of, 209–10; criteria for, 204–10; distribution of, 215–16,
paleovertisols, 286 223; extramural, 203, 204–5; and future research,
Pangaea, 272 221–23; and internal complexity, 210–12; in living ver-
Parabrontopodus, 256, 260, 273–74; P. mcintoshi, 257, tebrates, 201–4; and mass, 216–21; origin of, 207–9
264–65 Poland, 34–35, 267
Paralititan, 53, 68–69, 72, 73, 303 pollex, 245
paraoccipital process, 59–60, 158, 160 polytomies, 113–14, 116
paraphyly, 19, 54, 139, 244, 300; of prosauropods, 22, polytomy, 65, 67–68, 69, 71, 72
125, 141 Portugal, 256–57, 258, 260, 267–69, 274
parental care, 298 postacetabulum, 242
parietal, 59, 295–96, 297 postorbital, 158, 160, 296
Patagosaurus, 28, 115, 139, 205, 234 posture, 16, 186, 195; columnar, 21, 23–27, 40–41, 230;
patristic distance, 235–36 digitigrade, 29–30, 34, 237–38, 240; forefoot, 33–34;
PDEs (phylogenetic diversity estimates), 112, 121, 129, 137, graviportal, 40–41, 230, 273; hindfoot, 21, 28–32;
143, 147; disadvantages of, 109, 117, 127; and strati- and limb size, 232–33; neutral, 182; parasagittal, 39,
graphic range, 111, 113; vs. TDEs, 107–9, 114, 116, 118–19 273; semidigitigrade hindfoot, 30; wide-gauge limb,
pectoral girdle, 40–41, 180, 188, 192, 244 22, 39–41, 245. See also bipedalism; locomotion;
Peirópolis form, 53 quadrupedalism; semibipedalism
Pellegrinisaurus, 53, 115 preacetabulum, 241–42, 244–45
pelvic girdle, 64, 244, 245 predation, 179
pelvis, 180, 276 predentary, 172
Perú, 289 premaxilla, 24, 59, 160, 295; of Nigersaurus taqueti, 158,
pes, 28–29, 30–32, 35, 39–40, 64; and locomotion, 21, 159–63, 169
238–39, 240–41 preservation: artifacts of, 257; and diversity estimates, 106,
pes and tracks, 253–57; and claws, 255–56, 258–59, 109–11, 117, 119, 120–21, 129; poor, 141, 258, 275–76;
262–65, 273–74; digitigrade, 255, 258; and digits, and sea level, 137; and tracks, 257, 271–72, 276
255–56, 262–63; semidigitigrade, 258. See also primates, 203
manus-pes ratio proboscideans, 229, 239
phalanges, 28, 30, 31, 34, 41; and locomotion, 21–22, prosauropods (Prosauropoda), 2–3, 10, 27–28, 117, 146;
239–40, 245 bones of, 33, 37–38, 209, 263; and competition with
pharynx, 203 sauropods, 140–41; decline of, 143–44; distribution
Phuwiangosaurus, 72, 115, 214–15; phylogeny of, 53, 54, of, 4, 41; diversity of, 125–27, 129–48, 144; in Early
65, 67–71, 73, 234 Jurassic, 129, 143–44, 146; feet of, 31–32, 33, 263;
phylogeny, 5–6, 8; diagrams of, 19, 20; and growth, 310, limbs of, 24, 231; locomotion of, 237–40, 242;
312; and parsimony, 69, 73, 109, 114, 310, 318. See phylogeny of, 22–23, 234; TDEs for, 126–27, 129,
also cladograms; classification; diphyly; divergence; 130, 140, 146. See also names of specific genera
INDEX 345
Pseudotetrasauropus jaquesi, 262–63 Romer, Alfred S., 329
Psittacosaurus, 322 rostrum, 192, 297
PSP (postcranial skeletal pneumaticity), 203, 209, 216, Rotundichnus, 274; R. muenchehagensis, 257, 265–66
220–23. See also pneumaticity
pterosaurs, 204, 223 sacrum, 180, 245
pterygoids, 59–62, 160, 295 Salgado, L., 52
pubis, 63, 64, 245, 315 saltasaurids (Saltasauridae), 216; phylogeny of, 38, 52,
55–56, 67, 69–71, 73; posture of, 22, 40–42
quadrate, 59–60, 62, 158, 160, 295, 297–98 saltasaurines (Saltasaurinae), 233, 245, 297; phylogeny
quadratojugal, 59, 158–59, 160, 295–96, 297 of, 39, 52, 65, 67–72, 73, 296
quadrupedalism, 21, 23–27, 42, 236–39, 241, 255; and Saltasaurus, 115; bones of, 72, 214–15, 246; S. loricatus,
body size, 231–33; obligate, 237–38 22–23, 27, 32, 35, 38–39, 52, 246; phylogeny of,
quadrupeds, extant/modern, 180, 181, 183. See also 53–55, 56, 68–71, 234, 296; and pneumaticity, 202,
names of specific animals 204–5; size of, 303–4, 323
Quaesitosaurus, 72, 115; phylogeny of, 19, 53, 55, 68–71, sampling, 107, 113; biases in, 109–10, 117, 120, 122, 145;
73, 171 differences in, 116, 121; errors in, 118; insufficient,
quail, 311 119, 222
Santa Rosa indet., 53, 68–69, 72
radiation, 110, 111, 120, 126, 141; of diplodocoids, Santonian, 105, 115–16, 119–20, 130, 132–33, 135–36
134–35; of heterodontosaurids, 144; in Late saurischians (Saurischia), 2–4, 18, 22, 26, 42, 237;
Jurassic, 137–38, 233; in Late Triassic, 137; of locomotion of, 237, 239, 246; and pneumaticity,
neosauropods, 107, 118, 233; of ornithischians, 145; 209, 216. See also prosauropods; theropods
of plants, 147; of prosauropods, 129; of Sauropodichnus, 274; giganteus, 257, 265–66
sauropodomorphs, 148; of titanosaurs, 105, 137. See sauropodomorphs (Sauropodomorpha), 4, 22, 107, 125,
also diversity, increase in 237, 254; and interactions with ornithischians,
radius, 22, 41, 61, 63, 64, 307; Apatosaurus, 316–20 143–46; and interactions with plants, 146–47; and
ramus: alveolar, 159, 162–63; posterodorsal, 162, 169–70 pneumaticity, 215–16; radiation of, 148. See also
range: geographical, 138, 140, 142–43, 144; ghost, 108, diversity; prosauropods; sauropods
112, 113–14, 116–17, 119 Sauropodopus antiquus, 262–63
range, stratigraphic, 107–8, 109, 113–14; on sauropods, 23; adult, 303, 304; competition with,
charts/graphs, 105, 115–16, 130–32, 135–36, 261, 271; 140–41, 148; derived, 216, 240, 260; fossil record of,
and competition, 138, 140, 142 2–6, 10–11, 120; juvenile, 204, 210, 211, 299; size of,
Rapetosaurus, 170, 245; and embryos, 296, 297; R. krausei, 1, 6–7, 230, 235, 303–4, 312. See also eusauropods;
53, 55, 56–65, 67, 72, 73; phylogeny of, 68–71, 234 neosauropods
ratites, 189, 311 sauropods, basal/primitive, 5, 134, 139; and body size,
Rayosaurus, 115, 172, 234 230, 236; bones of, 30, 33–34, 207; locomotion of,
rearing, 41, 195, 245 237–39, 240, 243; and tracks, 255, 258. See also
rebbachisaurids (Rebbachisauridae), 38, 119, 137; in names of specific genera
Cretaceous, 10, 35, 171; phylogeny of, 32, 35, 42; Sauroposeidon, 206, 207, 209–11, 214, 222, 244
teeth of, 37, 172, 173–74. See also Nigersaurus taqueti; scapulae, 22, 61, 63, 64, 180–81; Apatosaurus, 316–18,
names of specific genera 320; of Nigersaurus taqueti, 158, 159
Rebbachisaurus, 172; R. garasbae, 35 scapulocoracoids, 180
reproduction, 10, 41, 212, 285, 292, 316. See also eggs, sea level, 106, 120–21, 137
Auca Mahuevo Seismosaurus, 243, 303
reptiles, 306, 312, 313, 322; modern, 298, 303, 315 semibipedalism, 231–33, 237
respiratory system, 216, 223. See also air sacs, pul- semidigitigrady, 30, 258, 264
monary; lungs; trachea Senonian, 66
Rhaetian, 23, 139; and diversity, 105, 115–16, 117, septa, 173; of Nigersaurus taqueti, 163, 168; vertebral,
129–32, 135 202, 207, 209, 210, 211
rhinocerotids, 239 Sereno, P. C., 18–19, 106–7
rhinos, 181, 190, 229 sexual dimorphism, 312, 316, 323
rhynchosaurs, 146 sexual maturity, 303, 316
ribs, 61, 63–64, 173, 180, 188, 204–5 SG (specific gravity), 217, 220, 221
Rocasaurus, 53, 68–71, 72, 234 shoulder, 192, 195, 239, 245, 246
Romania, 66, 300, 303 shoulder girdle, 40–41, 181
346 INDEX
Shunosaurus, 2, 115, 127; bones of, 24–25, 29; feet of, subunguligrady, 239
31–32, 33–34; S. lii, 27, 238, 240–41, 246; locomotion success rate, 9
of, 237–42, 246; phylogeny of, 18, 234; and tracks, Supersaurus, 205, 303
254–55; vertebrae of, 28, 37, 209, 211, 216 surangular, 59–60, 62, 158–59, 160, 163, 170
Shuvuuia, 322 sutures, 202, 203, 208
Simpson, George Gaylord, 1, 2 swimming, 275
simulations, 109, 113, 194 Switzerland, 261, 263, 267–68, 274
Sinemurian, 259; and diversity, 105, 111, 115–16, 130–32, sympatry, 138, 148, 194, 197
135; prosauropods in, 129–30 symphysis, 158, 165, 170
sirenians, 310, 318 symplesiomorphies, 244
sister-taxa, 54, 55, 65, 72, 108 synapomorphies, 3–4, 16, 20–23, 32–33, 41; appendicular,
sites. See formations and sites 39, 246; cranial, 36, 62; and diversity estimation, 107,
skeletochronology, 310, 313, 314, 317 113, 127; and embryos, 295, 296; of Eusauropoda, 27,
skeletons, 9, 15, 40, 185, 276, 311; appendicular, 180–81, 296; of titanosaurs, 52, 54, 55, 65; and tracks, 252,
237, 243; axial, 180, 181–83; complete, 16, 27; juvenile, 254, 255
56, 59–61; neutral pose of, 178, 180–83; theropod, 287; synapsids, 203
weight of, 220–22. See also bones; pneumaticity; skulls synonyms, 264, 266, 289
skin, 203, 275, 295. See also integument synovial capsule, 181, 185, 189–90, 191
skulls, 19, 56; embryonic, 295–98; evolution of, 36–37, Syntarsus, 309, 322
121; holotypic, 59–60; juvenile, 60–63; Nigersaurus systematics, 104, 139, 157
taqueti, 158–60, 169–70; of titanosaurs, 59–63, 295.
See also cranial remains tabefaction, 141, 142, 143, 146
slickensides, 286, 290, 291 Tadjikistan, 267, 269
snakes, 203 tails, 22, 37–38, 41, 252. See also vertebrae, caudal
snout, 28, 36, 172, 296; of Nigersaurus taqueti, 159 Tanzania, 66, 137
soft tissues, 304. See also adipose tissue; cartilage; taphonomic effect, 106, 120–21, 137
ligaments taphonomy, 196, 299
soil, 286 tarsal, 21, 27, 33
Somphospondyli, 38, 55, 62, 65, 70–71, 73 tarsus, 41
south, the, 157 taxa, 109; extant/modern, 309, 312, 318–19, 323; number
South Africa, 117, 144, 263 of, 105, 115–16, 142; sister, 54, 55, 65, 72, 108; termi-
South America, 5, 19, 142, 143, 146, 157; dicraeosaurids nal, 66–67. See also TDEs; names of specific taxa
in, 35–36, 172; Late Cretaceous of, 145, 170; taxonomy, 7, 51, 139. See also clades; cladograms;
titanosaurs in, 38, 54, 57; tracks in, 34, 274. See also classification
names of specific countries TDEs (taxic diversity estimates): for diplodocoids, 136–37,
South Korea, 261, 266, 269, 271–72 143; and feeding, 133, 138; and herbivory, 125–26; and
Spain, 66, 268–69 plants, 147; for prosauropods, 126–27, 129, 130, 140,
specializations: appendicular, 26; herbivorous, 21, 27–28, 146; for sauropods, 110, 117–20, 127, 130, 133; and
36–37, 42; locomotor, 230, 239–43, 247; presacral, 22, stratigraphic range, 111, 113; and teeth, 134–35; for
37; tail, 22, 38 titanosaurs, 136–37; vs. PDEs, 107–9, 114, 116, 118–19
speciation, allopatric, 106, 121, 137 tectonics, 137
species: and diversity estimation, 107, 109, 110, 112–13, teeth, 126, 287–88; abrasion of, 168, 172; batteries of,
127; number of, 1, 6 162, 167–69, 171–75; and classification, 18; columns
spongiosa, 205, 311 of, 163, 164, 168, 169, 172, 173; dentary, 162, 166,
squamates, 39, 220 167, 171, 172; of diplodocoids, 37; and diversity,
squamosal, 59, 295–96, 298 127–29; embryonic, 167–68; of Eusauropoda, 28,
squirrels, 314 36; lower, 166–67, 168, 172, 196; maxillary, 162–63,
sternal plates, 22, 40 164, 166–67, 168–69, 172; of Nigersaurus taqueti,
stratigraphic stages: on charts/graphs, 105, 115–16, 158–60, 162–69, 170–75; number of, 158, 166–68,
130–32, 135–36, 261, 271. See also Cretaceous; 171, 172; peglike, 194; pencil-shaped, 295–96, 297;
Jurassic; Mesozoic; Triassic; names of specific stages premaxillary, 163, 166–67, 169; replacement, 21,
stress, 306, 310 159, 163, 164, 172, 173; rows of, 21, 27, 168, 170;
stride length, 231, 239, 243, 244, 257, 265 spatulate, 16, 194–95; of titanosaurs, 60–62; upper,
subadults, 61, 316–17, 321 166–67, 168, 172, 196. See also crowns; dentary;
subclades, 16, 52 occlusion; wear facets
INDEX 347
Tehuelchesaurus, 115, 234, 246 topotypes, 264
Telmatosaurus transylvanicus, 300 trachea, 203, 216, 217–20, 221
temperature, body, 221 tracks, pneumatic, 204, 209
Tendaguria, 206 trackways/ichnofossils, 10; cladistic approach to,
tetrapods, 138, 178, 203, 240 254–55; and classification, 258–66; depth of, 273;
Tetrasauropus, 2, 23, 31, 35; tracks of, 262–63, 274; and diversity, 106; earliest, 30; and eggs, 286; gauge
T. unguiferous, 262–63 of, 276; ideal, 254, 258; juvenile, 259, 263, 272–73;
Texas, 265, 269, 275–76 and locomotion, 39, 239, 273–76; multiple, 275–76;
Thailand, 23, 67, 117, 268 paleoenvironmental distribution of, 271–72; paleolat-
Thecodontosaurus caducus, 215–16 itude of, 253; parallel, 275; prdicted, 252–58; size of,
theropods (Theropoda), 24, 27–28, 57; basal/primitive, 32, 258–59, 263, 266; spatiotemporal distribution of,
37, 239; bones of, 24, 182, 241; diversity of, 2, 121; feet 266–71; and speed, 265, 277; of titanosaurs, 38, 50,
of, 31–32, 33, 34; locomotion of, 239, 240, 241–42, 243; 106, 118, 245, 261; Triassic, 19, 23, 31, 34–35, 267,
phylogeny of, 4, 234; and pneumaticity, 204, 210, 214, 273–74. See also fossil formations and sites
216, 219, 223; tracks of, 40, 272, 275. See also birds trackways/ichnofossils, narrow-gauge, 39, 255–61, 264,
thin sections, 311, 316 273–75; Cretaceous, 268–69, 273; Jurassic, 106, 261,
thorax, 41, 180, 203 267–68, 273–74; Triassic, 263, 267, 273–74
thyreophorans, 2, 144, 231, 237 trackways/ichnofossils, wide-gauge, 39–40, 255–61,
Thyreophora, 3 263, 265–69, 273–75; titanosaur, 50, 106, 245, 261;
tibia, 24, 29, 40, 41, 64; and locomotion, 21–22, 239 titanosauriform, 258–61, 265–66
Titanosauridae, 138–39, 254–55, 274, 296 transverse process, 205, 222
Titanosauriformes: monophyly of, 54–56, 65, 67, 70, trees, 196. See also forests
73; phylogeny of, 38, 52, 62, 67–71, 296; possible, trees, phylogenetic, 19–20, 139, 234; Adams consensus,
139, 265–66; and tracks, 254–56, 258, 259, 260, 55, 67–71, 72; composite, 107; most parsimonious
265–66, 274; vertebrae of, 211, 219 (MPTs), 52, 55, 69, 73, 109, 114; reduced consensus,
Titanosaurinae, 51 114; strict consensus, 65, 70, 72, 73
Titanosaurinae indet., 53 Triassic, 4, 10, 125, 256; diversity in, 105, 115–16. See
Titanosauroidea, 51, 54, 55 also names of specific periods
titanosaurs/titanosaurians (Titanosauria), 50–73; back- Triassic, Late/Upper, 27, 30, 109, 134, 146; diversity in,
ground on, 51–56; basal, 55, 73, 244; derived, 56, 137, 117, 137; origin in, 2, 104, 125; prosauropods in, 19,
240; and diplodocoids, 55–56, 67, 68, 70, 73, 140; tracks in, 23, 32, 34–35, 261–64, 266–67,
142–43; eggs and embryos of, 285, 288, 292–97; 270–71, 273–74
feeding mechanism of, 128, 131; interrelationships tripodality, 128, 195
of, 65–70; locations of, 38–39, 50, 51; locomotion of, trochanter, lesser, 239, 242
10, 239, 240, 241–42, 243, 244–46; missing data on, trunk, 180, 192, 195, 243–44, 245
72; monophyly of, 52, 54–56, 65, 68, 69, 71; phy- turkeys, 182, 190
logeny of, 9, 19–20, 32, 38–39, 42, 67–71; and Turonian, 105, 115–16, 119–21, 131, 132, 135
plants, 147; and pneumaticity, 205, 212; radiation of, turtles, 318
105, 137; Rapetosaurus krausei, 53, 55, 56–65, 67, 72, Tyrannosaurus, 222
73; reproduction of, 298; size of, 39, 233, 244, 245,
303, 323; TDEs of, 136–37; teeth of, 18, 134, 167, 170, ulna, 41, 63, 64, 316–18, 320
171, 194, 296; time period of, 38, 50, 133, 139; and unguals, 21, 31, 237, 240, 264
tracks, 39–40, 106, 118, 253–54, 256–57, 260, 266. ungulates, 189, 219
See also Lithostrotia; Opisthocoelicaudiinae; United Kingdom, 118, 139, 256–58, 263, 267–68, 274.
saltasaurids; saltasaurines; names of specific genera See also England; Isle of Wight; Wales
‘Titanosaurus’, 54–55, 113, 115; T. colberti, 53, 55, 246; United States, 16, 66, 144, 267–69, 274. See also names
T. indicus, 50–51, 73; T. madagascariensis, 56–57, 59. of specific states
See also Isisaurus Upchurch, P., 18, 19–20
Tithonian: diplodocoids in, 136, 171; and diversity, 105, Utah, 259, 262, 267
112, 115–18, 120, 130, 135–36; feeding mechanisms Uzbekistan, 119
in, 132, 134
Toarcian: and diversity, 105, 115–18, 120–21, 130–32, 135; Valanginian, 105, 115–16, 119–20, 130, 132, 135–36
prosauropods in, 130–31, 140 variation, ontogenetic, 16
toes, 264 vascular canals, 158, 165, 306–7, 308–9, 317, 318. See
tooth combs, 126 also bone tissue, vascularization of
348 INDEX
vegetation. See plants presacral, 181, 192; ventriflexion of, 181, 184,
vertebrae: apneumatic, 205, 210, 218, 221; camellate, 194
202, 205, 210–12; complexity of, 204, 211–12, vertebrates, 230, 254, 306, 310
214–15; evolution of, 211–12, 215; fused, 204; num- vertebrates, extant/modern: growth of, 304–5, 311, 313,
ber of, 21–22, 28, 37, 171, 172, 189; polycamerate, 318, 320–22; necks of, 181, 182, 183, 189–90, 191;
211–12; presacral, 63, 204, 205, 211, 215–16; sacral, and pneumaticity, 201, 222. See also names of specific
37, 62, 64, 204, 215–16, 218; semicamellate, fauna
211–12; somphospondylous, 211; thoracic, 183–84, vigilance, 179–80
216; types of, 211. See also chambers, vertebral; viscera, 203
diverticula, vertebral; foramina, vertebral; fossae, Volkheimeria, 234
vertebral; laminae, vertebral; neural spines; septa, vomer, 160
vertebral Vulcanodon, 2, 115, 140, 231; bones of, 24–26, 28,
vertebrae, caudal, 15, 24, 215; number of, 37–38, 41, 62; 29–30; hindfeet of, 30–31; V. karibaensis, 241–42;
and pneumaticity, 204, 215–16, 218–19, 223; of locomotion of, 237–39, 241–42; phylogeny of, 23,
titanosaurs, 51, 52, 54, 56–57, 59, 62–63 234, 296; and tracks, 254–55
vertebrae, cervical, 37, 61, 189, 193; articulation of, Vulcanodontidae, 138–39
184–85, 187; diplodocoid, 171–73; number of, 21–22,
62, 171–72; opisthocoelous, 181, 184–85; and pneu- Wales, 263, 267, 274
maticity, 206, 207–8, 210, 213, 216, 218; wedge- water, 196, 271–73, 275, 276, 286. See also sea level
shaped, 183–84, 186–87 wear facets, 174, 196; buccal, 172; distal, 127–28; exter-
vertebrae, dorsal, 37, 64, 180; and head height, 183, nal, 166, 170–71, 172; internal, 166, 170, 171; mesial,
192; number of, 21–22, 62; and pneumaticity, 202, 127–28; of Nigersaurus taqueti, 127–28, 166, 167,
206–7, 213, 216, 218 168–69
vertebral centra, 181, 183, 186–87; and pneumaticity, Weishampel, D. B., 107, 113, 116–17
201–2, 204, 206, 207, 210, 215; volume of air in, whales, 315, 321
218
vertebral column: arching/curvature of, 180–81, 182, xenarthrans, 241
183, 185, 188, 192; caudal, 195; cervical, 182–92;
dorsal, 181, 182, 188; dorsiflexion of, 181, 184, zygapophyses, 206, 210; and feeding, 181, 182–85, 187,
193–94; neutral, 182; pneumaticity along, 215–16; 189–93
INDEX 349

The Sauropods Evolution and Paleobiology - Kristina Curry Rogers Amp Amp Jeffrey Wilson PDF

Uploaded by

Copyright:

Available Formats

The Sauropods Evolution and Paleobiology - Kristina Curry Rogers Amp Amp Jeffrey Wilson PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Sauropods Evolution and Paleobiology - Kristina Curry Rogers Amp Amp Jeffrey Wilson PDF

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What time period do sauropods originate from?

What time period do sauropods originate from?

THE SAUROPODS

Kristina A. Curry Rogers and Jeffrey A. Wilson

UNIVERSITY OF CALIFORNIA PRESS

University of California Press

University of California Press, Ltd.

© 2005 by the Regents of the University of California

Library of Congress Cataloging-in-Publication Data

The Sauropods: Evolution and Paleobiology / [edited by] Kristina A. Curry

The paper used in this publication meets the minimum requirements

Cover: Titanosaurid Nesting Grounds, painting by Mark Hallett.

Acknowledgments / ix 7 • POSTCRANIAL PNEUMATICITY IN SAUROPODS

Monoliths of the Mesozoic

Jeffrey A. Wilson and Kristina Curry Rogers

2 MONOLITHS OF THE MESOZOIC

CHAPTER PAGES SPECIES PAGES/SPECIES

1 Basal Saurischia 22 11 2.00

MONOLITHS OF THE MESOZOIC 3

4 MONOLITHS OF THE MESOZOIC

MONOLITHS OF THE MESOZOIC 5

6 MONOLITHS OF THE MESOZOIC

MONOLITHS OF THE MESOZOIC 7

SAUROPOD EVOLUTION groups. Wilson highlights the difficulties in trac-

8 MONOLITHS OF THE MESOZOIC

MONOLITHS OF THE MESOZOIC 9

10 MONOLITHS OF THE MESOZOIC

MONOLITHS OF THE MESOZOIC 11

12 MONOLITHS OF THE MESOZOIC

MONOLITHS OF THE MESOZOIC 13

14 MONOLITHS OF THE MESOZOIC

Overview of Sauropod Phylogeny

16 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

18 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 19

20 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

CHARACTER NUMBER, BY CLADISTIC ANALYSIS

SALGADO ET AL. UPCHURCH WILSON & SERENO WILSON

Columnar, quadrupedal posture (Sauropoda)

CHARACTER NUMBER, BY CLADISTIC ANALYSIS

SALGADO ET AL. UPCHURCH WILSON & SERENO WILSON

Presacral specializations (Flagellicaudata)

22 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 23

FORE:HIND MT III:TIBIA REFERENCE(S)

24 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

26 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 27

28 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

FIGURE 1.11. Right pes of Apatosaurus in proxi-

OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 29

30 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

(MINIMUM BREADTH) (LENGTH)

OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 31

Neosauropoda Bonaparte 1986b— Cretaceous neosauropods include rebbachisaurid

32 OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION

OVERVIEW OF SAUROPOD PHYLOGENY AND EVOLUTION 33