Encyclopedia of Dinosaurs - PDF Free Download (2022)

Encyclopedia of


Encyclopedia of


Philip J. Currie Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Kevin Padian Museum of Paleontology University of California, Berkeley




New York




Acaedemic Press is an imprint of Elsevier, Inc. 84 Theobald’s Road, London WC1X 8RR, UK Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands Linacre House, Jordan Hill, Oxford OX2 8DP, UK 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA Copyright © 1997 Elsevier Inc. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [emailprotected] Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material

Encyclopedia of Dinosaurs by Philip Currie ISBN 978-0-12-116810-6



Thematic Table of Contents





John R. Hutchinson and Kevin Padian

A Guide to Using the Encyclopedia Foreword


Australasian Dinosaurs






24 27

Ralph E. Molnar



Luis M. Chiappe



The Editors


A Abelisauridae


Kevin Padian


Fernando E. Novas

African Dinosaurs


Louis L. Jacobs

Age Determination



Gregory M. Erickson

Albany Museum

Barun Goyot Formation


Halszka Osmo´lska

Anusuya Chinsamy


Bastu´s Nesting Site


Bavarian State Collection for Paleontology and Historical Geology


Peter Dodson

American Museum of Natural History


Bayan Mandahu


Philip J. Currie


Bayn Dzak


Halszka Osmo´lska

Kenneth Carpenter



The Editors

Lowell Dingus



Jose´ L. Sanz and Jose´ J. Moratalla

Kevin Padian and John R. Hutchinson

American Dinosaurs




J. Michael Parrish

John R. Horner




vi Bernard Price Institute for Palaeontological Research

Carnosauria 50

Anusuya Chinsamy


Cedar Mountain Formation 51

Jean Le Loeuff




The Editors

Ralph E. Chapman and David B. Weishampel



Peter Dodson



Clive Trueman




68 71

Chinese Dinosaurs

Kevin Padian

Bird Origins Kevin Padian and Luis M. Chiappe

Bone Cabin Quarry


Brent H. Breithaupt

Braincase Anatomy


Philip J. Currie



Kevin Padian and John R. Hutchinson

105 106 106

Timothy Rowe, Ron Tykoski, and John Hutchinson

Chemical Composition of Dinosaur Fossils

Spencer G. Lucas


Mark A. Norell

R. McNeill Alexander



James I. Kirkland

Central Asiatic Expeditions 57


John R. Hutchinson and Kevin Padian


Herve´ Bocherens


Dong Zhiming

Chinle Formation


J. Michael Parrish

Cleveland–Lloyd Dinosaur Quarry


Joshua B. Smith

Cloverly Formation


W. Desmond Maxwell



John R. Hutchinson and Kevin Padian




Michael J. Ryan and Anthony P. Russell

Cabo Espichel


Martin Lockley

Cameros Basin Megatracksite

87 90 91

John S. McIntosh

Connecticut River Valley


Constructional Morphology


David B. Weishampel


Martin Lockley

Carnegie Museum of Natural History


Joanna Wright

Kenneth Carpenter


Computers and Related Technology Ralph E. Chapman and David B. Weishampel

Clive Coy

Can˜on City


Brent H. Breithaupt

Jose´ J. Moratalla and Jose´ L. Sanz

Canadian Dinosaurs

Como Bluff



Karen Chin


Craniofacial Air Sinus Systems Lawrence M. Witmer


Contents Cretaceous Period

vii 159

Eva B. Koppelhus



James M. Clark

Crystal Palace

Distribution and Diversity


Peter Dodson

Djadokhta Formation


Tom Jerzykiewicz


William A. S. Sarjeant

Dockum Group


Philip A. Murry and Robert A. Long



Philip J. Currie

D Dalton Wells Quarry

Dry Mesa Quarry 165

Brooks B. Britt and Kenneth L. Stadtman

Deccan Basalt


Brooks B. Britt and Brian D. Curtice



Michael J. Ryan


Ashok Sahni





Edmonton Group

The Editors

Denver Museum of Natural History Kenneth Carpenter

Devil’s Coulee Dinosaur Egg Site

David A. Eberth


Clive Coy


169 175 179 179 184 184 185

Martin Lockley


Erenhot Dinosaur Museum




European Dinosaurs




J. David Archibald


Martin Lockley



Eric Buffetaut

Don Lessem

Dinosaur Valley


The Editors

Martin Lockley

Dinosaur Society


Philip J. Currie

Clive Coy

Dinosaur Ridge


The Editors

Andrea B. Arcucci

Dinosaur Provincial Park

Eggs, Eggshells, and Nests

Philip J. Currie

Daniel J. Chure



Konstantin E. Mikhailov

Kevin Padian

Dinosaur National Monument

Egg Mountain John R. Horner

Michael J. Ryan and Matthew K. Vickaryous

Dinosauria: Definition


Extinction, Cretaceous


J. David Archibald


Extinction, Triassic Michael J. Benton





Graduate Studies


Philip J. Currie



Peng Guangzhao



David J. Varricchio


Martin Lockley

Feathered Dinosaurs

Growth and Embryology Growth Lines


Kevin Padian


Philip J. Currie

Footprints and Trackways



Martin Lockley

Forelimbs and Hands


Per Christiansen


Fruita Paleontological Area

Catherine A. Forster


Hell Creek Flora

James I. Kirkland

Fruitland Formation


Hell Creek Formation

Functional Morphology

Donald F. Lofgren



Kirk R. Johnson

Michael J. Ryan


Xiao-Chun Wu and Anthony P. Russell


302 303

Fernando E. Novas


G Gastralia

Heterodontosauridae 269 270



The Editors



J. Michael Parrish

Geologic Time


Carl C. Swisher

Ghost Ranch

Dale A. Winkler

Robin E. H. Reid

History of Dinosaur Discoveries: Early Discoveries


William A. S. Sarjeant

History of Dinosaur Discoveries: First Golden Period History of Dinosaur Discoveries: Quiet Times


Eric Buffetaut


History of Dinosaur Discoveries: Research Today

Kevin Padian

Glen Rose, Texas



Brent H. Breithaupt


Clive Coy

Glen Canyon Group


Per Christiansen

Histology of Bones and Teeth

Philip J. Currie


Joshua B. Smith

Hindlimbs and Feet

Leon Claessens



John A. Long and Kenneth J. McNamara

Louie Psihoyos



Contents Howe Quarry

ix 355



Keuper Formation

Brent H. Breithaupt



Hartmut Haubold

Hans-Dieter Sues



Martin Lockley

Kirtland Formation

I Iguanodontidae


Catherine A. Forster

Indian Dinosaurs



Ashok Sahni

Lameta Formation

Institute de Pale´ontologie, Muse´um National d’Histoire Naturelle, Paris, France

Ashok Sahni


The Editors

Institute of Vertebrate Paleontology and Paleoanthropology, Beijing, China

393 394

Brent H. Breithaupt


Spencer G. Lucas


Las Hoyas


Bernardino P. Pe´rez-Moreno and Jose´ L. Sanz


Dale A. Russell

Ischigualasto Formation

Lance Formation Land-Mammal Ages

Dong Zhiming



Michael J. Ryan


Raymond R. Rogers

Legislation Protecting Dinosaur Fossils


Vincent L. Santucci

Life History


David B. Weishampel



Martin Lockley

Long Necks of Sauropods

J Japanese Dinosaurs


Eberhard Frey and John Martin




Dong Zhiming

Yoichi Azuma and Yukimitsu Tomida

Judith River Wedge


David A. Eberth

Jurassic Park





Mary Schweitzer and Don Lessem

Jurassic Period Peter Dodson

Kevin Padian



x Maniraptoriformes




National Museum, Bloemfontein, South Africa

Kevin Padian

Marginocephalia The Editors



Anusuya Chinsamy

Kevin Padian

Natural History Museum, London


Paul G. Davis


Martin Lockley

Mesozoic Era

Nemegt Formation 417

Kevin Padian

Mesozoic Faunas


Mesozoic Floras


James F. Basinger

Mexican Dinosaurs


Rene Hernandez-Rivera

Microvertebrate Sites


Michael J. Ryan



Rinchen Barsbold

Morrison Formation


Kenneth Carpenter



J. Michael Parrish

Muse´e des Dinosaures, Espe´raza, Aude, France


Daniel J. Chure


Hans-Dieter Sues

O 481

Kevin Padian

Orlov Museum of Paleontology


Leonid P. Tatarinov



Matthew K. Vickaryous and Michael J. Ryan



Kevin Padian


Kevin Padian




The Editors






Kevin Padian


Rinchen Barsbold

Mary Schweitzer

Museums and Displays

Newark Supergroup

Halszka Osmo´lska


David K. Smith

Museum of the Rockies


The Editors


The Editors

Museum of Earth Science, Brigham Young University



Jean Le Loeuff

Museum of Comparative Zoology, Harvard University


Peter Dodson

Origin of Dinosaurs 444

Gregory S. Paul

Mongolian Dinosaurs



Judith A. Schiebout

Middle Asian Dinosaurs


Halszka Osmo´lska


Hans-Dieter Sues



Oxford Clay Joshua B. Smith




P Pachycephalosauria

Polish–Mongolian Paleontological Expeditions 511

Hans-Dieter Sues

Paleoclimatology Paleoecology

513 515

J. F. Lerbekmo



Emily A. Buchholtz

Paleontological Museum, Ulaan Baatar




Paul G. Davis



Pectoral Girdle

Problems with the Fossil Record


Pseudofossils Pseudosuchia


Clive Trueman

Petrified Forest

613 617

Kevin Padian

Purgatoire 542


Paul Sereno

Pelvis, Comparative Anatomy



Kevin Padian

Kevin Padian



Paul G. Davis

Pterosauria 536


Paul Upchurch

Kevin Padian

Diego Rasskin-Gutman


Ralph E. Molnar


Darren H. Tanke and Bruce M. Rothschild


Brooks B. Britt


Philip J. Currie


Peter Makovicky

Postcranial Pneumaticity 520


Donald F. Glut

Postcranial Axial Skeleton

Peter Dodson

Paleomagnetic Correlation

Halszka Osmo´lska

Popular Culture, Literature

Hartmut Haubold



Martin Lockley

J. Michael Parrish

Phylogenetic System


Kevin Padian

Phylogeny of Dinosaurs

546 Quadrupedality

Kevin Padian





Kevin Padian

Kevin Padian

Plants and Dinosaurs


Bruce H. Tiffney

Plate Tectonics


Ella Hoch

Polar Dinosaurs Thomas H. Rich, Roland A. Gangloff, and William R. Hammer


R Radiometric Dating


David A. Eberth

Raptors Philip J. Currie



xii Reconstruction and Restoration


Reproductive Behavior and Rates

630 637

Michael J. Benton

Rocky Hill Dinosaur Park


Joanna Wright

Royal Ontario Museum


Hans-Dieter Sues

Royal Tyrrell Museum of Palaeontology


Smithsonian Institution


Michael K. Brett-Surman

Gregory S. Paul


Skull, Comparative Anatomy The Editors

Sylvia J. Czerkas

Solnhofen Formation


Hartmut Haubold

South African Museum


Anusuya Chinsamy

South American Dinosaurs


Fernando E. Novas

Southeast Asian Dinosaurs


Eric Buffetaut


Bruce G. Naylor

Soviet–Mongolian Paleontological Expeditions


Clive Coy


S Samcheonpo

647 647

Spinosauridae and Baryonychidae


State Museum for Natural History, Stuttgart, Germany

Kevin Padian

Sauropoda John S. McIntosh


658 661






Peter M. Galton




Xiao-Chun Wu and Anthony P. Russell

Philip J. Currie

Sino–Swedish Expeditions


Angela Milner

Kevin Padian

Philip J. Currie

Sino–Soviet Expeditions


The Editors

Paul Upchurch

Sino–Canadian Dinosaur Project

Species J. David Archibald

Martin Lockley



J. David Archibald


Niall J. Mateer






Edwin H. Colbert

Size and Scaling R. McNeill Alexander

Skeletal Structures

Anthony R. Fiorillo


Kevin Padian

Skin Stephen A. Czerkas


Teeth and Jaws


P. Martin Sander


Tendaguru Gerhard Maier


Contents Tetanurae

xiii 727

John R. Hutchinson and Kevin Padian



University of California Museum of Paleontology


The Editors

Kevin Padian




Dale A. Russell


731 Variation

Philip J. Currie



Tooth Marks

738 739

Tooth Serrations in Carnivorous Dinosaurs



Paul G. Davis

Von Ebner Incremental Growth Lines

Gregory M. Erickson


Michael J. Benton

Vertebrate Paleontology

Aase R. Jacobsen

Tooth Replacement Patterns

Scott D. Sampson and Michael J. Ryan


Kenneth Carpenter



Gregory M. Erickson

William L. Abler

Tooth Wear


Anthony R. Fiorillo and David B. Weishampel

Trace Fossils


Kevin Padian

Triassic Period


J. Michael Parrish



W Wealden Group


David B. Norman

Willow Creek Anticline


John R. Horner

David J. Varricchio

Trophic Groups


James O. Farlow



David B. Weishampel

Two Medicine Formation


Y Yale Peabody Museum


Mary Ann Turner

Raymond R. Rogers



Kenneth Carpenter

Zigong Museum

U Ukhaa Tolgod Mark A. Norell

Z 790

Dong Zhiming






This Page Intentionally Left Blank


Ceratopsia Dryosauridae Euornithopoda Fabrosauridae Genasauria Hadrosauridae Heterodontosauridae Hypsilophodontidae Iguanodontidae Marginocephalia Neoceratopsia Ornithopoda Pachycephalosauria Psittacosauridae Stegosauria

Dinosaurs Around the World African Dinosaurs Asian Dinosaurs Australasian Dinosaurs Southeast Asian Dinosaurs Chinese Dinosaurs Indian Dinosaurs Japanese Dinosaurs Middle Asian Dinosaurs Mongolian Dinosaurs European Dinosaurs North American Dinosaurs American Dinosaurs Canadian Dinosaurs Mexican Dinosaurs

Pterosauria Pterosauromorpha

Polar Dinosaurs


South American Dinosaurs

Saurischia Abelisauridae Allosauroidea Arctometatarsalia Aves Avialae Bird Origins Enantiornithes Avetheropoda Bullatosauria Carnosauria Ceratosauria Coelurosauria Deinonychosauria Dromaeosauridae Elmisauridae Feathered Dinosaurs Maniraptora Maniraptoriformes Megalosaurus

Groups of Dinosaurs and Related Taxa Archosauria Thecodontia Ornithosuchia Ornithodira Pseudosuchia Crocodylia Dinosauria Dinosauromorpha Origin of Dinosaurs Phylogeny of Dinosaurs Herrerasauridae Staurikosauridae Ornithischia Ankylosauria Cerapoda


Thematic Table of Contents

xvi Neotetanurae Ornithomimosauria Oviraptorosauria Prosauropoda Raptor Sauropoda Sauropodomorpha Spinosauridae and Baryonychidae Tetanurae Therizinosauria Theropoda Thyreophora Troodontidae Tyrannosauridae Vertebrata

The Biology of Dinosaurs Age Determination Behavior Intelligence Reproductive Behavior and Rates Biogeography Biometrics Biomineralization Color Comparative Anatomy Axial Skeleton Skull, Comparative Anatomy Braincase Anatomy Craniofacial Air Sinus Systems Paleoneurology Teeth Teeth and Jaws Tooth Marks Tooth Replacement Patterns Tooth Serrations Tooth Wear Von Ebner Incremental Growth Lines Postcranial Axial Skeleton Long Necks of Sauropods Postcranial Pneumaticity Appendicular Skeleton Forelimbs and Hands Hindlimbs and Feet Postcranial Pneumaticity Pectoral Girdle Pelvis, Comparative Anatomy Gastralia

Gastroliths Histology of Bones and Teeth Growth Lines Musculature Ornamentation Paleopathology Skeletal Structures Skin Diet Coprolites Gastroliths Tooth Wear Distribution and Diversity Ecology Trophic Groups Eggs, Eggshells, and Nests Bastu´s Nesting Site Devil’s Coulee Dinosaur Egg Site Egg Mountain Evolution Extinction, Cretaceous Extinction, Triassic Genetics Speciation Species Systematics Variation Functional Morphology Biomechanics Constructional Morphology Size Size and Scaling Musculature Bipedality Quadrupedality Growth and Embryology Heterochrony Life History Locomotion Biomechanics Bipedality Musculature Quadrupedality Migration Physiology Plants and Dinosaurs Hell Creek Flora Mesozoic Flora

Thematic Table of Contents Size Size and Scaling Systematics Phylogenetic System Variation

Environments of the Past Paleoclimatology Paleoecology

Important Localities North American Sites Bone Cabin Quarry Can˜on City Cleveland-Lloyd Dinosaur Quarry Como Bluff Connecticut River Valley Dalton Wells Devil’s Coulee Dinosaur Egg Site Dinosaur Ridge Dinosaur Valley Dry Mesa Quarry Egg Mountain Fruita Paleontological Area Ghost Ranch Glen Rose, Texas Howe Quarry Judith River Wedge Purgatoire Willow Creek Anticline African Sites Tendaguru Asian Sites Bastu´s Nesting Site Bayan Mandahu Bayn Dzak Khodja-Pil-Ata Lufeng Samcheonpo Ukhaa Tolgod European Sites Cabo Espichel Cameros Basin Megatracksite Carenque Fatima Las Hoyas Lommiswil Trossingen

xvii Footprints and Trackways Dinoturbation Megatracksites Cabo Espichel Cameros Basin Megatracksite Carenque Connecticut River Valley Fatima Khodja-Pil-Ata Las Hoyas Samcheonpo Microvertebrate Sites Parks and Monuments Dinosaur National Monument Dinosaur Provincial Park Petrified Forest Rocky Hill Dinosaur State Park

Geology and Dinosaurs Formations and Groups North American Formations Late Triassic Chinle Formation Dockum Group Newark Supergroup Early Jurassic Glen Canyon Group Newark Supergroup Late Jurassic Morrison Formation Early Cretaceous Cedar Mountain Formation Cloverly Formation Late Cretaceous Edmonton Group Fruitland Formation Hell Creek Formation Judith River Wedge Kirtland Formation Lance Formation Two Medicine Formation Asian Formations Late Cretaceous Barun Goyot Formation Deccan Basalt Djadokhta Formation Lameta Formation Nemegt Formation South American Formations Late Triassic Ischigualasto Formation

Thematic Table of Contents

xviii European Formations Late Triassic Keuper Formation Late Jurassic Oxford Clay Solnhofen Formation Early Cretaceous Wealden Group Periods Mesozoic Era Cretaceous Period Jurassic Period Triassic Period Mesozoic Faunas Mesozoic Floras Biostratigraphy Fossil Record Chemical Composition of Dinosaur Fossils Biomineralization Extinction, Cretaceous Extinction, Triassic Paleomagnetic Correlation Permineralization Problems with the Fossil Record Pseudofossils Radiometric Dating Taphonomy Trace Fossils Paleontology Vertebrate Paleontology Geologic Time Plate Tectonics Land Mammal Ages

Institutions North American Institutions American Museum of Natural History Central Asiatic Expeditions Brigham Young University Carnegie Museum of Natural History Denver Museum of Natural History Dinosaur Society Museum of Comparative Zoology Museum of Earth Science Museum of the Rockies Royal Ontario Museum Royal Tyrrell Museum of Palaeontology Smithsonian Institution University of California Museum of Paleontology Yale Peabody Museum

African Institutions Albany Museum Bernard Price Institute National Museum, Bloemfontein South African Museum European Institutions Bavarian State Collection for Paleontology Crystal Palace Dinosaurs Muse´e des Dinosaures, Espe´raza Muse´um National d’Histoire Naturelle Natural History Museum, London Orlov Museum of Paleontology State Museum for Natural History, Stuttgart Asian Institutions Erenhot Dinosaur Museum Institute of Vertebrate Paleontology and Paleoanthropology Paleontological Museum, Ulaan Baatar Zigong Museum Museums and Displays

Expeditions Central Asiatic Expeditions Polish-Mongolian Paleontological Expeditions Sino-Canadian Dinosaur Project Sino-Soviet Expeditions Sino-Swedish Expeditions Soviet-Mongolian Paleontological Expedition

Research Techniques and Related Topics Computers and Related Technology Technological Advances Reconstruction and Restoration Legislation Protecting Dinosaur Fossils Jurassic Park

History of Dinosaur Discoveries Early Discoveries First Golden Period Quiet Times Research Today Popular Culture, Literature


William L. Abler

Emily A. Buchholtz

Oak Park, Illinois, USA

Wellesley College Wellesley, Massachusetts, USA

R. McNeill Alexander University of Leeds Leeds, United Kingdom

Eric Buffetaut University of Paris VI Paris, France

J. David Archibald San Diego State University San Diego, California, USA

Kenneth Carpenter Denver Museum of Natural History Denver, Colorado, USA

Andrea B. Arcucci Museo de Paleontologica Universidad Nacional de La Rioja La Rioja, Argentina

Ralph E. Chapman National Museum of Natural History Smithsonian Institution Washington, DC, USA

Yoichi Azuma Fukui Prefectural Museum Fukui City, Japan

Luis M. Chiappe American Museum of Natural History New York, New York, USA

Rinchen Barsbold Geology Institute Academy of Sciences of Mongolia Ulaan Baatar, Mongolia

Karen Chin

James F. Basinger

Anusuya Chinsamy

University of Saskatchewan Saskatoon, Saskatchewan, Canada

South African Museum Cape Town, South Africa

Michael J. Benton

Per Christiansen

University of Bristol Bristol, United Kingdom

University of Copenhagen Copenhagen, Denmark

Herve´ Bocherens

Daniel J. Chure

Universite´ Pierre et Marie Curie Paris, France

Dinosaur National Monument Jensen, Utah, USA

Brent H. Breithaupt

Leon Claessens

University of Wyoming Laramie, Wyoming, USA

Utrecht University Utrecht, The Netherlands

Michael K. Brett-Surman

James M. Clark

George Washington University Washington, DC, USA

George Washington University Washington, DC, USA

Brooks B. Britt

Edwin H. Colbert

Museum of Western Colorado Grand Junction, Colorado, USA

Museum of Northern Arizona Flagstaff, Arizona, USA

University of California, Santa Barbara Santa Barbara, California, USA



xx Clive Coy

Donald F. Glut

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

The Dinosaur Society Burbank, California, USA

Philip J. Currie

William R. Hammer

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Augustana College Rock Island, Illinois, USA

Brian D. Curtice

Hartmut Haubold

Earth Science Museum Brigham Young University Provo, Utah, USA

Martin Luther Universita¨t Halle, Germany

Stephen A. Czerkas

Instituto de Geologı´a UNAM Del. Coyoacan, Mexico

The Dinosaur Museum Monticello, Utah, USA

Sylvia J. Czerkas

Rene Hernandez-Rivera

Ella Hoch

The Dinosaur Museum Monticello, Utah, USA

Geological Museum University of Copenhagen Copenhagen, Denmark

Paul G. Davis

John R. Horner

National Science Museum Tokyo, Japan

Museum of the Rockies Montana State University Bozeman, Montana, USA

Lowell Dingus American Museum of Natural History New York, New York, USA

Peter Dodson University of Pennsylvania Philadelphia, Pennsylvania, USA

John R. Hutchinson Museum of Paleontology University of California, Berkeley Berkeley, California, USA

Louis L. Jacobs

Institute of Vertebrate Paleontology and Paleoanthropology Beijing, China

Shuler Museum of Paleontology Southern Methodist University Dallas, Texas, USA

David A. Eberth

Aase R. Jacobsen

Royal Tyrrell Museum of Paleaontology Drumheller, Alberta, Canada

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Gregory M. Erickson

Tom Jerzykiewicz

University of California, Berkeley Berkeley, California, USA

Geological Survey of Canada Calgary, Alberta, Canada

James O. Farlow

Kirk R. Johnson

Indiana University at Fort Wayne Fort Wayne, Indiana, USA

Denver Museum of Natural History Denver, Colorado, USA

Anthony R. Fiorillo

James I. Kirkland

Dallas Museum of Natural History Dallas, Texas, USA

Dinamation International Society Fruita, Colorado, USA

Catherine A. Forster

Eva B. Koppelhus

State University of New York at Stony Brook Stony Brook, Long Island, New York, USA

Geological Survey of Denmark and Greenland Copenhagen, Denmark

Eberhard Frey

Jean Le Loeuff

Staatliches Museum fu¨r Naturkunde Karlsruhe, Germany

Director, Muse´e des Dinosaures Espe´raza, Aude, France

Peter M. Galton

J. F. Lerbekmo

University of Bridgeport Bridgeport, Connecticut, USA

University of Alberta Edmonton, Alberta, Canada

Roland A. Gangloff

Don Lessem

University of Alaska Museum Fairbanks, Alaska, USA

Co-founder, The Dinosaur Society Waban, Massachusetts, USA

Dong Zhiming



Martin Lockley

Mark A. Norell

University of Colorado Denver, Colorado, USA

American Museum of Natural History New York, New York, USA

Donald F. Lofgren

David B. Norman

Raymond M. Alf Museum Claremont, California, USA

Sedgwick Museum Cambridge University Cambridge, England, UK

John A. Long Western Australian Museum Perth, Australia

Fernando E. Novas

Robert A. Long

Museo Argentino de Ciencias Naturales Buenos Aires, Argentina

Pleasanton, California, USA

Halszka Osmo´lska

Spencer G. Lucas

Institute of Paleobiology Warsaw, Poland

New Mexico Museum of Natural History Albuquerque, New Mexico, USA

Gerhard Maier Calgary, Alberta, Canada

Peter Makovicky University of Copenhagen Copenhagen, Denmark

John Martin

Kevin Padian Museum of Paleontology University of California, Berkeley Berkeley, California, USA

J. Michael Parrish Northern Illinois University DeKalb, Illinois, USA

Leicestershire Museum Leicestershire, United Kingdom

Gregory S. Paul

Niall J. Mateer

Peng Guangzhao

University of California, Berkeley Berkeley, California, USA

Zigong Dinosaur Museum Zigong, Sichuan, China

W. Desmond Maxwell

Bernardino P. Pe´rez-Moreno

New York College of Osteopathic Medicine Old Westbury, New York, USA

Universidad Auto´noma de Madrid Madrid, Spain

John S. McIntosh

Louie Psihoyos

Wesleyan University Middletown, Connecticut, USA

Boulder, Colorado, USA

Kenneth J. McNamara Western Australian Museum Perth, Australia

National Museum of Natural History Smithsonian Institution Washington, DC, USA

Konstantin E. Mikhailov

Robin E. H. Reid

Russian Academy of Sciences Moscow, Russia

Chesham Bois Amersham, United Kingdom

Angela Milner

Thomas H. Rich

Natural History Museum London, United Kingdom

Museum of Victoria Victoria, Australia

Ralph E. Molnar

Raymond R. Rogers

Queensland Museum Fortitude Valley, Queensland, Australia

Cornell College Mount Vernon, Iowa, USA

Jose´ J. Moratalla

Bruce M. Rothschild

Universidad Auto´noma de Madrid Madrid, Spain

Arthritis Center of Northeast Ohio Youngstown, Ohio, USA

Philip A. Murry

Timothy Rowe

Tarleton State University Stephenville, Texas, USA

University of Texas Austin, Texas, USA

Bruce G. Naylor

Anthony P. Russell

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

University of Calgary Calgary, Alberta, Canada

Baltimore, Maryland, USA

Diego Rasskin-Gutman


xxii Dale A. Russell

Carl C. Swisher

North Carolina State University Raleigh, North Carolina, USA

Berkeley Geochronological Institute Berkeley, California, USA

Michael J. Ryan

Darren H. Tanke

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Ashok Sahni

Leonid P. Tatarinov

Punjab University Chandigarh, India

Russian Academy of Sciences Moscow, Russia

Scott D. Sampson

Bruce H. Tiffney

New York College of Osteopathic Medicine Old Westbury, New York, USA

University of California, Santa Barbara Santa Barbara, California, USA

P. Martin Sander

Yukimitsu Tomida

Institute of Paleontology Universita¨t Bonn Bonn, Germany

National Science Museum Tokyo, Japan

Vincent L. Santucci

University of Bristol Bristol, United Kingdom

Slippery Rock University Slippery Rock, Pennsylvania, USA

Clive Trueman

Mary Ann Turner

Universidad Auto´noma de Madrid Madrid, Spain

Peabody Museum of Natural History Yale University New Haven, Connecticut, USA

William A. S. Sarjeant

Ron Tykoski

University of Saskatchewan Saskatoon, Saskatchewan, Canada

University of Texas Austin, Texas, USA

Judith A. Schiebout

Paul Upchurch

Museum of Natural Science Louisiana State University Baton Rouge, Louisiana, USA

Cambridge University Cambridge, United Kingdom

Mary Schweitzer

Old Trail Museum Choteau, Montana, USA

Jose´ L. Sanz

Museum of the Rockies Montana State University Bozeman, Montana, USA

Paul Sereno University of Chicago Chicago, Illinois, USA

David K. Smith Pima Community College Tucson, Arizona, USA

Joshua B. Smith University of Pennsylvania Philadelphia, Pennsylvania, USA

Kenneth L. Stadtman Earth Science Museum Brigham Young University Provo, Utah, USA

Hans-Dieter Sues Royal Ontario Museum and University of Toronto Toronto, Ontario, Canada

David J. Varricchio

Matthew K. Vickaryous Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

David B. Weishampel Johns Hopkins University School of Medicine Baltimore, Maryland, USA

Dale A. Winkler Shuler Museum of Paleontology Southern Methodist University Dallas, Texas, USA

Lawrence M. Witmer Ohio University Athens, Ohio, USA

Joanna Wright University of Bristol Bristol, United Kingdom

Xiao-Chun Wu University of Calgary Calgary, Alberta, Canada



he Encyclopedia of Dinosaurs is a complete source of information on the subject of dinosaurs, contained within the covers of a single volume. Each article in the Encyclopedia provides an overview of the selected topic to inform a broad spectrum of readers, from researchers to the interested general public. In order that you, the reader, will derive the maximum benefit from the Encyclopedia of Dinosaurs, we have provided this Guide. It explains how the book is organized and how the information within it can be located.

Subject Areas The Encyclopedia of Dinosaurs presents 275 separate articles on the whole range of dinosaur study. It includes information not only on the organisms themselves, of course, but also on all other aspects of this field. The articles in the Encyclopedia of Dinosaurs fall within nine general subject areas, as follows: ● ● ● ● ● ● ● ● ●

Kinds of Dinosaurs Around the World Groups of Dinosaurs and Related Taxa The Biology of Dinosaurs Environments of the Past Important Dinosaur Localities Geology and Dinosaurs Institutions of Dinosaur Study Dinosaur Expeditions Dinosaur Research and Techniques

placed according to subject area and in relation to other topics.

Organization The Encyclopedia of Dinosaurs is organized to provide the maximum ease of use for its readers. All of the articles are arranged in a single alphabetical sequence by title, from ‘‘A’’ (Abelisauridae, African Dinosaurs, etc.) to ‘‘Z’’ (Zigong Museum). An alphabetical Table of Contents for the articles can be found on p. v of this introductory section. As a reader of the Encyclopedia, you can use this alphabetical Table of Contents by itself to locate a topic. Or you can first identify the topic in the Thematic Table of Contents and then go to the alphabetical Table to find the page location. So that they can be more easily identified, article titles begin with the key word or phrase indicating the topic, with any descriptive terms following this. For example, ‘‘Pelvis, Comparative Anatomy’’ is the title assigned to this article rather than ‘‘Comparative Anatomy of the Pelvis,’’ because the specific term Pelvis is the key word.

Article Format

A Thematic Table of Contents appears in the introductory section of the Encyclopedia on page xv. It has a complete list of all the articles in the book,

Articles in the Encyclopedia of Dinosaurs are divided into three general categories. The first category includes concise entries that deal with highly focused topics, such as ‘‘Albany Museum,’’ ‘‘Can˜on City,’’ ‘‘Dinoturbation,’’ ‘‘Gastralia,’’ and ‘‘Tooth Marks.’’ These entries vary in length from one to several paragraphs. The second category, the bulk of the text, includes entries of the standard length of 1 to 4 pages,



xxiv such as ‘‘African Dinosaurs,’’ ‘‘Djadokhta Formation,’’ ‘‘Dry Mesa Quarry,’’ ‘‘Egg Mountain,’’ ‘‘Musculature,’’ ‘‘Paleoclimatology,’’ and ‘‘Trace Fossils.’’ Major articles constitute the third category. They deal with broad areas of dinosaur study, such as ‘‘Behavior,’’ ‘‘Bird Origins,’’ ‘‘Evolution,’’ and ‘‘Systematics.’’ These articles are of extended length, typically 5 to 15 pages, and are identified as to general theme by the following system of symbols:

Encyclopedia. (This reference appears here because the Museum was the sponsor of these expeditions.) An example of the third type, a cross-reference at the end of the article, can be found in the entry ‘‘Distribution and Diversity.’’ This article concludes with the statement: See also the following related entries: BIOGEOGRAPHY ● MIGRATION ● PLATE TECTONICS This reference indicates that these three articles all provide some additional information about the distribution and diversity of dinosaurs.

Kinds of Dinosaurs

Dinosaur Biology

Geology & Sites

Research Methods

Dinosaur Institutions

Articles in all three categories have been written by individual contributors, as indicated in the article heading. The only exception is that some shorter entries are unsigned. These have been prepared collectively by the general editors, Drs. Currie and Padian.

Cross-References The Encyclopedia of Dinosaurs has an extensive system of cross-referencing. Cross-references to other articles appear in three forms: as marginal headings within the A-to-Z article sequence; as designations within the running text of an article; and as indications of related topics at the end of an article. As an example of the first type of reference cited above, the following marginal entry appears in the A-to-Z article list between the entries ‘‘Archosauria’’ and ‘‘Arctometatarsalia:’’ Arctic Dinosaurs see POLAR DINOSAURS This reference indicates that the topic of dinosaurs of the Arctic region is discussed elsewhere, under the article title ‘‘Polar Dinosaurs.’’ An example of the second type, a cross-reference within the running text of an article, is this excerpt from the entry ‘‘Central Asiatic Expeditions’’: Discovery of non-avian dinosaurs in Central Asia was pioneered by the AMERICAN MUSEUM OF NATURAL HISTORY (AMNH) during a series of swashbuckling Central Asiatic Expeditions in the 1920s.

This indicates that the item ‘‘American Museum of Natural History,’’ which is set off in the text by small capital letters, appears as a separate article within the

Bibliography The bibliography section appears as the last element in an article, under the heading ‘‘References.’’ This section lists recent secondary sources that will aid the reader in locating more detailed or technical information. Review articles and research papers that are important to a more detailed understanding of the topic are also listed here. The bibliographic entries in this Encyclopedia are for the benefit of the reader, to provide references for further reading or research on the given topic. Thus they typically consist of a limited number of entries. They are not intended to represent a complete listing of all materials consulted by the author or authors in preparing the article.

Index The Subject Index for the Encyclopedia of Dinosaurs contains more than 5,840 entries. Within the entry for a given topic, references to general coverage of the topic appear first, such as a complete article on the subject. References to more specific aspects of the topic then appear below this in an indented list.

Encyclopedia Website The Encyclopedia of Dinosaurs maintains an editorial Web Page on the Internet at: http://www.apnet.com/dinosaur/ The Encyclopedia of Dinosaurs site provides information about this project and links to many related sites that feature dinosaurs. The site will continue to evolve as more information becomes available.



e human beings are fascinated by dinosaurs. The first reports of the giant bones of extinct creatures caused a worldwide sensation, and in the century and a half since then, our interest has never diminished. In every country of the world, children and adults are entranced by dinosaurs. It is often imagined that the current dinosaur mania is a recent phenomenon; in fact it is not. When I first started writing a novel about dinosaurs, back in 1981, I put the project aside because at that time, Americans seemed to be in the grip of an unprecedented dinosaur mania. There were dinosaur cups and saucers, dinosaur toys, dinosaur bedspreads; museums were having dinosaur shows; it seemed there were dinosaurs everywhere you looked. I did not want to write a book that exploited a fashion of the moment. So I waited. But year after year, the fashion never went away. Finally I realized that our fascination with dinosaurs is a permanent phenomenon. It is always there. Children, of course, have always been captivated by dinosaurs. To go to a museum and see young children, barely able to walk and talk, shrieking ‘‘Stegosaurus’’ and ‘‘Tyrannosaurus’’ as they view the creatures is a very striking thing. Why does it happen? What is going on in their minds as they shout out those complex Latin names? How do we explain the fact that dinosaurs excite the imagination of adults and children throughout the world? Over the years, I have entertained many theories. For a while, I thought the phenomenon might be characteristic of those countries, like the United States, where many fossils have been found—a kind of nationalistic interest, if you will. But dinosaurs are just as popular in countries

such as Japan and Italy, where few remains have been found. For a while, I thought it was primarily a childish interest. But in museums, you’ll notice that adults are equally fascinated. To be honest, it often seems that children are only an excuse for adults to visit the dinosaurs. Later on, I suspected the interest in dinosaurs might be something that children passed on to each other, a trait of a children’s subculture. But my own daughter showed a marked interest in dinosaurs long before she went to preschool—indeed, before she was even very verbal. Still later, I thought this enthusiasm was provoked by the great size of these creatures. But smaller dinosaurs excite just as much interest among children. Baby dinosaurs are very appealing. And in any case, the dinosaur toys are all small. . . . For a time, I wondered whether the interest had something to do with the fact that the dinosaurs had become extinct. But children are not clear about this. When my daughter was two years old, she asked to see dinosaurs at the zoo. She had been to the zoo several times, and apparently believed the dinosaurs were housed in some section we hadn’t visited yet. When she was told that she could not see dinosaurs, she gave a resigned shrug—parents never do what you want them to do! Perhaps, I thought, that was a clue. Children spend much of their lives powerless and frustrated. I began to entertain a Freudian notion that being able to pronounce the complex names of huge creatures afforded children a sense of control. In a child’s world, after


xxvi all, everything is big—parents, cars, everything. And naming things is a classic human procedure to reduce anxiety. (Patients are always relieved to hear that they have ‘‘idiopathic hypertension,’’ even though the term is literally meaningless.) But once again, careful observation cast doubt on this idea. When my daughter was four, I took her and two friends to Stan Winston’s workshop to see the dinosaurs being constructed for Jurassic Park. I thought they’d enjoy it, but they didn’t. Although the dinosaurs were then only sculpted in clay, the girls were distressed by what they saw. The animals were simply too big, and too real-looking. It is one thing to play with little dinosaur toys. It is quite another to walk beneath the enormous scaly legs of a towering tyrannosaur, or to touch the big claw of a Velociraptor. The kids were very uneasy. They wanted to leave. So in the end, I decided I just don’t understand why

Foreword children are fascinated by dinosaurs. And I don’t believe anybody understands. In the end, it is a mystery. And it may be that the mystery is part of the fascination. Certainly for adults, dinosaurs present an intriguing puzzle, in which fantasies are inevitably provoked. Although we know far more about dinosaurs than we did a few decades ago, the truth is that we still know very little. We don’t really know what these creatures looked like, or how they behaved. We have some bones, impressions of skin, some trackways, and many fascinating speculations about their biology and social organization. But what hard evidence remains of their long-vanished world is tantalizing and incomplete. And so they still provoke our dreams. And, probably, they always will. Michael Crichton



s it possible that the scientific understanding of dinosaurs has now come so far that a volume such as this, as big as the Manhattan telephone directory, is necessary to compile even a synopsis of what we know about them? Could Dr. Gideon Mantell and Dean William Buckland, when they described the first remains of what would become known as dinosaurs, have ever imagined the scientific attention that would be paid to them more than 170 years later? Could Richard Owen, who gave Dinosauria its name in 1842, have intended that his ‘‘fearfully great lizards’’ would be used as metaphors for both evolutionary success and obsolescence? As we sit today on three decades of the most awesome explosion of knowledge about dinosaurs in history, can we imagine how much more will be learned before most of us now active in the field reach retirement age? The answers to these questions are probably yes, no, yes, and no, perhaps in no particular order. This is not the first phonebook-sized compendium of dinosaur information, and we wish at the outset to acknowledge both our debt to and admiration for the already classic work, The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, eds.; University of California Press, 1990), which organizes so much of what is known in such an accessible way. This volume should be seen in many respects as a companion to that one, which was dedicated to exploring the individual dinosaurian groups in considerable depth. Our focus in this Encyclopedia is to provide background and a point of reference to the recent literature on dinosaurian subjects in general. The two books can be read in similar and completely different ways; in the

organization of the Encyclopedia we have tried to foresee the uses that it may serve for its audience. It is unlikely, of course, that Mantell and Buckland could have foreseen much of what has transpired around their Iguanodon and Megalosaurus, and hundreds of their reptilian kin, so far into the future; indeed, in the 1820s these men had themselves only the scarcest idea of the nature and importance of the fossil bones they were describing. As for Richard Owen, there was little that escaped his eye, and he was always looking to posterity. In his own time he regarded his creations as approaching the great mammals in physiological sophistication; yet, like all reptilian forms of the Secondary Era (Mesozoic), they were doomed to extinction and replacement. But even he, like us toiling in the fields today, can have no idea of what will come next. The possibilities seem almost limitless. A few words about what this book is and is not. (This part seems to be read only infrequently by reviewers, but we hope it will help the general reader.) An encyclopedia is designed to be a concise summary of knowledge, ideas, historical background, and current thinking on a general topic. The information for a given entry is not exhaustive; rather, through cross-references and citations to other literature, readers are invited to learn more and explore wider resources. Consequently, the name of every dinosaur, geological formation, quarry, museum, and idea about how dinosaurs lived and died will not be found under its own entry. Many of these names may be found in the indexes at the end of the book, subsumed under other entries, and still others will be gleaned by surfing the cross-references as one would on the Internet. We have assembled a most knowledgeable contingent of experts from all over the globe on various subjects, and we hope that they,


xxviii and you, enjoy and learn as much from perusing this book as we did in editing it. For more exhaustive listings of dinosaurian names and histories, a good start can be made in The Dinosauria or in Don Glut’s Dinosaurs, the Encyclopedia (McFarland & Company, Inc., 1997). For an entry into the primary literature about dinosaurs and all extinct backboned things, there is no finer source than the many volumes of the Bibliography of Fossil Vertebrates, produced over the years by A. S. Romer, Charles L. Camp, Joseph T. Gregory, Judy Bacskai, George Shkurkin, and a host of colleagues, under the most recent auspices of the American Geological Institute and the Society of Vertebrate Paleontology. Recent textbook-style works about dinosaurs include D. B. Norman’s The Illustrated Encyclopedia of Dinosaurs (Salamander, 1985), Spencer G. Lucas’s Dinosaurs: The Textbook (William C. Brown, 1996), and D. E. Fastovsky and D. B. Weishampel’s The Evolution and Extinction of the Dinosaurs (Cambridge University Press, 1996). Finally, for younger dinosaur enthusiasts, The Dinosaur Society (East Islip, New York) works with both professional paleontologists and educators to provide the best of what is new and what is known to the many children and adults who want to learn more. The taxonomic conventions of this book do not follow the venerable Linnean System, in place since 1763 (well before evolution was a mainstream scientific concept), but the newer Phylogenetic System, based on phylogenetic systematics or cladistics. The principles of this system are explained in this work and in others referenced herein, and need not be detailed at this point. Cladistic conventions hold that all taxa (named groups of organisms) must include a common ancestor and all its descendants in theory (i.e., monophyletic groups). All taxa must have both a definition of their ancestry and membership and a diagnosis of the uniquely shared evolutionary features by which they may be recognized. For the sake of stability, we have restricted definitions to node-based and stem-based kinds, as explained in the entry ‘‘Phylogenetic System.’’ We have tried to provide definitions for every taxon (stem- and node-based) and diagnoses for every node-based taxon in this book, though many will change with future research. A data

Preface matrix of the characteristics and taxa used in phylogenetic analyses is a sine qua non of formal systematic research, and it was our initial hope to include matrices in this work; however, the constant revision and expansion of these matrices would soon outdate any printed effort, and we are now more hopeful that these may be made available and updated on CD-ROM or in World Wide Web format in the future. Controversies are the business of science, which thrives by expanding, testing, or overturning what we think we already know. And dinosaurs are no strangers to debate; many questions from phylogeny to physiology have strong cases for divergent conclusions. In this book we present these controversies as they are seen today. Many are new, some are old; some may well be resolved with further work, and some may never be. The viewpoints of the authors of individual entries may not always coincide; the authors were not recruited because they agreed with each other’s conclusions, but because they had something intelligent to say. Of course, not all apparent controversies are real; some have been settled, at least to the satisfaction of the paleontological community’s consensus, and excessive attention is not paid to these here. Many apparent controversies regarding dinosaurs live more in the minds of the representatives of the press than in those of the scientists. Finally, we express our appreciation to our editorial crew at Academic Press, including Gail Rice, Chris Morris, and especially Chuck Crumly, who remained the driving force behind this volume’s realization; to Eva Koppelhus, John Hutchinson, and Leakena Au, for pragmatic assistance beyond the call of duty; to the staff of the Tyrrell Museum of Palaeontology, particularly Pat Bobra, and the support of the University of California Museum of Paleontology; to the various individuals, journal editors, publishers, and copyright holders who allowed us to reprint many of the wonderful illustrations in this book; and especially to all our authors, who met deadlines, extraordinary requests, and last-minute pleas with patience and cooperation. To everyone, our best thanks. Philip Currie and Kevin Padian Drumheller, Alberta, and Berkeley, California


To John H. Ostrom


n a career spanning some forty years in vertebrate paleontology, John Ostrom has worked on so many groups of dinosaurs, and on so many problems relating to the paleobiology of dinosaurs, that he has become the central figure in dinosaur research since the mid-1960s. Originally headed for a career in medicine, John became sidetracked under the spell of George Gaylord Simpson and Ned Colbert, from whom he learned vertebrate paleontology at the American Museum of Natural History while a student at Columbia University in the 1950s. He produced several studies of living and extinct amphibians and reptiles, and finally settled on a Ph.D. thesis project studying the skulls of North American hadrosaurs, mainly using the spectacular collections of the AMNH. It was not long before he challenged prevailing ideas about the paleoecology of hadrosaurs, too, suggesting that they were not aquatic but terrestrial. This approach of using anatomy and functional morphology to ask broader questions about paleobiology and behavior would become a hallmark of John Ostrom’s work. After a brief stint teaching at Beloit College, John joined the faculty at Yale, where he has spent the rest of his career. Undoubtedly his most important contributions to the collections of the Yale Peabody Museum were those from the Cloverly Formation of Wyoming, which included various remains of ornithopods, ankylosaurs, and particularly an ‘‘unusual’’ theropod dinosaur that John christened Deinonychus, or ‘‘terrible claw,’’ in 1969. This beast was to change

our concept of dinosaurs in more ways than one. John brought to life an animal that stalked its prey with jaws full of large teeth, long arms with prehensile hands and sharp talons, and a foot that bore a huge curved claw on its second toe, which could not have been used in walking. Deinonychus captured the imagination of dinosaur fans everywhere, notably in Michael Crichton’s Jurassic Park. But that was only a small part of the maelstrom of interest in dinosaurs that grew from John’s work. In 1970 John attended the First North American Paleontological Convention and presented an innocent-sounding paper called ‘‘Terrestrial Vertebrates as Indicators of Mesozoic Climates.’’ In it, he maintained that the zoogeography and behavior of living reptiles were inappropriately stereotyped, and that Mesozoic dinosaurs were at least as widespread and ecologically varied. His ideas on the subject were amplified and studied further by his student Robert Bakker, who was largely responsible for popularizing the ‘‘renaissance’’ of ideas about dinosaurian warmbloodedness. These debates thrived for a decade, and still survive today in modified form. Perhaps John’s greatest contribution to dinosaurs was to recognize that a whole group of them was largely unrecognized: namely, birds. In 1973 John first advanced a synopsis of the evidence that birds had descended from small coelurosaurian dinosaurs. He had gotten this idea (which can be traced back to T. H. Huxley, but was long abandoned) while in Germany studying pterosaurs in the company of his old friend Peter Wellnhofer. John had traveled to Haarlem, in The Netherlands, to look at their relatively small collection of Solnhofen pterosaurs, when he found that one specimen, known only from a hindlimb, had a very dinosaur-like foot. When he looked


xxx more closely, he also saw impressions of feathers. This was obviously not a pterosaur, but an Archaeopteryx. As John tells the story, he was faced with a dilemma. Should he ask to borrow the specimen, bring it home, and then ‘‘suddenly’’ discover its true identity? Or should he come clean from the start, risking the loss of ever seeing it again once its value became apparent? He took the latter course. The old, whitehaired curator was nonplussed; he gasped and wiped his brow. ‘‘You have made our museum famous,’’ he managed to say gratefully, and walked away with the newly precious relic. John forlornly began to gather his things, and was about to leave when the curator reappeared—with the world’s newest Archaeopteryx specimen wrapped in a shoebox, tied with string. He handed John the box, and history took its course. The Archaeopteryx specimen at Teyler, like the others, had reminded John not only of dinosaurs, but of theropod dinosaurs—particularly of the ones he had been recently studying at Yale, such as Deinonychus and Ornitholestes, and Compsognathus in Munich. Detailed comparisons of these animals, as well as living birds and more distantly related archosaurs, resulted in a series of papers in which John established that birds evolved from small coelurosaurs, probably sometime in the Middle or Late Jurassic. Controversy has flared around this issue intermittently for nearly 25 years, yet it is one of the more firmly established

Dedication hypotheses in vertebrate history. But John was not content with the origin of birds; he wanted to explore the origin of their flight. To him, the terrestrial, predatory habits of the theropod relatives of birds suggested their origin from the ground up, running and flapping, perhaps initially after small insects. John’s heuristic model of Archaeopteryx, using its wings as flyswatters, attracted support, intrigue, and brickbats, but stimulated a look into this question from many disciplines and a great deal of further research on the origin of major evolutionary adaptations. Perhaps one of the reasons for John’s pervasive influence, both in the professional field and among interested laymen, is the clarity and simplicity of his writing style. Had he chosen to obfuscate his ideas in a mountain of impenetrable scientific prose, they would not have gotten much of the attention they have. But he has always written directly, modestly, and accessibly, avoiding hyperbole and dogma. (How many paleontologists could have confined themselves to the understatement of calling Deinonychus ‘‘an unusual theropod’’?) He has been a model for his students and colleagues alike, and many entries in this book testify to the endurance and importance of his work and his thought. On behalf of all the authors, it is our great pleasure to dedicate the Encyclopedia of Dinosaurs to John Ostrom.

The Editors

A tures (Carnotaurus) or a dome-like prominence (Majungasaurus; Sampson et al., 1996); posterior surface of basioccipital wide and smooth below occipital condyle; dentary short, with convex ventral margin; and loose contacts among dentary, splenial, and postdentary bones. Abelisauridae seems to be related to the small theropod Noasaurus leali (Bonaparte and Powell, 1980) because both taxa share maxillae with subvertical ascending rami and cervical vertebrae with hypertrophied epipophyses and reduced neural spines (Bonaparte, 1991; Bonaparte et al., 1990; Novas, 1991, 1992). Novas (1991) coined the name Abelisauria to encompass Abelisauridae, Noasaurus, and their most recent common ancestor. Ligabueino andesi resembles abelisaurs in the morphology of the cervical vertebrae (Bonaparte, 1996). Interestingly, abelisaurids show close resemblances with the Cenomanian carcharodontosaurids Giganotosaurus carolinii and Carcharodontosaurus saharicus (Coria and Salgado, 1995; Sereno et al., 1996). These taxa exhibit many derived traits in common (e.g., antorbital fossae reduced, similar patterns of rugosities on external surfaces of nasals and maxillae, preorbital openings anteroposteriorly expanded, wide contacts between lacrimals and postorbitals forming thick ‘‘brows’’ above orbits, and eyes enclosed by sinuous orbital margins of lacrimals and postorbitals), suggesting that they are more closely related than previously thought (Coria and Salgado, 1995; Rauhut, 1994; Sereno et al., 1996). The resemblances noted among the different South American, Malagasian, and Indian taxa suggest Gondwanan origins for abelisaurs plus carcharodontosaurids, contrary to some recent proposals (e.g., Sereno et al., 1996). The phylogenetic relationships of Abelisauridae need to be studied in depth. Abelisaurus and Carnotaurus, at least, exhibit several apomorphic resemblances in the morphology of the dorsal and sacral vertebrae with the JURASSIC Ceratosaurus nasicornis,

Abelisauridae FERNANDO E. NOVAS Museo Argentino de Ciencias Naturales Buenos Aires, Argentina


constitutes a clade of CRETACEOUS theropods widely documented in the Gondwanan continents (South America, Madagascar, and India). As far as Argentina is concerned, abelisaurids are the best represented group of predatory dinosaurs from the Cretaceous deposits in both number of specimens and species diversity. Abelisaurids underwent a significant evolutionary radiation during the Cretaceous in South America, becoming active predators of large size (see SOUTH AMERICAN DINOSAURS). Abelisauridae was originally established by Bonaparte and Novas (1985) to include the Late Cretaceous Patagonian dinosaur Abelisaurus comahuensis as a mean of emphasizing the distinctness of this taxon with respect to the remaining theropods. The list of abelisaurids, however, rapidly increased: Carnotaurus sastrei (Bonaparte et al., 1990), Xenotarsosaurus bonapartei (Martı´nez et al., 1986), Indosaurus matleyi and Indosuchus raptorius (both taxa from the Lameta Group, Late Cretaceous of India; von Huene and Matley, 1933; Chatterjee, 1978; Bonaparte and Novas, 1985; Bonaparte, 1991), yet undescribed Carnotauruslike creatures (Coria and Salgado, 1993), and Majungasaurus crenatissimus (Maevarano Formation, Campanian; Madagascar; Lavocat, 1955; Bonaparte; 1991; Sampson et al., 1996) were added to the family. Abelisauridae is defined to include the previously listed taxa and all the descendants of their common ancestor. Diagnostic characters of Abelisauridae include craniocaudally short and deep premaxilla; dorsoventrally deep snout at the level of the narial openings; frontals dorsoventrally thickened resulting in a dorsal bulking (Abelisaurus), paired horn-like struc-


African Dinosaurs

2 and the taxon Neoceratosauria has been erected for this clade (Novas, 1991, 1992).


Sampson, S., Krause, D., Forster, C., and Dodson, P. (1996). Non-avian theropod dinosaurs from the Late Cretaceous of Madagascar and their paleobiogeographic implications. J. Vertebr. Paleontol. 16(3). [Abstract]

Bonaparte, J. F. (1991). The Gondwanan theropod families Abelisauridae and Noasauridae. Historical Biol. 5, 1–25.

Sereno, P., Dutheil, D., Iarochene, M., Larsson, H., Lyon, G., Magwene, P., Sidor, C., Varrichio, D., and Wilson, J. (1996). Predatory dinosaurs from the Sahara and Late Cretaceous faunal differentiation. Science 272, 986–991.

Bonaparte, J. F. (1996). Cretaceous tetrapods of Argentina. Mu¨nchner Geowissenchaftliche Abhandlungen (A) 30, 73–130.

von Huene, F., and Matley, C. (1933). The Cretaceous Saurischia and Ornithischia of the central provinces of India. Mem. Geol. Survey India 21, 1–74.

Bonaparte, J. F., and Novas, F. E. (1985). Abelisaurus comahuensis, n.g. et n.sp., Carnosauria del Creta´cico Tardı´o de Patagonia. Ameghiniana 21(2–4), 259–265. Bonaparte, J. F., and Powell, J. E. (1980). A continental assemblage of tetrapods from the Upper Cretaceous beds of El Brete, northwestern Argentina. Mem. Soc. Geol. France N. Ser. 139, 19–28. Bonaparte, J. F., Novas, F. E., and Coria, R. A. (1990). Carnotaurus sastrei Bonaparte, the horned, lightly built carnosaur from the middle Cretaceous of Patagonia. Contrib. Sci. Nat. History Museum Los Angeles County 416, 1–42. Chatterjee, S. (1978). Indosuchus and Indosaurus, Cretaceous carnosaurs from India. Journal of Paleontology 52, 570–580.

Academy of Natural Sciences, Philadelphia, Pennsylvania, USA Specimens collected by Joseph Leidy are a part of this collection, but specimens that were the basis of many of E. D. Cope’s publications are now at the American Museum of Natural History.




Coria, R. A., and Salgado, L. (1993). Un probable nuevo neoceratosauria Novas, 1989 (Theropoda) del Miembro Huincul, Formacio´n Rı´o Limay (Creta´cico presenoniano) de Neuquen. Ameghiniana 30(3), 327.

African Dinosaurs

Coria, R. A., and Salgado, L. (1995). A new giant carnivorous dinosaur from the Cretaceous of Patagonia. Nature 377, 224–226.

Southern Methodist University Dallas, Texas, USA

Lavocat, R. (1955). Sur une portion de mandibule de the´ropode provenant du Cre´tace supe´rieur de Madagascar. Bull. Musee Natl. d’Histoire Nat. Ser. 2 27, 256–259. Martı´nez, R., Ginmenez, O., Rodriguez, J., and Bochatey, G. (1986). Xenotarsosaurus bonapartei gen. et. sp. nov. (Carnosauria, Abelisauridae), un nuevo Theropoda de la Formacio´n Bajo Barreal, Chubut, Argentina. IV Congr. Argentino Paleontol. Bioestratigr. Actas 2, 23–31. Novas, F. E. (1991). Relaciones filogene´ticas de los dinosaurios tero´podos ceratosaurios. Ameghiniana 28(3/4), 410. Novas, F. E. (1992). La evolucio´n de los dinosaurios carnı´voros. In Los dinosaurios y su entorno bio´tico. Actas II Curso de Paleontologı´a en Cuenca (J. L. Sanz and A. Buscalioni, Eds.), pp. 125–163. Instituto ‘‘Juan de Valde´s,’’ Ayuntamiento de Cuenca, Espan˜a. Rauhut, O. (1994). Zur systematischen Stellung der afrikanischen Theropoden Carcharodontosaurus Stromer 1931 und Bahariasaurus Stromer 1934. Berliner Geowissenchaften Abhandlungen 16, 357–375.


The fossil record of dinosaurs in Africa extends from the Late TRIASSIC, over 200 million years ago, until the Late CRETACEOUS, presumably 65 million years ago, although the extinction event that ended the reign of dinosaurs has yet to be documented in Africa. Throughout this length of time, Africa remained relatively stable geologically, changing position only slightly by drifting and rotating northward. By contrast, Africa’s neighboring continents moved greatly, resulting in ocean barriers between what were once contiguous land masses. The changing geography of Africa and its neighbors throughout the MESOZOIC is fundamental to understanding the dinosaurs found there. During the Late Triassic through the Early JURASSIC, major continental land masses were united into the supercontinent of Pangaea. Because the land was not divided into separate continents, dinosaurs and

African Dinosaurs other animals were more or less free to expand across the entire area, not constrained by ocean barriers but rather by environmental and ecological differentiation of this large land area. Thus, the dinosaur fauna of the Late Triassic and Early Jurassic is generally similar across the globe because the globe had only one continent rather than several continents acting as separate theaters of evolution. Late Triassic dinosaur sites are found extensively in southern Africa (particularly South Africa, Lesotho, and Zimbabwe) and to a lesser extent in northern Africa (Morocco). Herbivorous prosauropods (Azendohsaurus, Blikanasaurus, Euskelosaurus, and Melanorosaurus) are the best known of African Triassic dinosaurs. Footprints and incomplete remains indicate the presence of small THEROPODS and ORNITHISCHIANS. The Triassic–Jurassic boundary is marked by extinctions globally, but the boundary has not been studied in detail in Africa. Early Jurassic localities, like those of the Late Triassic, are concentrated in southern (Lesotho, Namibia, South Africa, and Zimbabwe) and northern Africa (Algeria and Morocco). The northern record is predominantly footprints, although tracks are also found in the south. PROSAUROPODS, represented by Massospondylus, appear to be relatively abundant. Massospondylus and the ceratosaur Syntarsus are also known from North America. SAUROPODS are represented by Vulcanodon, a primitive genus known from Zimbabwe. ORNITHOPODS are small and primitive but apparently diverse, being represented by Abrictosaurus, Heterodontosaurus, Lanasaurus, Lesothosaurus, and Lycorhinus. The Middle Jurassic is poorly represented and poorly studied in Africa. Large sauropods, usually referred to Cetiosaurus, are known from Morocco and Algeria. There are few Late Jurassic localities. Theropod, sauropod, and ornithopod footprints are reported from Morocco and Niger. However, the most impressive concentration of Late Jurassic dinosaurs in Africa is TENDAGURU, Tanzania. This collection of sites was worked first by Germans, until they were disrupted by World War I, then by British. Although theropods, including the ornithomimosaur Elaphrosaurus, are present, by far the bulk of the material pertains to large sauropods (Barosaurus, Brachiosaurus, Dicraeosaurus, and Janenschia). ORNITHISCHIANS are represented by the ornithopod Dryosaurus and the stegosaur Kentrosaurus. Perhaps most surpris-

3 ing about the Tendaguru fauna is the similarity it shows to that of the Morrison Formation of North America. Madagascar has a history separate from that of Africa for the latter half of the Mesozoic Era. Bothriospondylus and Lapparentosaurus, both sauropods, are reported from the Jurassic. Those records are particularly important because Madagascar separated from Africa approximately 160–150 million years ago, at approximately the same time or slightly postdating the Jurassic fossils. They are perhaps among the last dinosaurs that could have inhabited a Madagascar connected to the African mainland. Although separated from Africa, in the Jurassic and Early Cretaceous Madagascar remained conjoined with India, and through India, to the land masses now known as Antarctica and Australia, and through Antarctica to South America. In the Early Cretaceous, Madagascar plus India separated from Australia and Antarctica. Then, in the Late Cretaceous, they separated from each other to drift to the configuration they have now achieved. This sequence of geographic events is important because it means that the biogeographic affinities of Late Cretaceous Madagascan dinosaurs may lie elsewhere than Africa. Recent work in the Late Cretaceous of Madagascar is greatly improving our knowledge of that island. Theropods are best represented by the probable abelisaurid Majungasaurus. Sauropods are best represented by a derived titanosaurid that was referred to as Titanosaurus in earlier literature, a genus first named from India. In addition to the biogeographic implications, titanosaurid remains in Madagascar include the first documented bony dermal armor from a sauropod. Of particular interest among recent finds are birds found in the quarries with the dinosaurs. Earlier studies indicated the presence of a pachycephalosaur, Majungatholus, but this animal is an ABELISAURID. Localities are more widespread across the continent during the Cretaceous Period, but with the Cretaceous lasting from 144 to 65 million years, with few radiometric dates in Africa, and with the density of localities sparse relative to the area and the time involved, it is not surprising that chronological resolution is poor. During the Early Cretaceous Africa remained connected to South America. By the end of the Early Cretaceous or in the early portion of the Late Cretaceous, Africa and South America split apart

Age Determination of Dinosaurs

4 with the completion of the South Atlantic. This was an important event because, with the completion of the Atlantic, new ocean current patterns were established, distributing heat across the globe and affecting climates. Besides the ecological changes this would bring about, the growing Atlantic formed a widening barrier, allowing the prediction that the similarity of South American and African dinosaur faunas decreases after the Early Cretaceous. Recent work suggests this may be the case. Most Early Cretaceous localities have yielded fragmentary theropod and sauropod material lacking detailed contextual data. Notable exceptions are the tetanuran Afrovenator from Niger, the primitive titanosaurid sauropod Malawisaurus from Malawi, the high-spined ornithopod Ouranosaurus from Niger, and the stegosaur Paranthodon from South Africa. The Late Cretaceous is equally in need of more field work and discovery, although numerous localities are scattered through northern Africa in particular. Notable taxa from the Late Cretaceous of Africa include the coelurosaur Deltadromeus, the tetanuran Spinosaurus, and the allosauroid Carcharodontosaurus, all from northern Africa, and Kangnasaurus, an ornithopod from southern Africa. In terms of collected and adequately described taxa, this is clearly unbalanced for both the Early and Late Cretaceous, indicating the fertile ground that Africa is for discovery. In the current geography of the earth, the Middle East is distinct from Africa. In the Mesozoic it was not. Therefore, indeterminate sauropod remains from Late Jurassic coastal deposits in Yemen and Late Cretaceous theropod footprints from Israel must be considered African. In addition, Croatian localities from a Mesozoic carbonate platform yield fragmentary bones and dinosaur footprints that may have been made on land that was originally a broad, intermittently submerged promontory of Africa or possibly a microplate that drifted northward to join Europe. Either way, the Croatian sites have great implications for the biogeography of African dinosaurs in the Cretaceous.


References Attridge, J., Crompton, A. W., and Jenkins, F. A. (1985). The southern African Liassic prosauropod Massospondylus discovered in North America. J. Vertebr. Paleontol. 5(2), 128–132. Dalla Vecchia, F. M. (1994). Jurassic and Cretaceous sauropod evidence in the Mesozoic carbonate platforms of the southern Alps and Dinarids. Gaia 10, 65–73. Jacobs, L. L., Winkler, D. A., and Gomani, E. M. (1997). Cretaceous dinosaurs of Africa: Examples from Cameroon and Malawi. In Gondwana Dinosaurs (R. Molnar and F. Novas, Eds.), in press. Memoirs of the Queensland Museum, Brisbane. Krause, D. W., Hartman, J. W., and Wells, N. A. (1997). Late Cretaceous vertebrates from Madagascar: Implications for biotic change in deep time. In Natural and Human-Induced Change in Madagascar (B. Patterson and S. Goodman, Eds.), in press. Smithsonian Institution Press, Washington, DC. Padian, K. (Ed.) (1986). The Beginning of the Age of Dinosaurs: Faunal Change across the Triassic–Jurassic Boundary, pp. 378. Cambridge Univ. Press, Cambridge, UK. Weishampel, D. B., Dodson, P., and Osmo´lska, H. (1990). The Dinosauria, pp. 733. Univ. of California Press, Berkeley.

Age Determination of Dinosaurs GREGORY M. ERICKSON University of California Berkeley, California, USA

Early attempts to estimate the longevity of dinosaurs used allometric scaling principles. Ages were determined by dividing individual mass estimates by rates of growth for extant taxa. For very large species, growth rates were extrapolated to dinosaur proportions using regression analysis. The results of these investigations have been extremely variable because they depend on mass estimates and growth rates that are highly disparate. For example, longevity estimates for the sauropod Hypselosaurus priscus range from a few decades to several hundred years (Case, 1978). Recently it has been shown that most dinosaur bones have growth lines that are visible in thin sectioned material viewed under polarized light (e.g., Reid, 1990; Fig. 1). Two types of growth lines exist: annuli and lines of arrested growth (Francillon-Viellot et al., 1990). Histological examinations have re-

Age Determination of Dinosaurs

FIGURE 1 Thin-sectioned tibia of the tyrannosaur Albertosaurus lancensis (LACM 23845) exhibiting a line of arrested growth (arrow) between zones of highly vascularized fibrolamellar bone. Scale ⫽ 1 mm.

vealed that annuli are composed of thin layers of avascular bone with parallel-aligned bone fibers. The growth line annuli are sandwiched between broad vascularized regions of bone with more randomly oriented fibrillar patterns known as zones (Fig. 1). Lines of arrested growth, like annuli, are found between zones and are avascular. They are, however, thinner and have relatively fewer bone fibers by volume (Fig. 1). Studies on extant vertebrates indicate that the vascularized zones form during moderate to rapid skeletogenesis, and that abrupt metabolic disruptions of bone formation trigger growth line deposition. Interruptions that significantly reduce bone growth cause the genesis of annuli, whereas lines of arrested growth form in response to near or complete cessations in bone formation (FrancillonViellot et al., 1990).

5 Both types of growth lines may be deposited in synchrony with endogenous biorhythms. For example, captive crocodilians exposed to constant temperature, diet, and photoperiod still exhibit the periodic and cyclical skeletal growth banding of their wild counterparts. In many extant vertebrates, including most actinopterygian fish, amphibians, lepidosaurian reptiles, and crocodilians, the growth lines have an annual periodicity of deposition (Castanet et al., 1993). Consequently, it is assumed by many paleontologists that the growth lines of dinosaurs reflect annual rhythms and that they can be used to determine individual ages. However, in the long bones of many taxa, resorption of internal and external bone proceeds even as new external cortical bone continues to be deposited, so growth lines deposited early in development may need to be inferred. The results of pioneering efforts to age dinosaurs using growth ring counts suggest that the longevity of the basal ceratopsian Psittacosaurus mongoliensis was 10 or 11 years (G. Erickson and T. Tumanova, unpublished data), the prosauropod Massospondylus carinatus 15 years (Chinsamy, 1994), the sauropod Bothriospondylus madagascariensis 43 years (Ricqle`s, 1983), the ceratosaur Syntarsus rhodesiensis 7 years (Chinsamy, 1994), and the maniraptor Troodon formosus 3–5 years (Varricchio, 1993). These data are being used in conjunction with mass estimates to infer the metabolic status and growth rates of dinosaurs and to reconstruct the trophic dynamics of Mesozoic ecosystems.

See also the following related entry: GROWTH LINES

References Case, T. J. (1978). Speculations on the growth rate and reproduction of some dinosaurs. Paleobiology 4, 320–328. Castanet, J., Francillon-Viellot, H., Meunier, F. J., and Ricqle`s, A. de (1993). Bone and individual aging. In Bone (B. K. Hall, Ed.), pp. 245–283. CRC Press, Boca Raton, FL. Chinsamy, A. (1994). Dinosaur bone histology: Implications and inferences. In Dino Fest (G. D. Rosenberg and D. L. Wolberg, Eds.), pp. 213–227. The Paleontological Society, Department of Geological Sciences, Univ. of Tennessee, Knoxville. Francillon-Viellot, H., de Buffre´nil, V., Castanet, J., Ge´raudie, J., Meunier, F. J., Sire, J. Y., Zylberberg, L.,

Albany Museum, Grahamstown, South Africa

6 and Ricqle´s, A. de (1990). Microstructure and mineralization of vertebrate skeletal tissues. In Skeletal Biomineralization: Patterns, Processes and Evolutionary Trends ( J. G. Carter, Ed.). Vol. 1, pp. 471–530. Van Nostrand– Reinhold, New York. Reid, R. E. H. (1990). Zonal ‘‘growth rings’’ in dinosaurs. Mod. Geol. 15, 19–48. Ricqle`s, A. de (1983). Cyclical growth in the long limb bones of a sauropod dinosaur. Acta Palaeontol. Polonica 28, 225–232. Varricchio, D. V. (1993). Bone microstructure of the Upper Cretaceous theropod dinosaur Troodon formosus. J. Vertebr. Paleontol. 13, 99–104.

Albany Museum, Grahamstown, South Africa ANUSUYA CHINSAMY South African Museum Cape Town, South Africa

The Albany Museum, established in 1855, has postcranial material of Euskelosaurus and Massospondylus. It also houses probable Paranthodon africanus fossils, including several bone fragments. In a recent expedition to the Kirkwood beds of the Algoa Basin, soonto-be described HYPSILOPHODONTID skeletal elements recovered included a partial jaw. Titanosaurid-like, brachiosaurid-like, and theropod teeth from the Kirkwood formation are represented in the collections. The paleontology section of the Albany Museum is currently being renovated. Dinosaurs from the Eastern Cape will be represented by recently collected brachiosaurid material and by full-scale models of Paranthodon africanus (the first dinosaur discovered in South Africa) and Syntarsus rhodesiensis.

Allosauroidea KEVIN PADIAN JOHN R. HUTCHINSON University of California Berkeley, California, USA


llosauroidea (Fig. 1) subsumes by content the concepts Allosauria, Allosauridae, and Allosaurus. As recently constituted, Allosauroidea includes the Allosauridae and Sinraptoridae (Currie and Zhao, 1993; Sereno et al., 1994; Holtz, 1994, 1996). Because CARNOSAURIA is defined as all TETANURAE closer to Allosaurus than to birds, Allosauroidea are by definition carnosaurs, and any potential members of Carnosauria must be evaluated against Allosaurus. Resolution of allosauroid phylogeny beyond this statement is a difficult matter, partly because many taxa that are clearly allied to the group are incompletely known or have had their systematic characters interpreted differently by different workers. Consequently, the membership of Allosauridae and Sinraptoridae is currently not agreed upon apart from their eponymous genera. Sereno et al. (1994), for example, place Sinraptor and Yangchuanosaurus in Sinraptoridae and Acrocanthosaurus, Allosaurus, Cryolophosaurus, and Monolophosaurus in Allosauridae. Holtz (1994) placed only Allosaurus and Acrocanthosaurus in Allosauridae. Holtz (1995) regarded Monolophosaurus as a tetanurine outside AVETHEROPODA, but in 1996 found it to be the sister taxon to Allosauroidea,


Algerian Dinosaurs Fragmentary remains of dinosaurs have been recovered from the Late Cretaceous of Algeria.


FIGURE 1 Phylogeny of Allosauroidea. For synapomorphies see text. Taxa of debated placement, not listed here, include Acrocanthosaurus, Monolophosaurus, Cryolophosaurus, Chilantaisaurus, Piatnitzkysaurus, Carcharodontosaurus, and Giganotosaurus. These are all recognized as Carnosauria and a possible phylogeny is given under that entry.



FIGURE 2 Skull of Allosaurus (after Madsen, 1976).

with Cryolophosaurus as a more basal carnosaur. Holtz (1996) also found Giganotosaurus and the Carcharodontosauridae to be allosauroids closer to Allosauridae than to Sinraptoridae. Meanwhile, Sereno et al. (1996) also found Carcharodontosauridae to be in Allosauroidea but suggested that Acrocanthosaurus and Giganotosaurus might belong in Carcharodontosauridae instead of in Allosauridae. Acrocanthosaurus was known until recently only from incomplete specimens, but more material has been discovered in the past few years and awaits formal description. Other poorly known taxa, such as Pianitzkysaurus and Chilantaisaurus, may be allosauroids as well, but their exact relationships have not yet been conclusively established. Sereno et al. (1994) provided four synapomorphies of Allosauroidea, including the participation of the nasal in the antorbital fossa, a flange-shaped lacrimal process on the palatine, the basioccipital excluded from the basal tubera, and the articular with a pen-

dant medial process (Fig. 2). The synapomorphies of Allosauridae include a short quadrate with the head level with the middle of the orbit, a deep anterior ramus of the surangular, and the small diameter of the external mandibular fenestra. Sinraptoridae (Currie and Zhao, 1993) is characterized by two accessory pneumatic excavations on the maxilla, an external nares with a marked inset of the posterior margin, a bulbous, anteriorly projecting rugosity on the postorbital, and a flange on the squamosal that covers the quadrate head in lateral view (Sereno et al., 1994). Holtz (1994) diagnosed Allosauridae by the possession of a pubic ‘‘foot’’ that is longer anteriorly than posteriorly and triangular in ventral view. Because the synapomorphies diagnosing other nodes in his 1995 and 1996 works have not yet been published, differences in phylogenetic conclusions cannot currently be evaluated. Given the current instability in diagnosing the content and hence the synapomorphies of the allosauroid groups, Allosauridae and Sinraptoridae can be defined only with reference to their eponymous genera. Hence, Allosauridae comprises Allosaurus and all Allosauroidea closer to it than to Sinraptor; Sinraptoridae comprises Sinraptor (Fig. 3) and all Allosauroidea closer to it than to Allosaurus. Allosauroidea is a nodebased taxon, diagnosed by the synapomorphies previously discussed, that includes Allosaurus and Sinraptor and all descendants of their most recent common ancestor. It will be noted that Allosauridae and Sinraptoridae are herein defined as stem-based taxa; neither Sereno et al. (1994) nor Holtz (1994) indicated node- or stem-based definitions but rather used characters and included taxa (see PHYLOGENETIC SYS-

FIGURE 3 Sinraptor (after Currie and Zhao, 1993).

8 TEM). Stem-based definitions are included here because currently the taxa are minimally monotypic: No consensus exists on the membership of more than one genus per taxon. The genus Allosaurus has had a slightly confusing history. Leidy (1870) assigned a partial caudal vertebra, collected from Colorado by the Ferdinand Hayden expedition, to Poicilopleuron [sic] valens, noting its putative similarity to the European genus Poikilopleuron (which has had its own tortured history and has usually been synonymized with, or allied to, MEGALOSAURUS). However, Leidy also provided the generic name Antrodemus, should his specimen eventually prove different from Poikilopleuron. Marsh (1877) described a tooth, two dorsal vertebrae, and a phalanx, collected by Benjamin Mudge from the Late Jurassic MORRISON FORMATION of Fremont County, Colorado (‘‘Garden Park Quarry’’), as Allosaurus fragilis, and Marsh eventually described some of the further remains from the same quarry, excavated by M. P. Felch, which included an almost complete skeleton, several partial skeletons, and other bones. Hence, Allosaurus came to be well known. Gilmore (1920), however, in describing the full material from Garden Park, decided that Leidy’s caudal half-centrum of Antrodemus valens had diagnostic characters also seen in the Allosaurus material, so he regarded all the material as belonging properly to Antrodemus. Gilmore’s judgment, however logical at the time, has not been sustained by character analysis, and the name Allosaurus is accepted today (Madsen, 1976). The greatest collection of Allosaurus material has been made at the CLEVELAND –LLOYD QUARRY, discovered in 1927 and worked intermittently by crews from the University of Utah, Princeton University, and the Earth Science Museum, Brigham Young University, where most of the material is now jointly stored (see Miller et al., 1996). Several thousand bones now provide a very full picture of this animal, including its osteology and ontogeny, which preserve bones of individuals (unfortunately, none articulated) ranging from approximately 3 to 12 m in length. A detailed study and map of the Cleveland–Lloyd Quarry (Miller et al., 1996) and recent Ph.D. work by David K. Smith at Brigham Young University on the morphometrics of the Allosaurus material testify to the continuing importance of this bonanza of skeletal material.

Allosauroidea Allosaurus was the largest well-known carnivore in the Morrison Formation, presumably feeding on SAUROPODS, HYPSILOPHODONTIDS, STEGOSAURS, ANKYLOSAURS, and probably other animals; Ceratosaurus, another large carnivore, is present but rarer, and an even larger, possibly allosauroid theropod, Saurophaganax (which apparently reached the size of some adult Tyrannosaurus specimens) has not yet been well described. Torvosaurus is a basal tetanurine theropod also from the Morrison Formation (see DRY MESA QUARRY) and it reached comparable size, plus a few other taxa have been reported from fragmentary material that may or may not be diagnostic; therefore, at least four large theropods are known from the Morrison Formation, and there may well have been a considerable diversity of large carnivores feeding on the other components of the Morrison fauna and perhaps on each other. Fragmentary specimens referable to Allosaurus, or at least to Allosauridae, are reported from the TENDAGURU Beds of Tanzania and the Strzelecki Group of Victoria, Australia, extending the survival of the allosaurid lineage into the Early CRETACEOUS Period (Molnar et al., 1981, 1985). The upper jaw of Allosaurus bears 20 or more trenchant, laterally compressed teeth; the dentary bears up to 13, but the lower tooth row does not extend as far posteriorly as the upper row, as in most theropods. An extensive system of pneumatic spaces characterizes the orbital ‘‘brow ridge’’ and the skull roof bones behind the orbit (see CRANIOFACIAL AIR SINUS SYSTEMS). The brow ridge in Allosaurus, Sinraptor, and apparently Yangchuanosaurus is centrally excavated in a particular way, but the function is unknown. One of the most important recent discoveries concerning Allosaurus is that it has a furcula (Chure and Madsen, 1996). This was first discovered during the excavation of a still undescribed, Allosauruslike theropod from DINOSAUR NATIONAL MONUMENT. Comparison of this specimen’s furcula with Allosaurus material from the Cleveland–Lloyd Quarry revealed that such elements were common at Cleveland–Lloyd Quarry but had been taken for median bones of the ventral cuirass, or gastralia (Madsen, 1976). Gilmore (1920) had figured them as proceeding down the length of the abdomen, when in fact there was only one per individual, situated properly between the pectoral girdles. Their anat-



omy is easily distinguished from that of the true median gastral elements (Chure and Madsen, 1996). As Holtz (1996) noted, the possession of this distinctive, boomerang-shaped furcular morphology is a synapomorphy of AVETHEROPODA (see also PECTORAL GIRDLE).

References Chure, D. J., and Madsen, J. H. (1996). On the presence of furculae in some non-maniraptoran theropods. J. Vertebr. Paleontol. 16, 573–577. Currie, P. J., and Zhao, X. J. (1993). A new carnosaur (Dinosauria: Theropoda) from the Jurassic of Xinjiang, People’s Republic of China. Can. J. Earth Sci. 30, 2037– 2081. Gilmore, C. W. (1920). Osteology of the carnivorous Dinosauria in the United States National Museum, with special reference to the genera Antrodemus (Allosaurus) and Ceratosaurus. Smithsonian Institution, United States National Museum Bulletin No. 110, pp. 1–159. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117. Holtz, T. R., Jr. (1995). A new phylogeny of the Theropoda. J. Vertebr. Paleontol. 15(Suppl. to No. 3), 35A.

Holtz, T. R., Jr. (1996). Phylogenetic analysis of the nonavian tetanurine dinosaurs (Saurischia: Theropoda). J. Vertebr. Paleontol. 16(Suppl. to No. 3), 42A. Leidy, J. (1870). [Remarks on Poicilopleuron valens, etc.] Proc. Acad. Nat. Sci. Philadelphia 1870, 3–5. Madsen, J. H. (1976). Allosaurus fragilis: A revised osteology. Bull. Utah Geol. Miner. Survey 109, 1–163. Marsh, O. C. (1877). Notice of new dinosaurian reptiles from the Jurassic formation. Am. J. Sci. 3(14), 513– 514. Miller, W. E., Horrocks, R. D., and McIntosh, J. H. (1996). The Cleveland–Lloyd Quarry, Emery County, Utah: A U.S. national landmark (including history and quarry map). Brigham Young Univ. Geol. Stud. 41, 3–24, 4 maps. Molnar, R. E., Flannery, T. F., and Rich, T. H. V. (1981). An allosaurid theropod dinosaur from the Early Cretaceous of Victoria, Australia. Alcheringa 5, 141– 146. Molnar, R. E., Flannery, T. F., and Rich, T. H. V. (1985). Aussie Allosaurus after all. J. Paleontol. 59, 1511– 1513. Sereno, P. C., Wilson, J. A., Larsson, H. C. E., Dutheil, D. B., and Sues, H.-D. (1994). Early Cretaceous dinosaurs from the Sahara. Science 265, 267–271.

A diorama of Albertosaurus over a prey item, Centrosaurus, perhaps a common scene from the Cretaceous of Alberta, Canada. (Photo by Franc¸ois Gohier.)

American Dinosaurs PETER DODSON University of Pennsylvania Philadelphia, Pennsylvania, USA


The explicit history of dinosaur paleontology in the United States extends back to 1856, when Joseph Leidy applied names to a collection of teeth from the JUDITH RIVER beds along the Missouri River in Montana that was sent to Philadelphia by Ferdinand Hayden. The four names are Deinodon, Trachodon, Paleoscincus, and Troodon. Unfortunately, the first three names are nomina dubia, as these teeth are diagnostic only at the family level (this being the rule for dinosaur teeth, making it generally unwise to name dinosaurs on that basis). Before this time, an interesting bone had turned up in Cretaceous deposits from Woodbury, New Jersey. Such material had been discussed as early as 1787 at the American Philosophical Society in Philadelphia, but dinosaurs had not yet been recognized scientifically, and the report was forgotten. (Donald Baird has proposed that a hadrosaur metatarsal in the collection of the Academy of Natural Sciences of Philadelphia is this specimen.) The discovery and description of Hadrosaurus foulkii from Haddonfield, New Jersey, by Leidy in 1858 marks the first time that a major portion of a dinosaur skeleton, including fore- and hind-limbs, had been found. This allowed Leidy to reconstruct Hadrosaurus as a biped, showing that the Owen–Hawkins reconstruction of Iguanodon, exhibited at the CRYSTAL PALACE since 1854, was incorrect. The reconstruction and exhibition of Hadrosaurus at the Academy of Natural Sciences in 1868 marked the first time that a dinosaur skeleton had ever been exhibited anywhere in the world. Casts of this specimen were exhibited at Princeton University Geology Museum, the SMITHSONIAN, and the Field Columbian Museum in Chicago, but it was not until the first decade of the 20th century that other dinosaur skeletons were exhibited at the AMERICAN MUSEUM OF NATURAL HISTORY, YALE PEABODY MUSEUM, and the Smithsonian. E. D. Cope described a partial skeleton of the enigmatic theropod Laelaps (preoccupied; renamed Dryptosaurus Marsh 1877) from New Jersey in 1868. Cope named the

he United States has a great diversity of dinosaurs spanning a wide stratigraphic range. Although the concept of dinosaur was born in England, it found fertile ground in the United States. The United States has more different kinds of dinosaurs than any other country by a wide margin. A recent tabulation based on data as of 1988 shows that the United States has 64 known genera of dinosaurs compared with 40 for Mongolia and 36 for China. Such figures rapidly become dated as new kinds from around the world are described. In 1993, for instance, four new dinosaurs were described from the United States: Shuvosaurus from Texas, Utahraptor from Utah, and Naashoibitosaurus and Anasazisaurus from New Mexico. In 1994, Mymoorapelta from Utah was added, and in 1995 the ceratopsids Einiosaurus and Achelousaurus from Montana were formally described (see VARIATION). Forthcoming are a theropod from the Early CRETACEOUS of Utah, an ANKYLOSAURIAN and an ORNITHOPOD from Texas, and a basal ornithischian from New Mexico. Thus, growth of knowledge of new kinds of dinosaurs continues at least as rapidly in the United States as in China and at a greater rate than in Argentina or in Mongolia. There are four fundamental reasons why the United States has so many different kinds of dinosaurs: stratigraphy, climate and geography, human resources, and history. Like Argentina and China, and unlike Canada and Mongolia, the United States has dinosaurbearing continental strata that span most of the stratigraphic interval in which dinosaurs may be expected from the Carnian stage of the Late TRIASSIC to the Maastrichtian stage of the Late Cretaceous. The United States has large areas of outcrop in semiarid climates, principally in the west, where erosion is relatively unencumbered by vegetation, unlike Canada, England, or the eastern United States, for example. There is also a large corps of professional, commercial, and amateur dinosaur collectors in this country, all of whom contribute to ongoing discoveries.


American Dinosaurs ceratopsids Agathaumas in 1872 and Polyonax in 1874 from Wyoming and Colorado, respectively, but these are nomina dubia based on fragmentary material. In 1876, he collected and named Monoclonius from the Judith River Formation of Montana, the first valid ceratopsid. Up to this point, dinosaur finds had been geographically widespread and generally of poor quality. Montana had produced the most dinosaurs up to this time, but most finds were not memorable. In 1877, dinosaurs were discovered in abundance for the first time anywhere in the world at three separate localities: CAN˜ ON CITY and Morrison, both in Colorado, and COMO BLUFF, Wyoming. The beds proved to be of Late Jurassic age and have produced a remarkable fauna dominated by large sauropods, with stegosaurs also important; theropods and ornithopods were less abundant; recently an ankylosaur was reported. Intensive examination of the Morrison fauna waned after 1885. Renewed interest in the Morrison at the turn of the century, after Marsh and Cope had died, produced further sauropods (Brachiosaurus and Haplocanthosaurus) and the small theropod Ornitholestes. Beds of Triassic age were documented with the description of Coelophysis bauri by Cope in 1889. The study of beds of latest Cretaceous age began with the description of Triceratops Marsh 1889, followed by Torosaurus Marsh 1891, and then Tyrannosaurus (1905) and Ankylosaurus (1908) early in this century. Lancian hadrosaurine species were described by Marsh in 1890 and 1892, but the proper generic assignments (to Anatotitan and Edmontosaurus) were not recognized until much more recently. The United States lacks a major dinosaur fauna correlative with the Horseshoe Canyon Formation (early Maastrichtian) of Alberta, Canada, although the TWO MEDICINE FORMATION of Montana, first studied by C. W. Gilmore beginning in 1914, contains an antecedent fauna, as do the FRUITLAND /KIRTLAND FORMATIONS of New Mexico and the Aguja Formation of Texas. A major fauna of Early Cretaceous age, very broadly correlative with the British WEALDEN fauna, was unknown in the United States until John Ostrom described the fauna of the CLOVERLY FORMATION of Wyoming and Montana in 1970. Lateral equivalents of the Cloverly (CEDAR MOUNTAIN FORMATION of Utah is partially equivalent; Trinity Group, TX) are now producing important specimens (Utahraptor; Proctor Lake ornithopod). Late Cretaceous dinosaurs from

11 New Mexico began to be described in 1910. The Late Triassic is sparsely productive of dinosaurs, the rich deposits of Coelophysis being a conspicuous exception. There are Early Jurassic dinosaurs in Connecticut and the southwest; the Middle Jurassic is essentially unknown. Primitive theropods are well represented, the most prominent being Coelophysis (known from scores of skeletons from the mass death assemblage at GHOST RANCH, NM), Dilophosaurus, and Ceratosaurus. Large theropods are represented by two principal taxa, the Allosauridae (Allosaurus) and the TYRANNOSAURIDAE. Tyrannosaurus now appears to be one of the most common large theropods. Good specimens of Albertosaurus are common in Canada but are very rare in the United States. The fossil record of maniraptorans in the United States is rather sparse, apart from the imperfect material of Ornitholestes and Coelurus. Deinonychus is the most important American maniraptoran, and recently the larger Utahraptor has been described. Ornithomimids are poorly represented but were surely present. Many MANIRAPTORAN taxa are documented principally by teeth (e.g., Aublysodon, Paronychodon, and Ricardoestesia) and thus are in perilous condition taxonomically. No segnosaurs have been confirmed. ‘‘Prosauropods’’ (basal SAUROPODOMORPHS) are somewhat sparse in the American fossil record. Anchisaurus and Ammosaurus are the principal taxa, although Massospondylus has been reported from the Early Jurassic of Arizona. No sauropods of Early and Middle Jurassic age are known, but the Late Jurassic Morrison Formation contains a sauropod assemblage that is rivaled in quality, quantity, and diversity only by the correlative assemblages from China. For nearly a century, these sauropods presented the basis for understanding sauropods everywhere in the world. The important taxa Camarasauridae (Camarasaurus), Brachiosauridae (Brachiosaurus), and Diplodocidae (Diplodocus, Apatosaurus, and Barosaurus) were established on Morrison sauropods. The taxa Cetiosauridae (Haplocanthosaurus) and Titanosauridae (Alamosaurus) are known but are much less important here. Basal ornithischians are poorly represented at present, but Technosaurus from Texas seems representative of such basal taxa. In addition, teeth of basal ornithischians have been documented in Late Triassic

12 sediments from Pennsylvania to Arizona. Scutellosaurus and Scelidosaurus are good basal thyreophorans. The Stegosauridae are magnificently characterized by Stegosaurus, but there is otherwise very low diversity of this family, in contrast to China. There are few basal ankylosaurians, but there are good representatives of the Nodosauridae (Sauropelta) and of the Ankylosauridae (Ankylosaurus), both taxa being established on American taxa. A very important recent discovery is that of the ankylosaur Mymoorapelta from the MORRISON FORMATION of Colorado. The nodosaur Edmontonia is now reported from Alaska. For both families, there are more skulls than skeletons, with no complete skeletons in either taxa having yet been collected. ORNITHOPODS are well represented in the United States. Hypsilophodontids are somewhat fragmentary (Othnielia and Orodromeus), although there are several good specimens of the enigmatic Thescelosaurus. Basal iguanodontians are also well represented by Dryosaurus and Tenontosaurus, the latter of which is particularly abundant and widespread with specimens being reported from Montana, Wyoming, Utah, Oklahoma, and Texas (possibly Maryland as well). Camptosaurus is an abundant American iguanodontian, and a fine skull of Iguanodon itself, named I. lakotensis, has been described. HADROSAURS are abundant in the United States, including both lambeosaurines and hadrosaurines. The former are represented only by Parasaurolophus from New Mexico and Utah and Hypacrosaurus from northern Montana. Hadrosaurines come from New Jersey (Hadrosaurus), Alabama (Lophorothon), New Mexico (Kritosaurus), and extensively from Wyoming, Montana, North Dakota, and South Dakota (Anatotitan and especially Edmontosaurus, which is one of the most abundant dinosaurs both in the United States and in the world). Edmontosaurus is also reported from the North Slope of Alaska. Pachycephalosaurs are principally represented by crania, particularly of Pachycephalosaurus itself. Protoceratopsids are documented by a few incomplete specimens of Leptoceratops and by a specimen of Montanoceratops. Ceratopsids include both centrosaurines (Monoclonius) from Montana and chasmosaurines, especially Triceratops, from Wyoming, Montana, North Dakota, South Dakota, and Colorado; Chasmosaurus from Texas; Torosaurus from Montana and South Dakota; and Pentaceratops from New Mexico. Ceratopsids are endemic to North

American Dinosaurs America. Triceratops is among the most abundant of all dinosaurs. There is a fragmentary occurrence of Pachyrhinosaurus from the North Slope of Alaska. Because dinosaurs are so diverse in the United States, it is tempting to think of this country as a center of evolution for worldwide faunas. This may not be so. In the Late Triassic, plateosaurids, common in Europe, Asia, and South America, are rare in the United States. Rare Early Jurassic sauropodomorphs have been found in Arizona and Connecticut. There are significant resemblances between Late Triassic Coelophysis of New Mexico and Early Jurassic Syntarsus of Zimbabwe and South Africa, but the resemblances between these relatively primitive theropods include few derived characters. There are essentially no Middle Jurassic beds in the United States to document the antecedents of the marvelous Late Jurassic sauropods, ornithopods, and stegosaurs of the Morrison Formation. Haplocanthosaurus may be presumed to be representative of the basal cetiosaurid radiation better documented in England, Europe, and South America. Brachiosaurus from Colorado has affinities with congeneric fossils from Tanzania. Other faunal elements having congeners in East Africa are Dryosaurus, probably Barosaurus, and less certainly Ceratosaurus and Allosaurus. It is significant that stegosaurs are much less diverse in the United States than they are in China, although Stegosaurus itself may be the most highly derived stegosaur. Camptosaurus is an important basal iguanodontian in the United States, with a sister species in the Middle Jurassic of England. In the Early Cretaceous, Tenontosaurus is an endemic ornithopod more basal than Camptosaurus. Although Iguanodon appears to have reached North America, it seems to have been uncommon there. Important new evidence suggests that Polacanthus from the WEALDEN of England also lived in Utah. In the Late Cretaceous, there is scant evidence for the titanosaurid sauropod fauna that dominated much of the world. It is postulated that Alamosaurus was a late migrant from South America, reintroducing sauropods which had been absent since the Early Cretaceous. There is evidence of faunal interchange with Asia based on similarities at the level of family and genus. A close relationship, possibly at the species level, of Tyrannosaurus with the Asiatic Tarbosaurus is recognized. Other evidence for exchange is better documented by Canadian dinosaurs, notably the ha-

American Dinosaurs drosaurine Saurolophus. Due to the relatively impoverished faunas of the Judith River Formation and dearth of early Maastrichtian dinosaurs in the United States, coupled with the relatively restricted area of late Maastrichtian strata in Alberta and Saskatchewan, faunal overlap between the United States and Canada is not as great as expected, and the greater diversity and completeness of specimens favors Canada. Although ceratopsids range from Alaska to Mexico, the only identifiable specimens of this family in Asia are teeth and horn core fragments from Uzbekistan. Mid-Cretaceous dinosaurs are found in Maryland, and Late Cretaceous dinosaurs are known from the eastern seaboard of the United States, from New Jersey to North Carolina, and also along the Gulf Coast and Mississippi embayment from Alabama to western Tennessee and Missouri. Few skeletons have been described (Hadrosaurus and Dryptosaurus from New Jersey), and the faunal relationship to dinosaurs in the West, across the Inland Sea, is not evident. Hadrosaurus appears to be a sister group of Kritosaurus from New Mexico and/or Gryposaurus from Alberta.

13 Dryptosaurus is a nonarctometatarsalian of unclear relationship to any other theropod. It is claimed that there is a specimen of Albertosaurus from Alabama, but it is undescribed. This would present the same biogeographic challenge that Hadrosaurus presents. The mechanism of faunal exchange across a 1000- to 1500-km inland sea has yet to be elucidated.


● ●

References Dodson, P. (1990). Counting dinosaurs: How many kinds were there? Proc. Natl. Acad. Sci. USA 87, 7608–7612. Russell, D. A. (1993). The role of Central Asia in dinosaurian biogeography. Can. J. Earth Sci. 30, 2002–2012. Weishampel, D. B., Dodson, P., and Osmolska, H. (Eds.) (1990). The Dinosauria, pp. 733. Univ. of California Press, Los Angeles.

Perhaps the largest dinosaur of all time, Seismosaurus dwarfs most other large sauropods. (Illustration by Donna Braginetz.)

American Museum of Natural History LOWELL DINGUS American Museum of Natural History New York, New York, USA


Field Expeditions of the Museum

he American Museum of Natural History (AMNH) in New York City is the largest private museum in the world, and it has made a similarly large contribution to vertebrate paleontology. The museum attracts millions of visitors each year to its famous dinosaur displays. Some of the most notable dinosaur discoveries have been made by field expeditions sponsored by the AMNH, including the excavations by Barnum Brown and others at COMO BLUFF and HELL CREEK in the 1890s, the Roy Chapman Andrews expeditions in the Gobi Desert in the 1920s, and the recent trips back to Gobi by AMNH curators Michael Novacek and Mark Norell. The AMNH was incorporated in 1869 in New York City. In 1877 the first permanent building opened at the current site, in midtown Manhattan west of Central Park. Albert S. Bickmore, a zoologist who studied under Louis Agassiz at Harvard, is regarded as the founder of the museum. He envisioned a natural science museum in the center of the metropolis of New York, comparable to the MUSEUM OF COMPARATIVE ZOOLOGY founded in Cambridge by Agassiz in 1859. Bickmore brought together a group of prominent New Yorkers who raised the money for the museum, including the financier J. P. Morgan. In its earliest years the AMNH had virtually no vertebrate fossils, a fact that museum president Morris K. Jesup set out to rectify in 1891 by his hiring of Henry Fairfield Osborn, a noted paleontologist and a faculty member at Princeton University. Osborn founded the museum’s Department of Vertebrate Paleontology and staffed it with an outstanding group of paleontologists, including Barnum Brown, William Diller Matthew, Jacob Wortman, Walter Granger, and Albert ‘‘Bill’’ Thomson. They were supplemented by preparators such as Peter Kaisen, Otto Falkenbach, and Adam Hermann.

During the Jesup–Osborn era, the museum initiated a series of highly productive field expeditions. Beginning in 1897, Wortman, Brown, Granger, and others explored the Jurassic fossil beds in the COMO BLUFF area of Wyoming. Although this area had already been extensively worked by the expeditions of Othniel Charles Marsh, the AMNH group made many additional important finds, including the sauropods Apatosaurus, Diplodocus, Allosaurus, and Ornitholestes. In 1898, museum paleontologists in this area located the distinctive BONE CABIN site. In 1902 Barnum Brown led an AMNH expedition to the Cretaceous beds of the HELL CREEK region of Montana. This resulted in the first known specimen of Tyrannosaurus rex, in 1902, and a second, more highly preserved specimen in approximately 1908. This second specimen is generally regarded as the most famous dinosaur fossil in the world and has long been a centerpiece of the AMNH. Brown went on to lead museum-sponsored expeditions in 1910–1915 to the Red Deer River region of Alberta, Canada. These also yielded rich discoveries, especially of hadrosaurs such as Saurolophus and Corythosaurus. In the 1930s another AMNH expedition led by Brown excavated a large collection of Jurassic fossils from the Howe Ranch site in Wyoming. What has been termed the golden age of the museum’s field expeditions extended from 1890 to 1930. In addition to Brown’s highly publicized efforts, various other paleontologists from the AMNH staff also made important finds in this period, including Matthew, Granger, Thomson, and Henry Fairfield Osborn himself. Osborn also obtained many specimens from independent collectors, such as Charles H. Sternberg.


American Museum of Natural History Expeditions of Roy Chapman Andrews The expeditions described previously all were situated in western North America, but the most renowned expeditions sponsored by the AMNH were carried out in central Asia. These took place in Mongolia in the period 1920–1930 under the leadership of the legendary Roy Chapman Andrews. Andrews originally went to Mongolia’s remote Gobi Desert region at the invitation of Henry Fairfield Osborn, who had succeeded Morris K. Jesup as president of the museum. Osborn believed that an investigation of the region could substantiate his theory that Asia, not Africa, was the original site of human habitation. Andrews found no human fossil evidence to support Osborn’s theory but did find many other significant vertebrate fossils, including the first dinosaur bone discovered in eastern Asia. The Flaming Cliffs region in particular produced important remains of dinosaurs such as Protoceratops, Oviraptor, Saurornithoides, Pinacosaurus, and Velociraptor, as well as Cretaceous mammals. The most noted single discovery was that of the predatory dinosaur Oviraptor lying on a clutch of supposed Protoceratops eggs.

Later Expeditions Both Barnum Brown and Roy Chapman Andrews achieved celebrity status as dinosaur hunters, and Andrews is now often cited as the model for the Hollywood character Indiana Jones. However, after 1930 a combination of reduced museum funds and the historic circumstances of the Depression and World War II meant that high-profile expeditions such as theirs were no longer feasible. The AMNH nevertheless continued to sponsor important expeditions such as Roland T. Bird’s examination of the GLEN ROSE, TEXAS, dinosaur trackway site in the 1930s and Edwin Colbert’s discovery of Coelophysis skeletons at GHOST RANCH, New Mexico, in the late 1940s. The AMNH has maintained this tradition of field explorations to the present day. The most notable recent example was a program to revisit the sites explored by Roy Chapman Andrews. Since 1990, field crews from the museum have participated in annual expeditions to the Gobi Desert in conjunction with colleagues from the Mongolian Academy of Sciences.

15 These efforts have resulted in the discovery of a new Late Cretaceous flightless bird, Mononykus, and the first known embryo of a theropod dinosaur.

Exhibits In the new Halls of Ornithischian and Saurischian Dinosaurs, the AMNH exhibits the largest collection of real dinosaur fossils anywhere in the world. More than 100 specimens are on display, and approximately 85% of them include real fossil material. Many new specimens have been added, and several of the older mounts have been modified, including those of Tyrannosaurus and Apatosaurus. In contrast to most exhibitions, the primary organizing framework is based on systematics rather than geologic time. Labels have intentionally been developed at different levels of technical difficulty to address the needs of a diverse audience. A main path down the center of each hall represents the trunk of an evolutionary tree. By walking down this path, one can see the most spectacular specimens and encounter labels addressing the major themes. Collection alcoves along the sides of the halls represent branches that contain fossil representatives of the principal dinosaurian clades. One system of computer interactives located in each alcove is utilized to present curatorial views about the evolutionary relationships of the dinosaurs on that branch. A second system of computer interactives is used to present the ‘‘walkthrough time’’ approach utilized in most exhibitions. The main point of the presentation is to illustrate for the visitor what we really know about these extinct animals. Many controversial issues are addressed by simply presenting the evidence for different ideas and letting the visitor make up his or her own mind about what to think. This is intentionally done in order to provide visitors with some insight into how the scientific process works. The two dinosaur halls are part of a loop of six halls on the fourth floor of the museum that are designed to tell the story of vertebrate evolution. In all, these halls contain 57,000 square feet of exhibition space. In emphasizing evolutionary relationships based on cladistic methods, the dinosaur exhibits also help make visitors aware of the kind of scientific research conducted in the museum’s Department of Vertebrate Paleontology. In terms of dinosaurs, the curatorial staff and associates actively pursue research and

16 fieldwork, including topics such as theropod evolution, the origin or birds, and the extinction of nonavian dinosaurs. Along with its prominence in field exploration and dinosaur displays as described previously, the museum is also a noted research center. In addition to research by the museum staff and students, each year scientists from around the world come to New York City to work in the museum’s dinosaur collections.


References Colbert, Ed. (1992). William Diller Matthew, Paleontologist. Columbia Univ. Press, New York. Dingus, L. (1996). Next of Kin: Great Fossils at the American Museum of Natural History. Rizzoli International, New York. Hellman, G. (1969). Bankers, Bones, and Beetles: The First Century of the American Museum of Natural History. Simon & Schuster, New York. Novacek, M. (1996). Dinosaurs of the Flaming Cliffs. Doubleday, New York. Preston, D. J. (1986). Dinosaurs in the Attic: An Excursion into the American Museum of Natural History. St. Martin’s Press, New York. Rogers, K. (1991). The Sternberg Fossil Hunters: A Dinosaur Dynasty. Mountain Press, LaHonda, CA. Wallace, J. (1994). American Museum of Natural History’s Book of Dinosaurs and Other Ancient Creatures. Simon & Schuster, New York.

Ankylosauria KENNETH CARPENTER Denver Museum of Natural History Denver, Colorado, USA


nkylosaurs are four-legged, armor-plated ornithischians that first appeared in the Middle Jurassic. Specimens range in size from 1-m-long juveniles of Pinacosaurus to 10-m-long adult Ankylosaurus. Ankylosaurs are united by several synapomorphic characters, among which are a low, wide skull; cheek teeth deeply inset from the sides of the jaws; fusion of

Ankylosauria armor to the skull masking the cranial sutures and covering the supratemporal fenestra; fusion of the last three or four dorsals with the sacrum into a rod of vertebrae or synsacrum; horizontal rotation of the ilium so the ilium faces upwards, not outwards; secondary closure of the acetabulum; reduction in size of the pubis; and a body encased in armor plates (Coombs and Maryanska, 1990). Ankylosauria may be defined as all thyreophoran ornithischians closer to Ankylosaurus than to Stegosaurus. The ankylosaurs have been placed into one of two families (Coombs, 1978) based primarily on the presence or absence of a bone club on the end of the tail (Fig. 1). Those with a club are placed in the family Ankylosauridae (or ankylosaurids in the vernacular), and those without in the Nodosauridae (or nodosaurids). Actually, there are many other differences between the two families. The skull of ankylosaurids seen in top view is triangular, with small ‘‘horns’’ at the upper and lower corners of the skull (Fig. 2). These horns are actually triangular armor plates. The entire surface of the skull is covered in a mosaic of small, irregularly shaped armor, except in Pinacosaurus, in which much of the cranial armor is secondarily lost. The front of the ankylosaurid skull is usually broad, with a wide beak suggesting nonselective cropping of low vegetation. The exception to this is the primitive ankylosaurid Shamosaurus from the Lower Cretaceous of Mongolia. It has a narrow, pointed muzzle. Ankylosaurid teeth are small for the size of the skull and have a swollen base or cingulum. In at least one form (Euoplocephalus) a bony eyelid was also present (Coombs and Maryanska, 1990). The external nares face forward in many ankylosaurids, with Ankylosaurus and Shamosaurus being the few exceptions (Tumanova, 1987). During respiration, the air moves through a complex air passage within the ankylosaurid skull, with at least one loop in it. The purpose for the complexity is puzzling, but it may have increased the surface area of the olfactory tissue or acted as a resonating chamber. In nodosaurids, the skull is elongated with rounded corners. A single large armor plate is present on the top center of the skull and pairs of regular shaped plates in front of this over the snout. The external nares often face laterally at the front of the snout and the air passage is more direct. The beak is narrow for selectively cropping vegetation, perhaps



FIGURE 1 The skeleton of an ankylosaurid, Dyoplosaurus (A), and nodosaurid, Sauropelta (B).

leaves. In some primitive forms, such as Silvisaurus, conical premaxillary teeth are present. The cheek teeth are larger than those in ankylosaurids and have a shelf-like cingulum at their base. In Edmontonia, an oval cheek plate is present. The vertebral column of ankylosaurs differs from that of many dinosaurs in that the neural spines are low. In many dinosaurs, the neural spines are often tallest on the posterior dorsals, sacrals, and anteriormost caudals, possibly to increase surface areas for ligaments holding the tail horizontally. In ankylosaurs, the armor may have functioned to hold the tail aloft as the dorsal armor does in crocodilians (Carpenter, 1997). Another major difference in the vertebral column between ankylosaurs and most dinosaurs is the coossification of the last three or four dorsals with the sacrals into a rod called the synsacrum. The ribs from these dorsals arch out to fuse to the underside of the ilium. Sometimes the first and second caudals may also fuse to the rear of the synsacrum. The ribs of the mid-dorsals may co-ossify with the vertebrae as well. In the tail, the caudal ribs are fused to the centra. These ribs in the anterior portion of the tail are long and slender and often curve downwards. In ankylo-

saurids, the pre- and postzygapophyses and chevrons are elongated in the posterior portion of the tail. The result is a rigid ‘‘handle’’ to the bone club that terminates the tail. This club is actually formed from enlarged bone plates that fuse together and to the vertebrae. In nodosaurids, there is no such modification of the tail vertebrae and no club. The shoulder girdle of ankylosaurs is massive, especially the scapula. This is probably due to the enlarged muscles associated with quadrupedality because the scapula of bipedal dinosaurs is proportionally much more slender. In both nodosaurids and ankylosaurids there is a knob near the scapula– coracoid suture for the scapulohumeralis muscle. This knob is large and occurs on a ridge, or pseudoacromion process, in nodosaurids. This ridge may have functioned much like the acromion process in the mammalian scapula—to divide muscle masses. The position of the pseudoacromion process relative to the glenoid varies among the different genera of nodosaurids, making the scapula a taxonomically important bone. The forelimb is short and stocky in many ankylosaurs and superficially resembles that of ceratopsians and stegosaurs. These similarities are unquestionably



FIGURE 2 The teeth and skull of an ankylosaurid, Euoplocephalus (A, C, E, G), and nodosaurid, Sauropelta (B, D, F, H).

Ankylosauria due to the quadrupedal stance of the three groups. The humerus of most ankylosaurs is short and stocky, although less so in some nodosaurids such as Sauropelta (Ostrom, 1970). The ulna has a very prominent olecranon process that can occupy more than a third of the length of the ulna. Such a well-developed olecranon indicates that the elbow was flexed and probably held near the body. The bones of the manus are short for bearing the weight of the animal. The pelvic girdle of ankylosaurs is considerably modified, with the ilium horizontal, the pubis reduced in size, and the ischium a nearly vertical rod. Furthermore, the neural spines are fused into a vertical sheet of bone, and the acetabulum is closed off by the ilium and ischium. Although both ceratopsians and stegosaurs show partial modification of the ilium into the horizontal plane, this is considerably less than seen in ankylosaurs. The preacetabular portion of the ilium is also expanded to provide a large surface area for the iliotibialis muscles (Coombs, 1979). It is puzzling, however, why this protractor muscle for the leg needed to be so large. Perhaps we have misinterpreted the muscle scars. The pubis is a small rectangular bone with a very short postpubic process in both ankylosaurids and nodosaurids. The ischium is a long, heavy bar of bone that projects downward, as in the saurischian pelvis, instead of horizontally as in the typical ornithischian pelvis. As with the forelimb, the hindlimb is adapted for carrying considerable weight. Not surprisingly, the ankylosaur hindleg also resembles that of stegosaurs and ceratopsians. The femur is robust, having a larger midshaft circumference compared to that in a bipedal ornithischian. The fourth trochanter is a large scar on the femur and does not project outwards like the flange seen in hadrosaurs. The tibia and fibula are short and robust. The distal end of the tibia has an enormous fibular process, thus preventing movement between the two bones. The pes is short for bearing weight. Primitively, in ankylosaurs there are four functional toes, but only three in advanced forms, such as the ankylosaurid Euoplocephalus (Carpenter, 1982). The armor of ankylosaurs is perhaps their most distinguishing character, separating them from all other dinosaurs. This armor consists of keeled plates of bone, short spines, and tall spikes. Ankylosaurids typically have plates arranged in transverse bands

19 along the neck, back, and tail. This armor may be supplemented by thin-walled, conical spines on the back, such as in Euoplocephalus (Carpenter, 1982). In addition, the neck armor of ankylosaurids is often fused to an underlying band of bone. In Pinacosaurus, and probably other ankylosaurids, this neck armor is the first to ossify in juveniles, the rest being cartilaginous. As stated previously, the characteristic tail club of ankylosaurids is made from the fusion of large terminal plates (Coombs, 1978). In nodosaurids, the keeled plates are co-ossified into a solid shield over the pelvis in some genera, such as Polacanthus. Nodosaurids also supplement their armor with outward-projecting spines along the sides of the body, except in Panoplosaurus. In Edmontonia, a pair of these spines is enlarged into forwardprojecting spikes on each shoulder, whereas in Sauropelta, four pairs of spikes project upwards from the neck. (Carpenter, 1984, 1990). The spines and spikes in nodosaurids probably played a dual role of defense and display. In Polacanthus a small club is present on the end of the tail, formed by a pair of enlarged plates [J. Kirkland (personal communication) suggests this character may indicate a third lineage of ankylosaurs]. Ankylosaurs are known from almost every continent, including Antarctica. Surprisingly, they have not been identified from South America. Keeled armor plates, once identified as ankylosaurian, are now known to belong to a titanosaur sauropod. Globally, ankylosaurs are rare, making up only a small percentage of any dinosaur fauna. Still, there are a few places where they occur in greater numbers, such as the sand dune deposits of the DJADOKHTA FORMATION of Mongolia and China (Tumanova, 1987). Nodosauridae and Ankylosauridae co-occur only in North America. Elsewhere, nodosaurids are known only from Europe and Antarctica and ankylosaurids from Asia. Another lineage of Ankylosauria is hinted by Minmi from Australia, which apparently shows a mixture of both ankylosaurid and nodosaurid features (R. Molnar, personal communication).

References Carpenter, K. (1982). Skeletal and dermal armor reconstruction of Euoplocephalus tutus (Ornithischia: Ankylosauridae) from the Late Cretaceous Oldman Formation of Alberta. Can. J. Earth Sci. 19, 689–697.


20 Carpenter, K. (1984). Skeletal reconstruction and life restoration of Sauropelta (Ankylosauria: Nodosauridae) from the Cretaceous of North America. Can. J. Earth Sci. 21, 1491–1498. Carpenter, K. (1990). Ankylosaur systematics: Example using Panoplosaurus and Edmontonia (Ankylosauria: Nodosauridae). In Dinosaur Systematics: Approaches and Perspectives (K. Carpenter and P. Currie, Eds.), pp. 281–298. Coombs, W. (1978). The families of the ornithischian dinosaur order Ankylosauria. Palaeontology 21, 143–170. Coombs, W. (1979). Osteology and myology of the hindlimb in the Ankylosauria (Reptilia, Ornithischia). J. Paleontol. 53, 666–684. Coombs, W., and Maryanska, T. (1990). Ankylosauria. In The Dinosauria (D. Weishampel, P. Dodson, and H. Osmo´lska, Eds.). pp. 456–483. Univ. of California Press, Berkeley. Ostrom, J. (1970). Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Bighorn Basin area, Wyoming and Montana. Peabody Museum Nat. History Bull. 35, 1–234. Tumanova, T. (1987). Pantsirnye dinozavry Mongolii. Sovmestnaya Sovetsko–Mongolskaya Paleontologischeskaya Ekspeditsiya. 32, 1–76. [In Russian]

Antarctic Dinosaurs see POLAR DINOSAURS

Antorbital Fenestra This opening is between the orbit and the nostril on the side of the skull and is a characteristic of dinosaurs and other archosaurs. It probably housed an airfilled sinus.


Archaeopteryx The first known bird (Aves), from the Late Jurassic (Solnhofen Formation) of Germany.


Ankylosauridae A member of the Ankylosauria along with Nodosauridae.



Northern Illinois University DeKalb, Illinois, USA

The Archosauria comprises one of the major radiaAnnie Riggs Museum, Texas, USA see MUSEUMS



Anniston Museum of Natural History, Alabama, USA see MUSEUMS



tions of terrestrial vertebrates, including such assemblages as the DINOSAURIA (including birds), the CROCODYLIA, and the PTEROSAURIA, along with a number of less familiar, extinct groups. The name Archosauria (‘‘ruling reptile’’) was erected by Cope in 1869 to include a somewhat different group of living and extinct reptiles. In recent decades, the Subclass Archosauria was considered to include the dinosaurs, crocodilians, pterosaurs, their common ancestors, and a number of other closely related TRIASSIC diapsid groups such as the Proterosuchia and Erythrosuchia. With the advent of wide use of phylogenetic systematics as a basis of classification in the 1980s, three significant changes occurred in the constitution of the

Archosauria Archosauria. First, the birds were included within the Archosauria (with the subclass designation dropped, along with all other nomenclature designating rank) because they evolved from dinosaur ancestors and thus form part of the monophyletic (single clade and continuous lineage) group including the dinosaurs and all their descendants. Second, the PERMIAN and Triassic archosaurs excluding dinosaurs, pterosaurs, and crocodilians had traditionally been placed into their own order, the THECODONTIA. One of the goals of phylogenetic systematics is to have classification reflect complete lineages of organisms (see SYSTEMATICS). The Thecodontia, by definition, consisted of only the basal parts of the archosaur lineage through excluding groups such as the dinosaurs and crocodilians that were derived from the basal archosaurs. The smallest monophyletic group that included all the ‘thecodonts’ was the Archosauria, so that name was retained and the Thecodontia discarded (see THECODONTIA). Finally, the Archosauria has been redefined as a crown group consisting of the last common ancestor of the two extant groups of archosaurs (birds and crocodiles) and all of its descendants. By this convention, animals that were formerly considered archosaurs but that appeared prior to the split of the crocodile and bird lineages are considered to be part of a newly erected monophyletic group, the Archosauriformes (Gauthier, 1984) but are excluded from the Archosauria (Fig. 1A). Looking at the broad pattern of evolution within the Amniota (the monophyletic group including birds, reptiles, and mammals and everything descended from their last common ancestor), three major lines diverged within that group (Fig. 1B). The first to split off, the Synapsida, include the mammals and a variety of other forms often informally called the ‘‘mammal-like reptiles’’ (an incorrect designation because they were never reptiles in the contemporary sense). The Anapsida include turtles and a number of fossil groups. The third group, the Diapsida, again consists of two main branches. The Lepidosauromorpha includes lizards, snakes, and the Sphenodontida (a mostly extinct group with one modern representative, the tuatara), along with a number of extinct clades. The Archosauromorpha includes the Archosauria, most of the marine reptile groups, and several extinct groups such as the beaked rhynchosaurs and the long-necked tanystropheids.


FIGURE 1 (A) Cladogram summarizing the current views of relationships among the Diapsida. (B) Cladogram summarizing the current views about the relationships among the major groups of amniotes with extant members.

The Archosauria proper appear to have originated near the end of the Early Triassic. The small archosauriform Euparkeria has often been cited as the best fossil example of what the common archosaurian ancestor might have looked like, although it lacks several key derived characters that diagnose the archosaurs proper. Known from a handful of specimens from a single locality in the Aliwal North Region of South Africa, Euparkeria was a small, agile terrestrial carnivore with a tall, mediolaterally compressed skull of a type that is common within several archosaur lineages. Euparkeria appears to be the first known archosauriform taxon with dermal armor, a feature that became widespread in the Archosauria. Another group that branched off near the base of the Archo-


FIGURE 2 Cladograms of archosaur relationships. (A) Cladogram combining Gauthier’s (1994) arrangement for the Ornithosuchia with Parrish’s (1993) phylogeny of the Crocodylotarsi (Pseudosuchia here). (B) Cladogram depicting Sereno’s (1991) views of archosaur relationships. Sereno did not present a phylogeny of relationships among the group he called Suchia.

sauria was the Proterochampsia, a group of mostly amphibious, crocodile-like quadrupeds that seem to be restricted to the Middle and Upper Triassic of South America. By its current definition, the Archosauria consists of two diverging lineages: one that leads to dinosaurs and ultimately to birds and another that leads to crocodilians. Several different names have been applied to these groups in recent years (Fig. 2). Gauthier (1986) called the archosaurs that are more closely related to birds than crocodilians, the ORNITHOSUCHIA, and those more closely related to crocodilians than birds, the PSEUDOSUCHIA. Benton and Clark (1988) and Parrish (1993) used Crocodylotarsi instead of Pseudosuchia, largely because of the philosophical para-

Archosauria dox of putting true crocodilians into a group with a name that means ‘false crocodile’ in Greek (Fig. 2A). Sereno (1991) used neither of these names, instead defining a group including most of the Crocodylotarsi and the Ornithosuchidae as the Crurotarsi (Fig. 2B). However, priority dictates that Pseudosuchia and Ornithosuchia are sister stem taxa of Archosauria; Crurotarsi is a node-defined taxon containing phytosaurs and crocodiles and all descendants of their most recent common ancestor Crocodylotarsi is a redundant junior synonym. A key issue in archosaurian systematics involves the interpretation of the structure of the proximal tarsus (ankle) in Archosauriformes. Crocodilians, and a number of related archosaur groups, have a distinctive tarsal pattern in which a ball on the astragalus articulates with a socket on the calcaneum, with the functional result that the calcaneum, structurally part of the foot, rotates on the astragalus, which is structurally united with the lower leg. One group of extinct archosaurs, the Ornithosuchidae, had a second mobile tarsal pattern, with a ball on the calcaneum that articulates with a socket on the astragalus. In basal archosauriforms and several other closely related groups, the proximal tarsals are united by a pair of facets that essentially prevent mobility between them. Instead, the main movement of the ankle takes place between the proximal tarsus and the distal tarsus, which is functionally connected to the foot. Most of the recent phylogenies of the Archosauria (Benton and Clark, 1988; Gauthier, 1994; Parrish, 1993; Sereno, 1991) agree about the constitution of the crown group Archosauria, although there are considerable differences of opinion about the composition of its constituent groups. A commonly recognized group is the Ornithodira, which comprises the Lagosuchidae, Pterosauria, and Dinosauria. The pterosaurs, the familiar clade of flying archosaurs, are discussed under their own entry. The Lagosuchidae consists of three small, long-legged archosaurs known from the Middle Triassic Chan˜ares Formation of Argentina. The best represented of these three taxa was originally placed into Lagosuchus by Romer (1971) but was later transferred to a new genus, Marasuchus, by Sereno and Arcucci (1994) (see DINOSAUROMORPHA). Gauthier (1986) united the ORNITHODIRA with the Ornithosuchidae, a group of Upper Triassic, quadru-

Archosauria pedal carnivores, in a larger group, the ORNITHOSereno (1991) considered the Ornithosuchia an invalid grouping. He instead erected a group he called the Crurotarsi that united the Ornithosuchia with Gauthier’s Pseudosuchia, with the most important character linking these groups being the presence of mobility in the proximal tarsus as opposed to the mesotarsal ankles in basal archosauromorphs and ornithodirans. However, regardless of the position of the Ornithosuchidae, Pseudosuchia cannot be redefined, and Crurotarsi is now recognized as a node group within Pseudosuchia (see above) (Padian and May, 1993). The rest of the non-dinosaurian archosaurs are now generally put into the Pseudosuchia, as noted previously. The position of one group, the Late Triassic Parasuchia, is also controversial. The parasuchians, or phytosaurs, were an abundant group of large, long-snouted archosaurs that appear to have been rough ecological equivalents of modern crocodilians. In most archosaurian phylogenies (e.g., Benton and Clark, 1988; Gauthier, 1986; Parrish, 1993), the phytosaurs occupy a basal position within the Pseudosuchia. Sereno (1993) places the phytosaurs as the sister group of his Crurotarsi because he interprets their tarsus as being immobile (but see Parrish, 1986). Several other taxa of pseudosuchians are recognized. The Stagonolepididae (aetosaurs) consist of a Late Triassic group of apparently herbivorous archosaurs that had a complete carapace of dermal armor and ranged in size from less than a meter to nearly 10 meters in length. They are distinctive in that they are the first archosaurian herbivores to appear in the fossil record. There are also several groups of tallsnouted, terrestrial pseudosuchians that have often been grouped together as the Rauisuchidae or Rauisuchia. Parrish (1993) divided members of this ecomorph into one group, the Prestosuchidae, that occurs near the base of the Pseudosuchia and a second clade, the Rauisuchia proper, that comprises the other members of the ecomorph plus the Crocodylomorpha. Within the Rauisuchia, two lineages were recognized, the Rauisuchidae (which includes several carnivores along with the enigmatic, beaked, amphibious Lotosaurus) and the Poposauria, a relatively poorly known group including carnivores, some of which may have been capable of running on their hind limbs. Poposaurs are recognized by most of the


23 recent phylogenies (e.g., Clark in Benton and Clark, 1988; Parrish, 1987, 1993; Sereno, 1991) as the closest relatives of the Crocodylomorpha. Benton in Benton and Clark (1988) restricted the Pseudosuchia to a monophyletic clade that he recognized comprising the Stagonolepididae and the Rauisuchidae, which he united on the basis of having ventrally facing hip sockets; however, given priority in the phylogenetic system, this constitution of Pseudosuchia is invalid. The Crocodylomorpha in its modern constitution was recognized by Walker (1970) as a clade including the Crocodylia and a group of long-legged, Triassic– Jurassic pseudosuchians that are often combined within a group, the Sphenosuchia, which may or may not be a monophyletic lineage (e.g., Clark in Benton and Clark, 1988; Walker, 1990). The Crocodylomorpha were initially terrestrial carnivores; they took on their current amphibious/aquatic habitus later in the Mesozoic (Parrish, 1987; Walker, 1970). During the Early Triassic, archosauriforms were still a relatively unimportant element in most terrestrial ecosystems. The Archosauria appeared early in the Triassic and became much more prominent, in terms of both abundance and diversity, during the Middle Triassic. The skeletal record of dinosaurs appeared near the beginning of the Late Triassic, although they did not become abundant parts of terrestrial ecosystems until well into the Norian (latest Triassic). By the end of the Triassic, all the archosaurs other than crocodylomorphs, pterosaurs, and dinosaurs became extinct.

See also the following related entry: THECODONTIA

References Benton, M. J., and Clark, J. M. (1988). Archosaur phylogeny and the relationships of the Crocodylia. in Phylogeny and Classification of Amniotes (M. J. Benton, Ed.), Systematics Association Special Vol. 35A, pp. 295–338. Clarendon Press, Oxford. Bonaparte, J. F. (1975). The family Ornithosuchidae (Archosauria: Thecodontia). Colloq. Int. CNRS 218, 485–501. Brinkman, D. L. (1981). The origin of the crocodiloid tarsi and the interrelationships of thecodontian reptiles. Breviora 464, 1–23.


24 Cruickshank, A. R. I. (1979). The ankle joint in some early archosaurs. South African J. Sci. 75, 168–178. Gauthier, J. A. (1984). A cladistic analysis of the higher systematic categories of the Diapsida, pp. 564. PhD dissertation, University of California, Berkeley. Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8,155. Gauthier, J. A. (1994). The diversification of the amniotes. In Major Features in Vertebrate Evolution. Short Courses in Paleontology 7 (D. R. Prothero and R. M. Schoch, Eds.), pp. 129–159. Paleontological Society, Knoxville, TN. Padian, K., and May, C. L. (1993). The earliest dinosaurs. New Mexico Museum Natl. History Bull. 3, 379–382. Parrish, J. M. (1986). Structure and function of the tarsus in the phytosaurs (Reptilia: Archosauria). In The Beginning of the Age of Dinosaurs (K. Padian, Ed.), pp. 35–43. Cambridge Univ. Press, New York. Parrish, J. M. (1987). The origin of crocodilian locomotion. Paleobiology 13, 396–414. Parrish, J. M. (1993). Phylogeny of the Crocodylotarsi and a consideration of archosaurian and crurotarsan monophyly. J. Vertebr. Paleontol. 13, 287–308. Romer, A. S. (1971). The Chan˜ares (Argentina) Triassic reptile fauna. X. Two new but incompletely known long-limbed pseudosuchians. Breviora 378, 1–10. Sereno, P. C. (1991). Basal archosaurs: Phylogenetic relationships and functional implications. Society of Vertebrate Paleontology Memoir 2. J. Vertebr. Paleontol. 11(Suppl. No. 4), 1–53. Sereno, P. C. (1994). Dinosaurian precursors from the Middle Triassic of Argentina: Marasuchus lilloensis, gen. nov. J. Vertebr. Paleontol. 14, 53–73. Sereno, P. C., and Arcucci, A. B. (1990). The monophyly of crurotarsal archosaurs and the origin of bird and crocodile ankle joints. Neues Jahrbuch Geol. Pala¨ontol. Abhandlungen 180, 21–52.

Arctometatarsalia JOHN R. HUTCHINSON KEVIN PADIAN University of California Berkeley, California, USA


rctometatarsalia (Fig. 1) was established by Holtz (1994b) to encompass all coelurosaurian theropods that shared the ‘‘arctometatarsalian’’ condition, a name given to the proximally pinched third metatarsal by Holtz (1994b; from the Latin arctus, meaning compressed) (see HINDLIMBS and FEET). He defined Arctometarsalia as the first theropod with this condition and all of its descendants, which included by his formulation Ornithomimosauria, Troodontidae, Tyrannosauridae, Elmisauridae, and Avimimus. However, Holtz (1996b) revised this definition because he recognized that an apomorphy-based definition was potentially unstable, and the condition had also been found in Mononykus (Perle et al., 1994), an early bird (see AVIALAE). Moreover, Elmisauridae may be closer to Oviraptoridae than to the other Arctometatarsalia, and ‘‘Avimimus’’ may be composed of several different taxa (Holtz, 1996a,b). Consequently, Holtz (1996b) amended the definition of Arctometatarsalia from an apomorphy-based to a stem-based taxon: the clade comprising Ornithomimus and all theropods closer to Ornithomimus than to birds. Currently, the Arctometatarsalia principally comprise Ornithomimosauria, Troodontidae (which together form BULLATOSAURIA), and Tyrannosauridae.

Walker, A. D. (1970). A revision of the Jurassic reptile Hallopus victor (Marsh). Philos. Trans. R. Soc. London Ser. B 257, 323–372. Walker, A. D. (1990). A revision of Sphenosuchus acutus Haughton, a crocodylomorph reptile from the Elliot Formation (Late Triassic or Early Jurassic) of South Africa. Philos. Trans. R. Soc. London Ser. B 330, 1–120.

Arctic Dinosaurs see POLAR DINOSAURS

FIGURE 1 Phylogeny of Arctometatarsalia, after Holtz and other sources.

Arctometatarsalia Tyrannosaurs are united to bullatosaurs in Arctometatarsalia by the contact of the iliac blades along most of their dorsal surface, a semicircular scar on the anterior of the ischium, the large surangular foramen, an elongate tibia and metatarsus, metatarsals that are deeper anteroposteriorly than mediolaterally, the loss of flexed cervical zygapophyses, and the pinched ‘‘arctometatarsalian’’ condition. In the derived arctometatarsalian pes, the proximal articular surfaces of metatarsals II and IV are mostly to entirely dominant at the proximal end of the metatarsal; metatarsal III is only expressed on the proximal end of the metatarsus in those taxa with a less derived arctometatarsus (e.g., Allosaurus). The solid, compressed shaft of the third metatarsal forms a rigid structure with metatarsals II and IV proximally, and in life must have been bound together with strong ligaments so that the whole metatarsus acted as a cohesive functional unit. In the distal portion of the metatarsus, metatarsal III is hollow, like metatarsals II and IV, and is expanded, often forming the bulk of the distal metatarsus. Metatarsals II and IV are heavily buttressed in this region for their articulation with metatarsal III, which completes the functional integration of the arctometatarsus (see Fig. 2; Holtz 1994a). Some arctometatarsalian-grade theropods, such as Elmisaurus, fuse some of the tarsal elements, including the metatarsals, which would add further rigidity to the pes. This condition is convergent with the condition in some other theropods, such as some ceratosaurs (e.g., Syntarsus) and avialians (e.g., enantiornithines and other birds). The arctometatarsalian pes has certain functional implications (Holtz, 1994a) (see Functional Morphology). Arctometatarsalian-grade theropods in general seem to have a more elongate and gracile pes (and hindlimb) than in other theropods. The biological significance of this is uncertain but it may indicate a degree of increased cursorial ability (Coombs, 1978; Holtz, 1994a). The precise biological action of the arctometatarsus has been a subject of some debate, but Holtz (1994a) and Wilson and Currie (1985) have convincingly argued that it is best interpreted as a forcetransducing structure, channeling and evenly distributing ground-reaction forces during locomotion proximally from the pes across the mesotarsal joint. There does not seem to be strong support for any alternate hypotheses, such as a pistoning action of


FIGURE 2 Representative theropod right metatarsals II–IV (metatarsal V illustrated in F and G; metatarsal I not included); dorsal (above) and anterior (below) views, with scale bar ⫽ 50 mm. From Holtz (1994a, Fig. 1, p. 481). Metatarsi A–E demonstrate the arctometatarsus morphology, whereas F–M exhibit less-derived morphology. Taxa pictured: A, Struthiomimus (Ornithomimidae); B, Albertosaurus (Tyrannosauridae); C, Tochisaurus (Troodontidae); D, Elmisaurus (Elmisauridae); E, ‘‘Avimimus’’ (validity questionable); F, Coelophysis (Coelophysidae); G, Dilophosaurus (Coelophysoidea); H, Ceratosaurus (Neoceratosauria); I, Allosaurus (Allosauridae); J, Chilantaisaurus (Tetanurae incertae sedis); K, Elaphrosaurus (Abelisauroidea); L, Ornitholestes (Coelurosauria incertae sedis); M, Deinonychus (Dromaeosauridae).

metatarsal III, ‘‘snap ligaments’’ (Coombs, 1978), or rotation of metatarsal III (Wilson and Currie, 1985; see Fig. 3 for further discussion). Holtz (1994a) has provided the only detailed functional analysis of the arctometatarsus, but further mysteries regarding the functional morphology of the pes during locomotion remain unresolved, such as the range of possible motion (joint angles) of the intrapedal and tarsal joints, the relationship between metapodial joint mobility and overall theropod hindlimb kinematics, and similarities or differences between nonavian and avian theropod locomotion. Further studies of theropod hindlimb functional morphology combined with trackway studies offer hope of further clarifying the matter. Holtz’s (1996a,b) revisions of the Arctometatarsalia have revealed that the characteristic arctometatarsal-



FIGURE 3 Three hypotheses of arctometatarsalian pes function during locomotion. From Holtz (1994a, Fig. 10, p. 499). Ranges of motion are exaggerated for clarity; force vectors are approximated. (A) ‘‘Pistoning’’ or ‘‘snap ligament’’ model from Coombs (1978). From left to right: (left) metatarsal III is suspended between metatarsals II and IV by elastic ligaments (L), which (middle) store potential energy from the ground-reaction force at footfall as metatarsal III moves proximally, and (right) release the stored potential energy at takeoff, adding thrust to the hindlimb. (B) ‘‘Rotational’’ model from Wilson and Currie (1985). From left to right: (left) elastic ligaments (L) running from the epipodium and the anterior projection (A) of the proximal end of metatarsals II and IV attach to the anterior surface of the proximal end of metatarsal III (shaded). The ligaments are stretched by the weight of the animal at footfall (middle) as metatarsal III rotates around the pivot point (P) and (right) release elastically stored potential energy at takeoff, returning metatarsal III to its relaxed position and adding thrust to the hindlimb. (C) ‘‘Force transmission’’ model from Wilson and Currie (1985) and Holtz (1994a). (Left) Underived theropod metatarsus example: Allosaurus metatarsus showing independent ground-reaction force transmission along each metatarsal, at proximal and distal sections. (Right) Arctometatarsus example: Albertosaurus metatarsus showing the transmission of forces from metatarsal III (distal metapodium; bottom section) to metatarsals II and IV (proximal metapodium; top section) via the wedge and buttress system (metatarsal III forms the ‘‘wedge’’; the surfaces of metatarsals II and IV facing metatarsal III along the distal half of the metatarsus form the ‘‘buttresses’’).

ian-grade pes has evolved several times within Theropoda—most likely independently in the Elmisauridae (including Caenagnathidae), AVIALAE (alvarezsaurids), Arctometatarsalia, and perhaps Avimimus as well. Examples of intermediate character states between the derived arctometatarsus and the primitive, noncompressed theropod metatarsus are few, although the third metatarsals of Alvarezsaurus (Alvarezsauridae), Chilantaisaurus (TETANURAE incertae sedis), Ornitholestes, and some basal Arctometatarsalia [e.g., Harpymimus (Ornithomimosauria)] seem to be less compressed than in more derived taxa within their respective clades (see Holtz, 1994a, pp. 494–496 for discussion). The third metatarsal in nonalvarezsaurid MANIRAPTORA, such as Deinonychus and Archaeopteryx, is not technically arctometatarsalian, but it is reduced compared to the third metatarsals of noncoelurosaurian theropods (cf. Fig. 2).

See also the following related entries: COELUROSAURIA ● ORNITHOMIMOSAURIA ● THEROPODA

References Coombs, W. P., Jr. (1978). Theoretical aspects of cursorial adaptations in dinosaurs. Q. Rev. Biol. 53, 393–418. Holtz, T. R., Jr. (1994a). The arctometatarsalian pes, an unusual structure of the metatarsus of Cretaceous Theropoda (Dinosauria: Saurischia). J. Vertebr. Paleontol. 14, 480–519. Holtz, T. R., Jr. (1994b). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117. Holtz, T. R., Jr. (1995). A new phylogeny of the Theropoda. J. Vertebr. Paleontol. 15(Suppl. to No. 3), 45A.

Australasian Dinosaurs Holtz, T. R., Jr. (1996a). Phylogenetic analysis of the nonavian Tetanurine dinosaurs (Saurischia, Theropoda). J. Vertebr. Paleontol. 16(Suppl. to No. 3), 42A. Holtz, T. R., Jr. (1996b). Phylogenetic taxonomy of the Coelurosauria (Dinosauria: Theropoda). J. Paleontol. 70, 536–538. Karhu, A. A., and Rautian, A. S. (1996). A new family of Maniraptora (Dinosauria: Saurischia) from the Late Cretaceous of Mongolia. Paleontol. J. 30, 583–592. Perle A., Chiappe, L., Barsbold, R., Clark, J. M., and Norell, M. A. (1994). Skeletal morphology of Mononykus olecranus (Theropoda: Avialae) from the Late Cretaceous of Mongolia. Am. Museum Novitates 3105, 1–29. Wilson, M. C., and Currie, P. J. (1985). Stenonychosaurus inequalis (Saurischia: Theropoda) from the Judith River Formation of Alberta: New findings on metatarsal structure. Can. J. Earth Sci. 22, 1813–1817.

Argentinean Dinosaurs Argentinean dinosaurs are found in Upper Triassic to Upper Cretaceous sediments and include prosauropods, theropods, herrerasaurids, sauropods, hadrosaurs, etc.


Armor Armor is a characteristic of ankylosaurs, some sauropods, and stegosaurs.


Asian Dinosaurs Dinosaurs are known from most of the continent of Asia. The best known sites are in Mongolia and China, but there are also well-known sites in middle Asia, Japan, and India.



Australasian Dinosaurs

Arizona Museum of Science and Technology, Arizona, USA see MUSEUMS



Arkansas Geological Commission, Arkansas, USA see MUSEUMS



Arkansas Museum of Science and History, Arkansas, USA see MUSEUMS



RALPH E. MOLNAR Queensland Museum Queensland, Australia

The unique modern mammals of Australia originated and evolved on a continent separated from all others since the beginning of the Cenozoic. During the Mesozoic Era, Australia and New Zealand were parts of the southern supercontinent, and so their dinosaurs might be expected to be similar to those from elsewhere. However, at least in part, there seems to have been an endemic fauna that shared some of the unusual features of the modern Australasian faunas.

Dinosaurs of Australia Australia is not only the smallest but also the flattest of the continents, having the lowest proportion of land with topographic relief. Few fossil-bearing rocks are well exposed. In the Mesozoic, terrestrial body

28 fossils (and the rocks that yield them) are known only from the Early Triassic (Scythian), Late Liassic through Bajocian, and Late ’Neocomian’ through Cenomanian. This represents about 20% of the Mesozoic but only about 10% can be considered wellknown. Many of these rocks are marine, deposited in the epeiric sea that periodically covered the northcentral part of the continent during the Early Cretaceous; only in Victoria are there continental deposits (Valanginian–Albian). Therefore, dinosaur bones are few and far between. However, the trackway record is better, extending (intermittently) from the Late Triassic through the Cenomanian and representing about 45% of the Mesozoic. Late Cretaceous terrestrial tetrapods and continental rocks are almost unknown in Australia. The best known dinosaurs and dinosaurian faunas come from Queensland, which has the greatest area of outcrop, and Victoria, due to the exceptional diligence and perseverance of Tom and Pat Rich. Other specimens are known from South Australia, Western Australia, and the Northern Territory, but only from New South Wales are there more than isolated bones of single taxa. The oldest reported Australian dinosaur, the prosauropod Agrosaurus, was described by Seeley in 1891. The specimen was reportedly collected during the voyage of H. M. S. Fly up the eastern coast of Australia, but extraordinarily there is no record of the event in the log. The specimen may have come from anywhere along the Fly’s route from England and back. However, tracks (probably Eubrontes sp.) show that dinosaurs were present in Late Triassic Queensland. Jurassic bones of any kind are rare, and at no locality is more than one taxon of terrestrial tetrapod known. A partial skeleton of the cetiosaurid Rhoetosaurus was found in southeastern Queensland in Bajocian rocks otherwise yielding only fossil plants. However, a contemporaneous caudal from Western Australia indicates that sauropods were widespread in Australia at this time (Long, 1992). The only other nonmarine tetrapods known are Liassic temnospondyls from southeastern Queensland. Tracks are more informative. They show that ornithopods were present already in the Liassic, and theropods not only were present but also had become quite large (to 10 m long) by Middle Jurassic times. A quadrupedal

Australasian Dinosaurs form, tentatively but probably incorrectly reported as stegosaur (Hill et al., 1966), was also present. Rhoetosaurus seems similar to the later Shunosaurus from Sichuan, China, where Jurassic temnospondyls have also been found. Four regions yield Cretaceous faunas: the southern coast of Victoria, where both a Valanginian–Aptian and an Aptian–Albian fauna are known; northern New South Wales (Albian); north-central Queensland (Albian and Cenomanian); and the northwestern coast of Western Australia (‘Neocomian’). Sauropod, theropod, and ornithopod tracks have recently been found in the Broome Sandstone of Western Australia (Long, 1990). Other tracks may represent stegosaurs, otherwise unknown from Australasia. New discoveries by Tony Thulborn indicate that the trackways are associated with a variety of habitats ranging from lagoons to swamp and forest (Thulborn et al., 1997). The two faunas from Victoria are basically similar in composition, as far as dinosaurs are concerned, but the earlier fauna also has temnospondyls, which are absent from the later (Rich and Rich, 1989). The most common dinosaurs are small hypsilophodontians. A moderately large theropod has been referred to Allosaurus. Unexpectedly, a ceratopsian and ornithomimosaur have been described (Rich and VickersRich, 1994), and a caenagnathid has been reported. The specimens, almost all isolated single bones, were deposited in the braided channels of a large river system flowing westward into the nascent rift valley opening between Australia and Antarctica. At least some of the dinosaurs probably inhabited the closed forests prevalent in the region (Dettmann et al., 1992). All seem to have been relatively small forms, the ornithopods about 1 or 2 m in length and even the Allosaurus only about 6 m long. This is consistent with a closed forest habitat, although the apparent small size may result from fluvial sorting of the isolated bones. Hypsilophodontids are also the most common dinosaurs from the opal mining district at Lightning Ridge, New South Wales (Molnar and Galton, 1986). Theropods, too, were present and miners report having observed what may be sauropod tracks. Other trackways are probably from large ornithopods (Molnar, 1991). The tracks indicate the presence of large dinosaurs, whereas only small forms are represented

Australasian Dinosaurs by the bony fossils, presumably the result of sorting. Their habitat appears to have been an estuary opening into the inland sea (Dettmann et al., 1992). Two kinds of sauropods, Austrosaurus and an unidentified species, are known from the Albian of Queensland, but the most common dinosaurs seem to have been ankylosaurs (Minmi) and large ornithopods (Muttaburrasaurus). Both forms are quite distinct and probably represent endemic lineages. Muttaburrasaurus has several unusual features relating to the feeding apparatus. The postorbital region of the skull is very broad and anteroposteriorly lengthened. All teeth in each toothrow are erupted to the same degree, suggesting that they were replaced simultaneously (Bartholomai and Molnar, 1981). Other character states were previously known only in pachycephalosaurs (transversely broad postorbital bar and contact between pterygoid and BRAINCASE). Minmi had but a single supraorbital and a broad dorsal ossification connecting the ilium to the sacral centra, seemingly in addition to the sacral ribs. Its pelvis is unexpectedly plesiomorphic, with a long postacetabular process on the ilium and a fabrosaurian-like pubis (see PELVIS, COMPARATIVE ANATOMY). These animals presumably lived in the open woodlands thought to have covered the coastal regions to the north and east of the epeiric sea (Dettmann et al., 1992). Sauropod bones, referred to the genus Austrosaurus, are the only Cenomanian body fossils of dinosaurs from Australia (Coombs and Molnar, 1981). However, large and small theropod and ornithopod tracks are known from the trackways at Lark Quarry, near Winton, central Queensland (Thulborn and Wade, 1984). None of the Cenomanian taxa represented by tracks have been found as bones and vice versa. The affinities of the Cretaceous sauropods are not known. They may yet prove to be primitive relatives of the titanosaurids. The dinosaurs of the Queensland Cretaceous seem quite distinct from those elsewhere, whereas those from Victoria appear more similar to other dinosaurs, especially if the identifications of a ceratopsian and ornithomimosaur are correct. On the other hand, these are based on single elements; therefore, maybe distinct but convergent lineages are represented. A single incomplete bone from marine deposits in Western Australia may represent a Late Cretaceous (Maastrichtian) theropod (Long, 1992).

29 Dinosaurs of New Zealand New Zealand has recently produced isolated, but identifiable, dinosaur bones (Molnar and Wiffen, 1997) serendipitously discovered during a research program carried out by amateur workers on the marine saurians of North Island. They were deposited on the bed of a Late Cretaceous estuary. A large but incomplete rib almost certainly represents a sauropod. Pedal phalanges of large and small theropods have been found, as has an incomplete ilium of a small dryosaur-like ornithopod (Wiffen and Molnar, 1989). A rib and caudals indicate a small, probably nodosaurid, ankylosaur. Although none of these taxa can confidently be identified to genus, they indicate that a fauna consisting of at least four and probably five taxa inhabited New Zealand during Late Campanian or Early Maastrichtian times, and thus provide information on a period not represented in Australia. Because New Zealand rifted away from Antarctica, these dinosaurs probably represent components of the Antarctic dinosaur fauna.

Biogeographic and Physiological Implications of Australasian Dinosaurs Australasia today is known for its unusual large tetrapods and prominent relict taxa. In New Zealand— until the coming of humans—the largest land-dwelling animals were not mammals but birds: This is a place where the descendants of Mesozoic dinosaurs still reigned supreme. The unusual nature of the Australian ornithischians Minmi and Muttaburrasaurus suggests that during the Early Cretaceous there was sufficient isolation between Australia and the other continents for these endemic lineages to develop. This is unexpected in view of the connection of Australia with Antarctica until the Eocene and the Patagonian aspect of the Eocene land mammal faunas from West Antarctica (Marenssi et al., 1994). Thulborn (1986) has argued that already in the Scythian, tetrapod faunas of Australia were significantly different— not in composition, but in proportions—from those elsewhere. These anomalies suggest some barrier to dispersal to (and from) Australia during the Mesozoic. This is supported by the occurrence of what seem to have been relict forms, Allosaurus among the dinosaurs, and more notably the temnospondyls (Rich et al., 1992).

Australasian Dinosaurs

30 During the Mesozoic, Australia was at ‘the end of the earth,’ situated at the end of what was basically a long peninsula made up of Africa, South America, and Antarctica. It was the region furthest from the Laurasian lands whose dinosaurs are well known. Therefore, it should be no surprise that some Australasian dinosaurs would be different from those of the Northern Hemisphere and that this isolation permitted the survival of forms that had become extinct elsewhere (Molnar, 1989, 1992). The Late Cretaceous New Zealand fauna lived there after it had rifted north from Antarctica to become insular. This adds another to the short list of insular dinosaurian faunas. The small size of the ankylosaur is consistent with insular ankylosaurs known from elsewhere (e.g., Late Cretaceous Europe), but the sauropod rib fragment shows that large animals did live on these islands. Furthermore, during Campanian–Maastrichtian times New Zealand was still near Antarctica. The fauna lived near the south polar circle, so an insular, near-polar fauna is represented. Marine temperatures suggest that these dinosaurs lived in a climate not significantly different from that of North Island today (mean annual temperature of 앑14⬚C. Although it has been suggested that Alaskan dinosaurs migrated south during the winter, this option was not open to the dinosaurs of New Zealand; it is unlikely that like giant lemmings they swam to Australia for the winter. No large reptiles today live in such climates. This point is made even more strongly by the Early Cretaceous dinosaurs of Victoria that lived south of the Antarctic circle. Paleotemperature measurements indicate a mean annual temperature of between ⫺1 and 9⬚C (Rich and Rich, 1989) (see PALEOCLIMATOLOGY). Although most Australian workers feel the latter is the more reliable of the temperatures, even this is significantly lower than that for New Zealand. These clearly indicate that dinosaurs were capable of inhabiting regions and climates not accessible to large, land-dwelling lepidosaurs, crocodilians, and chelonians. The Cretaceous dinosaurs of Australasia show both regional endemism and the ability to survive under climatic conditions that no later reptiles have been able to tolerate.

See also the following related entries: ASIAN DINOSAURS ● POLAR DINOSAURS

References Bartholomai, A., and Molnar, R. E. (1981). Muttaburrasaurus, a new iguanodontid (Ornithischia: Ornithopoda) dinosaur from the Lower Cretaceous of Queensland. Mem. Queensland Museum 20, 319–349. Coombs, W. P., Jr., and Molnar, R. E. (1981). Sauropoda (Reptilia, Saurischia) from the Cretaceous of Queensland. Mem. Queensland Museum 20, 351–373. Dettmann, M. E., Molnar, R. E., Douglas, J. G., Burger, D., Fielding, C., Clifford, H. T., Francis, J., Jell, P., Rich, T., Wade, M., Rich, P. V., Pledge, N., Kemp, A., and Rozefelds, A. (1992). Australian Cretaceous terrestrial faunas and floras: Biostratigraphic and biogeographic implications. Cretaceous Res. 13, 207–262. Hill, D., Playford, G., and Woods, J. T. (1966). Jurassic Fossils of Queensland, pp. 32. Queensland Palaeontographical Society, Brisbane. Long, J. A. (1990). Dinosaurs of Australia, pp. 87. Reed Books, Sydney. Long, J. A. (1992). First dinosaur bones from Western Australia. Beagle 9, 21–27. Marenssi, S. A., Reguero, M. A., Santillana, S. N., and Vizcaino, S. F. (1994). Eocene land mammals from Seymour Island, Antarctica: Palaeobiogeographical implications. Antarctic Sci. 6, 3–15. Molnar, R. E. (1989). Terrestrial tetrapods in Cretaceous Antarctica. Special publication, Geological Soc. Am. 47, 131–140. Molnar, R. E. (1991). Fossil reptiles in Australia. ln Vertebrate Paleontology of Australasia (P. V. Rich, J. M. Monaghan, R. F. Baird, T. K. Rich, E. M. Thompson, and C. Williams, Eds.), pp. 605–702. Pioneer Design Studio, Melbourne. Molnar, R. E. (1992). Paleozoogeographic relationships of Australian Mesozoic tetrapods. In New Concepts in Global Tectonics (S. Chatterjee and N. Hotton, III, Eds.), pp. 259–266. Texas Tech Univ. Press, Lubbock. Molnar, R. E., and Galton, P. M. (1986). Hypsilophodontid dinosaurs from Lightning Ridge, New South Wales, Australia. Geobios 19, 231–239. Molnar, R. E., and Wiffen, J. (1997). A Late Cretaceous polar dinosaur fauna from New Zealand Cretaceous. Cretaceous Res., in press.

Australasian Dinosaurs Rich, T. H. V., and Rich, P. V. (1989). Polar dinosaurs and biotas of the Early Cretaceous of southeastern Australia. Natl. Geogr. Res. 5, 15–53. Rich, T. H. V., Rich, P. V., Wagstaff, B., Mason, J. R. C. McE., Flannery, T. F., Archer, M., Molnar, R. E., and Long, J. A. (1992). Two possible chronological anomalies in the Early Cretaceous tetrapod assemblages of southeastern Australia. In Aspects of Nonmarine Cretaceous Geology (N. J. Mateer and P.-J. Chen, Eds.), pp. 165–176. China Ocean Press, Beijing.

31 Thulborn. R. A., Hamley, T., and Foulkes, P. (1997). Preliminary report on sauropod dinosaur tracks in the Broome Sandstone (Lower Cretaceous) of Western Australia. Gaia, in press. Thulborn, R. A., and Wade, M. (1984). Dinosaur trackways in the Winton Formation (mid-Cretaceous) of Queensland. Mem. Queensland Museum 21, 413–517. Wiffen, J., and Molnar, R. E. (1989). An Upper Cretaceous ornithopod from New Zealand. Geobios 22, 531–536.

Rich, T. H., and Vickers-Rich, P. (1994). Neoceratopsians and ornithomimosaurs: Dinosaurs of Gondwana origin. Res. Exploration 10, 129–131. Seeley, H. G. (1891). 0n Agrosaurus macgillivrayi (Seeley) a saurischian reptile from the N. E. coast of Australia. Q. J. Geological Soc. London 47, 164–165. Thulborn. R. A. (1986). Early Triassic tetrapod faunas of southeastern Gondwana. Alcheringa 10, 297–313.

Australian Museum see MUSEUMS



Aves LUIS M. CHIAPPE American Museum of Natural History New York, New York, USA


evidence. The available material of Protoavis is fragmentary and its association in the two specimens alleged by Chatterjee is not clear. In addition, some of the elements regarded as avian (e.g., furcula and carinate sternum) can be alternatively interpreted as something else. Likewise, the suggestion of bird-like footprints in Early Jurassic deposits, though interesting, is far from persuasive (Chiappe, 1995a). Discovered over a span of almost 140 years, the eight specimens (including an isolated feather) of the Bavarian Archaeopteryx constitute the most informative evidence of Late Jurassic birds. Many aspects of the biology of Archaeopteryx, however, are surrounded by controversy. Several anatomical characteristics (in particular, the braincase and temporal region of the skull) are matters of intense debate. Its mode of life and flying capability are also hotly debated. It is even unclear whether these specimens belong to a single taxon, Archaeopteryx lithographica, or to several closely related species. Until very recently, with the exception of an isolated feather from Kazakhstan (which some workers have regarded as a leaf), the specimens of Archaeopteryx were the only known Late Jurassic birds. This singular position of the spectacular specimens of Archaeopteryx has been challenged recently by new birds found in beds alleged to be of Late Jurassic age from northeastern China and North Korea. The former report includes several specimens of Confuciusornis from the Yixian Formation—a startling toothless bird with relatively short, clawed wings (Hou et al., 1995, 1996). Although Hou et al. (1995, 1996) have regarded these birds as Late Jurassic, palynological studies suggest a lowermost Cretaceous age for the Yixian Formation (Li and Liu, 1994) and recent 40Ar– 39Ar dates (앑121–122 million years) from the base and top of this formation indicate an even younger, Hauterivian to Aptian age, depending on the selected geological time scale (Smith et al., 1995). The specimen from North Korea appears to preserve portions of the skull, neck, and wing associated with feathers. Although

ves (birds) may be defined as Archaeopteryx plus extinct birds and all descendants of their most recent common ancestor (AVIALAE of Gauthier 1986). Aves are diagnosed by a suite of features, including flight features, hypertrophy of the forelimb to 120– 140% or more of the hindlimb in length, forearm more than 87% of humerus length, tail reduced to 23 or fewer free caudals, etc. (Gauthier 1986, ‘‘Avialae’’). The dinosaurian origin of birds is today strongly supported by the known evidence. This view of birds as feathered, flying theropods makes their treatment here meaningful. Morphological differences between nonavian dinosaurs and modern birds are, however, significant. For a century, studies of early avian evolution focused on the Urvo¨gel Archaeopteryx and the late Mesozoic (and much more derived) hesperornithiforms and ichthyornithiforms because these represented virtually all the available evidence (Figs. 1–3). In recent years, the series of transformations that occurred between the closest avian sister groups and their living representatives have been illuminated by many new finds, which in the past 5 years alone have doubled the number of known Mesozoic basal taxa (Figs. 1 and 3). The diversity of living dinosaurs is very large (almost 10,000 species of birds are usually recognized, and recent estimates are much greater), and their Cenozoic evolutionary history was complex. The current discussion, however, is restricted to their diversity and patterns of evolution in the Mesozoic. This is mainly due to limitations in space and because it is more closely related to other topics treated in this volume (for a discussion of Cenozoic diversity see Olson, 1985).

Mesozoic Avian Diversity The actual fossil record of birds starts in the Late Jurassic (Fig. 1). Claims for an older record are not supported by reliable evidence. The recent identification of Protoavis from the Late Triassic of Texas (Chatterjee, 1991) as a bird is not based on substantiated




FIGURE 1 Nonneornithine genera of the Mesozoic. Very fragmentary, nondiagnosable taxa have been excluded. The year of publication is listed on the right; note that nearly 50% of these taxa have been described after 1990. Patterns on the right refer to those in Fig. 2.

FIGURE 2 Proportions represented by the genera listed in Fig. 1. Patterns correspond to those given in Fig. 1.

(Video) Encyclopedia of Dinosaurs and Prehistoric Animals Edited by Dr Douglas Palmer

dubbed the ‘‘North Korean Archaeopteryx,’’ the manual proportions of this specimen are different from those of Archaeopteryx, in which the digits are proportionally shorter than the metacarpals. None of these new specimens have been fully described in the literature and their chronological significance has yet to be evaluated in light of clarification of their stratigraphic position. The record of birds is far more abundant in the Cretaceous (Fig. 1). Unquestionable osseous remains of birds have been found in all continents with the exception of Africa, where only footprints have been



FIGURE 3 Rate of discoveries of Mesozoic birds. Data based on the year of publication of the genera listed in Fig. 1.

recovered. The most primitive and bizarre looking are Mononykus (Fig. 4) and its allies (e.g., Alvarezsaurus; Fig. 1) from the Late Cretaceous of central Asia and southern Argentina. Bearing a short and robust forelimb that at first glance little resembles an avian wing, the general anatomy of the flightless

Mononykus supports its avian affinity. The functional meaning of the forelimb of Mononykus is puzzling. Although its overall appearance is suggestive of the morphology of digging mammals (e.g., moles), the long, gracile hindlimbs do not support such an idea. In fact, in contrast to digging mammals, in Mononykus the forelimbs were not used in locomotory activities. Mononykus differs from other flightless birds in that its forelimbs, instead of having simplified structures, have robust muscular attachments suggesting a particular function. The next branch of the cladogram illustrated in Fig. 4 is Iberomesornis. This sparrow-sized bird comes from the Early Cretaceous of Spain. In contrast to Archaeopteryx and Mononykus, Iberomesornis shows characters that strongly suggest an enhanced flying capability. Among these characters are the distal caudal vertebrae fused into a pygostyle; an elongate, strut-like coracoid; and a U-shaped furcula. Also from the Cretaceous of Spain but from significantly older deposits is Noguerornis. Although unquestionably basal within avian phylogeny, the fragmentary nature of the only known specimen prevents a precise

FIGURE 4 Cladogram of best known avian taxa. Synapomorphies diagnosing each node are listed on the left. Derived from Perle et al. (1993) and Chiappe (1995a,b).

Aves determination of the phylogenetic relationship of Noguerornis. Nevertheless, this taxon is important in that it documents the presence of a fused carpometacarpus, a U-shaped furcula with an enormous hypocleideum, and a well-developed propatagium as early as 10 million years after Archaeopteryx. The most diversified birds of the Mesozoic were the ENANTIORNITHES, of which more than a dozen valid species have been named (Figs. 1 and 2). First recognized in 1981 (although members of this group were collected as early as the 19th century), the Enantiornithes are known from the very Early Cretaceous to the end of this period and have been recorded in South America, North America, Europe, Asia, and Australia (Chiappe, 1995a,b). The Early Cretaceous enantiornithines include taxa such as Concornis from Spain and Sinornis and Cathayornis from China. These early enantiornithines were small toothed birds, showing characteristics indicating an enhanced flying ability and perching capabilities. More derived enantiornithines, such as the Mongolian Gobipteryx and other Late Cretaceous forms, were toothless and significantly larger. Some of the Argentine forms of the terminal Cretaceous, such as Enantiornis, had a wingspan of over a meter. Likewise, at the end of the Cretaceous, the Enantiornithes present different specializations. At El Brete, in northwestern Argentina, in addition to forms suggesting arboreal and perching specialization, taxa of wading and aquatic habits have been recorded. The Enantiornithes shared a common ancestor with the clade formed by Patagopteryx and the Ornithurae (Fig. 4), the group encompassing hesperornithiforms, ichthyornithiforms, and neornithines (socalled ‘‘modern birds’’). Patagopteryx was a flightless cursorial bird, of the size of a hen, known thus far only from the Late Cretaceous of southern Argentina. Although originally thought to be related to ratites (flightless birds such as the ostrich and its allies), the cladogram in Fig. 4 shows that flightlessness evolved independently in this taxon. The flightless foot-propelled divers, hesperornithiforms, are principally known from Late Cretaceous marine deposits of the North American western interior (Fig. 1), where excellent specimens were collected as early as the 1870s. Fragmentary remains of hesperornithiforms have been reported from the Late Cretaceous of eastern Europe and western Asia. Whether

35 this group was present in the Early Cretaceous is not clear because the affinity of Enaliornis, an alleged hesperornithiform from the Albian of England, is still controversial. The presence of hesperornithiform remains in estuarine deposits indicates that these birds were not exclusively oceanic. Historically (and in some ways both chronologically and geographically) linked to hesperornithiforms are the ichthyornithiforms, also known from the marine Late Cretaceous of North America and Asia. In contrast to hesperornithiforms, ichthyornithiforms were flying birds of the size of a tern. One of their most remarkable features is the notably large head. The early Cretaceous Ambiortus, from the Mongolian Gobi Desert, may be closely related to ichthyornithiforms. The question of which groups of modern birds were already differentiated in the Mesozoic and when they had their first occurrence in the fossil record is puzzling because their putative records (in most cases isolated bones ) are very fragmentary. It is quite clear, though, that forms related to shorebirds (Charadriiformes), loons (Gaviiformes), ducks (Anseriformes), and possibly petrels (Procellariiformes) were present at the end of the Cretaceous. Several fragmentary specimens from older deposits, such as the Early Cretaceous Gansus and Paleocursornis, have also been related to modern birds, although these need to be studied further.

Major Patterns of Avian Evolution The information provided by the numerous new Mesozoic findings have substantially overcome our ignorance about the ‘‘intermediate’’ evolutionary steps between Archaeopteryx and the much more advanced ichthyornithyforms and hesperornithiforms (Fig. 4). Clarification of the pattern of historical relationships of early birds has been coupled with a remarkable interest in physiological, morphofunctional, and developmental studies of nonavian theropods and extant birds. Bone Histology, Rates of Growth, and Physiology Extant birds are known to grow rapidly. Rates of growth in modern birds vary depending on their mode of development (altricial or precocial), but they normally reach adult size within the first year. The

36 bone tissue of extant birds is a fast-deposited, uninterrupted woven-like tissue, known as fibrolamellar bone. The bone microstructure of Mesozoic birds has been studied only in a few taxa. In Hesperornis, the bone microstructure resembles that of modern birds (Houde, 1987), consistent with a rapid rate of growth. Enantiornithines and Patagopteryx, however, have shown important differences with respect to their living counterparts (Chinsamy et al., 1994, 1995). In these basal birds, the bone deposition was cyclically interrupted by nondepositional pauses (lines of arrested growth; LAGs), a pattern of deposition typically absent in extant birds although known to occur in nonavian theropods (Chinsamy and Dodson, 1995). The bone tissue of Patagopteryx was highly vascularized and of fibrolamellar type, as in modern birds and Hesperornis. In contrast, that of the Enantiornithes was completely formed of lamellated bone, a slow-deposited, parallel-fibered bone tissue. The presence of LAGs in Enantiornithes and Patagopteryx, if annual, indicates cyclical pauses during postnatal growth. This pattern of bone deposition is typical of extant vertebrates in which each LAG is formed annually. Thus, the occurrence of LAGs in Enantiornithes and Patagopteryx suggests that their growth was not continuous throughout the year. These Cretaceous birds took more than a single year to acquire adult size. For example, the presence of four and five LAGs in the examined enantiornithine birds indicates that these birds were still growing 4 and 5 years, respectively, after hatching (Chinsamy et al, 1994, 1995) (see GROWTH LINES). Extant birds, along with mammals, are unique among tetrapods in being able to maintain steady rates of temperature-constrained physiological processes under variable climatic conditions. The presence of pauses in bone deposition and the associated slower rate of growth in both Patagopteryx and Enantiornithes may suggest physiological differences with respect to their modern counterparts. In the Enantiornithes, the presence of a compacta formed of only lamellated bone suggests that physiological differences between these and extant birds were even greater. As suggested by Chinsamy et al. (1994, 1995), if rates of growth are related to metabolic rates (see Ruben, 1995, for a skeptical view), these birds may not have been fully endothermic homeotherms in the sense that we think of extant birds. However, this is

Aves not to say that they were ‘‘cold-blooded’’ ectotherms. They might have had an intermediate thermal physiology within the spectrum of ectothermy– endothermy, as has been suggested for some nonavian dinosaurs (Chinsamy and Dodson, 1995). Moreover, mammals, which are considered endothermic homeotherms, show a broad range of thermal strategies and abilities to regulate body temperature. A ‘‘warm-blooded’’ metabolism has been proposed for basal birds such as Archaeopteryx and the enantiornithine Sinornis on the basis of allometric estimates, theoretical predictions, and aerodynamic capabilities, although neither of these birds has been studied histologically. In light of the new inferences discussed previously, the physiological interpretations previously made for Archaeopteryx and Sinornis should be revisited (see PHYSIOLOGY). Ontogeny and Developmental Modes Extant neornithine birds are known to have a broad range of hatchling appearance and conduct, ranging from one end to the other of the precocial–altricial spectrum. Precocial hatchlings are covered with down, they are capable of locomotion, and they leave the nest soon after hatching. In contrast, altricial hatchlings are usually naked and blind, they are incapable of locomotion, and they are fed by their parents in the nest. Several studies have shown that precociality—in particular the stages such as Precocial 1 and 2, in which the active and downy hatchling has a low rate of growth and follows its parents in search of food—is the ancestral condition for modern birds (Starck, 1993). Basal neornithines, such as paleognaths, anseriforms, and galliforms, typically have this type of early postnatal ontogeny (see BEHAVIOR). Mesozoic avian embryos are currently known only from the Late Cretaceous of Mongolia. Enantiornithine embryos regarded as Gobipteryx were likely precocial (Elzanowski, 1985). This interpretation is congruent with the low rates of growth inferred for the Enantiornithes because living precocial birds have slower rates of growth than altricial birds, which generally grow rapidly (Starck, 1993). The hypothesis that precociality is ancestral for modern birds is supported by the presence of this developmental mode in Enantiornithes (and likely in Patagopteryx) and also by recent findings suggesting that nonavian theropod bones were fully ossified (and likely preco-

Aves cial) at the time of hatching (Norell et al., 1994) (see also HETEROCHRONY). Locomotion and Habits Birds are unique among living vertebrates in that forelimbs and hindlimbs are involved in decoupled locomotory systems (see BIPEDALITY). The series of structural transformations from obligatory bipedal nonavian dinosaurs to the cursorial–aerial locomotion of modern birds has evolved differentially during avian evolution. Namely, the development of a modern flight system preceded the evolution of a modern system of cursorial locomotion (Chiappe, 1991, 1995b). Debates surround the aerial capabilities of Archaeopteryx, but soon after the occurrence of this taxon in the fossil record basal taxa such as Noguerornis and Iberomesornis show a series of structural transformations indicative of an enhanced flying capacity. Furthermore, the pedal morphology of Iberomesornis, Concornis, and Sinornis clearly suggests that these early birds were able to perch, indicating that an arboreal type of lifestyle was acquired very early in the evolution of birds. Basal birds inherited the bipedal capabilities of their theropodan ancestors, and a terrestrial habitat was probably ancestral for birds as shown by the anatomy of Archaeopteryx as well as Mononykus and its kin. Gatesy (1995) has shown that the pattern of hindlimb kinematics in basal theropods—retaining mainly the hip-extension mechanism characteristic of extant crocodiles and lizards—was significantly different from that of modern birds in which the hindlimb is moved through a knee-flexion mechanism. Gatesy interpreted the shortening and general reduction of the tail and associated musculature (M. caudifemoralis longus) from basal theropods to modern birds as the primary factor involved in the shift between these mechanisms. The transformation of this complex would have been correlated with the decoupling of the tail from hindlimb kinematics and its final linkage with the flight locomotor mechanism, and with the forward migration of the center of mass and the acquisition of the typical modern avian hindlimb posture. Interestingly, these differences in hindlimb kinematics appear to have had little effect on the actual gait because footprint evidence suggests that the bipedal gait of nonavian theropods was passed almost unaltered to their modern counterparts (Padian and Olsen, 1989). Most important for the cur-

37 rent discussion is that the final stages toward the typical modern pattern of hindlimb kinematics clearly occur rather late in avian history. Basal birds such as Archaeopteryx or Mononykus have relatively long tails and the pattern of limb kinematics was certainly more similar to that of nonavian maniraptoran theropods. Although basal ornithothoracine birds have a short tail with a distal pygostyle, they have retained primitive characters that might have played a role in their hindlimb kinematics. For example, ischiadic and pubic symphyses are preserved in Noguerornis and the enantiornithine Concornis, respectively. Furthermore, in Patagopteryx the tail was probably quite long, as suggested by the long neural spines of the mid-caudals. It is not until the rise of the Ornithurae that the modern morphology of the hindlimb and pelvis is ultimately acquired. Although the tail morphology of several basal birds is still not well known, it appears that a modern avian mechanism of hindlimb movement was not fully developed until the differentiation of the Ornithurae (Chiappe, 1991, 1995b). Minimal Ages and Taxonomic Dynamics The oldest fossils of a lineage provide evidence only for the minimal age of that particular taxon. Thus, a more precise picture of the temporal pattern of clade origination (and character evolution) emerges when the fossil record is calibrated with a phylogenetic hypothesis (Fig. 4), because sister groups originated at the same time. The minimal age for ornithurine birds is Early Cretaceous; although the phylogenetic relationships of Enaliornis and Ambiortus are not yet clear, they certainly belong to the Ornithurae. This implies that the lineage leading to Patagopteryx, the sister group of Ornithurae (Fig. 4), was already differentiated by the Early Cretaceous. Likewise, the earliest Enantiornithes and Iberomesornis are also of Early Cretaceous age. This predicts that early relatives of Mononykus had already arisen by this age as well. Therefore, calibration of the record of these Mesozoic avian clades with the information provided by their interrelationships indicates that all of them were already present at the beginning of the Cretaceous. Likewise, the fact that several neognaths (e.g., charadriiforms and anseriiforms) have reliable occurrences at the end of the Cretaceous implies that their extant outgroups (ratites and tinamous), so far unknown from the Mes-


38 ozoic, were differentiated in this time. In fact, if the Early Cretaceous Gansus is truly a neornithine, the origin of this clade (including ratites as well) is pushed back into the Early Cretaceous. Numerous new findings have shown that the Cretaceous was a time of active diversification for birds. Several basal lineages, in particular the Enantiornithes, evolved diverse lifestyles, and several modern bird lineages have their earliest occurrences. This period, however, was also one of dramatic extinction. The end of the Cretaceous (not necessarily the K–T boundary) documents the extinction of all the basal diversity, including Mononykus and its kin, Enantiornithes, Patagopteryx, hesperornithiforms, and ichthyornithiforms (Chiappe, 1995a).

See also the following related entries: AVIALAE ● BIPEDALITY ● BIRD ORIGINS HETEROCHRONY

References Chatterjee, S. (1991). Cranial anatomy and relationships of a new Triassic bird from Texas. Philos. Trans. R. Soc. (Biol. Sci.) 332(1265), 277–346. Chiappe L. M. (1991). Cretaceous avialian remains from Patagonia shed new light on the early radiation of birds. Alcheringa 15(3–4), 333–338. Chiappe, L. M. (1995a). The first 85 million years of avian evolution. Nature 378, 349–355. Chiappe, L. M. (1995b). Phylogenetic position of the Cretaceous birds of Argentina: Enantiornithes and Patagopteryx deferrariisi. In 3rd Symposium of the Society of Avian Paleontology and Evolution. Courier Forschungsinstitut, Senckenberg, Germany. Chinsamy, A., and Dodson, P. (1995). Inside a dinosaur bone. Am. Sci. 83, 174–180.

Elzanowski, A. (1985). The evolution of parental care in birds with reference to fossil embryos. Acta XVIII Congr. Int. Ornithol. 1, 178–183. Elzanowski, A., and Galton, P. M. (1991). Braincase of Enaliornis, an Early Cretaceous bird from England. J. Vertebr. Paleontol. 11(1), 90–107. Gatesy, S. M. (1995). Functional evolution of the hindlimb and tail from basal theropods to birds. In Functional Morphology in Vertebrate Paleontology (J. Thomason, Ed.), pp. 219–234. Cambridge Univ. Press, Cambridge, UK. Hou, L., Zhou, Z., Martin, L. D., and Feduccia, A. (1995). A beaked bird from the Jurassic of China. Nature 377, 616–618. Hou, L., Martin, L. D., Zhou, Z., and Feduccia, A. (1996). Early adaptive radiation of birds: Evidence from fossils from northeastern China. Science 274, 1164–1167. Houde, P. (1987). Histological evidence for the systematic position of Hesperornis (Odontornithes: Hesperornithiformes). Auk 104, 125–129. Li, W., and Liu, Z. (1994). The Cretaceous palynofloras and their bearing on stratigraphic correlation in China. Cretaceous Res. 15, 333–365. Norell, M. A., Clark, J. M., Dashzeveg, D., Barsbold, R., Chiappe, L. M., Davidson, A. R., McKenna, M. C., and Novacek, M. J. (1994). A theropod dinosaur embryo, and the affinities of the Flaming Cliffs dinosaur eggs. Science 266, 779–782. Olson, S. L. (1985). The fossil record of birds. In Avian Biology (D. S. Farner, J. King, and K. C. Parkes, Eds.), Vol. 8, pp. 79–238. Academic Press, New York. Padian, K., and Olsen, P. E. (1989). Ratite footprints and the stance and gait of Mesozoic theropods. In Dinosaur Tracks and Traces (D. D. Gillette and M. G. Lockley, Eds.), pp. 231–241. Cambridge Univ. Press, Cambridge, UK. Perle, A., Norell, M. A., Chiappe, L. M., and Clark, J. M. (1993). Flightless bird from the Cretaceous of Mongolia. Nature 362, 623–626. Ruben, J. (1995). The evolution of endothermy in mammals and birds: From physiology to fossils. Annu. Rev. Physiol. 57, 69–95.

Chinsamy, A., Chiappe, L. M., and Dodson, P. (1994) Growth rings in Mesozoic avian bones: Physiological implications for basal birds. Nature 368, 196–197.

Smith, P. E., Evensen, N. M., York, D., Chang, M., Jin, F., Li, J., Cumbaa, S., and Russell, D. (1995). Dates and rates in ancient lakes: 40Ar– 39Ar evidence for an Early Cretaceous age for the Jehol Group, northeast China. Can. J. Earth Sci. 32, 1426–1431.

Chinsamy, A., Chiappe, L. M., and Dodson, P. (1995). Mesozoic avian bone microstructure: Physiological implications. Paleobiology 21(4), 561–574.

Starck, J. M. (1993). Evolution of avian ontogenies. In Current Ornithology (D. M. Power, Ed.), Vol. 10, pp. 275– 366. Plenum Press, New York.



The theropod dinosaur taxon Avetheropoda (Fig. 1) was formally defined in the phylogenetic system by Holtz (1994; following use of the name by Paul, 1988) as the node within TETANURAE comprising the stem groups COELUROSAURIA and CARNOSAURIA, as redefined by Gauthier (1986). The node is diagnosed by loss of the obturator foramen, the proximally placed lesser trochanter of the femur, the basal half of metacarpal I closely appressed to metacarpal II, the cnemial process arising out of the lateral surface of the tibial shaft, the pronounced pubic ‘‘foot’’ or ‘‘boot,’’ the posterior tapering of the coracoid, the U-shaped premaxillary symphysis, and the asymmetrical tooth crowns of the premaxilla in cross section (Holtz, 1994). Its members ranged from as early as the late Middle or Late Jurassic through the latest Cretaceous and, in the case of Aves, to the present day. This node appears to be the same as NEOTETANURAE (Sereno et al., 1994), which is a junior synonym of Avetheropoda by priority of publication.

FIGURE 1 Phylogenetic relations of the major groups of Avetheropoda and its out-groups. For details see CARNOSAURIA; COELUROSAURIA.

References Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117.

39 Paul, G. S. (1988). Predatory Dinosaurs of the World. Simon & Schuster, New York. Sereno, P. C., Wilson, J. A., Larsson, H. C. E., Dutheil, D. B., and Sues, H.-D. (1994). Early Cretaceous dinosaurs from the Sahara. Science 265, 267–271.

Avialae KEVIN PADIAN University of California Berkeley, California, USA

Gauthier (1986) established the term Avialae (‘‘birdwings’’) to encompass Archaeopteryx plus ornithurine birds. Gauthier used the term ‘‘ornithurine’’ birds in a somewhat different sense than other workers, defining it as living birds and all taxa closer to them than to Archaeopteryx. In this work, Gauthier also restricted the use of the term ‘‘Aves’’ to crown group birds, that is, extant taxa of birds and all the descendants of their most recent common ancestor (see PHYLOGENETIC SYSTEM; SYSTEMATICS). His purpose in doing so was to maximize the information to taxonomists of soft parts and other structures not usually available in fossils. Moreover, Linneaus’s original concept of Aves did not include fossil forms, inasmuch as Linneaus did not know of them. Acceptance of this redefinition of Aves has been problematic. Archaeopteryx was recognized as a primitive but true bird by the 1880s (because it had feathers, a short tail, and other avian features) following several decades of dispute after its initial description in 1860. In standard textbooks in both ornithology and paleontology of the 20th century, Archaeopteryx has been treated unexceptionally as a bird, though usually in its own Subclass Archaeornithes, Order Archaeopterygiformes, and Family Archaeopterygidae to recognize its distinctness from other birds. Moreover, to millions of Spanish-speaking people around the world, the word Aves means ‘‘bird.’’ Hence, there seems to be strong reason as well as convention for retaining the term Aves to encompass Archaeopteryx, extant birds, and all the descendants of their most recent common ancestor (see BIRD ORIGINS). The term Neornithes is normally applied to the node defining



FIGURE 1 Phylogeny of Avialae.

crown group birds. Avialae (Fig. 1), then, is available to define the stem group consisting of Neornithes and all MANIRAPTORANS closer to them than to Deinonychus; the name DEINONYCHOSAURIA, used by Gauthier (1986) to include Dromaeosauridae plus Troodontidae (the latter since removed to propinquity with Ornithomimidae), can now be used as the stem-based sister taxon to Avialae, defined as Deinonychus and all maniraptorans closer to it than to birds. This is consistent with Gauthier’s original formulation of Deinonycho-

sauria and Avialae as sister stem taxa within Maniraptora. Despite the removal of Troodontidae from the former taxon, the meanings of the two groups remain the same because they can be interpreted as stem-based taxa (see Gauthier, 1986, Fig. 9).

Reference Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55.

B ephemeral streams. Thus, the climatic conditions were more humid than those of the Djadokhta. The Barun Goyot Formation is widely distributed in the Gobi Basin and occurs in the Pre-Altai (locality Udan Sayr), Trans-Altai (localities: Khulsan, southeast Nemegt, Khermeen Tsav, and Ingeni Tsav) and in the Eastern Gobi (localities: Shara Tsav and Khara Khutul). The Barun Goyot Formation yields a vertebrate assemblage similar to that of the Djadokhta Formation, and several species of lizards are shared by these formations. However, lizards are much more diverse in the Barun Goyot Formation. There are also aquatic (fish remains) and amphibious vertebrates (frogs), which are absent in the Djadokhta Formation. Uncommon but relatively diverse turtles and solitary representatives of two small terrestrial crocodiles (a gobiosuchid and a notosuchian) have also been reported from Barun Goyot deposits. The formation has yielded a small bird (Gobipteryx minuta) and numerous vertebrate eggs and eggshells. Except for one sauropod species, dinosaurs are represented by small- and medium-sized species, of which only one species (?Velociraptor mongoliensis) is shared with the Djadokhta. Dinosaur species found in the Barun Goyot Formation include the theropods Avimimus portentosus, Conchoraptor gracilis, Hulsanpes perlei, Ingenia yanshini, and ?Velociraptor mongoliensis; the sauropod Quesitosaurus orientalis; the pachycephalosaurid Tylocephale gilmorei; the ceratopsians Bagaceratops rozhdestvenskyi, Breviceratops kozlowskii, and Udanoceratops tschizhovi, and the ankylosaurs Saichania chulsanensis and Tarchia kielanae. Dinosaur eggs and eggshells include smooth and ornamented protoceratopsid eggs, and dendroolithid eggshells.

Barun Goyot Formation HALSZKA OSMO´ LSKA Polska Akademia Nauk Warsaw, Poland

The Barun Goyot Formation (previously referred to as the ‘‘Lower Nemegt Beds’’) was identified in 1948 as the ‘‘unfossiliferous lacustrine sandstones’’ or ‘‘barren deposits’’ in the Nemegt Basin, Mongolia, by the Palaeontological Expedition of the USSR Academy of Sciences. The Khulsan locality was subsequently designated as the type locality of the Barun Goyot Formation. The Barun Goyot Formation is directly overlaid by a layer of the intraformational conglomerate and the Nemegt Formation. The lower boundary of the formation is covered, and the physical contact with the presumably older Djadokhta Formation has not yet been discovered. The total thickness of the Barun Goyot Formation cannot be precisely determined, but it is not less than 110 m. The estimated age of this formation is not precisely determined. It has been considered as ?middle Campanian, Campanian, or even Maastrichtian by various authors; the last referral seems doubtful however. Red and brownish-red, fine-grained, poorly cemented sandstones dominate the Barun Goyot Formation more than the interstratifying sandy mudstones and sandy claystones. The sandstones are either not stratified or have characteristic large-scale cross-stratification. The dominant lithology of the Barun Goyot Formation resembles that of the underlying Djadokhta Formation but lacks the mature caliche paleosol horizons characteristic of the latter and has thicker, more common claystone layers. Sedimentation of the Barun Goyot Formation is interpreted to have occurred among eolian dunes and interdune deposits but also in small intermittent lakes and




Bastu´s Nesting Site

References Fox, R. C. (1987). Upper Cretaceous terrestrial vertebrate stratigraphy of the Gobi Desert (Mongolian People’s Republic) and western North America Special paper Geological Assoc. Can. 18, 571–594. Gradzinski, R., and Jerzykiewicz, T. (1974). Sedimentation of the Barun Goyot Formation. Palaeontol. Polonica 30, 11–146. Gradzinski, R., Kielan-Jaworowska, Z., and Maryanska, T. (1977). Upper Cretaceous Djadokhta, Barun Goyot and Nemegt formations of Mongolia, including remarks on previous subdivisions. Acta Geol. Polonica 27, 281–318. Jerzykiewicz, T., and Russell, D. A. (1991). Late Mesozoic stratigraphy and vertebrates of the Gobi Basin. Cretaceous Res. 12, 345–377. Osmolska, H. (1980). The Late Cretaceous vertebrate assemblages of the Gobi Desert, Mongolia. Me´m. Soc. Ge´ol. France 139, 145–150.

Bastu´s Nesting Site JOSE´ L. SANZ JOSE´ J. MORATALLA Universidad Auto´noma de Madrid Madrid, Spain


he Bastu´s dinosaur nesting site is found in the Arenisca de Are´n Formation (Maastrichtian, Upper Cretaceous) in the province of Le´rida (northeastern Spain) (Sanz et al., 1995). The sediments are red sandstones laid down in a shoreline environment. The outcrop has a volume of about 12,000 m3 and contains a huge number of eggshell fragments. The eggshell material represents about 0.5% of the whole deposit. Using the volume of the sandstone body, an average egg diameter of 20 cm, and eggshell that is 1.45 mm thick, the number of dinosaur eggs at the Bastu´s site can be estimated at 300,000. The outcrop contains the remains of 24 nesting structures, most of them having two or three eggs, with a maximum of seven. The eggs are subspherical in shape (Fig. 1), with spherulitic shell type. The dinosaurs were nesting in the exposed sandy sediments of a beach-ridge plain. Because the nesting structures are generally well preserved, postdeposi-

FIGURE 1: Dinosaur nest remains at the Bastu´s site (Le´rida province, Spain). Sections of the eggs are visible in the reddish sandstone surface.

tional transport can be excluded as the primary cause for the high level of fragmentation. This is better explained by the trampling and nesting activities of dinosaurs and also by subsequent paedogenesis. Some factor, possibly territorial behavior of the dinosaurs, prevented the destruction of the reasonably well-preserved nests. Two main conclusions may be inferred from evidence provided by the Bastu´s nesting site: (i) Remains of the huge number of eggs in the sandstone body indicate some kind of nesting fidelity, as has been reported in other dinosaur nesting localities (Horner, 1982); and (ii) Bastu´s represents unambiguous evidence of nesting behavior at a seashore locality.

Bayn Dzak See also the following related entries: BEHAVIOR ● EGG MOUNTAIN ● EGGS, EGGSHELLS, AND NESTS


Bayan Mandahu PHILIP J. CURRIE

References Horner, J. R. (1982). Evidence of colonial nesting and ‘‘site fidelity’’ among ornithischian dinosaurs. Nature 297, 675–676. Sanz, J. L., Moratalla, J. J., Dı´az-Molina, M., Lo´pez-Martı´nez, N., Ka¨lin, O., and Vianey-Liaud, M. (1995). Dinosaur nests at the sea shore. Nature 376, 731–732.

Bavarian State Collection for Paleontology and Historical Geology, Munich

The Bavarian State Collections, formerly the royal collections of the king of Bavaria, are housed in association with the University of Munich, Germany. There is a fine collection of material, including many important specimens from the Jurassic Solnhofen limestones and other Mesozoic formations of Germany. The Solnhofen collection is perhaps the finest in Europe and includes many pterosaurs, lizards, sphenodontidans, fishes, and invertebrates, as well as casts of all the Archaeopteryx specimens and the type specimen of the small coelurosaur Compsognathus. There is also a well-prepared collection of tetrapods from the Permian–Triassic deposits of South Africa and a fine specimen of Triceratops acquired from Maastrichtian beds of Montana. Displays highlight these fossils and many others, notably a fine pareiasaur skull and skeleton, some South American archosaurs recovered by von Huene in the 1930s, and restorations of a complete Pteranodon skeleton and a flock of Rhamphorhynchus. Recent acquisitions include fishes, pterosaurs, and other material from the Late Cretaceous Santana Formation of Brazil.

See also the following related entry: MUSEUMS AND DISPLAYS

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada Bayan Mandahu was made famous by the Sino– Canadian Dinosaur Project. Found in Inner Mongolia (China) approximately 300 km from Bayn Dzak in Mongolia, this locality has produced most of the dinosaurs that are characteristic of the Djadokhta Formation ( Jerykiewicz et al. 1993). Evidence from this site suggests that Pinacosaurus and Protoceratops were gregarious animals that sometimes died en masse in sandstorms. An Oviraptor found at Bayan Mandahu had died while laying its eggs in a nest (Dong and Currie 1996).


References Dong, Z. M. and P. J. Currie (1996). On the discovery of an oviraptorid skeleton on a nest of eggs at Bayan Mandahu, Inner Mongolia, People’s Republic of China. Canadian Journal of Earth Sciences 33:631–636. Jerzykiewicz, T., P. J. Currie, D. A. Eberth, P. A. Johnston, E. H. Koster and Zheng J. J. (1993). Djadokhta Formation correlative strata in Chinese Inner Mongolia: an overview of the stratigraphy, sedimentary geology, and paleontology and comparisons with the type locality in the pre-Altai Gobi. Canadian Journal of Earth Sciences 30:2180–2195.

Bayn Dzak HALSZKA OSMO´ LSKA Polska Akademia Nauk Warsaw, Poland

Bayn Dzak lies in the pre-Altai Gobi Desert, Mongolia. The Upper Cretaceous sediments crop out here in a 10-km-long escarpment, the general direction of

Bayn Dzak

44 which is WNW–ESE. Along its course, the escarpment is cut by numerous small canyons. Bayn Dzak is the type locality of the Djadokhta Formation, and it is famous for yielding Late Cretaceous fossil vertebrates, especially primitive mammals, lizards, dinosaurs, and dinosaur eggs. Bayn Dzak was discovered in 1922 by the CENTRAL ASIATIC EXPEDITION of the American Museum of Natural History and was called ‘‘Shabarakh Usu.’’ The American group also explored Bayn Dzak in 1923 and 1925, and they gave the name ‘‘The Flaming Cliffs’’ to the highest (up to 50 m) group of the cliffs. The majority of the Bayn Dzak fossil vertebrates were found at the base of these. Subsequently, Bayn Dzak was visited in 1948 by the PALAEONTOLOGICAL ExPEDITION OF THE USSR ACADEMY OF SCIENCES, and in 1963– 1971 by the POLISH –MONGOLIAN EXPEDITIONS. Although other localities that are richer in dinosaur remains were discovered in Asia later, Bayn Dzak was the first locality where representatives of two new dinosaur groups (the protoceratopsids and oviraptorids) were detected as well as the first Asian representatives of dinosaur groups known previously from other continents (ankylosaurids, troodontids, and dromaeosaurids) and the first known dinosaur hatchlings (of Protoceratops andrewsi). Additionally, the first known dinosaurian (and other vertebrate) eggs, the (then) oldest placental mammals and liz-

ards, also came from Bayn Dzak. Subsequent exploration of the Upper Cretaceous sediments at Bayn Dzak by the Polish–Mongolian Expedition increased the number of fossil vertebrate species, mainly of mammals, lizards, and crocodiles. The fossils found at Bayn Dzak include theropods (Velociraptor mongoliensis, Saurornithoides mongoliensis, and Oviraptor philoceratops), ceratopsians (Protoceratops andrewsi), ankylosaurs (Pinacosaurus grangeri), and at least four types of dinosaur eggs.


References Andrews, R. C. (1932). The New Conquest of Central Asia, Vol. 1. Gradzinski, R., Kazmierczak, J., and Lefeld, J. (1969). Geographical and geological data from the Polish– Mongolian Palaeontological Expeditions. Palaeontol. Polonica 19, 33–82. Granger, W., and Gregory, W. K. (1923). Protoceratops andrewsi, a preceratopsian dinosaur from Mongolia. Am. Museum Novitates 42, 1–9. Novacek, M. J. (1996). Dinosaurs of the Flaming Cliffs. Doubleday/Anchor Books, New York.

Behavior JOHN R. HORNER Montana State University Bozeman, Montana, USA


only conclusive evidence that can be used to determine the identity of particular eggs or egg clutches. Those that are known and have been described include Troodon cf. formosus (Horner and Weishampel, 1988, 1996), Oviraptor philoceratops (Norell et al., 1994), Maiasaura peeblesorum (Horner and Makela, 1979; Hirsch and Quinn, 1990), and Hypacrosaurus stebingeri (Horner and Currie, 1994). Among these four taxa, complete egg clutches are known only from Troodon, which laid approximately 24 eggs in a clutch (Horner, 1987; Varricchio et al., 1997), and Oviraptor, which laid an average clutch of 22 eggs (Sabath, 1991; Norell et al., 1994). Individual eggs of both Troodon and Oviraptor are elongated, and these elongated eggs are found in circular clutches with the tops of each egg pointing inward toward the clutch centers. The clutches of both taxa have average diameters of 50 cm. Also common to both Oviraptor and Troodon are indications of parental attention to the eggs. A skeleton of Oviraptor was found in a brooding position, sitting directly on a clutch of eggs from Mongolia (Norell et al., 1996), and a similar association between an adult Troodon skeleton and a clutch of eggs was discovered in Montana (Varricchio et al., 1997). A very well-preserved clutch of 24 Troodon eggs, discovered on Egg Mountain, also revealed a 12-cm-high, built-up sediment rim extending around the periphery of the clutch, about 12 cm out from the eggs (Varricchio et al, 1997). The rim indicates that the adult Troodon actually invested time and energy into the act of nest construction. Oviraptor clutches are very common in the Upper Cretaceous Djadokhta and Barun Goyot Formations of Mongolia (Sabath, 1991; Mikhailov, 1991; Norell et al., 1994), but because the eggs are found in eolian sandstones it has not been possible to determine whether these were actual Oviraptor nesting horizons or nesting grounds. In contrast, the Egg Mountain and Egg Island sites from the Upper Cretaceous WILLOW CREEK ANTICLINE of Montana reveal extensive nesting

nterpretation of the social behaviors of dinosaurs ranges from factual to highly speculative, depending on the completeness and specificity of the data and their explanation. We know, for example, that the dinosaur Oviraptor sat on its eggs in at least one instance (Norell et al., 1996) and that a Troodon individual constructed a rim around at least one clutch of eggs (Varricchio et al., 1997). We also know that the hadrosaurs Maiasaura and Hypacrosaurus and the theropod Troodon nested in colonies and used particular nesting areas for more than a single year (Horner, 1982; Horner and Weishampel, 1988, 1996; Currie and Horner, 1988; Horner and Currie, 1994). However, we only know with any degree of certainty that these specific animals accomplished these specific behaviors. We can only speculate that other individuals of the same species behaved similarly, or that other related taxa behaved similarly. Additional behaviors, including nesting, gregariousness and its derivatives, (such as herding, migrating, or pack hunting), parental care and its derivatives (such as protection and feeding of young), and display behaviors, are speculative and can only be hypothesized based on geological and paleontological evidence and comparisons to related living taxa (Coombs, 1989, 1990). In modern comparative biology, such hypotheses can be tested not only by analogy to living forms with structures of presumably similar form and function but also by phylogenetic analysis. Phylogenies can test and compare hypotheses about the presence, absence, and sequence of correlated progression of features related to particular behavior (Padian, 1987; Brooks and McLennan, 1991; Weishampel, 1995).

Nesting There is no evidence to suggest that any dinosaurs were born outside the confines of eggs, and yet very few eggs can be demonstrated to have been laid by particular dinosaur species (see EGGS, EGGSHELLS, AND NESTS). In situ, identifiable embryonic remains are the


46 horizons of Troodon (Horner, 1987; Horner and Weishampel, 1988, 1996). The two hadrosaurs, Maiasaura peeblesorum and Hypacrosaurus stebingeri, have less complete clutches of eggs, and none can be demonstrated to have had any particular clutch arrangement. Only one clutch of Maiasaura eggs has been prepared and it contains only 11 eggs, of which the maximum egg diameter does not exceed 12 cm. The eggs appear to have been either spherical or bluntly ovoid in shape, with a maximum volume of 1250 ml. The eggs of Hypacrosaurus, on the other hand, have an average volume of about 3900 ml and are very nearly spherical in shape (Horner and Currie, 1994). The eggs of a related taxon, an as yet unidentified lambeosaurine from the Judith River Formation of central Montana, produced 22 spherical eggs, each with a volume of 4000 ml. The clutch of this lambeosaur is oblong in shape, measuring 150 ⫻ 90 cm. These figures show that there was clearly a difference in egg size among very closely related species. Hadrosaur clutches are found in associated groups, indicating colonial nesting, and through relatively thick units of sediment suggesting prolonged use of the nesting areas. Egg clutches hypothesized to be dinosaurian in origin are found throughout the world (Carpenter and Alf, 1994; Currie, 1996) and are most commonly found on nesting horizons in association with other clutches of similar morphology, suggesting that numerous dinosaurian taxa probably nested in colonies. In addition, most clutches of eggs attributable to dinosaurs have geometric arrangements indicating that the adult invested time in nest construction.

Gregariousness Gregarious behaviors include those in which individuals of particular species congregate in groups, including nesting creches, herding, and pack hunting. In each of these situations interpretations are based on studies of TAPHONOMY, the context in which fossils have accumulated. Nesting creˆches are groups of juveniles that remain in their respective nesting areas for a period of time preceding hatching. Among living birds these creˆches are actually groups of juveniles protected by a number of adults. Abundances of juvenile bones, representing individuals between embryo and neonate size (2.0–2.5 times the linear dimensions of full-term embryos), are commonly found on the nesting horizons

Behavior of M. peeblesorum, H. stebingeri (Horner and Makela, 1979; Horner, 1994), and the unidentified Upper Cretaceous lambeosaurine from Montana. Comparative studies of the nesting grounds of colonial birds have shown that the greatest majority of bones found on a nesting ground are derived from babies that died during their nesting periods or are the remains of animals brought to the nesting area by parents feeding their young (Horner, 1994). The occurrences of abundant baby hadrosaur skeletal remains on horizons that yield hadrosaur eggs strongly suggest that the young posthatching hadrosaurs remained in their nesting areas for some time after eclosion (hatching). A group of 15 Maiasaura juveniles, all of equal size, found in a nest-like structure on a nesting horizon, has been used as one line of evidence to hypothesize that maiasaurs were born altricial and that the siblings remained together, attended to by one or both adults (Horner and Makela, 1979; Horner, 1984). This is elaborated on under Parental Care. Other groups of juveniles of similar size have been discovered (Gilmore, 1917, 1929; Dodson, 1971; Horner and Makela, 1979; Forster, 1990) and suggest that juvenile dinosaurs representing several species may have engaged in some fashion of congregational behavior (see NEOCERATOPSIA). Another kind of gregarious behavior in which some dinosaurs appear to have engaged is aggregation or herding, as evidenced by nearly monospecific bone beds of a variety of taxa (see Varricchio and Horner, 1993, for details). Interestingly, the largest bone beds, composed primarily of one dinosaur species, are those representing taxa of ceratopsians and hadrosaurians (Currie and Dodson, 1984; Nelms, 1989; Rogers, 1990; Varricchio and Horner, 1993). Most of these monospecific groups are represented by tens of individuals to several hundred individuals, and there is a rare instance in which the numbers of individuals appear to be in the thousands (Hooker, 1987). Most of these ceratopsian and hadrosaurian groups appear to represent animals that died in catastrophic events such as floods, droughts, or volcanicrelated events. Also, because the animals died in these catastrophes, it stands to reason that they were actually in groups or herds (Currie, 1981; Currie and Dodson, 1984) before their catastrophic deaths. Because both the hadrosaurs and the ceratopsians were plant eaters, it is reasonable to hypothesize that these animals migrated from one area to another seeking food

Behavior resources (Currie, 1989). Other nearly monotaxic groups of herbivorous dinosaurs that have been hypothesized to represent herds or aggregations include the Iguanodon assemblage of Bernissart, Belgium (Dupont, 1878) and the Plateosaurus assemblage of Trossingen, Germany (Huene, 1928). Both of these have recently been determined not to represent catastrophic events but rather time-averaged accumulations or some other more complex taphonomic situation (Norman, 1985; Weishampel, 1984). Additional evidence of herding and aggregating can be postulated from dinosaur trackways (Bird, 1944; Bakker, 1968; Ostrom, 1972, 1985; Thulborn, 1979; Lockley et al., 1983; Lockley and Hunt, 1995), although it is rare to be able to demonstrate that a group of animals actually traveled together (Currie, 1983) rather than in the same area over an extended period of time. Animals walking along the shores of lakes, rivers, or marine shores are most likely to travel in one of two directions, parallel to the body of water, regardless of whether they are together in a group (Lockley and Hunt, 1995). Packing or pack hunting is another form of gregariousness, and probably the most speculative. Pack hunting has been postulated, based on the discovery of a group of five Deinonychus skeletons in association with a partial Tenontosaurus skeleton (Ostrom, 1969), and more reliably on the basis of several associations of multiple shed Deinonychus teeth in aggregation with skeletons of Tenontosaurus that appear to have been preyed upon by the theropods (Maxwell and Ostrom, 1995). In one instance a Tenontosaurus skeleton is missing portions of its legs and its rib cage has obviously been pulled open. Eleven Deinonychus teeth found in close association with the herbivore skeleton imply that the animal was eaten by more than one carnivore. Because the tenontosaur shows no sign of having been scavenged, it can be hypothesized that more than one individual was involved in the actual kill. In addition to these occurrences are the nearly monospecific assemblages of Coelophysis from the GHOST RANCH QUARRY and Allosaurus from the CLEVELAND –LLOYD QUARRY. Both of these unusual assemblages contain numerous individuals representing a variety of ontogenetic stages. Farlow (1987) suggested that these assemblages might represent habitat preferences, and that the animals may have died during a particular part of their breeding season. Coombs

47 (1990) hypothesized that a catastrophic event might have driven the animals together. Additionally, a group of four associated specimens of Troodon from Montana, representing two juveniles, a subadult, and an adult, may represent a family unit that perished together on the shores of a freshwater lake (Varricchio, 1995).

Parental Care Parental care includes any investment that the parent or parents of a particular taxon make toward their offspring. Among dinosaurs it appears that several taxa made investments, from brooding (Norell et al., 1996; Varricchio et al., 1997) of eggs to protecting and possibly feeding helpless young (Horner and Weishampel, 1988; Horner et al., 1997). The only cases of direct evidence of parental care, however, are the association of the adult Oviraptor on the clutch of eggs discovered in Mongolia (Norell et al., 1996) and a partial Troodon skeleton found atop a clutch of Troodon eggs (Varricchio et al., 1997). The association of juvenile dinosaurs on the various hadrosaur nesting grounds, described previously, strongly suggests that the juveniles were being protected by adults (Horner, 1994), as occurs among living archosaurs. The group of 15 maiasaurs of equal size found in a bowl-shaped, nest-like structure may suggest that these dinosaurs were nestbound, and therefore in need of parental care and feeding. Additional evidence for this behavior is suggested by studies of the internal structure of the leg bones. A recent quantitative study (Horner et al., 1997) determined that the leg bones of a full-term embryonic hadrosaurs contained very little ossified bone, and that the percentages of ossified bone and cartilaginous tissues in these dinosaurs are equivalent to similar percentages found in extant altricial birds. These data contradict a previous qualitative study (Geist and Jones, 1996) suggesting that Maiasaura and other hadrosaurian taxa were born precocial.

Visual Display Another controversial aspect of behavior pertains to behaviors elicited by physical releasing mechanisms. These include morphological features that command behavioral responses through visualization. Examples include characteristics such as horns, frills, spikes, crests, bosses, plates, thickened skulls, long teeth, big eyes, and so on. (see ORNAMENTATION)


48 Some of the most obvious display features are the horns and crests found on the skulls of many dinosaurs and, in particular, dinosaurs hypothesized to have traveled in groups, such as the ceratopsians and hadrosaurs (Farlow and Dodson, 1975; Molnar, 1977; Spassov, 1979; Sampson, 1995a,b; Dodson, 1996). Farlow and Dodson (1975) and Sampson (1995b) have postulated that the horned dinosaurs used their horns and frills in frontal engagements, shoving and wrestling to settle their hierarchical disputes, much like horned mammals that use their horns to attract mates and determine hierarchical rank (Geist, 1966). Molnar (1977) extended this argument to include all dinosaurian taxa that have exceptional cranial characteristics, including caniniform teeth characteristic of some heterodontosaurs and jugal bosses characteristic of ceratopsians. The thickened skulls of pachycephalosaurs have also been suggested to be linked to hierarchical combat and display (Galton, 1971), and it is likely that the enlarged crest-like structures on the skulls of the oviraptors were also some kind of display feature. Other dinosaurs, such as Ceratosaurus and Monolophosaurus, have horns on the tops of their noses, and Allosaurus had horns over the upper front part of its orbits. Carnotarsus and the tyrannosaurids had horns over their orbits and apparently some kind of nasal ornamentation as well. The broad, flat dorsal plates or dorsal spikes of Stegosaurus and other STEGOSAURS could have served a dual purpose—in active defense mechanism and also in lateral display that would have made the animals look larger or more formidable. This could have acted as a passive defense mechanism or an attractant for mates. Some neotenous or paedomorphic morphological characters (see HETEROCHRONY) have been suggested as mechanisms that may have stimulated adults to respond to their juveniles in particular ways, such as care and feeding (Lorenz, 1971). Interestingly, baby hadrosaurs do not have the cranial characteristics of their adult counterparts, but instead have features, including large eyes, rounded heads, and shortened snouts, that are very similar to the features found in babies of extant birds and mammals that are cared for by parents. Although highly speculative, it may be reasonable to hypothesize that these neotenous features in the baby hadrosaurs triggered some form of parental care. Obviously, many other behaviors can be hypothesized as related to social interactions, including vocal-

ization (Weishampel, 1981), but such features or characters require considerable hypothetical reasoning and are based primarily on the notion that all related living forms possess the characters.


References Bakker, R. T. (1968). The superiority of dinosaurs. Discovery 3, 11–22. Bird, R. T. (1944). Did ‘‘brontosaurs’’ ever walk on land? Nat. History 53, 61–67. Brooks, D. R., and McLennan, D. A. (1991). Phylogeny, Ecology, and Behavior. Univ. of Chicago Press, Chicago. Carpenter, K., and Alf, K. (1994). Global distribution of dinosaur eggs, nests and babies. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 15–30. Cambridge Univ. Press, Cambridge, UK. Coombs, W. P., Jr. (1989). Modern analogues for dinosaur nesting and parental behavior. In Paleobiology of Dinosaurs (J. O. Farlow, Ed.), Spec. Pap. Geol. Soc. Amer. No. 238, pp. 21–53. Coombs, W. P., Jr. (1990). Behavior patterns of dinosaurs. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 32–42. Univ. of California Press, Berkeley. Currie, P. J. (1981). Hunting dinosaurs in Alberta’s huge bonebed. Can. Geogr. J. 101(4), 32–39. Currie, P. J. (1983). Hadrosaur trackways from the lower Cretaceous of Canada. Acta Palaeontol. Polonica 28, 63–73. Currie, P. J. (1989). Long distance dinosaurs. Nat. History 1989(6), 61–65. Currie, P. J. (1996). The Great Dinosaur Egg Hunt. Natl. Geogr. 189(5), 96–111. Currie, P. J., and Dodson, P. (1984). Mass death of a herd of ceratopsian dinosaurs. In Third Symposium on Mesozoic Terrestrial Ecosystems (W.-E. Reif and F. Westphal, Eds.), pp. 61–66. Attempto Verlag, Tubingen, Germany. Currie, P. J., and Horner, J. R. (1988). Lambeosaurine hadrosaur embryos (Reptilia: Ornithischia). J. Vertebr. Paleontol. 8, 13A. Dodson, P. (1971). Sedimentology and taphonomy of the Oldman Formation (Campanian), Dinosaur Provincial Park, Alberta (Canada). Palaeogeogr. Palaeoclimatol. Palaeoecol. 10, 21–74. Dodson, P. (1996). The Horned Dinosaurs, pp. 346. Princeton Univ. Press, Princeton, NJ.

Behavior Dupont, E. (1878). Sur la de´couverte d’ossements d’Iguanodon, de poissons et de ve´ge´taux dans la fosse Sainte Barbe du Charbonnage de Bernissart. Bull. Acad. R. Belgique Se´r. 2 46, 387–408. Farlow, J. O. (1987). Speculations on the diet and digestive physiology of herbivorous dinosaurs. Paleobiology 13, 60–72. Farlow, J. O., and Dodson, P. (1975). The behavioral significance of frill and horn morphology in ceratopsian dinosaurs. Evolution 29, 353–361. Forster, C. A. (1990). Evidence for juvenile groups in the ornithopod dinosaur Tenontosaurus tilletti Ostrom. J. Paleontol. 64(1), 164–165. Galton, P. M. (1971). A primitive dome-headed dinosaur (Ornithischia: Pachycephalosauridae) from the Lower Cretaceous of England, and the function of the dome in pachycephalosaurids. J. Paleontol. 45, 40–47. Geist, N. R., and Jones, T. D. (1996). Juvenile skeletal structure and the reproductive habits of dinosaurs. Science 272, 712–714. Geist, V. (1966). The evolution of horn-like organs. Behaviour 27, 175–214. Gilmore, C. W. (1917). Brachyceratops, a ceratopsian dinosaur from the Two Medicine Formation of Montana. United States Geological Survey Professional Paper No. 103, pp. 1–45.

49 Horner, J. R., and Makela, R. (1979). Nest of juveniles provides evidence of family structure among dinosaurs. Nature 282, 296–298. Horner, J. R., and Weishampel, D. B. (1988). A comparative embryological study of two ornithischian dinosaurs. Nature 332, 256–257. Horner, J. R., and Weishampel, D. B. (1996). Correction to: A comparative embryological study of two ornithischian dinosaurs (1988). Nature 383, 103. Horner, J. R., Horner, C. C., and Eschberger, B. (1997). Altricial dinosaurs: Evidence from embryonic and perinatal bone histology. Paleobiology, in press. Huene, F. von (1928). Lebensbild des Saurischier-Vorkommens imobersten Keuper von Trossingen in Wurttemberg. Palaeobiologica 1, 103–116. Lockley, M., and Hunt, A. P. (1995). Dinosaur Tracks, and Other Fossil Footprints of the Western United States, pp. 338. Columbia Univ. Press, New York. Lockley, M. G., Yang, S. Y., and Carpenter, K. (1983). Hadrosaur locomotion and herding behavior: evidence from footprints in the Mesaverde Formation, Grand Mesa Coal Field, Colorado. Mountain Geologist 20(1), 5–14. Lorenz, K. (1971). Studies in Human and Animal Behavior, Vol. 2, pp. 366. Harvard Univ. Press, Cambridge, MA.

Gilmore, C. W. (1929). Hunting dinosaurs in Montana. Explorations Field Work Smithsonian Inst. 1928, 7–12.

Maxwell, W. D., and Ostrom, J. H. (1995). Taphonomy and paleobiological implications of Tenontosaurus– Deinonychus associations. J. Vertebr. Paleontol. 15(4), 707–712.

Greene, H. W. (1986). Diet and arboreality in the emerald monitor, Varanus prasinus, with comments on the study of adaptation. Fieldiana 1370, 1–12.

Mikhailov, K. E. (1991). Classification of fossil eggshells of amniotic vertebrates. Acta Palaeontol. Polonica 36, 193–238.

Hirsch, K. F., and Quinn, B. (1990). Eggs and eggshell fragments from the Upper Cretaceous Two Medicine Formation of Montana. J. Vertebr. Paleontol. 10, 491–511.

Molnar, R. E. (1977). Analogies in the evolution of display structures in ornithopods and ungulates. Evolutionary Theor. 3, 165–190.

Hooker, J. S. (1987). Late Cretaceous ashfall and the demise of a hadrosaurian ‘‘herd.’’ Geol. Soc. Am. (Rocky Mountain Section), 19, 284. [Abstracts with Programs]

Nelms, L. G. (1989). Late Cretaceous dinosaurs from the North Slope of Alaska. J. Vertebr. Paleontol, 9, 34A.

Horner, J. R. (1982). Evidence of colonial nesting and ‘‘site fidelity’’ among ornithischian dinosaurs. Nature 297, 675–676. Horner, J. R. (1987). Ecologic and behavioral implications derived from a dinosaur nesting site. In Dinosaurs Past and Present, Vol. II (S. J. Czerkas and E. C. Olson, Eds.), pp. 51–63. Univ. Washington Press, Seattle. Horner, J. R. (1994). Comparative taphonomy of some dinosaur and extant bird colonial nesting grounds. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 116–123. Cambridge Univ. Press, Cambridge, UK. Horner, J. R., and Currie, P. J. (1994). Embryonic and neonatal morphology and ontogeny of a new species of Hypacrosaurus (Ornithischia, Lambeosauridae) from Montana and Alberta. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 312–336. Cambridge Univ. Press, Cambridge, UK.

Norell, M. A., Clark, J. M., Demberelyin, D., Rhinchen, B., Chiappe, L. M., Davidson, A. R., McKenna, M. C., Altangerel, P., and Novacek, M. J. (1994). A theropod dinosaur embryo and the affinities of the Flaming Cliffs dinosaur eggs. Science 266, 779–782. Norell, M. A., Clark, J. M., Chiappe, L. M., and Dashzeveg, D. (1996). A nesting dinosaur. Science 378, 774–776. Norman, D. B. (1985). The Illustrated Encyclopedia of Dinosaurs, pp. 208. Crecent, New York. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Peabody Museum Nat. History Bull. 30, 1–165. Ostrom, J. H. (1972). Were some dinosaurs gregarious? Palaeogeogr. Palaeoclimatol. Palaeoecol. 11, 287–301. Ostrom, J. H. (1985). Social and unsocial behavior in dinosaurs. Field Museum Nat. History 55(9), 10–21. Padian, K. (1987). A comparative phylogenetic and functional approach to the origin of vertebrate flight. In Re-

Bernard Price Institute

50 cent Advances in the Study of Bats (B. Fenton, P. A. Pacey, and J. M. V. Rayner, Eds.), pp. 3–22. Cambridge Univ. Press, Cambridge, UK. Rogers, R. R. (1990). Taphonomy of three dinosaur bonebeds in the Upper Cretaceous Two Medicine Formation of northwestern Montana: Evidence for drought-related mortality. Palaios 5, 394–413. Sabath, K. (1991). Upper Cretaceous amniote eggs from the Gobi Desert. Acta Palaeontol. Polonica 36, 151–192. Sampson, S. D. (1995a). Two new horned dinosaurs from the Upper Cretaceous Two Medicine Formation of Montana; with a phylogenetic analysis of the Centrosaurinae (Ornithischia: Ceratopsidae). J. Vertebr. Paleontol. 15(4), 743–760. Sampson, S. D. (1995b). Horns, herds, and heirarchies. Nat. History 104(6), 36–40. Spassov, N. B. (1979). Sexual selection and the evolution of horn-like structures in ceratopsian dinosaurs. Paleontol. Stratigr. Lithol. 11, 37–48. Thulborn, R. A. (1979). Dinosaur stampede in the Cretaceous of Queensland. Lethaia 12, 275–279. Varricchio, D. J. (1995). Taphonomy of Jack’s Birthday Site, a diverse dinosaur bonebed from the Upper Cretaceous Two Medicine Formation of Montana. Palaeogeogr, Palaeoclimatol, Palaeoecol. 114, 297–323. Varricchio, D. J., and Horner, J. R. (1993). Hadrosaurid and lambeosaurid bone beds from the Upper Cretaceous Two Medicine Formation of Montana: taphonomic and biologic implications. Can. J. Earth Sci. 30, 997–1006. Varricchio, D. J., Jackson, F., Borlowski, J., and Horner, J. R. (1997). Nest and egg clutches for the theropod dinosaurs Troodon formosus and evolution of avian reproductive traits. Nature, in press. Weishampel, D. B. (1981). Acoustic analysis of potential vocalization in lambeosaurine dinosaurs (Reptilia: Ornithischia). Paleobiology 7(2), 252–261. Weishampel, D. B. (1984). Trossingen: E. Fraas, F. von Huene, R. Seemann and the ‘‘Schwa¨bische Lindwurm’’ Plateosaurus. In 3rd Symposium Mesozoic Terrestrial Ecosystems, Short Papers (W.-E. Reif and F. Westphal, Eds.), pp. 249–253. Attempto Verlag, Tu¨bingen, Germany. Weishampel, D. B. (1995). Fossils, function, and phylogeny. In Functional Morphology in Vertebrate Paleontology (J. J. Thomason, Ed.), pp. 34–54. Cambridge Univ. Press, Cambridge, UK.

Beijing Natural History Museum, People’s Republic of China see MUSEUMS



Bernard Price Institute for Palaeontological Research, University of the Witwatersrand, Johannesburg, South Africa ANUSUYA CHINSAMY South African Museum Cape Town, South Africa

The Bernard Price Institute for Palaeontological Research, established in 1945, houses a fairly extensive collection of cranial and postcranial elements of the prosauropod Massospondylus and fewer postcranial elements of Euskelosaurus. Of the heterodontosaurids, specimens of Heterodontosaurus tucki and Lycorhinus angustidens (⫽Lanasaurus scalpridens) can be found in the Johannesburg collections. Theropods are represented by skull and postcranial elements of Syntarsus rhodesiensis, as are eggs and eggshell fragments from the Early Jurassic of South Africa. The recently completed museum includes a mounted, partial skeleton of Euskelosaurus and an articulated skull and postcranial skeleton of Massospondylus that is still embedded in its original matrix.


Bernissart Museum, Belgium see MUSEUMS






Biogeography JEAN LE LOEUFF Muse´e des Dinosaures Espe´raza, France


s land animals incapable of crossing wide water barriers, dinosaurs yield important information about the ancient distribution of land masses and seas (paleogeography) during the Mesozoic Era. When previously continuous dinosaur populations were isolated from each other, they evolved on their own and produced so-called endemic forms. As a modern example of an endemic terrestrial fauna, Australia shows a completely original assemblage of marsupial mammals because it became isolated from other continents at the end of the Mesozoic Era before placental mammals became dominant in the rest of the world. The ancestors of modern Australian marsupials had no contact with placentals and evolved on their own while their relatives on other lands were largely replaced by placentals. To the contrary, when two endemic dinosaur faunas were reassociated because of the regression of a sea, they mixed together and many taxa became extinct. A review of the dinosaur distribution patterns from the Late Triassic to the Late Cretaceous adds new information to the paleogeographic reconstructions of the world established by geologists. The distribution of extinct land animals and its relationship to former geographies is not a new subject. It was discussed as early as 1750 by the French naturalist Buffon, who considered that Europe and America had been separated recently because fossil elephants are found on both continents. In Buffon’s mind, it was clear that elephants had colonized a large area that was later divided by water. The first contribution to dinosaur biogeography itself is probably that of the American paleontologist Richard Swann Lull in 1910. Of course, Lull’s contribution predates the establishment of the theory of plate tectonics, and it reflects a now old-fashioned view. Like most geologists of his time, Lull believed the continents had always occupied their present position, and that variations in sea level resulted in geographies only slightly different from those of today. Although claiming that ‘‘the significance of terrestrial verte-

brates . . . in throwing light upon the isolation and connection of the continents is becoming more and more appreciated’’ and that ‘‘the dinosaurs, with their known geological range throughout the entire Mesozoic period, and of almost world wide distribution, are the most significant vertebrates of Secondary times,’’ Lull did not understand the importance of his data for changing concepts of geology. This vision of fixed paleogeography became the model for later reflections on dinosaur biogeography and culminated in the 1960s with the paleogeographic atlases of the French paleontologists H. and G. Termier, who drew ‘‘intercontinental bridges’’ of a mysterious nature reaching several thousand kilometers in length and allowing transoceanic migrations of various animals. With the establishment of a universally accepted model of plate tectonics in the 1960s, these kinds of explanations became untenable, and later papers were written in a completely different way. A single and major exception to the stability dogma of the first half of the century was the brilliant Hungarian paleontologist Franz Nopcsa’s posthumous paper, published in 1934, that incorporated Wegener’s contemporary theory of continental drift with his incomparable knowledge of the fossil record outside North America. The mechanisms of the observed migrations (‘‘The formation of an Equatorial belt of folding, helping the migration of reptiles from one block of sial to another’’) and observed differences (‘‘The gradual cleavage of old masses of sial, beginning in the Permian and lasting until the Tertiary period. This cleavage accentuated more and more the separation of the different faunas. . . . Transitory inundations of old masses (transgressions) which isolated some part of the firm land and, when receding, opened again new connections, thereby disturbing the ‘economic equilibrium of Nature’’) were also analyzed by Nopcsa in a very modern way. The differences between Nopcsa’s work and the hypotheses advanced below rely mainly on a better fossil record and new advances in



52 plate tectonics rather than on a fundamental change in the concepts. With increasing worldwide information about dinosaurs, new patterns of dinosaur biogeography are emerging (Bonaparte and Kielan-Jaworovska, 1987; Rage, 1981; Buffetaut et al., 1988; Le Loeuff, 1991; Russell, 1993). A paleobiogeographic study relies at first on the identification of the different faunal provinces (paleobioprovinces) at a given time, characterized by their own land animals. Initially, one can consider that each continental mass is occupied by a particular biota. However, other nonoceanic barriers may prove to be uncrossable for terrestrial vertebrates because of latitudinal, altitudinal, ecological, climatic, or marine (epicontinental seas) barriers. Although latitudinal variations seem to have had little influence on the global distribution of dinosaurs, epicontinental seas on the different land masses have provoked dramatic long-time separations between the terrestrial faunas separated by the barriers. An epicontinental sea (called the Uralian Sea) separated Europe from Asia during the Jurassic and the Cretaceous, preventing any exchange between the two provinces and leading to an important endemism in central Asia during the Jurassic and Early Cretaceous. As a result, paleobioprovinces do not fit exactly with the continental plates recognized by geologists. On the other hand, when two plates are converging, land communication may be possible before the collision across island arcs; in this case, plates are still unique in a geographic sense until complete suture, but a single paleobioprovince exists.

Late Paleozoic and Early Mesozoic paleogeography is characterized by the progressive unification of the main land masses into a single continent called Pangaea (Fig. 1). This enormous continent was not divided by epicontinental seas, and the Early Triassic fauna is rather uniform throughout the world. Pangaea then began to break into different continents, and dinosaur biogeography reflects both this dislocation of Pangaea and the later adventures of its different pieces, which first diverged and then sometimes touched each other again. In the Late Triassic, the supercontinent Pangaea was indented to the east by an equatorial ocean, called Tethys, which existed until the Tertiary (its closure led to the formation of alpine mountains from Central Europe to the Himalayas). This triangular ocean extended from southern Europe and Africa to the west and joined the megaocean Panthalassa between Australia and Asia. The most abundant dinosaurs of Pangaean times (Late Triassic to Middle Jurassic; 230–165 Ma) are the herbivorous prosauropods, of which six families are now recognized. Several families had a large distribution, such as the Plateosauridae (North and South America, Europe, and China) and the Melanorosauridae (South America, Europe, and Africa). Others appear to be endemic, such as the North American Anchisauridae, the Chinese Yunnanosauridae, and the South African Blikanasauridae. The theropod Syntarsus is known from South Africa and North America. It is difficult during these early times of the dinosaur era to recognize paleobioprovinces. Using other land

FIGURE 1 Late Triassic reconstruction of the paleobioprovinces (after Dercourt et al., 1993; drawing by Guy Le Roux).



vertebrates, Martin (1981) and Rage (1988) defined a ‘‘Peritethysian’’ province including Europe, North Africa, North America, India, Madagascar, and perhaps Southeast Asia. The Peritethysian province is characterized by phytosaurs (crocodile-like reptiles) and metoposaurian amphibians. However, despite regional differences that are difficult to analyze, Gondwanan and Laurasiatic faunas were not different in the Late Triassic, although one can suggest that a kind of latitudinal variation existed. Russell (1993) has demonstrated that two paleobioprovinces might be defined during the Jurassic: a central Asian province with marked endemicity and a ‘‘Neopangean’’ one corresponding to Pangaea without central Asia. The isolation of central Asia from the rest of the world is possibly linked with the transgression of an epicontinental sea between Europe and Asia, called the Uralian Sea. From the Bathonian (165 Ma) to the Tertiary, it seems that this Uralian Sea acted as an uncrossable barrier for land vertebrates. Central Asia was isolated from Neopangaea (165–145 Ma) and later from Euramerica (145–110 Ma) until the Middle Cretaceous, when a land route opened between North America and Asia across the Bering Strait. During the Jurassic, the Tethys ocean continued opening to the west, between North America to the north and South America and Africa to the south (Fig. 2); the separation became effective during Callovian or Kimmeridgian times (155–141 Ma), when seas invaded the rifting corridor south of North

America. The Neopangaean province ceased to exist at this time. In the Kimmeridgian Stage (146–141 Ma), similarities still existed between Africa and North America; the rich Tanzanian locality of Tendaguru has yielded several dinosaur genera that are also recorded in the North American Morrison Formation (Brachiosaurus, Barosaurus, and Dryosaurus). This confirms that land communications were available between Africa and North America until the Late Jurassic. The absence in Africa of the Euro-American Tithonian (141–135 Ma) families Camarasauridae and Camptosauridae, as well as the absence in the north of titanosaurid and dicraeosaurid sauropods, may reflect the beginning of separate evolution in the different parts of breaking Neopangaea, i.e., Euramerica (Europe and North America) and Gondwana (South America, Africa, India, Australia, and Antarctica). To the east, Chinese dinosaurs had no close relationships with their western counterparts, as was shown by Vale´rie Martin (1995): The well-known Chinese sauropods Mamenchisaurus, Euhelopus, Shunosaurus, and so on, which were usually referred to western families, in fact constitute an endemic Chinese family (Euhelopodidae) with no clear relationships to other sauropods. Russell (1993) reached nearly the same conclusions about sauropods, adding that theropods (Yangchuanosaurus and Sinraptor, which constitute the new family Sinraptoridae), stegosaurs (Chialingosaurus, Chungkingosaurus, and Tuojiangosaurus), and even mammals were endemic. This

FIGURE 2 Late Jurassic reconstruction of the paleobioprovinces (after Dercourt et al., 1993; drawing by Guy Le Roux).

54 endemicity indicates that central Asia was isolated from the rest of the world. The distribution pattern of Early Cretaceous dinosaurs is more clear than that of the Jurassic. With the opening of the central Atlantic Ocean (approximately 141 Ma), communications became impossible between North America and the southern continents, leading to the differentiation of Gondwanan (South America, Africa, India, Australia, and Antarctica) and Euramerican (Europe and North America) faunas. The transgression of epicontinental seas occurred 110 Ma ago as the North Atlantic Ocean broke the faunal continuity between North America and Europe (end of the Euramerican province), while the establishment of a land bridge across the Bering Strait allowed communications between Asia and America (birth of the Asiamerican province). A Gondwanan fauna was well defined in South America and Africa by titanosaurid and dicraeosaurid sauropods and abelisaurid theropods. The Argentinian Early Cretaceous fauna of La Amarga is close to the Late Jurassic Tendaguru fauna, with the dicraeosaurid Amargasaurus. During this period Gondwana began to break into different plates corresponding to the current continents. The eastern Gondwanan continents (India, Australia, and Madagascar) had begun to separate from Africa and South America as early as the Late Jurassic. In the Early Cretaceous, India and Australia separated from each other. These geological interpretations are problematic because they imply a long separation between India and west Gondwana (Africa–South America) that is not at all confirmed by terrestrial vertebrate distribution. The Late Cretaceous Indian fauna looks like the South American one, with abelisaurid and titanosaurid dinosaurs and madtsoiid snakes. During the Middle Cretaceous, rifting began between Africa and South America, leading to the opening of the South Atlantic Ocean and the isolation of the continental faunas (end of West Gondwana). North American and European dinosaurs were very similar during the earliest Cretaceous. Several examples of intercontinental genera (the ornithopods Iguanodon and Hypsilophodon and the nodosaurid ankylosaur Polacanthus) are recorded from the Lakota Formation of Dakota and the Wealden of England. These indicate that continuous dinosaur populations lived on both continents 120 million years ago. Other Euramerican dinosaurs include ornithomimids (Pelecanimimus from Spain and an unnamed ornithomimid

Biogeography from the Cloverly Formation) and troodontids (unnamed species from the Isle of Wight, England). It is likely that this Euramerican paleobioprovince no longer existed when North American and Asian faunas merged at the end of the Early Cretaceous. Euramerica was ‘‘broken’’ by the invasion of the future North Atlantic Ocean by seas, possibly before Aptian times (113 Ma). The European province also shows Gondwanan components such as the titanosaurid Iuticosaurus (Isle of Wight, England) and the spinosaurid Baryonyx. The ornithopod Valdosaurus is recorded from Africa (Niger) and Europe (England). These similarities indicate that Europe had land connections with both North America and Africa during the Early Cretaceous. Connections with Africa existed across the Mediterranean. To the east, central Asia came into contact with North America by Aptian–Albian times, forming a new Asiamerican province. It seems that Euramerican taxa, such as iguanodontids, dromaeosaurids, and troodontids, invaded central Asia at this time (Russell, 1993). Before this important event, the Early Cretaceous Asian vertebrates were still provincial, with psittacosaurids and peculiar sauropods such as Phuwiangosaurus from Thailand. With the dislocation of Gondwana in the Cretaceous, and high marine levels leading to the inundation of lands by epicontinental seas, the Late Cretaceous is the most complex period for dinosaur biogeography (Fig. 3). Unfortunately, little is known of the Late Cretaceous ‘‘Gondwanan’’ faunas, which theoretically should have undergone many independent evolutionary events. Western North America, with its well-known dinosaurs (including Tyrannosaurus, Triceratops, and Edmontosaurus), was isolated from eastern North America by the ‘‘Western Interior Sea’’ from the Cenomanian until the uppermost Cretaceous. Possibly because of the northern position of the Bering ‘‘bridge,’’ which may have acted as a climatic or ecological filter, exchanges with Asia were limited because there were few common genera but many common families (such as the Tyrannosauridae with Tyrannosaurus in North America and Tarbosaurus in Asia). Russell (1993) suggested that most of the North American Late Cretaceous families, including Tyrannosauridae, Ceratopsidae, Protoceratopsidae, and Troodontidae, originated in Asia during the Middle Cretaceous.



FIGURE 3 Late Cretaceous reconstruction of the paleobioprovinces (after Dercourt et al., 1993; drawing by Guy Le Roux).

Despite the advanced stage of dislocation of Gondwana in the Late Cretaceous, the known faunas from its different pieces show strong affinities until the Late Cretaceous. However, this is possibly due to a poor fossil record outside South America. South American dinosaurs of the Late Cretaceous include abelisaurids, titanosaurids, noasaurids, and basal hadrosaurids. Despite the fragmentation of Gondwana, titanosaurids and abelisaurids are found at this time in Africa, India, and Madagascar. Europe, as Nopcsa showed 60 years ago, supported both conservative Euramerican taxa (Telmatosaurus, Rhabdodon, Dromaeosauridae, and Struthiosaurus) and Gondwanan elements (the abelisaurid

Tarascosaurus and the titanosaurids Ampelosaurus and Magyarosaurus) (Fig. 4). The Asiamerican taxa show many affinities to Albian North American dinosaurs: Rhabdodon is close to Tenontosaurus, and the dromaeosaurid looks like Deinonychus (Fig. 5). One can postulate that their Early Cretaceous ancestors (unknown in Europe) belonged to the same intercontinental populations. During the very Late Cretaceous, collisions between North America and South America to the west, and India and Asia to the east, led to faunal exchanges well documented in America with the arrival of southern taxa such as the titanosaurid Alamosaurus. Due to a poor fossil record, it is not possible to know

FIGURE 4 A Late Cretaceous ‘‘Gondwanan’’ element of the European fauna: the titanosaurid sauropod Ampelosaurus atacis (drawing by Guy Le Roux).



FIGURE 5 Rhabdodon priscus, an ‘‘Euramerican’’ Late Cretaceous ornithopod from Europe (drawing by Guy Le Roux).

about the extinctions that probably followed these collisions, a few million years before the final extinction of dinosaurs. From this review of dinosaur biogeography, one can conclude that it fits rather well with modern paleogeographic reconstitutions (which, admittedly, are also based on paleontological data). Of course, the main source of error in dinosaur biogeography is still linked to the imperfection of the fossil record. As suggested by Russell (1993), ‘‘it is conceivable that dinosaurian diversity in Gondwana will one day be found to have surpassed that in Paleolaurasia’’ (Asiamerica plus Europe); from a paleobiogeographical point of view, it is not only conceivable but predictable. Overall, however, a study of dinosaur paleobiogeography shows the essential influence of geological and tectonic factors in the evolution of continental ecosystems.

See also the following related entries: DISTRIBUTION AND DIVERSITY ● PLATE TECTONICS

References Bonaparte, J. F., and Kielan-Jaworowska, Z. (1987). Late Cretaceous dinosaur and mammal faunas of Laurasia and Gondwana. In Fourth Symposium on Mesozoic Terres-

trial Ecosystems, Short Papers (P. J. Currie and E. H. Foster, Eds.), pp. 24–29. Tyrrel Museum of Paleontology, Drumheller, Alberta, Canada. Buffetaut, E., Me´chin, P., and Me´chin-Salessy, A. (1988). Un dinosaure the´ropode d’affinite´s gondwaniennes dans le Cre´tace´ supe´rieur de Provence. Comptes Rendus Acad. Sci. Paris Se´r. 2 306, 153–158. Dercourt, J., Ricou, L. E., and Vrielynck, B. (1993). Atlas Tethys palaeoenvironmental maps, 14 maps, 1 plate, pp. 307. Gauthier-Villars, Paris. Le Loeuff, J. (1991). The Campano–Maastrichtian vertebrate faunas from Southern Europe and their relationships with other faunas in the world: Palaeobiogeographical implications. Cretaceous Res. 12, 93–114. Lull, R. S. (1910). Dinosaurian distribution. Am. J. Sci. 29, 1–39. Martin, M. (1981). Les dipneustes me´sozoı¨ques malgaches, leurs affinite´s et leur inte´reˆt pale´obioge´ographique. Bull. Soc. Ge´ol. France 23, 579–585. Martin, V. (1995). Nopcsa, F. (1934). The influence of geological and climatological factors on the distribution of non-marine fossil reptiles and Stegocephalia. Q. J. Geol. Soc. London 90, 76–140. Rage, J. C. (1981). Les continents pe´ri-atlantiques au Cre´tace´ supe´rieur: Migration des faunes continentales et proble`mes pale´obioge´ographiques. Cretaceous Res. 2, 65–84. Russell, D. A. (1993). The role of Central Asia in dinosaurian biogeography. Can. J. Earth Sci. 30, 2002–2012.

Biomechanics R. MCNEILL ALEXANDER University of Leeds Leeds, United Kingdom


iomechanics is the application of engineering methods to the study of animals and plants. It deals with the strength and elasticity of skeletons, with mechanisms of movement, and with the energy that is needed. One of the landmarks in the subject’s history is a book, On Growth and Form, published in 1917 by D’Arcy Thompson, who was professor of natural history at St. Andrew’s University, Scotland. In that book he compared the skeleton of a sauropod dinosaur to a nearby bridge that was at the time the most remarkable feat of engineering in his country. He compared the dinosaur’s legs to the bridge’s piers, the muscles and ligaments of the back to the bridge’s tension members, and so on. Since then biomechanics has become immensely more sophisticated and has been applied in far more detail to dinosaurs, which present such obvious problems in engineering. Some of the problems are discussed in another article, SIZE AND SCALING. It shows that larger animals need relatively thicker bones if they are to be strong enough to be as athletic as small animals. It uses a biomechanical approach to assess the athleticism of large dinosaurs. It concludes that large sauropods were amply strong enough to support their weight on land and could probably have moved much like elephants, which can run at moderate speeds but cannot gallop or jump. Many biochemical problems require estimates of the mass of the body or of parts of the body. Methods for obtaining these are described in the article titled SIZE AND SCALING. Many problems also require an estimate of the position of the body’s center of gravity. This is generally most easily obtained by experiments with scale models of uniform composition—for example, the solid plastic models that can be bought at many museums. A rough-and-ready way to find a model’s center of gravity is to hold it between two pin points, one on each side of the body. If the model balances with the body horizontally the center of mass must be vertically below the line joining the two points. There

is a more precise method that depends on photographing the model as it hangs suspended from various parts of its body. Models suitable for this purpose are made of uniform material throughout, but real dinosaurs consisted partly of muscle and guts, which are slightly denser than water; partly of bone, which is more than twice as dense; and partly of air-filled lungs. The heavy bones are distributed throughout the body and probably do not greatly alter the position of the body’s center of gravity, but the light lungs are all in the chest. Because the lungs are placed well forward in the body, and have such low density, the center of gravity of a living dinosaur would be slightly behind the position corresponding to the center of gravity of a solid plastic model. The effect is easily calculated and is small, moving the center of gravity of a 25-m Diplodocus only about 20 cm. Figure 1 shows the positions of the centers of gravity of two sauropods determined in this way. Diplodocus, with its very long tail, has its center of gravity far back in the trunk, close to the hip joints. In contrast, Brachiosaurus, with a shorter tail, large forelegs, and heavy neck, has its center of gravity much further forward. This tells us that the hindlegs must have supported most of Diplodocus’ weight but that the weight of Brachiosaurus was more evenly shared between fore- and hindlegs. We can be more precise. In Fig. 1, Diplodocus has its left hindfoot well forward, directly under the center of gravity, and its right hindfoot well back, about 1.4 m behind the center of gravity. At the mid-point of its step, each foot would be about 0.7 m behind the center of gravity. Each forefoot, at the mid-point of its step, would be about 2.5 m in front of the center of gravity. This tells us that the load supported by the hindfeet must have been 2.5/0.7 times the load supported by the forefeet: The hindlegs supported 78% of body weight and the forelegs only 22%. Similarly, in the case of Brachiosaurus, the hindlegs supported 52% and the forelegs 48% of body weight.



FIGURE 1 The stars show the positions of these dinosaurs’ centers of gravity as estimated by experiments with models. From Dynamics of Dinosaurs and Other Extinct Giants by R. McNeill Alexander. Copyright © 1989 by Columbia University Press. Reprinted with permission of the publisher.

It cannot be claimed that these (or any other calculations in dinosaur biomechanics) are very accurate estimates—the results would be modified by changes in the shape and posture of the models—but they seem unlikely to be too inaccurate. They enable us to assess the strengths of leg bones in relation to the loads they have to bear. In particular, they were used in the calculations of ‘‘strength indicators’’ described in the article SIZE AND SCALING. Bone is a fairly uniform material about equally strong whether it comes from a large animal or from a small one or from a bird or a mammal. (I have no data for the strength of reptile bone.) It seems reasonable to assume that dinosaur bone was about as strong as bone from modern animals. Calculations of the loads that leg bones can bear must take account of the angles at which they are held. A bone held erect like a pillar will support a much greater load than if it is held horizontally, with a vertical load on its end. Modern lizards stand and run with their feet well out to either side of the body, with the humerus (upper arm bone) and femur (thigh bone) horizontal: The forelimbs are positioned like the arms of a person doing press-ups. We know from fossil footprints, however, that dinosaurs kept their feet directly under the body, and we believe that the large ones walked with their legs nearly straight, like elephants. Simple calculations tell us that the leg bones of, for example, Apatosaurus were amply strong enough for walking like this, but that the femur would have broken if it had tried to walk like a lizard. The necks of sauropods raise another biomechanical question: How were they supported? Figure 2 shows the neck of Diplodocus in the horizontal posi-

Biomechanics tion in which it is usually restored. There are deep notches in the neural spines of the vertebrae that seem likely to have housed a ligament, similar to the large ligament in the necks of cattle. It seems likely that (as in cattle) the weight of the head and neck was supported largely by tension in this ligament. In cattle, the ligament is made mainly of elastin, a rubberlike protein that stretches to allow the animal to lower its head to graze or drink and recoils elastically when the head is raised. Could Diplodocus also have had an elastin ligament? The mass of the head and neck of a Diplodocus was estimated as 1.3 tons from measurements of a model’s neck. (There is room for doubt here because of uncertainty about the mass of muscle in the neck). Think of the neck as pivoting about the joint in the backbone where the joint reaction force is shown acting in Fig. 2. The weight acts 2.2 m in front of that joint and the ligament is 0.42 m above it; hence, the ligament force is 2.2/0.42 times the neck’s weight or 7 tons force (70 kN). The dimensions of the ligament can be estimated from the size of the notches in the vertebrae, and it has been calculated that (if it was elastin) it was just strong enough to withstand this force. At least it was strong enough to take a large part of the load, relieving the neck muscles. It has been suggested that sauropods such as Diplodocus were not limited to feeding like giraffes, keeping all their feet on the ground, but could rear up on their hindlegs to reach leaves on high branches. Is this biomechanically feasible? The strength of the hindlegs presents no problem: The bones were amply strong enough to carry the whole weight of the body and in any case (as we have seen) they carried most of the body’s weight even in four-footed standing. However, could the animals have reared up? To do that, they must get their hindlegs directly under the

FIGURE 2 A diagram of the neck of Diplodocus showing how it may have been supported by tension in a large ligament. From Dynamics of Dinosaurs and Other Extinct Giants by R. McNeill Alexander. Copyright © 1989 by Columbia University Press. Reprinted with permission of the publisher.



center of gravity so that no load would remain on the forelegs. That, too, presented no difficulty. In Fig. 1, the left hindfoot is already under the center of gravity. If the right one were brought alongside it, the animal would be able to rear up. That does not, of course, prove that Diplodocus did rear up, and in any case there is another biomechanical problem to be considered: Is the heart likely to have been strong enough to keep blood flowing to the brain if the head were raised 6 m above the heart (as it might be when the dinosaur reared up). The blood pressures of sauropod dinosaurs are discussed in the article titled PHYSIOLOGY. There are a wide variety of other problems concerning dinosaurs that have, or could be, tackled by biomechanical methods. However, we will finish with just one more example. The thick skull roofs of dome-headed dinosaurs have been interpreted as adaptations for fighting by head-butting, similar to the contests between bighorn rams. The interpretation is controversial but let us consider its implications. Imagine two 20-kg males colliding at 3 m per second (a jogging speed for a human). Each would have 90 J kinetic energy (1/2 ⫻ 20 ⫻ 32). The force needed to bring the males to a halt would be (kinetic energy)/(stopping distance). If both kept their necks rigid, they would be halted in a few centimeters, the force would be enormous, and they would surely break their necks. If, on the other hand, they allowed their necks to bend in the impact, using their neck muscles to absorb the energy, the stopping distance would be much longer, perhaps 0.3 m. If it were that, the force required would be only 90/0.3 ⫽ 300 N, or 30 kg force, which could probably be tolerated. As often happens in biomechanics, a very simple calculation can be illuminating.


Biometrics RALPH E. CHAPMAN National Museum of Natural History Smithsonian Institution Washington, DC, USA

DAVID B. WEISHAMPEL Johns Hopkins University School of Medicine Baltimore, Maryland, USA

Biometrics or biometry is the measurement of life. In their classic textbook, Biometry: The Principles and Practice of Statistics in Biological Research, Sokal and Rohlf (1995) define biometry as the application of statistical methods to the solution of biological problems. Webster’s Ninth New Collegiate Dictionary similarly defines biometry as the statistical analysis of biological observations or phenomena. We would define biometry more broadly to include other forms of measurement and quantification from simple counting to advanced areas within subjects such as geometry, mathematics, and computer science. As such, we would modify Sokal and Rohlf’s (1995) definition as follows: Biometry is the application of statistical, geometrical, mathematical, and/or algorithmic approaches to the solution of biological problems.

As thus defined, biometry is not only at the heart of all theoretical studies of dinosaurs but also an important contributor to more descriptive studies, and its importance increases with time. Classically, biometry is associated with studies of shape and size within morphometric and allometric contexts. Following the definition given previously, however, there are also a number of other research areas in which biometrical methods not only are used but also are major factors in the research. These include phylogenetic analysis, distributional analyses, and functional morphology.

References Alexander, R. McN. (1989). Dynamics of Dinosaurs and Other Extinct Giants. Columbia Univ. Press, New York. McGowan, C. (1994). Diatoms to Dinosaurs. The Size and Scale of Living Things. Island Press, Washington, DC.

Morphometrics and Allometry Morphometrics is the quantitative measurement of shape and includes a broad range of techniques from the simple measurement of bone or footprint dimen-

60 sions to the application of very sophisticated statistical and geometrical methods for comparing the shapes of these structures (see Chapman, 1990, for a detailed discussion). Allometry is the study of size and its consequences and is usually carried out using bivariate (two-variable) and multivariate (many variable) morphometric analyses comparing the relative rate of growth of biological structures. Through time, morphometric analyses have progressed from relatively simple methods to the very complex approaches used today. Early studies consisted of taking simple measurements and presenting tables of them. This progressed to presenting comparative tables among specimens and taxa and noting differences. Two-dimensional analyses, in which two different sets of measurements are used for the same specimens, were introduced early on as well, mostly as unquantified observations (e.g., noting that one vertebrae was relatively broader than another) that later were quantified as ratios of these variables. Differences in ratios typically were used as evidence for differences among specimens and as evidence for the erection of new taxa. Morphometric analyses became more sophisticated as these analyses of two variables progressed to bivariate allometric analyses in which the investigator was able to note relative growth rates and correlations between variables. From two-dimensional analyses the next step was to multivariate analyses, those utilizing many variables simultaneously (e.g., principal components analysis), and to early techniques of shape analysis such as D’Arcy Thompson’s (1942) transformation grids. These two approaches then converged toward techniques of modern shape analysis (Chapman, 1990). The morphometric study of dinosaurs followed the same progression as the general field of morphometrics, although typically with some lag time. This progression can be seen in the analyses of many different dinosaur groups, such as pachycephalosaurs and hadrosaurs, and even in footprints. Studies of saurischians are less common or typically less advanced due to the relatively small number of specimens available for all but a few taxa. As an example, here we will concentrate on studies done on the ceratopsians. Morphometric analyses of dinosaur material began early. These began as very simple analyses

Biometrics amounting to measuring and comparing the simple dimensions of bones (e.g., vertebrae). In one of the first papers on dinosaurs, even predating the term dinosaur, Mantell (1833) measured the dimensions of bones of the fossil reptiles found from southeast England, including various marine reptiles and dinosaur material referred to Megalosaurus, Iguanodon, and the ankylosaur Hylaeosaurus. He presented these dimensions and even made note of differences among the different specimens and in comparison with modern analogs such as crocodiles and iguanas. One of his most interesting uses of this data was an attempt to estimate the size of Iguanodon by noting the ratio of the size of its various skeletal parts to those of Iguana and using their average to estimate the total size of the dinosaur. The level of sophistication of the analysis of dinosaurs increased at a slow pace during the 19th century and into the first part of this century, and then increased rapidly. In early studies of ceratopsians, typified by Marsh (1890), most of the biometrical data were presented as selected measurements (e.g., skull length) and observations that some proportions supposedly separated new taxa from others closely related to it. For example, Marsh (1890, p. 82), in his description of Triceratops prorsus, suggests it is different from T. horridus in various ways including a more compressed rostral bone (a ratio of length to width), a less broad parietal crest (another ratio), and a larger frontal horn core (a volume estimate compared with an implied overall size for the specimens). These differences were noted but, as was typical for the time, no real data were presented. Hatcher et. al, (1907) went a little further by providing tables of measurements of specimens of the same taxon and, later still, Lull (1933) provided more comparative data, including comparisons of limb proportions between foreand hindlimbs of a single taxon. Brown and Schlaikjer (1940) increased the level of sophistication slightly with their detailed analysis of the growth of Protoceratops. They provided tables of measurements as well as long lists of differences in shapes between juveniles and adults. They even published an interesting summary diagram showing the proportionate changes that take place in the skull of Protoceratops from juvenile to adult. It is surprising, however, that they never attempted any bivariate (two-variable) allometric analyses, which were very

Biometrics popular at that time because of the publication of Sir Julian Huxley’s (1932) Problems of Relative Growth. This was finally done for known ceratopsian dinosaurs in a pioneering paper by Gray (1946) who looked at growth trajectories, within both phylogenetic and ontogenetic contexts, and noted that they seemed to support Lull’s phylogenetic analysis of the Ceratopsia. Later, Lull and Gray (1949) used Thompson’s (1942) grids to analyze shape differences among ceratopsians. In this method, a rectangular grid is superimposed on a starting specimen and the shape change is shown by distorting the grid to show the second specimen. The two papers by Gray (1946) and Lull and Gray (1949) were the most sophisticated morphometric analyses ever done on dinosaurs until the work by Dodson on the growth and taxonomic structure of lambeosaurine hadrosaurs (Dodson, 1975) and, again, Protoceratops (Dodson, 1976). In these studies, Dodson used detailed bivariate allometric analyses in combination with multivariate procedures, those that analyze many variables at the same time, to analyze the growth patterns and shape variations in these groups and note the implications of them for the analysis of taxonomic structure and sexual dimorphism. This work was developed further by Dodson (1990a), who combined allometric and shape analysis with character analysis within a phylogenetic context to explore the relationships between Monoclonius and Centrosaurus. At the same time Lehman (1990) used detailed allometric analyses combined with some basic shape analysis to study systematics and sexual dimorphism within the chasmosaurine ceratopsians. Finally, Dodson (1993) has applied a high-level morphometric approach, RFTRA (see Chapman, 1990, for a detailed discussion), to analyze changes in the shapes of ceratopsian heads in great detail by noting changes in the position of anatomical landmarks on the skulls. Additional studies of ceratopsians using other advanced shape analysis methods are also in development (C. A. Forster, 1994, personal communication). Biometry is also an important part of many other types of analyses; space limitations allow us to mention only three here. In phylogenetic analyses various techniques are used to reconstruct the relationships among various organisms. Biometrical procedures are used to docu-

61 ment many of the characters used in reconstructing these phylogenies. One dinosaurian example among the many possible is Russell and Zheng’s (1993) study of sauropodomorphs. In their phylogenetic analysis they use counts, ratios, and overall dimensions to define their character states. These include the number of vertebrae of certain types, the relative lengths of different parts of the vertebral column, the shapes of teeth defined by ratios, and many other biometrically defined characters. These data are assembled and run through various algorithmic procedures using computer programs such as PAUP (Swofford and Begle, 1993) to provide the resulting phylogenetic trees showing these relationships. Distributional analyses attempt to analyze the distribution of organisms through space and time. Doing this incorporates biometrical procedures ranging from simple counting to detailed multivariate analyses. Studies of dinosaurs through time and space are rather rare because of the small number of specimens, but initial studies have been published (see Dodson, 1990b; Dodson and Dawson, 1991; Weishampel, 1990; Weishampel and Norman, 1989) documenting the number of species found and documenting species and higher level taxonomic diversity, as well as literature citations, through time. More sophisticated analyses are currently in progress (R. Chapman and D. Weishampel, manuscript in preparation). Studies of functional morphology attempt to determine the function of structures given their form and context within the organism. Here again, biometrical procedures are very important and many important studies have been published and are in development. For example, in Weishampel’s (1981) acoustical analysis of the nasal system in lambeosaurines, analyzing and reconstructing the systems depends strongly on detailed measurements of the heads and nasal systems of the lambeosaurines. Similarly, Alexander’s (1989) detailed analyses of the movements of large dinosaurs utilize limb proportions, estimates of dinosaur mass and bone strength, and mechanical equations to assess the abilities of different taxa to move around at various speeds, especially with an eye toward determining maximum running speeds.

See also the following related entries: BIOMECHANICS ● COMPUTERS AND RELATED TECHNOLOGY



References Alexander, R. McN. (1989). Dynamics of Dinosaurs and Other Extinct Giants, pp. 167. Columbia Univ. Press, New York. Brown, B., and Schlaikjer, E. M. (1940). The structure and relationships of Protoceratops. Ann. N. Y. Acad. Sci. 40(3), 133–266. Chapman, R. E. (1990). Shape analysis in the study of dinosaur morphology. In Dinosaur Systematics— Perspectives and Approaches (K. Carpenter and P. J. Currie, Eds.), pp. 21–42. Cambridge Univ. Press, New York. Dodson, P. (1975). Taxonomic implications of relative growth in lambeosaurine hadrosaurs. Systematic Zool. 24, 37–54. Dodson, P. (1976). Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. J. Paleontol. 50, 929–940. Dodson, P. (1990a). On the status of the ceratopsids Monoclonius and Centrosaurus. In Dinosaur Systematics— Perspectives and Approaches (K. Carpenter and P. J. Currie, Eds.), 231–244. Cambridge Univ. Press, New York. Dodson, P. (1990b). Counting dinosaurs: How many kinds were there? Proc. Natl. Acad. Sci. USA 87, 7608– 7612. Dodson, P. (1993). Comparative craniology of the Ceratopsia. Am. J. Sci. 293A, 200–234. Dodson, P., and Dawson, S. D. (1991). Making the fossil record of dinosaurs. Mod. Geol. 16(1/2), 3–15. Gray, S. W. (1946). Relative growth in a phylogenetic series and in an ontogenetic series of one of its members. Am. J. Sci. 244(11), 792–807. Hatcher, J. B., Marsh, O. C., and Lull, R. S. (1907). The Ceratopsia, Monograph No. 49, pp. 300. U. S. Geological Survey, Washington, DC. Huxley, J. (1932). Problems of Relative Growth, pp. 276. Methuen, London. Lehman, T. M. (1990). The ceratopsian subfamily Chasmosaurinae: Sexual dimorphism and systematics. In Dinosaur Systematics—Perspectives and Approaches (K. Carpenter and P. J. Currie, Eds.), pp. 211–230. Cambridge Univ. Press, New York. Lull, R. S. (1933). A revision of the ceratopsia or horned dinosaurs. Mem. Peabody Museum Nat. History 3(3), 1–175. Lull, R. S., and Gray, S. W. (1949). Growth patterns in the Ceratopsia. Am. J. Sci. 247(7), 492–503. Mantell, G. (1833). The Geology of the South-East of England, pp. 415. Longman, London. Marsh, O. C. (1890). Description of new dinosaurian reptiles. Am. J. Sci. Third Ser. 39(229), 81–86.

Russell, D. A., and Zheng, Z. (1993). A large mamenchisaurid from the Junggar Basin, Xinjiang, People’s Republic of China. Can. J. Earth Sci. 30(10/11), 2082–2095. Sokal, R. R., and Rohlf, F. J. (1995). Biometry: The Principles and Practice of Statistics in Biological Research, Third Edition. Freeman, San Francisco. Swofford, D. L., and Begle, D. P. (1993, March). User’s Manual for PAUP: Phylogenetic Analysis Using Parsimony, Version 3.1. Laboratory of Molecular Systematics, Smithsonian Institution, Washington, DC. Thompson, D’A. W. (1942). On Growth and Form: Revised and Enlarged Edition, pp. 1116. Cambridge Univ. Press, New York. Webster’s Ninth New Collegiate Dictionary. Merriam-Webster, Springfield, MA. Weishampel, D. B. (1981). Acoustic analysis of potential vocalization in lambeosaurine dinosaurs (Reptilia: Ornithischia). Paleobiology 7, 252–261. Weishampel, D. B. (1990). Dinosaur distribution. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 63–139. Univ. California Press, Berkeley. Weishampel, D. B., and Norman, D. B. (1989). Vertebrate herbivory in the Mesozoic; Jaws, plants, and evolutionary metrics [Special paper]. Geological Soc. Am. 238, 87–100.

Biomineralization CLIVE TRUEMAN University of Bristol Bristol, United Kingdom

Biomineralization is the process by which organisms produce mineral or inorganic tissues. Bone is the most common mineralized tissue in vertebrates. It is used as a structural support and as a reservoir for physiologically important ions such as calcium. It must therefore be both strong and soluble in body fluids.

Growth and Mineralization of Bone The general sequence of bone mineralization is synthesis and extracellular assembly of the organic matrix framework, followed by mineralization (nucleation and growth of bone mineral) within the framework (Lowenstam and Weiner, 1989). On a macroscopic level, new bone forms either on a scaffolding of preexisting cartilage (endochondral



ossification) or directly onto outer periosteal bone surfaces (membranous ossification).

still much controversy surrounding the nucleation of apatite crystallites in bone.

Bone Mineral

Nucleation of HAP

The inorganic or mineral component of bone is usually considered to be some form of hydroxyapatite (HAP), although most sites within the apatite lattice may be subject to substitutions; therefore, the formula may be better represented as

When hydroxyapatite is synthesized from supersaturated solutions, an amorphous precursor phase is found with lower Ca:P ratio. This spontaneously converts slowly to hexagonal hydroxyapatite. This has led to the theory that an amorphous calcium phosphate may form as a precursor to true bone mineral. However, other precursor phases have been suggested, such as brushite (CaHPO4⭈2H2O), octacalcium phosphate [Ca8H2(PO4)6⭈5H2O], and also direct precipitation of hydroxyapatite (Posner, 1985). The identification of the precursor phase has great implications for the conditions of initial biomineralization of bone because each phase suggests different conditions of biomineralization, particularly in terms of pH and saturation of Ca and PO4 ions. The processes leading to nucleation of HAP are poorly understood, with many unresolved questions. A number of problems can be highlighted (e.g., Simkiss and Wilbur, 1989);

(Ca,Sr,Mg,Na,H2O,REE,[ ])2 (PO4 ,HPO4 ,CO3 ,P2O7)6 (OH,F,CO3 ,Cl,H2O,O,[ ])2, where ([ ]) represents unfilled lattice positions. Bone is formed of extremely small crystallites, giving bone a very high surface area. A large amount (3–5%) of structural carbonate substituting for phosphate is also found in bone. The carbonate substitution causes lattice defects and increases the reactivity of bone. These two features of bone apatite ensure that it has a relatively high solubility and can be resorbed fairly easily. This is essential to bone growth and repair; in addition, the vertebrate skeleton serves as an important store for calcium and many trace elements.

Bone–Collagen Relationships There is a well-documented intimate relationship between bone apatite crystallites and collagen fibers. The apatite crystallites are oriented with their long (c) axes parallel to the collagen fibers. This orientation can be seen in polarized light, XRD, or TEM studies, and has been seen in dinosaur bone fragments from Seismosaurus (Gillette, 1994) and Allosaurus (Zocco and Schwartz, 1994). Bone collagen is a type 1 collagen, and the fibers are linked to form a triple-helix structure. This crosslinking occurs in a stepped fashion between the molecules, and space considerations mean that a gap of 35 nm exists between the head and tail of individual collagen molecules. This forms a regular spacing of holes within collagen fibers. The intimate linking of bone apatite and collagen, the small size of bone apatite crystallites, and the presence of gaps in the collagen structure in which bone crystals may occur lead to many suggestions that primary bone nucleation occurs within the collagen molecule. Direct evidence of this has not been found, however, and there is

1. To create a sound structural unit, the bulk of mineralization must occur extracellularly, but must be under cellular control. 2. Extracellular fluid levels of Ca2⫹ and P must be kept only slightly supersaturated with respect to precursor phases to avoid unintentional ectopic calcification. 3. Even though Ca and P levels must be slightly supersaturated, some method of overcoming the nucleation barrier (i.e., activation energy) must be found to initiate ectopic calcification. 4. Nucleation must be confined to specific sites and therefore some inhibitory mechanism must be present on the extracellular fluid. 5. Mineralization must be at least as rapid as growth to maintain structural integrity. Stability of apatite appears to grow at the expense of speed—enamel forms much more slowly than bone—and produces very large, well-ordered stable crystals.

Inhibitor Molecules Many body fluids are supersaturated with respect to Ca and P but do not precipitate any form of calcium phosphate. This has lead to the suggestion that certain


64 inhibitory molecules may be present in body fluids, which may be selectively removed at sites of bone formation, thus allowing a site-specific mechanism of precipitating calcium phosphates. The first of these inhibitor molecules to be recognized was pyrophosphate (Fleisch and Neuman, 1961). This molecule is found in plasma, urine, and saliva and is a successful inhibitor of most forms of calcification. In addition, it may be hydrolyzed by a number of enzymes to orthophosphate, thus providing a biological ‘‘on–off switch’’ for calcium phosphate precipitation. Since this discovery was made, several alternative inhibitors have been suggested, many of which may work in conjunction or separately under differing conditions.

Matrix Vesicles A second potential mechanism of apatite nucleation involves small extracellular microstructures known as matrix vesicles. These were initially found within mineralizing growth cartilage but subsequently discovered in primary mineralizing dentine and bone. These vesicles contain high concentrations of enzymes as well as high Ca and P concentrations. It has been inferred that the matrix vesicles could act to raise the ion concentrations to the point where direct precipitation of HAP or a precursor phase occurs or could release enzymes that act to destroy inhibitor molecules, and hence allow precipitation of apatite. This mechanism is attractive because it would allow precipitation of bone mineral extracellularly but within cellular control.

Overcoming the Energy Barrier Slightly supersaturated solutions may precipitate if certain materials are present. The ion concentration of the solution will determine the sensitivity of the precipitation to the exact catalyst. As the solution becomes more supersaturated, more substances may act as catalysts. Assuming condition 2 (see previous list) to be true, then condition 3 may be met by a variety of mediating substances found in mineralizing bone. Initially, collagen was inferred to mediate the nucleation process, principally due to its presence in abundance in bone, the intimate association of apatite and collagen, and the apparent association of newly formed apatite crystallites with collagen molecule periodicity. Regular spacing of bone crystallites within mineralizing collagen fibers has been found, and it is postulated that hydroxyapatite is nucleated at a specific site within this gap. Initial positive studies were later found to be misleading because the association between primary apatite crystallites and collagen molecule periodicity was not upheld. Many theories of bone apatite mineralization now suggest that initial primary nucleation of apatite crystallites is not directly associated with the collagen molecule, but subsequent growth and organization of the apatite crystallites is controlled or at least influenced by the regular spacing in the collagen molecule. Most other noncollagenous proteins found in bone, such as osteonectin and osteocalcin, have also been implicated as nucleation triggers.

Summary ●

Bone mineral is some form of HAP. There is still debate as to whether there is a precursor phase and, if so, what that phase is. Bone matrix is composed largely of collagen fibers, and it is likely that after initial nucleation the collagen spaces play an important role in biomineralization, forming sites for collagen–protein–apatite bonds. A number of substances may act as inhibitors to nucleation and must be removed from the sites of biomineralization. The major roles of bone cells are known, but cellular influences on formation of extracellular material are still poorly understood.

Paleoecological Information from Conditions of Biomineralization The physical and chemical environment in which bone mineral forms can be recorded by a number of chemical signatures. For instance, the temperature at which bone mineral formed may be recorded by the ratio of oxygen isotopes contained within the phosphate group in HAP. If fossil bone phosphate can be shown to be unaffected by diagnoses, then information concerning the body temperature of dinosaurs may be retrieved. This has been attempted by Barrick and Showers (1995). Comparing the temperatures recorded from oxygen isotope signatures of internal bones and bones from extremities, they produced an



approximation of the homogeneity of the body temperature of different dinosaurs and compared this to that recorded by a varanid lizard. The trace element signature of fossil bone may also record paleobiological signals. Differing diets (e.g., shellfish vs meat vs plant) have different trace element contents, and the levels of these elements within bones may reflect the diet of that animal. It is more difficult to retrieve dietary signals from fossil bones because most of the diagnostic trace elements reside in the Ca site of bone apatite, and it is this site that is most susceptible to diagenetic ion exchange.

Eggshell Formation The biomineralization of eggshell has been well documented in a series of papers by K. Simkiss. This summary is principally taken from that in Simkiss and Wilbur (1989). In the oviduct of reptiles and birds, the epithelial layer is used as the basis for biomineralization. In birds, the oocyte passes down the oviduct and is held at the end for 15-30 hr (depending on the species) while calcite is deposited on the egg membrane. In birds, eggshell formation is extremely rapid. In domestic fowl, 5 g of calcite is deposited in 20 hr (Simkiss and Wilbur, 1989). This requires a huge reserve of Ca, which is held in bone. CaCO3 production [(Ca2⫹ ⫹ CO2 ⫹ H2O ⫺ CaCO3 ⫹ ⫹ H2 )] produces protons that can induce an acidosis in the bird. The urine of laying hens acidifies during eggshell formation, and respiratory activity is often increased by panting. Within the eggshell itself, calcite crystals form from a number of discrete nucleation sites and grow outward from the membrane in a progressively more uniform direction because lateral growth is inhibited by abutting with neighboring crystals. The directional growth of eggshell can be studied by XRD (SilynRoberts and Sharp, 1986) and different growth patterns determined between turtles and ostrich related to the distribution and nature of organics within the eggshell. This is in turn related to the strength of the eggshell.


References Particular attention is drawn to the excellent reviews by Simkiss and Wilbur, and Lowenstam and Weiner, from which much of this summary was drawn. Barrick, R. E., and Showers, W. J. (1995). Oxygen isotope variability in juvenile dinosaurs (Hypacrosaurus): Evidence for thermoregulation. Paleobiology 21(4), 552–560. Fleisch, H., and Neuman, W. F. (1961). Mechanisms of classification: Role of collagen, polyphosphates and phosphatase. Am. J. Physiol. 200, 1296–1300. Gillette, D. G. (1994). Seismosaurus—The Earth Shaker. Columbia Univ. Press, New York. Lowenstam, H. A., and Weiner, S. (1989). On Biomineralization. Oxford Univ. Press, New York. Posner, A. S. (1985). The mineral of bone. Clin. Orthopaedics 200, 87–99. Silyn-Roberts, H., and Sharp, R. M. (1986). Crystal growth and the role of the organic network in eggshells. Proc. R. Soc. London Ser. B 227, 303–324. Simkiss, K., and Wilbur, K. M. (1989). Biomineralization. Academic Press, San Diego. Zocco, T. G., and Schwartz, H. L. (1994). Microstructural analysis of bone of the sauropod dinosaur Seismosaurus by transmission electron microscopy. Palaeontology 37(3), 493–503.

Biostratigraphy SPENCER G. LUCAS New Mexico Museum of Natural History and Science Albuquerque, New Mexico, USA

Biostratigraphy identifies and distinguishes strata (layers of sedimentary rock) by their fossil content (Salvador, 1994). Strata with distinctive fossils are the units of biostratigraphy. The most basic unit is the biostratigraphic zone, called biozone, or zone for short. Extensive collecting of fossils over the vertical and lateral extent of a stratum, or strata, allows biozones to be defined.

Historical Development ●

Biostratigraphy was born in the early 1800s in western Europe. William Smith, a British civil engineer, realized that a given stratum usually contains distinctive


66 fossils. Smith’s recognition that many of these distinctive fossils could be traced over large areas, even when the nature of the enclosing rocks changed, allowed him to publish the first geological map of England. Parallel to Smith’s work, French comparative anatomist Georges Cuvier and geologist Alexander Brongniart independently discovered that distinctive kinds of fossils are often associated with specific strata. Whereas Smith did not attach geologic time significance to his observations, Cuvier and Brongniart, and other French paleontologists of the early 18th century (especially A. d’Orbigny), did. They argued that each stratum with its distinctive fossils represents a particular ‘‘stage’’ in the history of life. This conclusion forms the basis for biochronology.

Biostratigraphy vs Biochronology Biozones, by themselves, have no necessary time significance. They are simply bodies of rock characterized by their fossil content. As such, zones can be mapped, and their thicknesses can be measured. The time equivalent to a biozone is a biochron. Biochronology is thus the use of fossils to delineate intervals of geological time. Many American geologists and paleontologists do not distinguish biostratigraphy from biochronology; they use the term biostratigraphy to encompass the delineation both of rock units and of intervals of geologic time by fossils. The distinction between biostratigraphy and biochronology, however, has long been made by Canadian and many European scientists. They do so because it is conceptually useful to distinguish the biozone, which may be of varying ages over its lateral extent, from the biochron, which theoretically is of the same time value everywhere (Fig. 1).

FIGURE 1 A biostratigraphic zone may represent different time intervals at four different locations, but the biochron based on the zone is of one duration everywhere.

FIGURE 2 Biostratigraphers recognize four principal kinds of zones: range zones, interval zones, assemblage zones, and abundance zones.

Different Kinds of Biozones The four most commonly identified kinds of biozones are range, interval, assemblage, and abundance (acme) zones (Fig. 2). Recognition of one or more kinds of zones depends largely on the nature of the fossil record being studied and the purpose of the investigation. Range zones are strata that encompass the range (vertical distribution) of a particular kind of fossil. For example, in the western United States, the stratigraphic range of fossils of the Late Cretaceous dinosaur Triceratops defines a Triceratops range zone. Interval zones are the strata between two biostratigraphically significant horizons. Often they are the intervals between the last record of one kind of fossil and the first record of another kind. For example, the last record of the thyreophoran dinosaur Scelidosaurus well predates the first record of Stegosaurus, and an interval zone could be defined between these two records.

Biostratigraphy Assemblage zones are strata with a characteristic assemblage (association) of fossils. Many kinds of fossils define an assemblage zone. For example, the Late Cretaceous dinosaurian and other vertebrate fossils from the Judith River Formation of Montana could be considered to define an assemblage zone. Indeed, this assemblage zone has time significance as a biochronological unit, the Judithian land-vertebrate ‘‘age.’’ Abundance zones are strata recognized by the abundance (acme) of a kind (or kinds) of fossil, regardless of range or association. In western North America, hadrosaurs reached their maximum abundance between about 75 and 68 million years ago. This abundance of hadrosaur fossils could thus define a hadrosaur abundance zone.

Correlation The principal goal of biostratigraphy and biochronology is to correlate rock strata and the physical and biological events that the strata record. To correlate is to establish the equivalence of age or stratigraphic position of strata in separate areas. Broadly speaking, to correlate is to establish the contemporaneity of events in the geological histories of separate regions. Understanding geological and biological history is central to all geological and paleontological investigations, and stratigraphic correlation is critical to this understanding.

The Geological Timescale The Phanerozoic (last 570 million years) geological timescale is based largely on biostratigraphic data and biochronological correlations (Berry, 1987). This is because the intervals of the timescale are rooted in biozones that form the basis of the biochrons used in global correlation. Biostratigraphic events thus are critical to the geological timescale. For example, the Mesozoic Era, during which the dinosaurs lived, was originally defined by two biostratigraphic events—the global mass extinctions at the end of the Paleozoic and at the end of the Mesozoic. Today, the boundaries of the Mesozoic are generally recognized by smaller scale biostratigraphic events. Thus, the beginning of the Mesozoic corresponds to the first record of the ammonoid (extinct cephalopod) Otoceras, a biostratigraphic event.

67 Index Fossils and Facies Fossils Fossils are fundamental to biostratigraphy and biochronology; therefore, they are essential to most stratigraphic correlation. Fossils valuable to correlation are called index fossils because they identify and determine the age of the strata in which they are found. A good index fossil has a short stratigraphic range, is geographically widespread, and is easy to identify. In contrast, fossils of animals and plants that were very sensitive to environmental conditions are termed facies fossils (facies refers to a specific kind of environment). Facies fossils are usually specific to a certain kind of rock that represents a particular ancient environment. Therefore, facies fossils usually have much longer stratigraphic ranges and more restricted geographic ranges than do index fossils. In reality, the dichotomy between index and facies fossils is somewhat artificial, and both terms should be seen as endpoints of a spectrum. All organisms of the past had a definite stratigraphic range; those with the shortest ranges make the best index fossils. Also, all organisms of the past lived in particular environments; those most restricted environmentally produce the best facies fossils.

Dinosaurs and Biostratigraphy Dinosaur fossils are often distributed very unevenly in strata. They rarely are extremely numerous and dense through a stratigraphic interval. For this reason, dinosaurs have been little used by biostratigraphers until recently. Extensive collecting and a desire to apply the dinosaur fossil record to solving new problems has led to new efforts in the areas of dinosaur biostratigraphy and biochronology. Detailed documentation of dinosaurian range zones has been undertaken to examine problems of dinosaur microevolution (Horner et al., 1992) and extinction (Sullivan, 1987). Dinosaur-dominated assemblage zones have formed the basis for new biochronological units (Jerzykiewicz and Russell, 1991). Extensive regional and intercontinental correlations based on dinosaurs are being advocated (Lucas, 1993). Dinosaur footprints have also proven biochronological significance (Lockley, 1991). As Lucas (1991) argued, many kinds of dinosaurs make excellent index fossils. Further biostratigraphic organization of the dinosaurian fossil record is

68 needed so they can assume their rightful place as important fossils for Mesozoioc biochronology.

See also the following related entries: PALEOMAGNETIC CORRELATION ● RADIOMETRIC DATING

References Berry, W. B. N. (1987). Growth of a Prehistoric Time Scale (rev. ed.), pp. 202. Blackwell, Palo Alto, CA. Horner, J. R., Varrichio, D. J., and Goodwin, M. B. (1992). Marine transgressions and the evolution of Cretaceous dinosaurs. Nature 358, 59–61. Jerzykiewicz, T., and Russell, D. A. (1991). Late Mesozoic stratigraphy and vertebrates of the Gobi basin. Cretaceous Res. 12, 345–377. Lockley, M. (1991). Tracking Dinosaurs: A New Look at an Ancient World, pp. 238. Cambridge Univ. Press, Cambridge, UK. Lucas, S. G. (1991). Dinosaurs and Mesozoic biochronology. Mod. Geol. 16, 127–138. Lucas, S. G. (1993). Vertebrate biochronology of the Jurassic–Cretaceous boundary, North American Western Interior. Mod. Geol. 18, 371–390.

Bipedality Their metatarsals (sole bones) are longer than those in typical reptiles and nearly equal in length, with the middle toe the longest (Fig. 1). These features are considered typical of bipedal animals and are clearly seen in birds (Coombs, 1978). It is generally accepted that dinosaurs were digitigrade; that is, they walked on their toes, holding their sole bones off the ground almost without exception, even in the largest sauropods (Alexander, 1985). The bones of the dinosaurian ankle are also modified (see HINDLIMBS AND FEET). Typically in tetrapods, especially amniotes, the ankle consists of two rows of tarsal (ankle) bones: proximal and distal (Fig. 2). In dinosaurs, the proximal tarsals (the astragalus, which is distal to the tibia, and the calcaneum, which is distal to the fibula) do not rotate against each other or against the tibia and fibula during locomotion. In adult forms these bones are often fused to each other and to the leg bones. The distal tarsals, meanwhile, have a similar functional relationship to each other and to the metatarsals: They tend to cap, as opposed to rotate against, the sole bones. The ankle joint flexes

Salvador, A. (Ed.) (1994). International Stratigraphic Guide, 2nd ed., pp. 214. International Union of Geological Sciences and Geological Society of America, Boulder, CO. Sullivan, R. M. (1987). A reassessment of reptilian diversity across the Cretaceous–Tertiary boundary. Nat. History Museum Los Angeles County Contrib. Sci. 391, 1–26.

Bipedality KEVIN PADIAN University of California Berkeley, California, USA

Bipedality, or the habit of walking on two legs, was a characteristic of basal dinosaurs, which were small (up to 2 m in length) and lightly built (see ORNITHODIRA; PTEROSAURIA; DINOSAUROMORPHA). Their hindlimbs were considerably longer than their forelimbs, so much so that it is doubtful that the forelimbs in most early dinosaurs could have been used for walking. Their femora (thigh bones) were shorter than their tibiae (shin bones), though this eventually reversed when large forms evolved in several lineages.

FIGURE 1 Hindlimb of a generalized theropod dinosaur (after Coelophysis).


FIGURE 2 Crurotarsal and mesotarsal ankle joints, simplified. (a) Generalized archosauriform ankle (Euparkeria); (b) crurotarsal ankle (the crocodylomorph Notochampsa); (c) mesotarsal ankle (Lagosuchus). Bold line represents the general plane of flexion between leg and foot: The crurotarsal joint flexes obliquely, whereas the mesotarsal joint flexes as a hinge.

between the two rows of tarsals in a hinge joint; this configuration is called mesotarsal (‘‘mid-ankle’’). Changes in the form of the hindlimb bones also accompany bipedality (Fig. 1). The femoral head becomes offset from the shaft; the head is inturned to some degree, ranging from approximately 60 to 90⬚. The shaft is not sigmoid (curved in both dorsoventral and mediolateral planes) but rather is straight (most ornithischians) or only dorsally bowed (all theropods except the largest ones and some ornithischians), reflecting the restriction of movement to the dorsoventral plane. The distal condyles are more strongly developed ventrally and are relatively similar in size,

69 whereas in sprawling reptiles the medial condyle is appreciably larger. The tibia and fibula are no longer of similar size because they are not needed as opposable columns against which the limb muscles rotate the lower leg. Instead, the fibula is reduced (sometimes to a splint) and virtually no rotation of the lower leg occurs at the knee. The positions of trochanters and other muscle attachments are adjusted accordingly. In particular, the calcaneum lacks the prominent heel of crocodilians: The M. gastrocnemius no longer needs to rotate the foot against the leg because the mesotarsal ankle acts as a hinge, as noted previously (Gatesy and Dial, 1996). All the features discussed so far are considered typical not only of dinosaurs but also of their closest relatives, including other dinosauromorphs, Pseudolagosuchus, Lagosuchus (Marasuchus), Lagerpeton, and pterosaurs. This branch of archosaurs is collectively called the ORNITHOSUCHIA, and the distribution of these features among its members suggests that the common ancestor of all these animals had bipedal abilities. In contrast, the PSEUDOSUCHIAN branch of the archosaurs, those closer to crocodiles, flexed their ankles between the proximal tarsals: The astragalus is functionally connected to the tibia, whereas the calcaneum flexes with the foot (Parrish, 1986). This accounts for the sprawling walk of crocodiles, in which the feet splay slightly to the side. This functional arrangement is called crurotarsal (‘‘crossankle’’). However, when crocodiles execute the ‘‘high walk’’ (in which, instead of sprawling, they tuck their hindlimbs under their bodies and move the legs more parasagittally), their ankles are considerably stiffer and crurotarsal motion is less emphasized (Brinkman, 1980; Parrish, 1987; Gatesy, 1991; see ARCHOSAURIA; FUNCTIONAL MORPHOLOGY; PSEUDOSUCHIA). It should be noted that in dinosaurs bipedal stance and parasagittal gait were both primitive and obligatory: Dinosaurs could not sprawl. Some living lizards can run bipedally but this is primarily a consequence of high rates of the step cycle coupled with a high disparity between the forelimb and hindlimb lengths. That is, the forelimb is lifted off the ground at high speeds because it is so much shorter than the hindlimb that it would have to move at a prohibitively high rate in order to keep up with the hindlimb. Also, lizards never approach the true parasagittal step cycle of birds and other ornithodirans (Christian et al., 1994).


70 Footprints and trackways of dinosaurs reveal patterns in their locomotion. THEROPODS, apparently like other bipeds, typically placed one foot virtually in front of the other, and usually the axis of the foot pointed inward slightly; this ‘‘pigeon-toed’’ feature is retained in most avian descendants of these dinosaurs, despite the loss of the fleshy tail and the evolution of flight (Padian and Olsen, 1989). Even quadrupedal dinosaurs retained a narrow lateral distance between left and right feet. QUADRUPEDALITY in dinosaurs is secondary and evolved at least four times. All theropods were bipedal. Within SAUROPODOMORPHS, basal forms such as Anchisaurus were bipedal and this may have been true of juveniles of larger forms. Larger ‘‘PROSAUROPODS’’ such as Plateosaurus were at least facultatively quadrupedal, and the larger melanorosaurids were mostly or entirely quadrupedal. SAUROPODS were all quadrupedal. On the view that sauropods evolved from melanorosaurid or closely related basal sauropodomorphs, this quadrupedality was a continuation of an inherited condition. However, if sauropods did not evolve from prosauropods, they must have become quadrupedal independently from unknown relatives because their saurischian outgroups (prosauropods and theropods) were bipedal, at least originally (Sereno, 1991). Some groups appear to have been facultatively quadrupedal, depending on the situation. Within ORNITHISCHIA, the basal forms (Fabrosaurus, Lesothosaurus, and related taxa) were bipeds, and this is true for basal members of the major ornithischian branches. In the THYREOPHORA, Scutellosaurus, a lightly built armored biped, is the outgroup to Scelidosaurus and the ankylosaurs and stegosaurs, which were habitually quadrupedal. Heterodontosaurus, in turn, is the corresponding basal taxon to Scutellosaurus for the Ornithischia, and it is clearly a bipedal, small and long-legged form. Conventionally, all the typical ORNITHOPODA are considered to have been bipedal, although some larger forms, notably the hadrosaurids and some iguanodontids, were facultatively quadrupedal, especially while foraging. Some recent analyses have considered these larger forms quadrupedal most of the time, given the hoof-like unguals

that are borne on the hand as well as the foot. The MARGINOCEPHALIA (ceratopsians and pachycephalosaurs) are sometimes considered a third branch of Ornithischia and sometimes an offshoot from Ornithopoda, but in either case the common marginocephalian ancestor was almost certainly bipedal: All pachycephalosaurs were and the basal ceratopsians such as Psittacosaurus were as well.


References Alexander, R. McN. (1985). Mechanics of posture and gait of some large dinosaurs. Zool. J. Linnean Soc. 83, 1–25. Brinkman, D. (1980). The hind limb step cycle of Caiman sclerops and the mechanics of the crocodile tarsus and metatarsus. Can. J. Zool. 58, 2187–2200. Christian, A., Horn, H.-G., and Preuschoft, H. (1994). Biomechanical reasons for bipedalism in reptiles. AmphibiaReptilia 15, 275–284. Coombs, W. P. (1978). Theoretical aspects of cursorial adaptations in dinosaurs. Q. Rev. Biol. 53, 393–418. Gatesy, S. M. (1991). Hind limb movements of the American alligator (Alligator mississipiensis) and postural grades. J. Zool. (London) 224, 577–588. Gatesy, S. M., and Dial, K. P. (1996). Locomotor modules and the evolution of avian flight. Evolution 50, 331–340. Padian, K., and Olsen, P. E. (1989). Ratite footprints and the stance and gait of Mesozoic theropods. In Dinosaur Tracks and Traces (D. D. Gillette and M. Lockley, Eds.), pp. 231–241. Cambridge Univ. Press, New York. Parrish, J. M. (1986). Locomotor adaptations in the hindlimb and pelvis of the thecodontians. Hunteria 1(2), 1–35. Parrish, J. M. (1987). The origin of crocodilian locomotion. Paleobiology 13, 396–414. Sereno, P. C. (1991). Basal archosaurs: Phylogenetic relationships and functional implications. Soc. Vertebr. Paeontol. Mem. 2, 1–53.

Bird Origins KEVIN PADIAN University of California Berkeley, California, USA

LUIS M. CHIAPPE American Museum of Natural History New York, New York, USA


ish in 1916). Heilmann’s book was an exceptionally thorough consideration of avian biology, including skeletal anatomy, embryology, musculature, pterylosis, paleontology, and many other subjects. Heilmann found that theropod dinosaurs were most similar of all fossil groups to Archaeopteryx and other birds, but he rejected a theropod ancestry because theropods lacked clavicles; hence, under his interpretation of Dollo’s law of evolutionary irreversibility, the clavicles (furcula) of birds could not have reevolved from a theropod precursor. He concluded that the origin of birds must have been among more archaic ARCHOSAURS (see also ORNITHOSUCHIA; PSEUDOSUCHIA), perhaps forms related to Ornithosuchus or Euparkeria, which had clavicles. Clavicles have since been found in the basal ceratosaurian theropods Coelophysis and Segisaurus, and a fully formed furcula has been recently discovered in tetanuran theropods ranging from Allosaurus and tyrannosaurs to Velociraptor, Oviraptor, and Ingenia (Fig. 2; see PECTORAL GIRDLE). Some critics contend that the avian furcula is a neomorph not homologous to the reptilian clavicles, partly because the latter are apparently lost in orni-

ince the 1970s it has come to be nearly universally accepted that birds evolved from small carnivorous dinosaurs most closely related to DROMAEOSAURIDS, probably sometime in the Middle to early Late Jurassic. Archaeopteryx (Fig. 1) is the first known bird, now represented by seven skeletons and a feather from the Late Jurassic SOLNHOFEN limestones of Germany. Other records of Late Jurassic birds so far have been questionable or apocryphal, although research in the past decade continues to unearth new Early Cretaceous birds that are only slightly more derived than Archaeopteryx (Chiappe, 1995; Padian and Chiappe, 1997; see AVES). These new finds help to fill the stratigraphic and morphological gaps between Archaeopteryx and more derived, Late Cretaceous birds such as Hesperornis and Ichthyornis, which have been known for well over a century.

Hypotheses of Bird Origins As reviewed by Gauthier (1986) and Witmer (1991), there are three major hypotheses of bird origins. One is that they evolved from an unspecified group of basal archosaurs characterized by the disused wastebasket term ‘‘THECODONTS.’’ A second is that they share an immediate common ancestor with crocodylomorphs. A third is that they evolved from small THEROPOD dinosaurs. Other suggestions have been made, including common ancestry with lizards, pterosaurs, or mammals, but these ideas were based on only superficial resemblances in a few features and were discredited long ago (Gauthier, 1986). ‘‘Thecodont’’ Hypothesis This can be traced to the early 1900s but reached its most detailed statement in 1926 with the English-language publication of Gerhard Heilmann’s classic The Origin of Birds (an earlier version, Fuglenes Afstamning, was published in Dan-

FIGURE 1 Archaeopteryx, the first known bird, restored by J. H. Ostrom.



FIGURE 2 The clavicles in several theropods, including (a) the ceratosaur Segisaurus, (b) a new allosauroid, the maniraptoran coelurosaurs, (c) Ingenia, and (d) Archaeopteryx.

thodirans (they are absent in pterosaurs and not known in any nontheropodan dinosauromorph) and partly because in some recent birds the furcula seems to be composed of both dermal and endochondral bone. Regardless of these facts, there is no doubt about theropod monophyly, so the homology of the ceratosaurian clavicles and the tetanuran furcula would not seem to be in question; also, there is no mistaking the identity in shape and position of the boomerang-shaped furcula in nonavian tetanurans, Archaeopteryx, and other birds. Heilmann’s retreat to a thecodont hypothesis was a default argument; as he recognized, no features that linked any particular basal archosaur to birds are also not found in theropods, usually with greater similarity. Since Heilmann’s work, other authors have advocated a thecodont hypothesis, and the approach is very much the same. No specific candidate among basal archosaurs has been presented as the direct ancestor or the closest known animal to birds; rather, a range of forms with one or two supposedly birdlike characters is advanced, even though most of their character states are far more primitive than those in theropods (see Witmer, 1991, for a thoughtful review). Tarsitano (1991) has advanced most forcefully the idea that the origin of birds is to be found among such ‘‘avimorph thecodonts,’’ but cladistic analyses have found that all these animals are more closely related to other forms quite distant from birds. For example, Cosesaurus and Megalancosaurus are aquatic prolacertiform archosauromorphs, Scleromochlus is the closest known sister group to pterosaurs, and Lagosuchus and Lagerpeton are closest to basal dinosaurs (Gauthier, 1986; Sereno, 1991; Benton, 1988; Padian and Chiappe, 1997). Protoavis (Chatterjee,

Bird Origins 1991, 1995, 1997) has been advanced as a Triassic bird but has been met with substantial skepticism (reviewed in Padian and Chiappe, 1997); there are apparently even more differences of opinion about the interpretation of its morphology than there are about Archaeopteryx. The question has been made more difficult by the circumstance that, in the more than two decades since this controversy was renewed, no advocate of a thecodont ancestry has produced a cladogram incorporating all, or even any, of the available evidence that would support such a case. Cladistic analyses (see SYSTEMATICS; PHYLOGENETIC SYSTEM) are not infallible but at least they are explicit: If the weight of evidence supports a different interpretation than the several independent analyses that have placed birds within the theropods, then this result would be very interesting to see expressed in cladistic terms. Crocodilian Hypothesis This should more properly be termed the ‘‘crocodylomorph’’ hypothesis because its advocates regard the ancestry of birds as complete before true CROCODILES evolved. Indeed, it is within the sphenosuchian crocodylomorphs, outside Crocodylia, that bird-like characters appear to have been most pronounced. A. D. Walker (e.g., 1977) has been the chief advocate of this view, based on his detailed studies of the braincase, quadrate, ear region, and other features of Sphenosuchus, an Early Jurassic crocodylomorph from South Africa. His view has been generally supported by L. D. Martin and his students (e.g., Martin, 1983), although they have tended to draw similarities to birds from true crocodilians as much as from crocodylomorphs. Many of these similarities are valid but have been shown to apply either to a more general level among archosaurs or to have evolved convergently in certain crocodilians and early birds (but not present in the hypothesized common ancestor of both groups) (Gauthier, 1986). Again, no cladogram incorporating all the available evidence has to date supported a crocodylomorph origin of birds. Theropod Hypothesis This had its roots in the 1860s with T. H. Huxley, who noted in a series of papers (e.g., 1870) a suite of 35 characters shared uniquely by birds and theropod dinosaurs (reviewed by Gauthier, 1986, pp. 4–6). Many of these are still considered valid today, whereas others have turned

Bird Origins out to be more general to dinosaurian or other archosaurian groups. Desmond (1982) and Gauthier (1986) have both noted that Huxley’s hypothesis, presented to the Geological Society of London in 1870, was contested by Harry Govier Seeley, who ‘‘thought it possible that the peculiar structure of the hinder limbs of the Dinosauria was due to the functions they performed, rather than to any actual affinity with birds’’ (See DINOSAURIA: DEFINITION). This shadow of potential convergence, though not explicitly tested either by Seeley or anyone since, nonetheless not only frustrated the acceptance of Huxley’s views at the time but also continues to be contended by opponents of the theropod hypothesis (e.g., Feduccia, 1996). In the 1970s John Ostrom, in a series of papers (e.g., 1975a,b, 1976b), demonstrated the detailed similarities of Archaeopteryx to theropod dinosaurs. Although he did not specify a taxon within Theropoda to which birds might be directly connected, his comparisons tended to run to the dromaeosaurid Deinonychus, which he had recently described in a monograph (Ostrom, 1969). As it turned out, dromaeosaurids were found to be the closest sister group to birds in several independent cladistic analyses beginning in the early 1980s by Padian, Gauthier, Benton, Sereno, Holtz, Perle et al., and others (see Padian and Chiappe, 1997). Synapomorphies that link dromaeosaurids and Archaeopteryx include the presence of dorsal, caudal, and rostral tympanic recesses, the semilunate carpal, thin metacarpal III, longer pubic peduncle, posteroventrally directed pubis with only a posteriorly projecting foot, shortened ischium, and other features of the skull, pectoral girdle, and hindlimb (Gauthier, 1986; Padian and Chiappe, 1997). A great many characters classically considered ‘‘avian’’ apply to more general levels within Theropoda. Basal theropods have lightly built bones and a foot reduced to three main toes, with the first usually held off the ground and the fifth lost. Closer to birds, the fifth and fourth digits of the hand are progressively reduced and lost, the skeleton (especially the vertebrae) becomes lighter, and the tail becomes shorter as its vertebrae partially interlock through the elongation of zygapophyses to reinforce its stiffness. In COELUROSAURS (sensu Gauthier, 1986, the sister taxon to CARNOSAURS), contrary to the picture suggested by opponents of the theropod hypothesis, the forelimbs become progressively longer until they are nearly as long as the hindlimbs in some dromaeo-


FIGURE 3 The forelimbs compared to the hindlimbs in (a) the ceratosaur Coelophysis, (b) the ostrich dinosaur Struthiomimus, (c) the maniraptoran coelurosaur Deinonychus, and (d) Archaeopteryx. Compare these to the phylogeny in Fig. 4.

saurs (Fig. 3); the first toe (hallux) begins to rotate behind the metatarsus, although it does not descend to the point seen in perching birds; the metatarsals become longer; and the scapular blade becomes longer and more strap-like. The presence of the furcula may turn out to be a general tetanuran character, and it is not yet clear how general the calcified sternum in adults may be (it is known, for example, in some oviraptorids, dromaeosaurs, tyrannosaurs, and sinraptorids). Holtz (1994, 1996) reevaluated the phylogenetic relationships of Theropoda, and his conclusions uphold Gauthier’s (1986) comprehensive analysis, with some adjustments that do not affect the position of birds within Theropoda (Fig. 4).

Clavicles, Digits, Feathers, and Stratigraphy Individual characters are sometimes advanced as conclusive evidence that birds could not have descended from theropod dinosaurs. For example, it is claimed that the digits of the bird hand are II–III–IV, whereas they are I–II–III in theropods; that the semilunate carpal bones of maniraptorans and other theropods are different from those of Archaeopteryx and the birds; and that the ascending process of the astragalus is not the same in theropods and birds. [Feduccia (1996) reviews these claims favorably, and to obviate

Bird Origins


FIGURE 4 Phylogeny of Theropoda, after Holtz (1994, 1996).

extensive literature citations readers are referred to his work for historical background and strong advocacy.] These statements amount to hypotheses that the characters are not homologs but homoplasies. Since Darwin’s day, homologies have been recognized as features inherited from common ancestors; even to non-Darwinians, such as Richard Owen, homologies were established by similarity of morphology, position, development, and histological structure. Roth (1988) proposed that, like monophyletic taxa, homologies are defined by ancestry but diagnosed by features such as the criteria just listed. In comparative biology, the recognition of homologous structures in two or more organisms can be tested using the phylogenetic distributions of other, presumably independent characters and by including more taxa in the analysis (see SYSTEMATICS). Proceeding according to this method, the hypothesis that the furcula of birds is not homologous to the clavicles of theropod dinosaurs is weak both because undoubted nonavian tetanuran theropods have boomerang-shaped clavicles in the same position as those of Archaeopteryx and other basal birds and because basal theropods have structures that are manifestly

similar in morphology and position to clavicles in other tetrapods. Digits of the Bird Hand The three remaining digits of the bird hand are sometimes regarded as I, II, and III and sometimes as II, III, and IV. The theropod hand is unquestionably I, II, and III, and opponents of the theropod hypothesis of avian descent are unanimous in their contention that the bird hand is II, III, and IV (therefore, the digits could not be homologous between the two groups). This is often inaccurately portrayed as a difference between paleontologists and ornithologists: Feduccia (1996, p. 2), for example, reproduced a figure of the avian skeleton by Lucas and Stettenheim with the digits numbered II, III, and IV, but reproduced on page 7 of the same work two figures, one by Van Tyne and Berger and another by Burton and Milne, in which the digits are numbered I, II, and III. These illustrations are all by ornithologists. Classical ornithologists from W. K. Parker to Proctor and Lynch (1993) have agreed, and Heilmann (1926) examined the problem at length and reached the same conclusion: I, II, and III. The evidence for II, III, and IV comes entirely from

Bird Origins some interpretations of the ontogeny of the hand in some living birds. As Heilmann (1926) noted, interpretations on this basis have varied. Some have been based on an assumption of Morse’s ‘‘law’’ that digits must be lost from the both sides inward (as in the bird foot: 5, then 1; and in horses’ feet: 5, then 1, 4, and 2), although this pattern must surely be related to the weight-bearing function of the locomotory structures. Some authors have even claimed to find Anlagen of four digits in the bird hand. Hinchliffe and Hecht (see Hecht and Hecht, 1994) have strongly advocated the modern developmental case for II, III, and IV based on the presence of an ephemeral ‘‘element X’’ medial and palmar to the wrist in early ontogeny that is taken for a remnant of digit I. However, as Shubin (1994) and Padian and Chiappe (1997) note, Anlagen do not appear with labels on them but have to be interpreted. To accept the II–III–IV view, the first digit, including its carpal and metacarpal, has to be lost, and digits II, III, and IV have to assume the precise forms, articulations, and proportions of the original digits I, II, and III. If this is possible, developmental biology has so far not provided examples or mechanisms from living tetrapods that support this process. Unfortunately, we do not have similar embryological stages for Archaeopteryx or any other Mesozoic birds or dinosaurs, so we cannot examine phylogenetically the hypothesis of element X outside living birds. In favor of the I–II–III hypothesis, the theropod

75 hand follows a consistent pattern of reduction and loss of elements from the lateral side medially (Shubin, 1994; Padian and Chiappe, 1997) (Fig. 5). Crocodilians and other ornithodiran outgroups have a five-fingered manus in which the third digit is longest. In Dinosauria the fourth and fifth digits are reduced in size, and the phalangeal formula is reduced to 2-3-4-3-2. In Saurischia the second digit becomes the longest, and the fourth digit’s phalanges are reduced. In Theropoda the fifth digit is reduced to a nubbin of the metacarpal or lost altogether, and the fourth digit is quite small. In tetanurans all trace of the outer two digits is lost. In tyrannosaurs [which von Huene (1914, 1920, 1926) and Novas (1992) showed are not carnosaurs but actually gigantic coelurosaurs], the third digit is lost, thus proving the invalidity of Morse’s law at least in the case of dinosaurs. Given the suite of dozens of synapomorphies from other parts of the skeleton that also support the placement of birds within theropods, it is difficult not to accept the manifest similarities of other features of the hand of Archaeopteryx as homologous to those of theropods, including 14 characters related to the form and proportion of hand and wrist elements in theropods (Gauthier, 1986: characters 21–26, 43–46, 61, 62, 75, and 76). Semilunate Carpal In maniraptorans, and perhaps at a more general level within tetanurans, a halfmoon-shaped wrist element overlaps the bases of

FIGURE 5 The reduction of the hand in theropods, including birds, from Padian and Chiappe (1997).

76 metacarpals I and II (Ostrom, 1969, 1975a,b, 1976b; Gauthier, 1986). The rounded proximal surface allows the hand to swivel sideways, a feature that appears to have been related to predation in basal maniraptorans and exapted for the flight stroke in birds (Gauthier and Padian, 1985). There appears to be little question about the morphological similarity of this bone in Archaeopteryx and troodontids, oviraptors, and dromaeosaurs such as Deinonychus and Velociraptor; the question surrounds its homology. Carpal elements are often incomplete or unknown for Mesozoic theropods: They may not have been preserved because they were not ossified, perhaps because the specimen in question was not adult; or they may have been removed or destroyed by taphonomic processes, preserved out of position and so unrecognized and not collected, or collected and not correctly identified. Furthermore, as in other amniotes, during ontogeny many of these elements attained better definition and frequently fused, and phylogenetically fusion appears to have increased (Gauthier, 1986). There is no simple answer to the identity of some elements of the wrist in theropods (Feduccia, 1996), but some points are clear. Contrary to some inferences (Feduccia, 1996), theropods have not ‘‘lost’’ a row of carpals, but sometimes they are incompletely ossified or preserved. Ostrom (1969 et passim) identified the semilunate carpal in Deinonychus and Archaeopteryx as the radiale and the element next to it as the ulnare, but this cannot be because the radiale contacts the radius and does not contact the metacarpals in any tetrapod. Comparison to basal theropods, such as Coelophysis and Syntarsus, in which the manus is preserved, demonstrates a row of carpals between those contacting the radius and ulna (de facto the radiale and ulnare) and the metacarpals (Fig. 6). A single distal carpal, identified by both Colbert and Raath as a fusion of distal carpals 1 and 2, overlaps metacarpal II, and this element is exactly in the position of the semilunate carpal of tetanurans (Gauthier, 1986). In birds, the radiale and ulnare have become more tightly associated with the radius and ulna as the sideways flexion of the wrist has evolved between the proximal and distal carpal rows (Padian and Chiappe, 1997); the distal carpals have become associated immovably with the metacarpals (Ostrom, 1976a). Hence, the semilunate carpal of birds and other tetanurans is the result of the fusion of distal carpals

Bird Origins

FIGURE 6 The wrists of Coelophysis, Deinonychus, and Archaeopteryx (modified from Colbert and Ostrom).

1 and 2; the radiale and ulnare are not lost but are not always preserved and should be sought in association with the radius and ulna; and a third distal carpal is usually present, if ossified, in tetanurans, but the fourth distal carpal, like the fifth, has been lost in tetanurans along with these digits. Pubic Foot The end of the pubis is unexpanded in basal dinosaur groups, including all ornithischians, sauropodomorphs, and basal theropods (ceratosaurs, including Coelophysis, Syntarsus, Ceratosaurus, etc.). In tetanurans it is expanded fore and aft: This can be seen both in carnosaurs such as Allosaurus and in coelurosaurs such as ornithomimids and tyrannosaurs (Fig. 7). In maniraptorans, such as the dromaeosaurs Deinonychus, Velociraptor, and Adasaurus, the anterior projection of this foot is severely reduced or lost, as it is in Archaeopteryx and the birds. Moreover, in dromaeosaurids, as in birds, the pubis itself is retroverted (and convergently in therizinosaurids), although apparently not to the extent seen in living birds. Herrerasaurids, seen variously as basal theropods, basal saurischians, or a sister taxon to dinosaurs, also have a development of the distal pubis similar in some respects to the theropod pubic foot, but this is expected to be a convergence because they are not otherwise closer to birds than are dromaeo-

Bird Origins


FIGURE 7 The anatomy of the pubic foot varies and can be used to establish affinities among and between lineages of dinosaurs.

saurids. Likewise, the Triassic archosaur Postosuchus (like other poposaurids) appears to have a similar expansion of the distal pubis, but this is clearly a convergence because Postosuchus is closely allied not to birds or dinosaurs but rather to crocodylomorphs (see PSEUDOSUCHIA). Ascending Process of the Astragalus In ornithodiran ornithosuchians, posture and gait have changed from the general sprawled or semierect reptilian condition to a more upright stance and a parasagittal gait (Fig. 8). The ankle flexes mesotarsally so that the proximal ankle elements (astragalus and calcaneum) are associated with the lower leg and the distal tarsals with the metatarsals and phalanges. Because parasagittal gait virtually eliminates rotation at the knee, and skeletal mass is concentrated medially, the tibia and fibula have the same action. The tibia becomes the dominant element and the fibula is reduced, and the same happens to their corresponding proximal tarsals. Astragalus and calcaneum frequently fuse in adults, and in no cases do they rotate against each other, as in crurotarsal archosaurs (see PSEUDOSUCHIA). The astragalus expands transversely in basal ornithodirans such as pterosaurs (in which the

proximal tarsals are always fused to each other and to the tibia and the fibula is reduced to a splint), Marasuchus/Lagosuchus, and Lagerpeton. Hence, for the first time, the astragalus articulates with both the tibia and fibula. A dorsal process of the astragalus separating these two articulations begins in basal ornithodirans and is known in all taxa in which the elements are preserved. It is continuous with an ascending process that is posterior to the lower leg bones in Lagerpeton, medial in Marasuchus/Lagosuchus, but anterior in dinosaurs and much more expanded in theropods (especially tetanurans) than in any other taxa. Based on congruence with other characters, it would appear parsimonious to conclude that birds carry on this basal ornithodiran feature because it is seen to trend through Theropoda. The identity of this ossification, however, is at issue; it is held by some to be different from the avian ‘‘pretibial’’ bone (summarized in Feduccia, 1996), which is a separate ossification. Its shape and position on the tibia also varies, but generally it is centered anterolaterally, as is the pretibial bone. Feduccia (1996, p. 75) misinterpreted the variation in the ascending process described by Welles and Long (1974): They identified five morphological types, but these

FIGURE 8 The ankle region in (a) Coelophysis, (b) Deinonychus, (c) Archaeopteryx, (d) the Cretaceous bird Baptornis (after Martin, 1983, with permission), and (e) the hoatzin (after Martin, Stewart, and Whetstone).

Bird Origins

78 are not of independent phylogenetic origin, and in fact the types they identified are mostly not characteristic of any natural taxa within theropods; they are simply morphological types that vary for reasons apparently connected with relative size or functional features. As Gauthier (1986, p. 29) noted, the ascending process may be a separate ossification in Dilophosaurus (as S. P. Welles first discovered) and other theropods, as it is in birds. Moreover, although the pretibial bone fuses with the precociously developed calcaneum in neognath birds, this is not the case in ratites and tinami in which, as in other theropods, it fuses with the astragalus, and in all cases it is located on the anterolateral side of the tibia. Hence, the pretibial bone of birds appears to be homologous to the ascending process of the astragalus in theropods. Stratigraphic Disjunction A difficulty regarded as insurmountable by opponents of the theropod origin of birds is the presumption that the taxa identified as closest to Archaeopteryx among theropods—the dromaeosaurids—do not appear in the fossil record until Albian–Aptian times (perhaps 110 mya: Deinonychus, Cloverly Formation, Wyoming), whereas Archaeopteryx comes from Late Jurassic (Tithonian) times (approximately 150 mya). The apparent absence of earlier records of dromaeosaurids, although puzzling, is not unusual in the Mesozoic fossil record: For example, although stegosaurs and ankylosaurs are regarded as sister taxa that must have diverged by the late Early Jurassic, stegosaurs are not known before the Bathonian–Callovian (approximately 170 mya), whereas before the 1980s, ankylosaurs were not known before the Aptian–Albian (approximately 110 mya; Weishampel et al., 1990). The situation is not unique to dinosaurs. No one doubts today that marsupials and placentals are sister taxa within mammals, and monotremes are their sister taxon. Hence, the split between therians (marsupials ⫹ placentals) and monotremes must have taken place before the first recognizable marsupials and placentals evolved. However, the first marsupials and placentals are known from Early Cretaceous times (approximately 100 mya), whereas until recently, monotremes were not known until the Oligocene (approximately 20 mya), a disjunction of 80 million years—over twice that between Archaeopteryx and Deinonychus (Carroll, 1988)! Moreover, small maniraptorans are not at all

absent from Late Jurassic sediments: Jensen and Padian (1989) described a collection of bones pertaining to small maniraptorans from the Dry Mesa Quarry (Late Jurassic: ?Tithonian; Morrison Formation, Colorado). These bones unfortunately could not be identified to the generic level but nonetheless indicated that if they are not bones of birds, then they are certainly those of their sister taxon, the dromaeosaurids. These arguments would appear to dispose of the fatality of the stratigraphic argument to the theropod hypothesis. In summary, birds, as Gauthier (1986) pointed out, must be considered dinosaurs because phylogenetic analysis clearly indicates that they evolved from dinosaurs. They are not only dinosaurs but also saurischian, theropodan, tetanuran, and maniraptoran dinosaurs. Arguments to the contrary have been proposed for 20 years since the theropod hypothesis was advanced by Ostrom, but these can no longer be considered matters of evidence. Rather, it is a question of whether one uses the methods of modern comparative biology (see SYSTEMATICS). Issues related to the origin of flight in birds and other topics with starkly contrasting viewpoints are discussed at length in Hecht et al. (1985), Schultze and Trueb (1991), Feduccia (1996), and Padian and Chiappe (1997).

See also the following related entries: AVES ● AVIALAE ● COELUROSAURIA ● DROMAEOSAURIDAE

References Benton, M. J. (Ed.) (1988). The Phylogeny and Classification of the Tetrapods, Volume 1. Clarendon, Oxford. Carroll, R. L. (1988). Vertebrate Paleontology and Evolution. Freeman, New York. Chatterjee, S. (1991). Cranial anatomy and relationships of a new Triassic bird from Texas. Philos. Trans. R. Soc. London B 332, 277–346. Chatterjee, S. (1995). The Triassic bird Protoavis. Archaeopteryx 13, 15–31. Chatterjee, S. (1997). Protoavis and the early evolution of birds. Palaeontographica, in press. [Abstract A]. Chiappe, L. M. (1995). The first 85 million years of avian evolution. Nature 378, 349–355. Desmond, A. J. (1982). Archetypes and Ancestors: Palaeontology in Victorian London, 1850–1875. Muller, London.

Bird Origins


Feduccia, A. (1996). The Origin and Evolution of Birds. Yale Univ. Press, New Haven, CT.

Ostrom, J. H. (1976b). Archaeopteryx and the origin of birds. Biol. J. Linnean Soc. 8, 91–182.

Gauthier, J. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55.

Padian, K., and Chiappe, L. M. (1997). The early evolution of birds. Biol. Rev., in press.

Gauthier, J., and Padian, K. (1985). Phylogenetic, functional, and aerodynamic analyses of the origin of birds and their flight. In The Beginnings of Birds (M. K. Hecht, J. H. Ostrom, G. Viohl, and P. Wellnhofer, Eds.), pp. 185–197. Freunde des Jura-Museums, Eichstatt.

Proctor, N. S., and Lynch, P. J. (1993). Manual of Ornithology: Avian Structure and Function. Yale Univ. Press, New Haven, CT.

Hecht, M. K., and Hecht, B. M. (1994). Conflicting developmental and paleontological data: The case of the bird manus. Acta Paleontol. Polonica 38(3/4), 329–338. Hecht, M. K., Ostrom, J. H., Viohl, G., and Wellnhofer, P. (1985). The Beginnings of Birds. Freunde des Jura-Museums, Eichstatt. Heilmann, G. (1926). The Origin of Birds. pp. 210. Appleton, New York. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68(5), 1100–1117. Holtz, T. R., Jr. (1996). Phylogenetic taxonomy of the Coelurosauria (Dinosauria: Theropoda). J. Paleontol. 70, 536–538. Huxley, T. H. (1870). Further evidence of the affinities between the dinosaurian reptiles and birds. Q. J. Geol. Soc. London 26, 12–31. Jensen, J. A., and Padian, K. (1989). Small pterosaurs and dinosaurs from the Uncompahgre Fauna (Brushy Basin Member, Morrison Formation: ?Tithonian), Late Jurassic, Western Colorado. J. Paleontol. 63, 364–373. Martin, L. D. (1983). The origin of birds and of avian flight. In Current Ornithology (R. F. Johnston, Ed.), Vol. 1, pp. 106–129. Plenum, New York. Novas, F. E. (1992). La evolucio´n de los dinosaurios carnı´voros. In Los Dinosaurios y su entorno bio´tico (J. L. Sanz and A. D. Buscalione, Eds.), pp. 125–163. Instituto ‘‘Juan Valdes,’’ Cuenca, Spain. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Bull. Peabody Museum Nat. History Yale Univ. 30, 1–165. Ostrom, J. H. (1974). Archaeopteryx and the origin of flight. Q. Rev. Biol. 49, 27–47. Ostrom, J. H. (1975a). The origin of birds. Annu. Rev. Earth Planet. Sci. 3, 35–57. Ostrom, J. H. (1975b). On the origin of Archaeopteryx and the ancestry of birds. Proc. CNRS Colloq. Int. Prob. Act. Paleontol.—Evol. Verte´br. 218, 519–532. Ostrom, J. H. (1976a). Some hypothetical anatomical stages in the evolution of avian flight. Smithsonian Contrib. Paleobiol. 27, 1–27.

Roth, V. L. (1988). The biological basis of homology. In Ontogeny and Systematics (C. J. Humphries, Ed.), pp. 1–26. Columbia Univ. Press, New York. Schultze, H.-P., and Trueb, L. (Eds.) (1991). Origins of the Higher Groups of Tetrapods. Cornell Univ. Press, Ithaca, NY. Sereno, P. (1991). Basal archosaurs: phylogenetic relationships and functional implications. J. Vertebr. Paleontol. 11(Suppl. 4), 1–53. Shubin, N. (1994). History, ontogeny, and evolution of the archetype. In Homology: The Hierarchical Basis of Comparative Biology (B. K. Hall, Ed.), pp. 248–271. Academic Press, New York. Tarsitano, S. (1991). Archaeopteryx: Quo Vadis? In Origins of the Higher Groups of Tetrapods (H.-P. Schultze and L. Trueb, Eds.), pp. 541–576. Cornell Univ. Press, Ithaca, NY. von Huene, F. (1914). Das naturliche System der Saurischia. Zentralblatt Mineral. Geol. Pala¨ontol. B 1914, 154–158. von Huene, F. (1920). Bemerkungen zur Systematik und Stammesgeschichte einiger Reptilien. Zeitschrift Indukt. Abstammungslehre Vererbungslehre 24, 162–166. von Huene, F. (1926). The carnivorous Saurischia in the Jura and Cretaceous formations, principally in Europe. Revista Museo de La Plata 29, 35–167. Walker, A. D. (1977). Evolution of the pelvis in birds and dinosaurs. In Problems in Vertebrate Evolution (S. M. Andrews, R. S. Miles, and A. D. Walker, Eds.), pp. 319– 357. Academic Press, New York. Walker, A. D. (1985). The braincase of Archaeopteryx. In The Beginnings of Birds (M. K. Hecht, J. H. Ostrom, G. Viohl, and P. Wellnhofer, Eds.), pp. 123–134. Freunde des Jura-Museums, Eichstatt. Weishampel, D. B., Dodson, P., and Osmo´lska, H. (Eds.) (1990). The Dinosauria. Univ. of California Press, Berkeley. Welles, S. P., and Long, R. A. (1974). The tarsus of theropod dinosaurs. Ann. South African Museum 64, 191–218. Witmer, L. (1991). Perspectives on avian origins. In Origins of the Higher Groups of Tetrapods (H.-P. Schultze and L. Trueb, Eds.), pp. 427–466. Cornell Univ. Press, Ithaca, NY.

Bone Cabin Quarry


Birmingham Museum, United Kingdom see MUSEUMS



Black Hills Museum of Natural History, South Dakota, USA see MUSEUMS



Bloemfontein National Museum see NATIONAL MUSEUM, BLOEMFONTEIN

Bolivian Dinosaurs Late Cretaceous dinosaur footprints have been discovered in Bolivia.



Bone Cabin Quarry BRENT H. BREITHAUPT University of Wyoming Laramie, Wyoming, USA


lthough Cope and Marsh carried their bitter feud to their deaths in 1897 and 1899, respectively, by the 1890s the ‘‘bone wars’’ were over in the West and

other institutions had begun collecting there. In 1897 Henry Fairfield Osborn at the American Museum of Natural History sent collecting crews to the area of Wyoming where Marsh and Cope’s crews had collected dinosaurs years earlier. The American Museum crews, under the direction of Jacob Wortman, found that much of Como Bluff was barren of vertebrate fossils. However, they did discover and collect partial skeletons of Diplodocus and Apatosaurus the first year (see HISTORY OF DISCOVERY: EARLY YEARS). Patience and persistence resulted in American Museum of Natural History crews finding several new dinosaur quarries in the region. One of the most spectacular of these was the famous Bone Cabin Quarry found in 1898 on a hill approximately 15 km north of Como Bluff on the Little Medicine Bow Anticline. Taking its name from the remains of a sheepherder’s cabin foundation made entirely of dinosaur bone fragments, Bone Cabin Quarry preserved more than 50 partial dinosaur skeletons (e.g., Diplodocus, Camarasaurus, Apatosaurus, Allosaurus, Ornitholestes, Camptosaurus, Dryosaurus, and Stegosaurus) in a total area of only 1,529 m2. Major collections were made for the American Museum from this and other nearby Morrison Formation quarries (e.g., Nine Mile Quarry and Quarry R) between 1898 and 1905 (McIntosh, 1990), primarily under the direction of Walter Granger after Wortman took a position with the Carnegie Museum in 1899. A large number of complete articulated sauropod fore- and hindlimbs, in some cases with complete feet, were found (Osborn, 1899). Many long tail segments were also found, in addition to several skulls. Famed dinosaur collector William Harlow Reed wrote that he and Frank Williston had come across this site decades earlier but because of the deteriorated condition of many of the bones (called ‘‘head cheese’’) they decided to ignore it (Breithaupt, 1990). Reed was independently working in this region in 1901 when American Museum crews were collecting fossils there. Because Reed had recently resigned from the field crew of the Carnegie Museum, he proposed working for the American Museum of Natural History. Reed’s Quarry R (located to the southeast of Bone Cabin Quarry on the Prager Anticline) and the Nine Mile Quarry (located south of Bone Cabin Quarry on the Little Medicine Bow Anticline) proved to be quite rich in dinosaur remains. Camarasaurus and Allosaurus bones were found at Quarry R and a

Braincase Anatomy partial skeleton of Apatosaurus was located at Nine Mile Quarry. During the major field seasons in Wyoming (1898– 1903), crews from the American Museum of Natural History collected approximately 275 boxes weighing more than 68 metric tons (Colbert, 1968). More than 500 specimens representing approximately 69 animals (i.e., 44 sauropods, 3 stegosaurs, 4 ornithopods, 9 theropods, 4 crocodiles, and 5 turtles) were found (McIntosh, 1990; Osborn, 1904). American Museum of Natural History exhibition specimens from this region of Wyoming collected between 1897 and 1905 included Apatosaurus (Nine Mile Quarry and Bone Cabin Quarry; 1898), Stegosaurus (Como Bluff; 1901), Ornitholestes (Bone Cabin Quarry; 1901), and Camptosaurus (Bone Cabin Quarry; 1905) (Norell et al., 1995). Another exhibit sauropod skeleton was found by American Museum crews in Bone Cabin Quarry. This specimen of Diplodocus was subsequently sent to Germany (McIntosh, 1990). In 1905 the Apatosaurus remains from the Nine Mile and Bone Cabin quarries were used by the American Museum of Natural History to mount the first sauropod skeleton in the world (Norell et al., 1995). In the 1990s Bone Cabin Quarry was reopened and important remains of Allosaurus, Diplodocus, Camarasaurus, Stegosaurus, and an ankylosaur were recovered.

See also the following related entries: BEHAVIOR ● MIGRATION ● MORRISON FORMATION ● TAPHONOMY


Bone Chemistry In addition to information on the chemical composition of bone and other preservable material, there are also issues relevant to preservation of interest to paleontologists.


Brain Studies of the brains of dinosaurs are done by making computerized axial tomography scans of well-preserved, relatively complete, and undeformed skulls, by making endocranial casts, and by examining natural endocranial casts.



Braincase Anatomy PHILIP J. CURRIE Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada


The braincase is generally one of the most poorly

Breithaupt, B. H. (1990). Biography of William Harlow Reed: The story of a frontier fossil collector. Earth Sci. History 9, 6–13.

understood regions of the dinosaur skeleton. It is often partially or completely obscured by other skull bones. Parts of it do not ossify or are fragile and easily destroyed, and it is a complex of numerous bones pierced by nerves, blood vessels, and pneumatic diverticula. Nevertheless, braincases have been described for each of the major dinosaurian lineages. Increased use of noninvasive computerized tomography (CT) scanning (see COMPUTERS AND RELATED TECHNOLOGY) has also revealed details that could previously only have been visible through serial sectioning, which would invariably result in the destruction of most of the specimen. Because the braincase is not directly subject to the same selective pressures as parts of the skeleton in-

Colbert, E. H. (1968). Men and Dinosaurs, pp. 283. Dutton, New York. McIntosh, J. S. (1990). The second Jurassic dinosaur rush. Earth Sci. History 9, 22–27. Norell, M. A., Gaffney, E. S., and Dingus, L. (1995). Discovering Dinosaurs in the American Museum of Natural History, pp. 204. Knopf, New York. Osborn, H. F. (1899). Fore and hind limbs of carnivorous and herbivorous dinosaurs from the Jurassic of Wyoming. Am. Museum Nat. History Bull. 5, 161–172. Osborn, H. F. (1904). Fossil wonders of the West. The dinosaurs of Bone Cabin Quarry. Century Magazine 68, 680–694.

82 volved in the acquisition and processing of food (teeth, limb proportions, etc.), in the protection of the animal from predators (i.e., defensive horns, spikes, and armor), or in sexual or other display structures, its morphology tends to be conservative within a lineage. Comparison of braincases among taxa can therefore provide clues to relationships that may otherwise be obscured in more rapidly evolving parts of the body. The braincase houses the brain; study of the endocranial cavity can approximate the overall size of the brain (see INTELLIGENCE) and show the relative development of different parts of the braincase. Because most cranial nerves and blood vessels pass through foramina and canals in the braincase, and the positions of these openings are conservative in all tetrapods, most of the openings can be identified in fossil skulls. The positions and sizes of these openings can provide information on interrelationships and can give clues about sensory abilities. Because the braincase also forms the inner walls of the middle ear, adjacent bones are often invaded by pneumatic diverticula from the middle ear air sac. These openings are less regular than those of the nerves and blood vessels and can be asymmetrical. The ossified braincase can develop from as many as 23 separate centers of ossification, from bones of both dermal and endochondral origin. Some of the braincase components are paired, some are medial and singular, and others are complexes of several bones that are co-ossified in even the youngest animals. Braincases usually fuse up completely in mature animals, and most of the sutures are difficult to see. Dermal bones include the frontals and parietals on the skull roof and the parasphenoid, which fuses with the basisphenoid. Specific endochondral bones ossify in particular regions of the chondrocranium and can be identified by their consistent relationships to the cranial nerves and to the inferred positions of the cartilages. The braincase is roofed by the frontals and parietals in a consistent manner. The ventral surface of the frontal usually has well-defined impressions of the olfactory tract and the cerebral hemispheres, and depressions in the ventral surface of the paired parietals show what the top of the back part of the brain looked like. In dinosaurs such as troodontids (Russell, 1969), the enlarged cerebral hemispheres mostly covered the midbrain so that it left no impression in the frontals. Hadrosaurs were relatively large-brained

Braincase Anatomy animals (Hopson, 1979), which is reflected by the doming of the frontal bones. The occiput is normally formed by four bones (supraoccipital, basioccipital, and a pair of exoccipitals), all of which form the margins of the foramen magnum. The supraoccipital generally makes a contribution to the dorsal margin of the foramen magnum, the sides are formed by the paired exoccipitals, and the basioccipital makes a small contribution to the ventral margin. The supraoccipitals of Lesothosaurus (Weishampel and Witmer, 1990a), heterodontosaurids (Weishampel and Witmer, 1990b), hypsilophodonts (Galton, 1989), stegosaurs (Gilmore, 1914), and protoceratopsians (Brown and Schlaikjer, 1940) are larger than normal and contribute to the entire dorsal margin of the foramen magnum. In iguanodonts (Taquet, 1976; Norman, 1986; Galton, 1989), hadrosaurs (Langston, 1960; Weishampel and Horner, 1990), ceratopsids (Hatcher et al., 1907), and possibly Pachycephalosaurus (Maryanska and Osmo´lska, 1974), the supraoccipital is excluded from the margin of the foramen magnum by the exoccipitals, which meet on the midline. The supraoccipital normally makes only a narrow contribution to the dorsal margin of the foramen magnum in the vast majority of dinosaurs, including Camptosaurus (Gilmore, 1909), sauropods (Madsen et al., 1995), and most pachycephalosaurids (Maryanska and Osmo´lska, 1974). An anterior extension of the supraoccipital contacts the prootic and laterosphenoid in all neoceratopsians. The two epiotic bones probably form from separate centers of ossification than the supraoccipital. They can make a small contribution to the occiput, but in most specimens they have fused indistinguishably to the supraoccipital (Currie and Zhao, 1993a). In all known dinosaurs the exoccipital is fused without a trace of sutures to the opisthotic, and together they form a conspicuous paroccipital process. The paroccipital process meets the squamosal, parietal, and quadrate in a loose butt joint in most dinosaurs but fuses to the quadrate and squamosal in nodosaurid ankylosaurs. In neoceratopsians, the distal end of the paroccipital process is expanded and is embedded in a slot in the squamosal (Brown and Schlaikjer, 1940). The passages of the 12th cranial nerves are completely enclosed within the exoccipital, which also forms the posteromedial margin of the metotic fissure, through which passed the internal jugular vein and cranial nerves X and XI. The jugular and associated nerves can be diverted posteriorly to

Braincase Anatomy exit on the occiput in some dinosaurs (Currie and Zhao, 1993b). The skull has a prominent knob-like process, known as the occipital condyle, that articulates with the first cervical vertebra (the atlas). It is formed by the basioccipital and both exoccipitals. Usually more than two-thirds of the condyle is formed by the basioccipital, which separates the dorsolateral contributions from the exoccipitals. In neoceratopsians (Hatcher et al., 1907; Brown and Schlaikjer, 1940), the basioccipital is excluded from the foramen magnum by the exoccipitals, which can form up to two-thirds of the ball-like occipital condyle in ceratopsids. The otic capsule is formed from three centers of ossification, although only the outer two—the prootic and opisthotic—can generally be distinguished. As previously mentioned, the epiotic ossifies with the supraoccipital in the tectum synoticum of the chondrocranium. In protoceratopsians and probably ceratopsids (Brown and Schlaikjer, 1940), part of the epiotic is exposed on the lateral surface of the braincase between the prootic, opisthotic, laterosphenoid, and parietal. The facial nerve (VII) and the anterior (vestibular) and posterior (cochlear) branches of the eighth cranial nerve pass through the prootic, which also forms the posterior margin of the exit for cranial nerve V. The prootic sends a tongue-like process posteriorly to extensively overlap the opisthotic above the stapedial recess. The crista prootica often extends ventrolaterally into a wing-like process that forms the anterior wall of a pneumatic cavity in the side of the basisphenoid. This process, referred to by S. Welles (personal communication, 1996) as the preotic pendant, is usually formed in part by the basisphenoid, but in some cases can be formed almost entirely by the basisphenoid (in which case it is called the ala basisphenoidalis). The opisthotic co-ossifies with the exoccipital within the occipital arch of the chondrocranium. It forms the anteroventral borders of the opening for the ninth and tenth cranial nerves and for the stapedial recess. The otic capsule encloses the inner ear and semicircular canals and forms a conspicuous bulge on the inner wall of the endocranial cavity. The floccular recess invades the capsule anteromedially and tends to be relatively large in small animals. These and other bones adjacent to the middle ear sac are invaded by pneumatic diverticula in theropods such as ornithomimids (Osmo´lska et al., 1972), troodontids (Currie, 1985; Currie and Zhao, 1993b), and tyrannosaurs (Russell, 1970) (Fig. 1).


FIGURE 1 Braincase of Dromaeosaurus albertensis from right posteroventral (A) and medial (B) views. ac, fossa auriculae cerebelli; bt, basal tubera; de, ductus endolymphaticus; f, fenestra ovalis and fenestra pseudorotunda; ic, internal carotid; ls, laterosphenoid; m, metotic fissure; ms, metotic strut; oc, occipital condyle; pn, pneumatic space, s, stapes; vcma, anterior canal of middle cerebral vein. Roman numerals represent cranial nerves. (From Currie (1995) with permission.)

The floor of the braincase ossifies within the basal plate of the chondrocranium into the basioccipital posteriorly and the basisphenoid anteriorly. The basioccipital forms the floor of the metotic fissure and effectively makes up the lower margins of the exit foramina for cranial nerves IX to XI. Beneath the occipital condyle are a pair of processes called the basal tubera, formed primarily by the basioccipital but supported anteriorly by the basisphenoid. Primitively, a large pneumatic sinus (known by many names, including the basisphenoidal recess and Rathke’s Pouch) opens ventrally. Although bounded mostly by the basisphenoid, the posterior wall is formed primarily by the basioccipital. The basicranial complex is pierced on the midline by a eustachian opening in most neoceratopsians (Dodson and Currie, 1990), except in Bagaceratops in which the opening passes between the basioccipital and basisphenoid. The sixth cranial nerve passes anteroventrally through the ba-

Braincase Anatomy

84 sisphenoid behind the dorsum sellae. In coelurosaurian dinosaurs (Currie and Zhao, 1993b), including birds, cranial nerve VI exits the basisphenoid lateral to the pituitary, whereas in most dinosaurs (including allosaurids, sauropods, hadrosaurs, and pachycephalosaurids) it enters the hypophyseal fossa in primitive fashion. The pituitary is nested in the fossa (which is also called the hypophyseal recess or pituitary fossa) anterior to the dorsum sellae. The basisphenoid and parasphenoid are indistinguishably fused in all dinosaurs, with the possible exception of pachycephalosaurids (Maryanska and Osmo´lska, 1974). Anteriorly they taper into the distinctive cultriform process (parasphenoid rostrum), which supports the interorbital septum. In ornithomimids, therizinosaurids, and troodontids (Barsbold, 1983), the parasphenoid is expanded into a pneumatic, balloon-like structure referred to as the bulbous parasphenoid. Ventrally the same bones form the paired basipterygoid processes, which articulate with the pterygoids. Posterodorsal to the basipterygoid process, the lateral wall of the basisphenoid is invaded by the internal carotid. This artery passes anteromedially to meet its counterpart on the midline in the hypophyseal recess. Like other archosaurs, including crocodiles and birds, the laterosphenoid (pleurosphenoid of some authors) of dinosaurs is ossified in the pila antotica. It extends dorsolaterally in a wing-like process that contacts the frontal, parietal, and postorbital. Ventrally it forms the margin of the foramen for the various branches of the fifth cranial nerve. One of these branches—the ophthalmic—is usually expressed on the laterosphenoid either by a groove on the lateral surface or by a canal enclosed within the bone. The ophthalmic branch of the trigeminal is separate in Allosaurus (Hopson, 1979), protoceratopsids (Brown and Schlaikjer, 1940), ceratopsids (Brown, 1914), troodontids (Currie and Zhao, 1993b), and tyrannosaurids (Bakker et al., 1988), but the branches exit from a single opening in all other dinosaurs. Anteriorly, the laterosphenoid forms the posterior border of the foramen for the third cranial nerve, and the fourth passes through its upper regions. Anterior to the laterosphenoid is a series of ossifications within the interorbital cartilages that show great ontogenetic and taxonomic variability. They are either absent or unknown in most smaller and primitive dinosaurs, including ceratosaurs (Welles, 1984;

Raath, 1985), protoceratopsids (Brown and Schlaikjer, 1940), and psittacosaurids (Sereno, 1987). There is also considerable confusion concerning what each of these ossifications should be called. The orbitosphenoids, which develop late in the pila metotica of the chondrocranium, are a pair of small ossifications that form around the common (carcharodontosaurids, troodontids, and tyrannosaurids) or separate openings (allosaurids) on the midline for the second (optic) cranial nerves. The orbitosphenoid also borders the opening for the third cranial nerve and probably the fourth in at least some cases. Other than theropods, the orbitosphenoid is also known in sauropods (Madsen et al., 1995), hadrosaurs (Lull and Wright, 1942), iguanodontids (Norman, 1986), and pachycephalosaurids (Maryanska and Osmo´lska, 1974). Dorsally, the sphenethmoid can ossify to form an elongate tube or pair of tubes beneath the frontal for the olfactory tracts and bulbs. According to S. Welles (personal communication, 1996), there is a separate ossification, which he calls the septosphenoid, in some theropods beneath the parietal and behind the sphenethmoid. The interorbital septum can also ossify into a thin, vertical sheet of bone between the sphenethmoid and the cultriform process (Maryanska, 1977; Madsen et al. 1995; Coria and Currie, 1997). Despite the complexity of braincases and lack of much comparative information about them, they provide much information about individual dinosaurs. One research area where they have been particularly useful in elucidating relationships has been in the origin of birds debate. Differences in terminology are slowly being resolved, and with the advent of CT scanning, braincases have become more accessible. It is therefore highly likely that the number of publications describing braincases will steadily continue to increase in coming decades.

See also the following related entries: PALEONEUROLOGY ● SKULL, COMPARATIVE ANATOMY

References Bakker, R. T., Williams, M., and Currie, P. J. (1988). Nanotyrannus, a new genus of pygmy tyrannosaur from the latest Cretaceous of Montana. Hunteria 1(5), 1–30. Barsbold, R. (1983). Carnivorous dinosaurs from the Cre-

Braincase Anatomy


taceous of Mongolia. Joint Soviet–Mongolian Paleontol. Expedition Trans. 19, 5–120. [In Russian]

Selma Formation of Alabama. Part VI, the Dinosaurs. Fieldiana Geol. Mem. 3, 313–360.

Brown, B. (1914). Anchiceratops, a new genus of horned dinosaur from the Edmonton Cretaceous of Alberta. With discussion of the origin of the ceratopsian crest and the brain casts of Anchiceratops and Trachodon. Am. Museum Nat. History Bull. 33, 539–548.

Lull, R. S., and Wright, N. F. (1942). Hadrosaurian dinosaurs of North America. Geol. Soc. Am. Spec. Papers 40, pp. 200.

Brown, B., and Schlaikjer, E. M. (1940). The structure and relationships of Protoceratops. Proc. N. Y. Acad. Sci. 40, 133–266. Coria, R. A., and Currie, P. J. (1997). The braincase of Giganotosaurus carolinii (Dinosauria: Theropoda) from the Upper Cretaceous of Argentina. Manuscript in preparation.

Madsen, J. H., Jr., McIntosh, J. S., and Berman, D. S. (1995). Skull and atlas–axis complex of the Upper Jurassic sauropod Camarasaurus Cope (Reptilia: Saurischia). Carnegie Museum Nat. History Bull. 31, 1–115. Maryanska, T. (1977). Ankylosauridae (Dinosauria) from Mongolia. Palaeontol. Polonica 37, 85–151. Maryanska, T., and Osmo´lska, H. (1974). Pachycephalosauria, a new suborder of ornithischian dinosaurs. Palaeontol. Polonica 30, 45–102.

Currie, P. J. (1985). Cranial anatomy of Stenonychosaurus inequalis (Saurischia, Theropoda) and its bearing on the origin of birds. Can. J. Earth Sci. 22, 1643–1658.

Norman, D. B. (1986). On the anatomy of Iguanodon atherfieldensis (Ornithischia: Ornithopoda). Inst. R. Sci. Nat. Belgique Bull. 56, 281–372.

Currie, P. J. (1995). New information on the anatomy and relationships of Dromaeosaurus albertensis (Dinosauria: Theropoda). J. Vertebr. Paleontol. 15, 576–591.

Osmo´lska, H., Roniewicz, E., and Barsbold, R. (1972). A new dinosaur, Gallimimus bullatus, n. gen. n. sp. (Ornithomimidae) from the Upper Cretaceous of Mongolia. Palaeontol. Polonica 27, 103–143.

Currie, P. J., and Zhao, X. J. (1993a). A new large theropod (Dinosauria, Theropoda) from the Jurassic of Xinjiang, People’s Republic of China. Can. J. Earth Sci. 30, 2037–2081. Currie, P. J., and Zhao, X. J. (1993b). A new troodontid (Dinosauria, Theropoda) braincase from the Dinosaur Park Formation (Campanian) of Alberta. Can. J. Earth Sci. 30, 2231–2247.

Raath, M. A. (1985). The theropod Syntarsus and its bearing on the origin of birds. In The beginnings of Birds (M. K. Hecht, J. H. Ostrom, G. Viohl, and P. Wellnhofer, Eds.), pp. 219–227. Freunde des Jura-Museums Eichsta¨tt, Willibaldsburg, Eichsta¨tt. Russell, D. A. (1969). A new specimen of Stenonychosaurus from the Oldman Formation of Alberta. Can. J. Earth Sci. 6, 595–612.

Dodson, P., and Currie, P. J. (1990). Neoceratopsians. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 593–618. Univ. of California Press, Berkeley.

Russell, D. A. (1970). Tyrannosaurs from the Late Cretaceous of western Canada. National Museum of Natural Sciences, Publ. Palaeontology No. 1, pp. 34.

Galton, P. M. (1989). Crania and endocranial casts from ornithopod dinosaurs of the families Dryosauridae and Hypsilophodontidae (Reptilia: Ornithischia). Geol. Palaeontol. 23, 217–239.

Sereno, P. C. (1987). The ornithischian dinosaur Psittacosaurus from the Lower Cretaceous of Asia and the relationships of the Ceratopsia. Ph.D. thesis, Columbia University, New York.

Galton, P. M. (1990). Stegosauria. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 435–455. Univ. of California Press, Berkeley.

Taquet, P. (1976). Ge´ologie et Pale´ontologie du gisement de Gadoufaoua (Aptien du Niger), pp. 180. Cahiers de Pale´ontologie, CNRS, Paris.

Gilmore, C. W. (1909). Osteology of the Jurassic reptile Camptosaurus, with a revision of the species of the genus, and descriptions of two new species. U.S. Natl. Museum Proc. 36, 197–332. Gilmore, C. W. (1914): Osteology of the armoured Dinosauria in the United States National Museum, with special reference to the genus Stegosaurus. Bull. U.S. Natl. Museum 89, 1–136. Hatcher, J. B., Marsh, O. C., and Lull, R. S. (1907). The Ceratopsia. U.S. Geol. Survey Monogr. 49, 1–300. Hopson, J. A. (1979). Paleoneurology. In Biology of the Reptilia, Volume 9 (C. Gans, Ed.), pp. 39–146. Academic Press, London. Langston, W., Jr. (1960). The vertebrate fauna of the

Weishampel, D. B., and Horner, J. R. (1990). Hadrosauridae. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 534–561. Univ. of California Press, Berkeley. Weishampel, D. B., and Witmer, L. M. (1990a). Lesothosaurus, Pisanosaurus and Technosaurus. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 416–425. Univ. of California Press, Berkeley. Weishampel, D. B., and Witmer, L. M. (1990b). Heterodontosauridae. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 486–497. Univ. of California Press, Berkeley. Welles, S. P. (1984). Dilophosaurus wetherilli (Dinosauria, Theropoda), osteology and comparisons. Palaeontographica A 185, 85–180.



Brazilian Dinosaurs Theropods and sauropods have been found mostly in Campanian–Maastrichian sediments.


British Dinosaurs The first scientific reports of dinosaurs were based on specimens from Britain. New specimens and taxa continue to be described and analyzed.




British Museum of Natural History see NATURAL HISTORY MUSEUM, LONDON

Bullatosauria KEVIN PADIAN JOHN R. HUTCHINSON University of California Berkeley, California, USA

FIGURE 1 Phylogeny of Bullatosauria.

beneath and behind the orbit (see BRAINCASE ANATOMY), which is considered a valid synapomorphy that unites the clade. Other diagnostic characteristics of Bullatosauria include an enlarged braincase and eyes and a ventrally deflected occipital region. Braincase size in these animals is four to seven times higher than expected in a crocodile of the same size (‘‘encephalization quotient’’: see INTELLIGENCE). These characteristics suggest but do not necessarily indicate some degree of heightened cerebral function compared to most other nonavian theropods. Both ornithomimosaurs and troodontids have brain:body size ratios among the highest for nonavian Reptilia, that is, encephalization comparable to that of an ostrich or an early mammal.

See also the following related entries: ARCTOMETATARSALIA ● COELUROSAURIA

Bullatosauria (Fig. 1) refers to the clade of arctometa-


tarsalian theropods recognized originally by Kurzanov (1976), then by Currie (1985), and formalized by Holtz (1994), who later (Holtz, 1996) clarified it to represent a node-based clade including Ornithomimus and Troodon (⫽ Stenonychosaurus) and all descendants of their most recent common ancestor. The monophyly of the taxon Bullatosauria (composed of TROODONTIDAE and ORNITHOMIMOSAURIA) has been well supported by recent analyses (e.g., Holtz 1995) and seems to have its origin in the Early Cretaceous (Holtz, 1994). Tyrannosaurids appear to be their closest relatives. The term Bullatosauria (Latin bullatus, meaning inflated) refers to the bulbous parasphenoid capsule

Currie, P. C. (1985). Cranial anatomy of Stenonychosaurus inequalis (Saurischia, Theropoda) and its bearing on the origin of birds. Can. J. Earth Sci. 22, 1643–1658. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117. Holtz, T. R., Jr. (1995). A new phylogeny of the Theropoda. J. Vertebr. Paleontol. 15(Suppl. to No. 3), 35A. Holtz, T. R., Jr. (1996). Phylogenetic taxonomy of the Coelurosauria (Dinosauria: Theropoda). J. Paleontol. 70, 536–538. Kurzanov, S. M. (1976). Braincase structure in the carnosaur Itemirus n. gen. and some aspects of the cranial anatomy of dinosaurs. Paleontol. Zhurnal 1976, 127–137. [In Russian]

C Cabo Espichel

Cameros Basin Megatracksite



University of Colorado at Denver Denver, Colorado, USA

Universidad Auto´noma de Madrid Madrid, Spain

Situated southwest of Lisbon, Portugal, in the cliffs

The Cameros Basin is located in the north branch

below the famous monastery of Cabo Espichel, are a series of about 10 Late Jurassic, track-bearing layers. Most of the trackways are attributable to sauropods, but one was made by a limping theropod (Lockley et al., 1994). One layer reveals evidence of a small herd of seven juvenile sauropods heading southwest. Not only is this the richest area for sauropod trackways in all of Europe but also local folklore reveals that the tracks have been known since the 13th century.

of the Iberian Range, including part of the Spanish provinces of Burgos, Soria, and La Rioja. The basin is about 8000 km2 wide and the sediments are approximately 9 km thick. The sediments of the Cameros Basin are Upper Jurassic (Tithonian)–Lower Cretaceous (Aptian) in age. Most of the dinosaur tracks are located in the ‘‘Hue´rteles’’ Alloformation (Middle Berriasian) and the ‘‘Enciso’’ group (Berriasian– Lower Aptian) (Moratalla, 1993). The deposits of the Hue´rteles Alloformation come from a fluvial channel system discharging into a shallow saline lake. During the dryness periods playa-like environments appeared. The sediments of the Enciso group indicate fluvial channels and lacustrine environments. The scarce sauropod tracks at Cameros Basin have been identified as Brontopodus and Parabrontopodus. The hindlimb prints are up to 50–60 cm in length. Four theropod track morphotypes can be found. Among the nonavian theropod prints the ichnogenus Buckeburgichnus is one of the most abundant. This ichnogenus ranges up to 70 cm in length. The digits are broad and robust, and the length/width ratio ranges between 0.95 and 1.2. Buckeburgichnus is well represented at the ‘‘Los Cayos’’ tracksite (La Rioja) (Fig. 1) in which 425 tracks, including 36 trackways, have been identified. Some footprints from Los Cayos occasionally present a clear impression of the hallux. The second theropod track morphotype ranges between 15 and 30 cm in length. The digits are robust. The heel is occasionally elongated and there is a strong medial indentation. The length/width ratio

See also the following related entries: EUROPEAN DINOSAURS ● FOOTPRINTS AND TRACKWAYS

Reference Lockley, M. G., Santos, V. F., Meyer, C. A., and Hunt, A. P. (Eds.) (1994). Aspects of Sauropod Biology. Gaia: Geoscience Magazine of the National Natural History Museum, Lisbon, Portugal.

California Academy of Sciences see MUSEUMS



Cameroon Dinosaurs see AFRICAN DINOSAURS



FIGURE 1 Los Cayos tracksite (La Rioja province, Spain): several theropod trackways. The site has been protected, fenced, and roofed by Iberdrola Co.

ranges between 1.3 and 1.5. Similar dinosaur tracks have been reported from Moab, Utah (Lockley, 1991). The most slender nonavian theropod tracks from Cameros have a length range between 12 and 30 cm. Digits are long and thin and the length/width ratio is 1–1.2. Finally, some tracks from the Los Cayos (Moratalla and Sanz, 1992) and Serrantes tracksite (Soria) (Fuentes, 1997) have been interpreted as avian in origin. Ornithopod tracks from the Cameros Basin are represented by four morphotypes. The first one comprises broad footprints of large size, between 40 and 60 cm in length. The digits are short and robust and two indentations (lateral and medial) delimit the heel area. Most of these tracks belong to bipedal trackways, but occasional quadrupedal ones have

Cameros Basin Megatracksite been reported. This is the case for the sites known as Cabezo´n de Cameros (La Rioja) (Fig. 2) (Moratalla et al., 1992) and Regumiel de la Sierra (Burgos) (Moratalla et al., 1994). The largest ornithopod footprints have a broad heel surface and shorter and more robust digits than the first morphotype referred to previously (Fig. 3). The length/width ratio ranges from 0.9 to 1.12. The trackway gauge is wide and the stride length is short. All the available evidence seems to indicate that the trackmaker is a stout, graviportal iguanodontid with a foot skeleton close to that of hadrosaurs (Moratalla et al., 1988). At Valdete´ tracksite (La Rioja), another ornithopod track morphotype has been found. The length/width ratio is larger than 1.2, and the trackway is narrow. The trackmaker has been tentatively identified as being close to camptosaurs. Finally, the smallest ornithopod tracks, 10–15 cm in length, have been found at Valdevajes site (La Rioja). The trackmaker has been tentatively assigned to hypsilophodontids. Most of the Cameros Basin trackway ichnorecord has been produced by theropods (80%). Ornithopods are represented by 16% and, finally, sauropod tracks are scarce, representing about 4% (Moratalla, 1993). These percentages are similar to those of the Glen Rose Formation (Texas), and, to a lesser degree, to those of the Sousa Formation (Brazil) (Lockley and Conrad, 1989). The great differences between the Cameros Basin and the English Wealden are striking, taking into account the facies similarity and the synchronicity of both areas. These differences are difficult to explain, but it is possible that many Wealden tracks, supposedly ornithopod in origin, can actually be identified as theropod footprints. The Cameros Basin is characterized by an evident change in the dinosaur ichnofauna from the Middle Berriasian to the Upper Berriasian–Lower Aptian. Two main features can be distinguished: (i) The oldest ichnofauna is more clearly dominated by theropods than the youngest one—the ichnogenus Buckeburgichnus is predominantly recorded in this youngest ichnofauna; and (ii) both the record and the diversity of ornithopod and sauropod trackways clearly increase from the Mid-Berriasian to the Upper Berriasian– Lower Aptian. The Cameros Basin ichnorecord suggests a gregarious behavior within theropods, iguanodontids, and

Cameros Basin Megatracksite


FIGURE 2 Cabezo´n de Cameros tracksite (La Rioja province, Spain): quadrupedal ornithopod trackway.

small ornithopods (hypsilophodontids?). The relative speeds in all the recorded trackways have the following ratio: ␭ /h ⬍ 2 (Alexander, 1976). The values range from 0.40 to 1.84. The average ratio for theropods is 1.18, suggesting a larger biodynamic index than that of ornithopods. The lowest value is that of the iguanodontid trackway of Regumiel de la Sierra. The Cameros Basin track evidence indicates an inversely proportional relationship between the ␭h ratio and the size of the trackmaker. Thus, if the smallest theropod footprints of the Los Cayos (and other sites) actually belong to the same species of those of the largest ones, juvenile individuals had a greater biodynamic index. This evidence suggests that immature individuals would be more agile and mobile than the adults. This conclusion could also be suggested for other evidence involving structurally similar (but distinct) species of different sizes.

See also the following related entries: EUROPEAN DINOSAURS ● FOOTPRINTS AND TRACKWAYS

References Alexander, R. McN. (1976). Estimates of speeds of dinosaurs. Nature 261, 129–130.

FIGURE 3 La Magdalena tracksite (La Rioja province, Spain): graviportal ornithopod trackway.

Fuentes, C. (1997). Primeras huellas de aves en el Jura´sico de Soria (Wealdense, Formacio´n Cameros, Grupo

Canadian Dinosaurs

90 Oncala). Descripcio´n del nuevo icnoge´nero ‘‘Archaeornithipus.’’ Estudios Geolo´gicos, in press. Lockley, M. G. (1991). Tracking Dinosaurs, pp. 250. Cambridge Univ. Press, Cambridge, UK. Lockley, M. G., and Conrad, K. (1989). The paleoenvironmental context, preservation and paleoecological significance of dinosaur tracksites in the western USA. In Dinosaur Tracks and Traces (D. D. Gillette and M. G. Lockley, Eds.), pp. 121–134. Cambridge Univ. Press, Cambridge, UK. Moratalla, J. J. (1993). Restos indirectos de dinosaurios del registro espan˜ol: Paleoicnologı´a de la Cuenca de Cameros (Jura´sico superior–Creta´cico inferior) y Paleoologı´a del Creta´cico superior, pp. 729. Unpublished PhD thesis, Univerisidad Auto´noma de Madrid. Spain. Moratalla, J. J., and Sanz, J. L. (1992). Icnitas aviformes en el yacimiento del Creta´cico inferior de Los Cayos (Cornago, La Rioja, Espan˜a). Zubı´a 10, 153–160. Moratalla, J. J., Sanz, J. L., and Jime´nez, S. (1988). Nueva evidencia icnolo´gica de dinosaurios en el Creta´cico inferior de La Rioja (Espan˜a). Estudios geolo´gicos 44, 119–131. Moratalla, J. J., Sanz, J. L., Jime´nez, S., and Lockley, M. G. (1992). A quadrupedal ornithopod trackway from the Lower Cretaceous of La Rioja (Spain): Inferences on gait and hand structure. J. Vertebr. Paleontol. 12(2), 150–157. Moratalla, J. J., Sanz, J. L., and Jime´nez, S. (1994). Dinosaur tracks from the Lower Cretaceous of Regumiel de la Sierra (province of Burgos, Spain): Inferences on a new quadrupedal ornithopod trackway. Ichnos 3, 89–97.

Canada–China Dinosaur Project see SINO –CANADIAN DINOSAUR PROJECT

Canadian Dinosaurs CLIVE COY Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Canada has a well-deserved reputation as a productive field for dinosaur discoveries. The western prairie provinces and their extensive modern river valleys have produced thousands of dinosaur specimens that

are housed in institutions around the world. The first dinosaur bones found in Canada were discovered by G. M. Dawson of the Canadian Geological Survey near Morgan Creek, Saskatchewan, in 1874. In 1884, Joseph Tyrrell discovered dinosaurs along the Red Deer River valley near the present city of Drumheller, Alberta. Between 1910 and 1917 extensive, valuable collections of dinosaur skeletons were made along the Red Deer River. Since this period of intense collection on the prairies, dinosaurs have been discovered from Canada’s chilling Arctic to the pounding surf of the Bay of Fundy in Nova Scotia. Late Triassic vertebrates have been recovered from the Heilberg Formation of the Northwest Territories and from several formations on the east coast. Arctosaurus osborni from the Carnian–Rhaetian is sometimes thought to be an indeterminate theropod. The Lower Wolfville Formation (Carnian) has produced indeterminate prosauropod and ornithischian remains, whereas the Upper Wolfville Formation (Norian) has ornithischian footprints. The Blomidon Formation (Norian) has yielded theropod footprints. Early Jurassic dinosaurs have been recovered from the McCoy Brook Formation (Hettangian) of Nova Scotia and include saurischian footprints referred to the dinosaurs cf. Anchisaurus sp. and cf. Ammosaurus sp. and ornithischian footprints referred to the osteotaxon Scutellosaurus sp. Early Cretaceous dinosaurs from the Gething and equivalent formations (Barremian–Aptian) of British Columbia and Alberta include theropod, ankylosaur, and ornithopod footprints. By far the majority of dinosaurs that have been recovered from Canada lived during Late Cretaceous times. In the north, hadrosaur bones have been found in the Maastrichtian Bonnet Plume Formation of the Yukon and the Kangguk Formation of the Northwest Territories. Theropod bones have also been recovered from the latter, whereas the Summit Creek Formation (Late Maastrichtian) has produced indeterminate ceratopsid bones. Cenomanian footprints (theropods, ankylosaurs, and ornithopods) are known from the Dunvegan Formation of Alberta and British Columbia, along with some poorly preserved bones and teeth. Early Campanian dromaeosaurids, tyrannosaurids, hadrosaurs, ceratopsids, and nodosaurids are represented mostly by teeth in the Milk River Formation. Some of the richest dinosaur deposits in

Can˜on City the world are located in DINOSAUR PROVINCIAL PARK and nearby sites in southern Alberta. These middle to late Campanian beds (Judithian LAND MAMMAL AGE) have produced a huge range of dinosaurs including indeterminate theropods (Richardoestesia), dromaeosaurids (Dromaeosaurus and Saurornitholestes), caenagnathids (Chirostenotes), elmisaurids (Elmisaurus), avimimids, tyrannosaurids (Aublysodon, Gorgosaurus, and Daspletosaurus), ornithomimids (Struthiomimus, Dromicieomimus, and Ornithomimus), troodontids (Troodon), a possible therizinosaur (cf. Erlikosaurus), many hadrosaurs (Brachylophosaurus, Gryposaurus, Prosaurolophus, ?Maiasaura sp., Corythosaurus, Hypacrosaurus, Lambeosaurus, and Parasaurolophus), hypsilophodonts (Orodromeus), pachycephalosaurids (Stegoceras, Gravitholus, Pachycephalosaurus, and Ornatotholus), protoceratopsids (Leptoceratops and Montanoceratops), ceratopsids (Anchiceratops, Chasmosaurus, Centrosaurus, and Styracosaurus), and ankylosaurs (Edmontonia, Panoplosaurus, and Euoplocephalus). The late Campanian–early Maastrichtian beds of the Horseshoe Canyon Formation near Drumheller are also rich with skeletons of indeterminate theropods (Richardoestesia), dromaeosaurids (Dromaeosaurus and Saurornitholestes), caenagnathids (Chirostenotes), tyrannosaurids (Aublysodon, Albertosaurus, and Daspletosaurus), ornithomimids (Struthiomimus, Dromicieomimus, and Ornithomimus), troodontids (Troodon), hypsilophodontids (Parksosaurus), hadrosaurs (Edmontosaurus, Saurolophus, and Hypacrosaurus), pachycephalosaurids (Stegoceras), ceratopsids (Anchiceratops, Arrhinoceratops, and Pachyrhinosaurus), and ankylosaurs (Edmontonia and Euoplocephalus). Latest Maastrichtian dinosaurs of the Scollard Formation include Tyrannosaurus rex, Thescelosaurus, Edmontosaurus, Leptoceratops, Triceratops, Torosaurus, and Ankylosaurus. In Saskatchewan, Judithian beds have yielded the remains of tyrannosaurids, ornithomimids, caenagnathids (Chirostenotes), dromaeosaurids (Saurornitholestes and Dromaeosaurus), hadrosaurs, and ceratopsians. Dromaeosaurus, Saurornitholestes, Chirostenotes, Tyrannosaurus, Thescelosaurus, Edmontosaurus, Triceratops, Torosaurus, and Ankylosaurus have all come from the higher Lancian age beds. This is not a complete listing of the localities, formations, and genera of dinosaurs found in Canada but serves to show the richness of these resources. Traditional sites such as DINOSAUR PROVINCIAL PARK

91 (see DINOSAUR PARK FORMATION) and Drumheller (see HORSESHOE CANYON FORMATION; ROYAL TYRRELL MUSEUM OF PALAEONTOLOGY) continue to produce new specimens every year, while new sites such as DEVIL’S COULEE continue to be found.


Can˜on City KENNETH CARPENTER Denver Museum of Natural History Denver, Colorado, USA

The Upper Jurassic dinosaur beds of the Morrison Formation near Can˜on City, Colorado, were first excavated in 1877 by crews working for O. C. Marsh and E. D. Cope (e.g., Cope, 1877; Marsh, 1877). Some of the best known dinosaurs from the Morrison Formation, such as Camarasaurus supremus, Diplodocus longus, Allosaurus fragilis, Ceratosaurus nasicornis, and Stegosaurus stenops, were first named from specimens collected from there. Many other genera and species were also named but most of these are no longer considered valid taxa. One exception is the little enigmatic Nanosaurus agilis, the first dinosaur named from Can˜on City. The most prolific site, Marsh Quarry 1, has produced more than 12 species of dinosaurs, including the sauropods ‘‘Morosaurus’’ agilis, Diplodocus longus, Haplocanthosaurus priscus, and Brachiosaurus; the theropods Allosaurus fragilis, Ceratosaurus nasicornis, Coelurus agilis, and Elaphrosaurus sp.; the ornithopods Dryosaurus altus and Othnelia rex; and the stegosaurs Stegosaurus armatus and Stegosaurus stenops (Carpenter, 1997a). From Cope’s sites, at a stratigraphically higher level, the sauropods Camarasaurus supremus, Amphicoelias altus, and the giant Amphicoelias fragillium were named. Amphicoelias fragillium is based on a partial vertebra that, whole, would have been more than 2.4 m tall! The entire animal would have been about 45 m long, making it the biggest dinosaur yet discovered (Carpenter, 1997a).


92 Recently, another species of Haplocanthosaurus, H. delfsi, has been named from a level lower than the Marsh Quarry (Carpenter, 1997a; McIntosh and Williams, 1988). Also at this level, the oldest dinosaur eggs from North America, called Prismatoolithus coloradensis, have recently been found (Hirsch, 1994). Another recent discovery is that of a nearly complete skeleton of Stegosaurus stenops with most of the armor preserved in place (Carpenter, 1997b). This specimen showed the plates on the back in two alternating rows on both sides of the neural spines. In addition, the tail spikes were found projecting posterolaterally, numerous small, keeled disks of armor were found over the pelvis, and a wattle of small ossicles was found under the neck. Recent work has shown that the Morrison Formation at Can˜on City can be divided into a lower Haplocanthosaurus zone and an upper Camarasaurus zone (Carpenter, 1997a). The replacement of Haplocanthosaurus by Camarasaurus corresponds to an environmentally linked facies change. With dinosaur specimens distributed throughout the formation, the Can˜on City dinosaur beds offer a unique place to study the evolution of dinosaur faunas during the Upper Jurassic.

See also the following related entry: MORRISON FORMATION

References Carpenter, K. (1997a). Vertebrate biostratigraphy of the Morrison Formation near Can˜on City, Colorado. Modern Geology, in press. Carpenter, K. (1997b). Armor of Stegosaurus stenops, and the taphonomic history of a new specimen from Garden Park, Colorado. Modern Geology, in press. Cope, E. D. (1877). On reptilian remains from the Dakota beds of Colorado. Proc. Am. Philos. Soc. 17, 193–196. Hirsch, K. F. (1994). Upper Jurassic eggshells from the Western Interior of North America. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. Horner, Eds.), pp. 137–150. Cambridge Univ. Press, New York. Marsh, O. C. (1877). Notice of some new dinosaurian reptiles from the Jurassic formation. Am. J. Sci. 14, 514–516. McIntosh, J. S., and Williams, M. (1988), A new species of sauropod dinosaur, Haplocanthosaurus delfsi sp. nov., from the Upper Jurassic Morrison Formation of Colorado. Kirtlandia 43, 3–26.

Canterbury Museum, New Zealand see MUSEUMS



Carenque MARTIN LOCKLEY University of Colorado at Denver Denver, Colorado, USA

The Carenque dinosaur tracksite is situated in the suburb of Lisbon that bears that name. It reveals what, at the time of discovery in 1992, was the longest trackway in the world (127 m), which was made by a large bipedal dinosaur in Cretaceous (Cenomanian) sediments. As detailed in an entire book on the subject (Galopim, 1994), the site was saved from destruction by diverting a freeway through a tunnel beneath the tracks. During the course of this project the trackway was further excavated to a length of 141 m. (see FATIMA and KHODJA-PILATA entries for further information on the world’s longest trackways).

See also the following related entries: EUROPEAN DINOSAURS ● FOOTPRINTS AND TRACKWAYS

Reference Galopim, A. M. de C. (1994). Dinosaurios e a batalha de Carenque, pp. 291. Editorial Noticias, Lisbon, Spain.

Carnegie Museum of Natural History JOHN S. MCINTOSH Wesleyan University Middletown, Connecticut, USA

The Carnegie Museum of Natural History in Pittsburgh, Pennsylvania, was one of the first institutions in the United States to amass a major collection of

Carnegie Museum of Natural History dinosaur fossils. Its collection of SAUROPOD dinosaurs remains today the greatest in the United States. On the basis of a newspaper article reporting the discovery of a large dinosaur in Wyoming, Andrew Carnegie, steel magnate and philanthropist, charged Dr. William J. Holland, director of his new museum, to purchase the animal. Holland hired experienced paleontologists such as Jacob Wortman and later John Bell Hatcher, collectors including the legendary William H. Reed, and preparators such as Arthur Coggeshall to carry out major expeditions to the West, beginning in 1898 and continuing unabated until 1923. The first of these succeeded in collecting the greater portion of a skeleton of the JURASSIC sauropod Diplodocus carnegii from the MORRISON FORMATION of Sheep Creek, Wyoming. A second skeleton of that animal was collected from the same quarry in the following year, and together they formed the basis of a composite mounted skeleton, only the second such of a sauropod dinosaur. From 1899 until his untimely death in 1904, Hatcher broadened the scope of dinosaur exploration, dispatching parties to a number of quarries in Wyoming, on Sheep Creek and also to the Freezeout Hills and the Red Fork of the Powder River. He also reopened the famous quarry in Garden Park, Colorado, that had provided Professor O. C. Marsh of Yale with most of the Jurassic dinosaur skulls collected before 1900, including those of Diplodocus, Allosaurus, Ceratosaurus, and Stegosaurus. Nearly complete skeletons of the latter three of these ended up in the National Museum of Natural History in Washington, DC, where they are now mounted and on exhibition. This quarry also yielded Hatcher two partial skeletons of the new and very rare sauropod Haplocanthosaurus. From a second quarry on Sheep Creek, a party led by C. W. Gilmore collected a fine skeleton of Apatosaurus (Brontosaurus ), which Holland planned to mount next to the Diplodocus. A partial skeleton of a very young individual of the same animal (originally called Elosaurus) was found with it. These plans were disrupted by the discovery in August 1909 of the greatest of all Jurassic dinosaur quarries by Earl Douglass in what is now the DINOSAUR NATIONAL MONUMENT, north of Jensen, Utah. The first specimen taken from this quarry proved to

93 be the most complete skeleton of Apatosaurus ever found. Named A. louisae by Holland for Andrew Carnegie’s wife, this skeleton was mounted next to the Diplodocus in 1913, and the Wyoming Apatosaurus was eventually sent to the University of Wyoming in Laramie, where it is now on exhibition. The Carnegie Quarry at Dinosaur National Monument continued to yield skeleton after skeleton, many of which, including those of Allosaurus, Camptosaurus, Dryosaurus, and Stegosaurus, were eventually mounted in the Carnegie Museum. Perhaps the finest specimen was a nearly complete and articulated skeleton of a juvenile Camarasaurus, which formed the basis of a panel mount just as it lay in the quarry. For the first time many sauropod skulls were recovered, including three of Diplodocus, several of Camarasaurus, and, until very recently, the only known skull of Apatosaurus. The Carnegie collection from the quarry includes a large number of skeletal elements of all three of these animals as well as the neck and part of the thorax of the huge, long-necked, and very rare sauropod Barosaurus. Also among the non-sauropods were a partial skeleton of a very young Dryosaurus and much Stegosaurus material. Although not nearly as numerous, important dinosaur specimens from the CRETACEOUS period are also present in the Carnegie collections. Among these are mounted skeletons of Tyrannosaurus rex (the original or type specimen obtained from the American Museum of Natural History), the duck-billed Corythosaurus, and the primitive Mongolian Protoceratops. Also on exhibit is a fine skull of Triceratops. The collections further contain a skeleton of Edmontosaurus, casts of many bones of the primitive European ankylosaur Struthiosaurus, as well as many scattered specimens of tyrannosaurids, hadrosaurids, and ceratopsids. The only TRIASSIC dinosaur material in the collection consists of specimens of Coelophysis from the famous GHOST RANCH QUARRY in New Mexico and the cast of a complete hindfoot of the prosauropod Plateosaurus from Trossingen, Germany.

See also the following related entries: DINOSAUR NATIONAL MONUMENT ● MUSEUMS AND DISPLAYS



Carnosauria JOHN R. HUTCHINSON KEVIN PADIAN University of California Berkeley, California, USA

Carnosauria (Figs. 1 and 2) was a name coined by F. von Huene (1914, 1920, 1926) to include a variety of large theropod dinosaurs with great skulls and enormous teeth. It was distinguished from CoELUROSAURIA, which were generally the smaller, lightly built theropods. Since that time both terms have been used in variously formal and informal senses, but it has often been thought that the large forms comprised a more or less natural group, the lineage beginning with Early Jurassic forms such as Dilophosaurus and extending through the Late Jurassic Ceratosaurus, Megalosaurus, and Allosaurus to the Late Cretaceous TYRANNOSAURS (though von Huene did not regard tyrannosaurs or Ceratosaurus as members of this group). Gauthier (1986, p. 26) redefined Coelurosauria in cladistic terms as a stem-based taxon comprising birds and all theropods closer to birds than to Carnosauria. Hence, Coelurosauria and Carnosauria were

FIGURE 1 Phylogeny of Carnosauria, drawn mostly from Holtz (1994, 1995, 1996). Sereno et al. (1996), in contrast, find a monophyletic group composed of Giganotosaurus, Acrocanthosaurus, and Carcharodontosaurus, with the other taxa in this diagram collapsed into a polytomy together comprising Allosauridae.

sister taxa within what Gauthier (1986) called TETAa stem-based taxon. Holtz (1994) formalized AVETHEROPODA as the node-based taxon within Tetanurae comprising Coelurosauria and Carnosauria. Gauthier, however, did not formally define Carnosauria apart from listing included taxa: the genera Allosaurus, Acrocanthosaurus, Indosaurus, Alectrosaurus, Dryptosaurus, Albertosaurus, Alioramus, Daspletosaurus, Indosuchus, Tarbosaurus, and Tyrannosaurus. By implication, Carnosauria would be defined as these taxa and all others closer to them than to birds within Tetanurae. Gauthier had formally defined Carnosauria with some misgivings, acknowledging that many characters that appeared to distinguish the group were probably size-related convergences, as they were with large CERATOSAURS and ‘‘MEGALOSAURS.’’ He also noted that tyrannosaurids were further derived within the group, approaching in some respects the features of elmisaurids, ornithomimids, and Hulsanpes. These misgivings turned out to be prophetic. Novas (1992) realized that tyrannosaurs did not belong in Carnosauria, and an analysis by Holtz (1994) indicated that they were the sister group to BULLATOSAURIA (Troodontidae ⫹ Ornithomimidae) within Coelurosauria. Hence tyrannosaurs, despite their great size, are not carnosaurs but rather coelurosaurs, as von Huene (1914, 1920, 1926) first recognized. In short, the implication of this taxonomic shift is that large body size evolved many times within Theropoda rather than only once within the traditional ‘‘Carnosauria.’’ The reassignment of tyrannosaurids to the Coelurosauria removes Albertosaurus, Alectrosaurus, Alioramus, Daspletosaurus, Tarbosaurus, and Tyrannosaurus from Carnosauria. As Molnar (1990) summarized and Bonaparte (1991) elaborated on, some large forms often considered carnosaurs in a more or less formal sense, such as Carnotaurus, Indosaurus, Indosuchus, Majungasaurus, and Xenotarsosaurus, are related to Abelisaurus as members of the ABELISAURIDAE (Bonaparte and Novas, 1985), neoceratosaurian theropods, and many others are represented by fragmentary remains that defy precise taxonomic assignment. This leaves Allosaurus, Acrocanthosaurus, and Dryptosaurus. Dryptosaurus may actually be a basal coelurosaur rather than a carnosaur (Denton, 1990; Holtz, 1995), so the taxonomic composition of Gauthier’s cladistically defined Carnosauria reduces to two genera, Allosaurus NURAE,



FIGURE 2 Carnosaur skulls. Allosaurus in lateral (a) and palatal (c) views; Sinraptor in lateral (b) and dorsal (d) views; (e) Monolophosaurus; (f) Cryolophosaurus.

and Acrocanthosaurus, that are already regarded as closely related. Holtz (1994) included both in the Allosauridae, distinguished by the form of the pubic foot (longer anteriorly than posteriorly, triangular in ventral view). This rather weak unity is by default of incomplete or incompletely described material in other taxa, including Acrocanthosaurus, Chilantaisaurus, Piatnitzkysaurus, and Szechuanosaurus (Molnar et al., 1990). Holtz (1995) regarded the last two genera as possible members of this branch but outside Allosauridae. Accordingly, the definition of Carnosauria can be formally amended to include Allosaurus and all Avetheropoda closer to Allosaurus than to birds. Re-

cent finds, such as Giganotosaurus in Argentina (Coria and Salgado, 1995) and new Carcharodontosaurus material from Morocco (Sereno et al., 1996), have added to the known diversity of the Carnosauria. They have also provided evidence for new biogeographical hypotheses (Currie, 1996) and revealed the huge sizes that some carnosaurs attained (the latter two taxa may have reached larger body sizes than any known tyrannosaurs). Currie (1996) and Rauhut (1995) pointed out that, during the Late Cretaceous Period, different continental regions were dominated by different clades of large theropods (abelisaurids in South America, tyrannosaurids in North America and Asia, and carcharodontosaurs in Africa), although there

96 were obviously numerous regions of overlap where these predators must have come into contact. Current versions of the Carnosauria eloquently demonstrate the changing views of theropod phylogeny. Holtz (1995) included many recently described taxa in his analysis and found that many of them were grouped as a monophyletic, globally distributed Carnosauria, including taxa from Antarctica (Cryolophosaurus), China (Monolophosaurus and the Sinraptoridae: Sinraptor and Yangchuanosaurus), Africa (the Carcharodontosauridae: Carcharodontosaurus and Bahariasaurus), South America (Giganotosaurus, also apparently a carcharodontosaurid), and North America [the Allosauridae: Allosaurus, Acrocanthosaurus (possibly a carcharodontosaurid; Sereno et al., 1996), and related taxa]. The stem-based ALLOSAUROIDEA includes the latter three node-defined clades (Sinraptoridae, Carcharodontosauridae, and Allosauridae). The tremendous sizes reached by some carnosaurs (estimated at up to 8 tons; Coria and Salgado, 1995) raise many interesting biological questions (Molnar and Farlow, 1990). How did such gigantic bipeds manage to support their own weight? Scaling constraints must have limited their locomotory capabilities somewhat, but exactly how much is a difficult question to answer, as Molnar and Farlow (1990) have noted. It is, however, unlikely that giant carnosaurs were multiton speedsters, as some authors have depicted them (e.g., Paul, 1988); known trackways suggest that their gaits were moderate, though speeds more rapid than those recorded or calculated on the basis of available evidence are certainly possible (e.g., Alexander, 1985). At least one study concludes that if a large theropod fell at even a moderate speed, it would have risked breaking its leg, with the resultant disability soon proving fatal (Farlow et al., 1995). Biomechanical problems are not the only issues; ecological (especially trophic) considerations have been the focus of some amount of controversy. Could gigantic bipeds such as Carcharodontosaurus have been efficient predators, or were their primary ecological roles more as scavengers? The behavior and ecological interactions of extinct organisms cannot, of course, be monitored directly. Studies of large extant predators suggest that most carnivorous terrestrial vertebrates are opportunistic feeders, taking live or dead prey as they need it to sustain their metabolic de-

Carnosauria mands. This does not preclude the existence of giant scavenging theropods, but it does cast some doubt on the ecological restriction of such animals to a wholly scavenging lifestyle. All known carnosaurs are from localities also inhabited by large herbivores that would have been adequate fodder for them, and in many cases these faunas curiously lack other potential predators. It may be more fruitful to attempt to ask what kind and how much of an impact their predation had on large herbivore populations rather than whether they hunted them or not. Most predators do not engage the most robust individuals of prey populations but rather feed on the very young, old, diseased, isolated, or injured. Nor do they usually risk their own safety more than necessary, which is why prolonged combative engagements are generally avoided in favor of quick strikes that produce locomotor disability, bleeding, shock, and lacerations that encourage disease (note that the TOOTH SERRATIONS of theropods have pockets that might have harbored infective bacteria, as in the living Komodo monitor lizard). Injuries both to the skeletons of allosaurids and to those of their likely prey suggested to earlier investigators such as Gilmore and Lambe that carnosaurs focused their aggression both on the animals they hunted and on each other (see Molnar and Farlow, 1990). In carnosaurs and other large theropods, the forelimbs are short compared to the hindlimbs— perhaps as low as one-third the hindlimb length in allosaurids and independently as low as one-fifth in tyrannosaurids. Part of the hindlimb disparity is a carryover from the condition in basal tetrapods, as well as the ancestral bipedal habit of dinosaurs and other ornithodirans, which put less emphasis on the forelimb (see BIPEDALITY). It has often been assumed that reduction of the forelimb was a characteristic of theropods in general, which has been claimed as evidence against the possibility of birds, with their long forelimbs, evolving from theropods (see BIRD ORIGINS). However, this trend did not hold for coelurosaurs, particularly the maniraptorans, which include birds. Forelimb reduction has also been considered an inherent developmental consequence of large size, in which heterochronic processes favored the robusticity of the jaws and hindlimbs at the expense of the forelimbs, which had a reduced predatory function (see HETEROCHRONY). Developmental processes alone, however, cannot have driven this trend for large ther-

Carnosauria opods in general, as witnessed by the 2.5-m-long arms of the giant ornithomimid Deinocheirus. The negative allometry that describes the small forelimbs of large theropods may or may not have been the work of selective forces. It is possible that the immediate forerunners of these animals used their forelimbs less in predation than their great jaws and tearing teeth. With increased body size, the predatory function of the forelimb would have been lessened further or eliminated, hence reducing its selective value as well as its unnecessary mass. Phylogenetic analysis of the basal members of the clades of large theropods could test this prediction, but currently these animals are not sufficiently known.


97 Holtz, T. R., Jr. (1995). A new phylogeny of the Theropoda. J. Vertebr. Paleontol. 15(Suppl. to No. 3), 35A. Holtz, T. R., Jr. (1996). Phylogenetic taxonomy of the Coelurosauria (Dinosauria: Theropoda). J. Paleontol. 70, 536–538. Molnar, R. E. (1990). Problematic Theropoda: ‘‘Carnosaurs.’’ In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 280–305. Univ. of California Press, Berkeley. Molnar, R. E., and Farlow, J. O. (1990). Carnosaurian paleobiology. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 210–224. Univ. of California Press, Berkeley. Molnar, R. E., Kurzanov, S. M., and Dong, Z. (1990). Carnosauria. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 169–209. Univ. of California Press, Berkeley. Novas, F. E. (1992). La evolucio´n de los dinosaurios carnivo´ros. In Los Dinosaurios y su Entorno Biotico. Actas II, Curso de Paleontolo´gia en Cuenca ( J. L. Sanz and A. Buscalione, Eds.), pp. 123–163. Instituto ‘‘Juan de Valdes,’’ Cuenca, Spain. Paul, G. S. (1988). Predatory Dinosaurs of the World. Simon & Schuster, New York.

References Alexander, R. McN. (1985). Mechanics of posture and gait of some large dinosaurs. Zool. J. Linnean Soc. London 83, 1–25. Bonaparte, J. F. (1991). The Gondwanan theropod families Abelisauridae and Noasauridae. Historical Biol. 5, 1–25. Bonaparte, J. F., and Novas, F. E. (1985). Abelisaurus comahuensis, n.g., n. sp., Carnosauria del Cretacico tardio de Patagonia. Ameghiniana 21, 259–265. Coria, R. A., and Salgado, L. (1995). A new giant carnivorous dinosaur from the Cretaceous of Patagonia. Nature 377, 224–226.

Rauhut, O. W. M. (1995). Zur systematischen Stellung der afrikanischen Theropoden Carcharodontosaurus Stromer 1931 und Bahariasaurus Stromer 1934. Berliner Geowissenschaftlichen Abhandlungen E16, 357–375. Sereno, P. C., Dutheil, D. B., Iarochene, M., Larsson, H. C. E., Lyon, G. H., Magwene, P. M., Sidor, C. A., Varrichio, D. J., and Wilson, J. A. (1996). Predatory dinosaurs from the Sahara and Late Cretaceous faunal differentiation. Science 272, 986–991. von Huene, F. (1914). Das naturliche System der Saurischia. Zentralblatt Mineral. Geol. Pala¨ontol. B 1914, 154–158.

Currie, P. J. (1996). Out of Africa: Meat-eating dinosaurs that challenge Tyrannosaurus rex. Science 272, 971–972.

von Huene, F. (1920). Bemerkungen zur Systematik und Stammesgeschichte einiger Reptilien. Zeitschrift Indukt. Abstammungslehre Vererbungslehre 24, 162–166.

Denton, R. K., Jr. (1990). A revision of the theropod Dryptosaurus (Laelaps) aquilunguis (Cope 1869). J. Vertebr. Paleontol. 9(Suppl. to No. 3), 20A.

von Huene, F. (1926). The carnivorous Saurischia in the Jura and Cretaceous formations, principally in Europe. Revista Museo de La Plata 29, 35–167.

Farlow, J. O., Smith, M. B., and Robinson, J. M. (1995). Body mass, bone ‘‘strength indicator,’’ and cursorial potential of Tyrannosaurus rex. J. Vertebr. Paleontol. 15, 713–725. Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117.

Carter County Museum, Montana, USA see MUSEUMS




Cedar Mountain Formation JAMES I. KIRKLAND Dinamation International Society Fruita, Colorado, USA

The Lower Cretaceous of the Colorado Plateau is represented by the Cedar Mountain Formation in Utah and the largely correlative Burro Canyon Formation east of the Colorado River. When first named, the Cedar Mountain Formation was characterized as differing from the underlying Upper Jurassic MORRISON FORMATION in having an abundance of dinosaur gizzard stones (see GASTROLITHS) and in lacking preserved dinosaur bone. Its age was based on palynology and it was assumed to be largely correlative to the Aptian to middle Albian Cloverly Formation of Wyoming and Montana. In recent years, research has shown that there are considerable dinosaur remains in the Cedar Mountain Formation and that instead of preserving only one dinosaur fauna correlative with the Cloverly Formation, it preserves three successive distinct dinosaur faunas (Table I). The basal fauna ranges upward from a regionally persistent calcrete, which marks the contact between the underlying Morrison Formation and the Jurassic– Cretaceous unconformity, upward to a regionally persistent ledge forming sandstone. This fauna is best developed in the region around Arches National Park. The fauna is characterized by a dominance of nodosaurid ankylosaurs such as ‘‘Gastonia burgei’’; two species of iguanodont, Iguanodon ottingeri and a large undescribed sail-backed species; two undescribed species of sauropod, one with large spatulate teeth (cf. Ornithopsis); a small undescribed theropod; and the large dromaeosaur Utahraptor as the dominant meat-eating dinosaur. This fauna has close ties with the Upper Wealden fauna of the Isle of Wight in southern England and with a poorly known fauna from the Lakota Formation of the Black Hills region of South Dakota, which suggests a Barremian Age for the fauna. Apparently this is the oldest faunal level in the Cretaceous of North America and indicates that at least 25 million years of the earliest Cretaceous terrestrial record is missing in North America. This fauna also helps to establish the final time when terrestrial dispersal routes between Europe and Utah were open during the Cretaceous. The fauna predates

Cedar Mountain Formation TABLE I Dinosaurs Unearthed from the Cedar Mountain Formation of Utah Cedar Mountain Formation (Bodily 1969; Jensen 1970; Galton and Jensen 1979b; Nelson and Crooks 1987; Madsen pers. comm.; Stadtman pers. comm.; Nelson pers. comm.; Britt pers. comm.) Theropoda Theropoda indet. Troodontid indet. Dromaeosauridae cf. Deinonychus sp. Dromaeosaurid indet. Sauropoda indet. Ornithischia indet. Ornithopoda Iguanodontia Tenontosaurus tilletti ?Iguanodontidae Undescribed ?iguanodontid (?2 species; Britt pers. comm.) ?Iguanodontid indet. (⫽ Iguanodon ottingeri) ?Hadrosaurid indet. Ankylosauria Nodosauridae Sauropelta edwardsi (⫽ Hoplitosaurus sp.) Ankylosaurid indet. Dinosaur eggshell Age: Albian (Tschudy et al. 1984) ?Dakota Formation (Marsh 1899; Carpenter pers. comm.; not figured) Sauropoda Diplodocidae ?Barosaurus lentus Ornithopoda Hypsilophodontid indet. Iguanodontid indet. Age: late Aptian-early Cenomanian (Tschudy et al. 1984) NOTE. From Weishampel (1990).

the appearance of common flowering plants. Carbonate soil nodules indicate that the climate was dry and seasonal, but a greater abundance of fossil fishes, turtles, and crocodilians suggest that it was somewhat wetter than during the Late Jurassic. The cliff-forming sandstone separating this fauna from the overlying middle fauna may represent as much as 10 million years. Within this sandstone interval a very large undescribed nodosaurid ankylosaur was excavated by Jim Jensen during the mid-1960s. Although the middle Cedar Mountain fauna is the most widespread, being recognized in both the

Cedar Mountain Formation Arches region and in the area of the San Rafael Swell to the west, it also is the most poorly known. It occurs from the top of the cliff-forming sandstone up to a thin discontinuous sandstone horizon that marks a sharp break between sediments containing abundant calcareous soil nodules below with sediments completely lacking carbonate soil nodules above. To date, the meager dinosaur remains excavated indicate the presence of the generalized ornithopod Tenontosaurus, the basal nodosaurine ankylosaur Sauropelta, the sauropod Pleurocoelus, the dromaeosaur Deinonychus, and the giant sail-backed theropod Acrocanthosaurus. This fauna was apparently unique to North America. Flowering plants were beginning to come into their own, signaling one of the greatest floristic changes in the earth’s history. Although the climate had not changed appreciably, sea levels were rising globally and tectonic activity increased worldwide. The upper fauna ranges to the top of the Cedar Mountain Formation and is associated with lignitic intervals that preserve a diversity of flowering plants and highly smectitic sediments, indicating increased volcanic activity. A diversity of freshwater actinopterygians, elasmobranchs, turtles, and crocodilians indicate considerably wetter climatic conditions. This upper interval is best developed in the San Rafael Swell region toward the Sevier Thrustbelt but may extend eastward as far as the east side of Arches. Radiometric dates of 98.5 million years old indicate that this fauna lived during the earliest Cenomanian on the western margin of the Mowry Sea. It is domi-

99 nated by an undescribed species of basal hadrosaur, ‘‘Eohadrosaurus caroljonesi.’’ The fauna also includes a small iguanodontid, hypsilophodontids, nodosaurs, cf. Alectrosaurus, and a diversity of small theropods including cf. Troodon, cf. Paranycodon, cf. Richardoestei, and dromaeosaurs. This fauna appears to have Asian affinities, where the probable ancestors of many of these dinosaurs resided. With the draining of the Dawson Strait across northwestern Canada at the end of the Lower Cretaceous, migration of dinosaurs from the contiguous Siberian–Alaskan land area could have led to the replacement of much of the preexisting Cloverly fauna. What is truly remarkable is that there is less change in the western North American dinosaur fauna over the following 35 million years of the Cretaceous than there had been between each of these three faunas in the Cedar Mountain Formation and that of the underlying Upper Jurassic Morrison Formation. This apparently reflects the changes in North America’s ecosystems with the rapid diversification and spread of flowering plants and the transition from faunas with European affinities to those with Asian affinities.


Central Asiatic Expeditions MARK A. NORELL American Museum of Natural History New York, New York, USA


he use of powdered fossil bones as a key ingredient in traditional Asian medicine implies that the ‘‘dragon bones’’ of central Asia have been known for centuries. They have been known to Western science for only about 75 years. Discovery of non-avian dinosaurs in central Asia was pioneered by the American Museum of Natural History (AMNH) during a series of swashbuckling Central Asiatic Expeditions (CAE) in the 1920’s (Fig. 1) (Andrews, 1932). Conceived by the AMNH’s Roy Chapman Andrews, the work was pitched by Director H. F. Osborn as a search for human ancestors, based on AMNH President Henry Fairfield Osborn’s conviction that Asia, not Africa, was the cradle of humanity. On this account the expedition failed, but its successes highlighted the region as one of the most important dinosaur hunting grounds on the planet. The CAE spent field seasons in 1922, 1923, and 1925 in Mongolia and several more in China. More expeditions were planned,

but the volatile political climate of the times impeded further exploration. Incidents concerning the accidental discovery of the Flaming Cliffs at the end of the 1922 field season have been recounted several times (Andrews, 1932). Lost, the CAE caravan stopped near a couple of gers (the Mongolian noun for the Russian yurt—the familiar dwelling of central Asian nomads) to ask directions. While Andrews went to the ger, J. B. Schackelford (the expedition photographer) went in the other direction to check out some promising red rocks. As he closed in on the outcrop, he found himself standing on the edge of a large cliff, bordering an extensive area of badlands. Within minutes, Schackelford began to find fossils and alerted the rest of the party to his discovery. By the end of the afternoon a wealth of fossil material had been collected, including a Protoceratops skull and a fossil egg. Later this exposure was named the Flaming Cliffs, after the intense red

FIGURE 1 Important dinosaur localities discovered by the CAE and MAE.


Central Asiatic Expeditions color that they ‘‘take on’’ during sunset at Shabarakh Usu (shabarak ⫽ muddy; usu ⫽ water: referring to the spring at the base of the cliffs). In recent times the locality has become known by its traditional Mongolian name Bayn Dzak (bayn ⫽ many; dzak ⫽ a small tree). The size of the badlands and the large amount of bone made it a priority for continued work. However, because the season was running short, the expedition needed to push on immediately toward Kalgan and Beijing. The following year they hot-footed it toward the locality and spent nearly the entire summer collecting fossils at this locality, returning again in 1925 (Fig. 2). Their total haul included the first definitive dinosaur eggs and nests ever discovered, complete skeletons of Protoceratops (Granger and Gregory, 1923), Pinacosaurus (Gilmore, 1933a), and remains of the three major Djadokhta theropods, Oviraptor, Velociraptor, and Saurornithoides (Osborn, 1923). The expedition also collected the first skulls of Cretaceous mammals (Gregory and Simpson, 1926) and lizards (Gilmore, 1943), discovered the nearby Paleocene locality of

101 Gashato (khashat ⫽ corral) (Matthew and Granger, 1925), and collected archeological materials (including beads made from dinosaur egg-shell) from stabilized sand dunes at the cliff base (Berkey and Nelson, 1926). In addition to the finds at the Flaming Cliffs, CAE paleontologists discovered several other sites that provided important specimens. Ondai Sair and Oshih are located in what is known as ‘‘Valley of the Lakes’’ in central Mongolia. These Early Cretaceous sites produced the type specimens of Psittacosaurus and its synonym Protiguanodon (Osborn, 1924). Although fossils are sparse at these localities, the CAE succeeded in collecting juvenile remains of Psittacosaurus (Andrews, 1932), dinosaur eggs, and fragmentary sauropod and theropod remains (Osborn, 1924). In Inner Mongolia, under Chinese control, the CAE also made important dinosaur finds. In 1922 the first dinosaur fossils collected by the CAE were found at the Early Cretaceous locality of Iren Dabasu. At this locality remains of the primitive ornithomimid Archeornithomimus, the tyrannosaurid Alectrosaurus, and

FIGURE 2 Sketch map of the Flaming Cliffs (Bayn Dzak) made after the 1925 expedition. X, fossil occurrence, O, egg or nest occurrence.

102 the hadrosaurs Gilmoreosaurus and Bactrosaurus were collected (Gilmore, 1933b). Sauropod material was collected in 1928 at other Inner Mongolian localities (Gilmore, 1933a). Because of political realignment, American Museum dinosaur collecting expeditions retreated from work in Asia at the end of the 1920s. However, this was only the end of the first chapter in Mongolian dinosaur collecting. In subsequent years important Mongolian, Russian (see Lavas, 1993), and Polish expeditions (Kielan-Jaworowska, 1969; see also Lavas, 1993) picked up where AMNH paleontologists left off and made important discoveries. When the Cold War ended, AMNH expeditions were invited to resume field-work in 1990. These expeditions are organized in collaboration with the Mongolian Academy of Sciences and have been informally called the Mongolian American Museum Expedition (MAE). They differ substantially from the CAE in that they are truly collaborative and include a number of Mongolian scientists including Demberilyin Dashzeveg, Altangarel Perle, and Rinchen Barsbold. Another significant difference is that specimens collected during the MAE remain the property of the Mongolian Academy of Sciences and will be returned to Mongolia. Like the CAE, the MAE concerns itself with all aspects of the fossil fauna and is not restricted to dinosaurs. The MAE expeditions are still in progress; however, it is not too early to propose that results from the Mesozoic part of the work have equaled or exceeded those of the CAE (Novacek, 1996). In 1990, AMNH paleontologists Michael Novacek, Malcolm McKenna, and Mark Norell traveled to Mongolia to negotiate plans for this new set of expeditions. During this visit they signed an agreement initiating the MAE. A preliminary field trip was taken in 1990, and full-scale expeditions commenced in 1991 and have continued through 1995. At this writing additional field excursions and visits to New York by our Mongolian colleagues have been approved through several years. Most of the work has concentrated on localities near the Flaming Cliffs and in the Nemegt Basin. During the early years, classic localities discovered by American, Russian, Polish, and Mongolian scientists were visited. Several excellent specimens were recovered, especially from Khulsan (Norell et al., 1992a) and Tugrugeen Shireh (Novacek and McKenna, 1993; Norell et al., 1992b).

Central Asiatic Expeditions A major achievement of these reconnoiterings was the description of the primitive bird Mononykus (Perle et al., 1993). The original description was based on two specimens, one collected during the Mongolian– Russian expedition of 1987 at Bugin Tsav and a second by the MAE at Tugrugeen Shireh. Since then, Mononykus specimens have been collected by MAE expeditions at several localities (Chiappe et al., 1997). A Mononykus specimen collected at Bayn Dzak during the 1923 expedition, labeled only as ‘‘bird-like dinosaur,’’ was identified in the AMNH collection (Norell et al., 1993) (see AVES). Thus far, the most significant result of the MAE is the discovery of Ukhaa Tolgod in July 1993 (Novacek et al., 1994 Dashzeveg et al. 1995); (Fig. 3). Like most of the other Djadokhta (Bayn Dzak and Bayan Mandahu) and Djadokhta-like (red beds of Kheerman Tsav and Khulsan variously considered to be referable to the Barun Goyot Formation) localities, Ukhaa Tolgod is composed of predominantly red sandstones. The major unit at these localities is a massive cross-bedded eolian sandstone, sandwiched between two fluvial units (Eberth, 1993; Fastovsky et al., 1994; Dashzeveg et al., 1995). At Ukhaa Tolgod, most of the fossil specimens occur in the thick eolian unit. The site is unparalleled in the abundance of Cretaceous vertebrate fossil remains in all of Asia and perhaps the world (Novacek et al., 1994; Dashzeveg et al., 1995). The most significant dinosaur discovery yet from Ukhaa Tolgod is a dinosaur nest (Norell et al., 1994; Clark, 1995). Among the fragmented eggs, one contained a nearly complete embryo of a near-hatchling oviraptorid (Fig. 4). These are the first definitive remains of a theropod embryo. Curiously, in the same nest lay skulls of two neonate dromaeosaurids. The oviraptorid eggs are identical to eggs collected by the CAE in 1923 at the Flaming Cliffs. The CAE eggs were referred to Protoceratops based on the abundance of this taxon at the site, although no direct association between these eggs and this ornithischian taxon was ever demonstrated. Since then, the protoceratopsian affinities of these eggs have been challenged (Sabath, 1991); however, the myth that these were the eggs of the small ceratopsian dinosaur continued to be propagated by museum displays, popular books, and the scientific literature. The egg and embryo discovered by the MAE provide definitive evidence of this misidentification.

Central Asiatic Expeditions


FIGURE 3 Map of the Ukhaa Tolgod basin with sublocalities indicated. Numbers correspond to section in Dashzeveg et al. (1995). The oviraptorid embryo (Norell et al., 1993) was found at sublocality 7 (Xanadu).

FIGURE 4 An embryo of an oviraptorid theropod collected from Ukhaa Tolgod. Scale bar ⫽ 1 cm.

104 In conclusion, the American Museum of Natural History has had a major impact on our knowledge of Asian dinosaurs. These expeditions discovered many of the classic dinosaur sites and paved the way for Mongolian, Russian (see Lavas, 1993), Polish (Kielan-Jaworowska, 1969; see also Lavas, 1993), Chinese, Canadian (Currie, 1994), Japanese (Watabe, 1994), and later generations of AMNH paleontologists (Novacek et al., 1994; Norell et al., 1995). Some of these, like Bayn Dzak and Ondai Sair, are still producing important dinosaur remains. AMNH expeditions through the early 1990’s have been extremely successful in collecting excellent specimens from many of the classic Mongolian localities and discovering important new localities such as Ukhaa Tolgod. This work is still in its infancy, but important new information on dinosaurs is still being produced from old and new AMNH collecting efforts.


References Andrews, R. C. (1932). The New Conquest of Central Asia. A Narrative of the Central Asiatic Expeditions in Mongolia and China, 1921–1930, pp. 678. American Museum of Natural History, New York. Berkey, C. P., and Nelson, N. C. (1926). Geology and prehistoric archeology of the Gobi Desert. Amer. Mus. Novitates 222, 1–16. Chiappe, L., Norell, M. A., and Clark, J. M. (1997). Phylogenetic position of Mononykus olecranus from the Upper Cretaceous of the Gobi Desert. Submitted for publication. Clark, J. M. (1995). The egg thief exonerated. Nat. History June, 56–57. Currie, P. J. (1994). Hunting ancient dragons in China and Canada. In Dino Fest (R. S. Spencer, Ser. Ed.; G. D. Rosenberg, and D. L. Wolberg, Eds.), Spec. Publ. No. 7, pp. 387–396. Paleontolgical Society. Dashzeveg, D., Novacek, M. J., Norell, M. A., Clark, J. M., Chiappe, L. M., Davidson, A., McKenna, M. C., Dingus, L., Swisher, C., and Perle, A. (1995). Unusual preservation in a new vertebrate assemblage from the Late Cretaceous of Mongolia. Nature 374, 446–449.

Central Asiatic Expeditions Eberth, D. A. (1993). Depositional environments and facies transitions of dinosaur-bearing Upper Cretaceous redbeds of Bayan Mandahu (Inner Mongolia, People’s Republic of China). Can. J. Earth Sci. 30(10/11), 2196– 2213. Fastovsky, D. E., Badamgarav, D., Ishimoto, H., and Watabe, M. (1994). Paleoenvironments of Tugrikin-Shire (Late Cretaceous: Mongolia) and Protoceratops (Dinosauria: Ornithischia). J. Vertebr. Paleontol. 14(3), Suppl. 24A. Gilmore, C. W. (1933a). Two new dinosaurian reptiles from Mongolia with notes on some fragmentary specimens. Amer. Mus. Novitates 679, 1–20. Gilmore, C. W. (1933b). On the dinosaurian fauna of the Iren Dabasu Formation. Bull. Am. Museum Nat. History 67, 23–78. Gilmore, C. W. (1943). Fossil lizards of Mongolia. Bull. Am. Museum Nat. History 81, 361–384. Granger, W., and Gregory, W. K. (1923). Protoceratops andrewsi, a pre-ceratopsian dinosaur from Mongolia. Amer. Mus. Novitates 72, 1–9. Gregory, W. K., and Simpson, G. G. (1926). Cretaceous mammal skulls from Mongolia. Amer. Mus. Novitates 22, 1–20. Kielan-Jaworowska, Z. (1969). Hunting for Dinosaurs, pp. 177. MIT Press, Cambridge, MA. Lavas, J. R. (1993). Dragons from the Dunes, pp. 137. Published by the author. Matthew, W. D., and Granger, W. (1925). Fauna and correlation of the Gashato Formation of Mongolia. Amer. Mus. Novitates 189, 1–12. Norell, M. A., McKenna, M. C., and Novacek, M. J. (1992a). Estesia mongoliensis, a new fossil varanoid from the Cretaceous Barun Goyot Formation of Mongolia. Amer. Mus. Novitates 3045, 1–24. Norell, M. A., Clark, J. M., and Perle, A. (1992b). New dromaeosaur material from the Late Cretaceous of Mongolia. J. Vertebr. Paleontol. 12(3), Suppl. 45A. Norell, M. A., Chiappe, L. M., and Clark, J. M. (1993, September). New limb on the avian family tree. Nat. History, 38–43. Norell, M. A., Clark, J. M., Dashzeveg, D., Barsbold, R., Chiappe, L. M., Davidson, A. R., McKenna, M. C., Perle, A., and Novacek, M. J. (1994). A Theropod dinosaur embryo and the affinities of the Flaming Cliffs dinosaur eggs. Science 266, 779–782. Norell, M. A., Dingus, L., and Gaffney, E. S. (1995). Discovering Dinosaurs, pp. 225. Knopf, New York. Novacek, M. J. (1996). Dinosaurs of the Flaming Cliffs. Anchor/Doubleday, New York. Novacek, M. J., and McKenna, M. C. (1993). Therian mammals from the Late Cretaceous of Mongolia. J. Vertebr. Paleontol. 13(3), Suppl. 51A.



Novacek, M. J., Norell, M. A., McKenna, M. C., and Clark J. (1994). Fossils of the Flaming Cliffs. Sci. Am. December, 60–69. Osborn, H. F. (1923). Two Lower Cretaceous dinosaurs from Mongolia. Amer. Mus. Novitates 95, 1–10. Osborn, H. F. (1924). Three new Theropoda, Protoceratops zone, central Mongolia. Amer. Mus. Novitates 144, 1–12. Perle, A., Norell, M. A., Chiappe, L. M., and Clark, J. M. (1993). Flightless bird from the Cretaceous of Mongolia. Nature 362, 623–626. Sabath, K. (1991). Upper Cretaceous amniotic eggs from the Gobi Desert. Acta Paleontol. Polonica 36(2), 151–192. Watabe, M. (1994). Results of the Hayashibara Museum of Natural Sciences–Institute of geology, Academy of Sciences of Mongolia Joint Paleontological Expedition to the Gobi Desert in 1993. J. Vertebr. Paleontol. 14(3), Suppl. 51A.

Central Geological and Prospecting Museum, Russia see MUSEUMS





Cerapoda (Fig. 1) was established by Sereno (1986) to represent EUORNITHOPODA (ORNITHOPODA of other authors; see PHYLOGENY OF DINOSAURS) ⫹ MARGINOCEPHALIA, a major branch of ornithischia. The node is supported by several synapomorphies, including the substantial diastema between the premaxillary and maxillary teeth, the asymmetrical enamel on upper and lower teeth, reduction to five or fewer premaxillary teeth, and other features. An alternate view of

FIGURE 1 Phylogeny of Cerapoda, after (a) Sereno 1986, (b) Norman 1984.

ornithischian interrelationships (Fig. 1b; see NEOCERATOPSIA and PHYLOGENY OF DINOSAURS) is that the origin of Marginocephalia is within Ornithopoda, perhaps among Hypsilophodontidae (Norman 1984; See Benton 1990 for review); if so, Cerapoda would be a problematic taxon.


References Benton, M. J. (1990). Origin and interrelationships of dinosaurs. pp. 11–29 in D. B. Weishampel, P. Dodson, and H. Osmolska (eds.), The Dinosauria. Berkeley: University of California Press.


106 Norman, D. B. (1984). A systematic appraisal of the reptile order Ornithischia. pp. 157–162 in W.-E. Reif and F. Westphal (eds.), Third Symposium on Mesozoic Terrestrial Ecosystems: Short Papers. ATTEMPTO Verlag, Tu¨bingen. Sereno, P. C. (1986). Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). Natl. Geogr. Res. 2, 234–256.


both characterized by an incipient parietosquamosal frill overhanging the back of the skull.

See also the following related entries: MARGINOCEPHALIA ● NEOCERATOPSIA ● PSITTACOSAURIDAE


Ceratopsia PETER DODSON University of Pennsylvania Philadelphia, Pennsylvania, USA

The Ceratopsia are the horned dinosaurs, important herbivores of the Late Cretaceous. They are defined as all Marginocephalia closer to Ceratopsidae than to Pachycephalosauria. Diagnostic features include one rostral bone, a maxilla at least two-thirds as tall as its length, a broad immobile mandibular symphysis, and a tall snout with relatively broad premaxilla (Sereno, 1990). The name Ceratopsia, meaning ‘‘horn faced,’’ was coined by O. C. Marsh in 1890, who understood them only to include the large, quadrupedal, frilled horn-bearers of western North America. With the description of Leptoceratops from Alberta in 1914, and of Protoceratops from Mongolia in 1923, the concept of Ceratopsia was extended to include not only Marsh’s Ceratopsidae, but also the PROTOCERATOPSIDAE. Protoceratopsians have flaring jugals and at least rudimentary parieto-squamosal frills; most lack true horn cores, but all share a number of derived characters with ceratopsids, first among which is the rostral bone in front of the premaxilla, a character found in no other dinosaur. PSITTACOSAURUS from the Early Cretaceous of Mongolia was a small, ornithopod-like biped that lacked both a frill and horns of any kind, but whose face was otherwise remarkably ceratopsian in appearance, including a toothless beak and flaring jugals. In 1975, the existence of a rostral in Psittacosaurus was definitively demonstrated by Teresa Marya´nska and Halszka Osmo´lska, requiring the admission of Psittacosaurus into the Ceratopsia. In 1986, Paul Sereno created the NEOCERATOPSIA, comprising the Protoceratopsidae and the Ceratopsidae, to stand as a sister group to the Psittacosauria as monophyletic clades within the Ceratopsia. Sereno also created the MARGINOCEPHALIA, comprising the Ceratopsia plus its sister group the PACHYCEPHALO-

Dodson, P. (1990a). Ceratopsia. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), p. 578. Univ. of California Press, Berkeley. Dodson, P. (1990b). Marginocephalia. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 562–563. Univ. of California Press, Berkeley. Dodson, P., and Currie, P. J. (1990). Neoceratopsia. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 593–618. Univ. of California Press, Berkeley. Sereno, P. C. (1990). Psittacosauridae. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 579–592. Univ. of California Press, Berkeley.

Ceratosauria TIMOTHY ROWE RON TYKOSKI University of Texas Austin, Texas, USA

JOHN HUTCHINSON University of California Berkeley, California, USA

Ceratosauria is a stem lineage of dinosaurs that comprises the sister lineage of Tetanurae within Theropoda. Many of the taxa now recognized as members of Ceratosauria have a long history of controversial taxonomic assignment. Before the advent of phylogenetic systematics and the recognition of lineages based on shared derived characters, size alone was the principal criterion used by systematists to assign theropods to major taxonomic categories (see CARNOSAURIA; COELUROSAURIA). As a result the larger theropod ceratosaurs, such as Ceratosaurus nasicornis and Dilophosaurus wetherilli, were grouped as ‘‘carnosaurs’’ or ‘‘megalosaurs,’’ whereas the smaller taxa, such as Coelophysis bauri, were grouped as ‘‘coelurosaurs’’ or ‘‘podokesaurs.’’ Recently, the recognition


FIGURE 1 (a) A tree showing proposed names and relationships among the better known ceratosaurs (from Holtz, 1994) and (b) a consensus tree that may be a more accurate representation of current knowledge (based on Rowe and Gauthier, 1990; Holtz, 1994; Sereno et al., 1994).

of uniquely derived characters has led to the recognition of a monophyletic Ceratosauria (Figs. 1a and 1b) that, like Tetanurae, includes a highly diversified assemblage of large and small theropods. Ceratosaurs first appear in the Late Triassic fossil record with Coelophysis (Figs. 2 and 3) in North America and the poorly known Liliensternus in Europe (Rowe and Gauthier, 1990). Currently, only approximately 20 named theropod species have been assigned to Ceratosauria; several are incomplete, leaving doubts about their assignments. Collectively, these 20 or so taxa may document a lineage extending from the Triassic into the Late Cretaceous, with a global distribution. The known fossil record of ceratosaurs spans approximately 170 million years. The most extensive records of ceratosaurs are from Late Triassic and Early Jurassic rocks, where they are the most common theropods in all currently known faunas. Several ceratosaur taxa are known from burials that preserve multiple individuals, and some offer rare opportunities to study posthatching development in early dinosaurs. One of the richest Mesozoic dinosaur burials ever discovered is the Coelophysis quarry at GHOST RANCH, in northwestern New Mexico. Hundreds of Coelophysis individuals were buried

107 together en masse in sediments of the Triassic Chinle Formation (Colbert, 1995; Schwartz and Gillette, 1994). Despite the spectacular preservation in the case of Coelophysis, the ceratosaur lineage as a whole is only very poorly represented in the fossil record. O. C. Marsh (1884) described the first known member and namesake of the lineage, C. nasicornis, based on a nearly complete skeleton from the Late Jurassic Morrison Formation. A few years later, E. D. Cope (1889) described the first fragments of C. bauri, but it was another 60 years before a complete skeleton was found. Relatively complete skeletons are also known for the Early Jurassic D. wetherilli (Welles, 1954, 1970), two species of Syntarsus (Raath, 1969; Rowe, 1989), and the Late Cretaceous Carnotaurus sastrei (Bonaparte, 1985). The remaining ceratosaur taxa are known from only fragmentary specimens. Coelophysis is the oldest known ceratosaur, but it already exhibits many skeletal peculiarities of its own, which evolved following its divergence from the an-

FIGURE 2 Skeletons of three representative ceratosaurs: (a) Coelophysis (after Colbert, 1995), (b) Dilophosaurus (after Welles, 1984), and (c) Carnotaurus (after Bonaparte, 1985).



FIGURE 3 Skulls of some of the best known ceratosaurs.

cestral ceratosaur. This implies that the earliest stages of ceratosaur history remain as yet undiscovered. The Late Jurassic C. nasicornis is the last known member of the lineage in the Northern Hemisphere, but in Argentina, Madagascar, and India ceratosaurs evidently persisted until the Late Cretaceous. Because documentation of the entire ceratosaur lineage rests on the scant remains of only 20 or so taxa, our knowledge of it is obviously very incomplete. We expect many features of the ceratosaur phylogeny, and especially the diagnoses presented below, to change with new discoveries and we urge caution in making sweeping generalizations with so little evidence. The first studies to recognize a monophyletic Ceratosauria defined its name in reference to a stembased lineage that includes C. nasicornis (Marsh, 1884; Gilmore, 1920) and its closest relatives among Theropoda (Gauthier, 1984; Rowe and Gauthier, 1990). To date, the relationships among ceratosaurs have not been studied rigorously. Initial studies included within Ceratosauria the taxa C. nasicornis, Sarcosaurus woodi, Segisaurus halli, D. wetherilli, Liliensternus lilliensterni, C. bauri, Syntarsus rhodesiensis, and Syntarsus kayentakatae (Gauthier, 1984; Rowe and Gauthier, 1990). Later authors have generally supported a monophyletic Ceratosauria composed of two clades, Coelophysoidea and Neoceratosauria (Novas, 1991; Holtz, 1994; Sereno et al., 1994, 1996). Coelophysoidea is composed of Dilophosaurus and Coelophysidae (⫽ Coelophysis ⫹ Syntarsus; see Figs. 1a and 1b). Neoceratosauria (Novas, 1991) is made up of Ceratosaurus, Abelisauridae (Bonaparte, 1985), and Elaphrosaurus bambergi (Holtz, 1994). These authors reject a conflict-

ing suggestion that Ceratosauria is paraphyletic, and that C. nasicornis is more closely related to Tetanurae than to coelophysoids (Bakker et al., 1988; Currie, 1995). Diagnosing Ceratosauria is hindered by incompleteness of the record as well as our state of understanding among basal theropods. Owing to recent discoveries of Triassic dinosauromorphs, several characters previously thought to be diagnostic of the lineage now have equivocal distribution and might or might not be diagnostic of Ceratosauria (Novas, 1996). In addition, some diagnostic ceratosaur features, such as those involving fusion among skeletal elements, are only expressed in mature individuals, and some workers have been inadvertently misled by their absence in subadult specimens. With those caveats, Ceratosauria is diagnosed by the presence of two pairs of pleurocoels in the cervical vertebrae, perforation of the pubic plate by two fenestrae, fusion of sacral vertebrae and ribs in adults, fusion of the astragalus and calcaneum in adults, and a flange of the distal end of the fibula that flares medially to overlap the ascending process of the astragalus anteriorly. The characters involving fusion are convergent upon conditions seen in ornithurine theropods. Additionally, ceratosaurs are distinguished from most other theropods by the retention of a number of plesiomorphic features, including a lack of maxillary fenestration, and the retention of four fingers in the hand. Due to their fragmentary nature, most referrals of taxa to Ceratosauria are only based on a few characters and will warrant reevaluation as more complete specimens are discovered.

Ceratosauria The most distinctive and strongly supported lineage within Ceratosauria is Coelophysoidea, which includes small to medium-sized, lightly built species. These are the most common theropod remains from Upper Triassic and Lower Jurassic deposits of North America, Europe, and southern Africa. They lack axial pleurocoels and the transverse processes of their dorsal vertebrae are roughly triangular in dorsal view. The premaxilla and maxilla meet along a loose contact that creates a gap or incisure in the tooth row termed the ‘‘subnarial gap’’ (Welles, 1984). The premaxillary teeth are subcircular in cross section and are not serrated. Two species, D. wetherilli and S. kayentakatae, sport thin, paired crests on the skull (Welles, 1984; Rowe, 1989). These crests are far too fragile to have served in combat or any other mechanically demanding function and were likely used for visual display. Whether the crests were dimorphic or possessed by only one sex is unknown. Preserved sclerotic rings in Syntarsus (Fig. 3) indicate a large eyeball that filled the orbit. The skulls in coelophysoids are delicately built and set on long necks. The arms are of moderate length, compared to Tyrannosaurus at one extreme and Deinonychus at the other. The hands are equipped with large claws on digits I–III. Digit IV consists of a reduced metacarpal and one or two tiny phalanges and was so reduced in overall size that it was nonfunctional as a separate digit. The ankle of coelophysoids was primitive in being equipped with only a short ascending process of the astragalus and in the retention of participation by all five metatarsals in the ankle joint. Since its first recognition (Bonaparte et al., 1990), both the diagnosis and composition of Neoceratosauria have been debated. A thorough reanalysis of basal theropods will be needed to resolve the question, but in lieu of such a study the popular opinion among most experts is that Abelisauridae and C. nasicornis both belong to this lineage. Some authors have also included E. bambergi (Holtz, 1994). These are all medium- to large-sized theropods known from the Late Jurassic of western North America and East Africa and the Late Cretaceous of South America, India, and Madagascar. One purported synapomorphy is that the premaxilla is as tall or taller than it is long, with nearly vertical anterior and posterior borders. In addition, the quadrate is tall and posteroventrally angled, which places the quadrate/articular joint posteroventrally as well, yielding a large infratemporal fenestra. The effects of size alone may explain

109 some of these features. Another purported synapomorphy is that the femoral head is directed anteromedially, in contrast to the medially directed femoral head of tetanurine theropods. However, all ceratosaurs have anteromedially directed femoral heads, so the monophyly of Neoceratosauria remains at best only weakly defended. The name Abelisauroidea was suggested (Bonaparte, 1991) for the most inclusive clade within Neoceratosauria. The name was coined in reference to a stem-based lineage that includes Abelisauridae and all taxa closer to abelisaurids than to the North American C. nasicornis. Currently, Abelisauroidea includes only Elaphrosaurus and Abelisauridae. Purported synapomorphies include pleurocoelous dorsal vertebrae, a cnemial process arising from the lateral surface of the tibia shaft, a pronounced pubic boot, an aliform anterior (⫽ lesser) trochanter, and more than five sacral vertebrae (Holtz, 1994). Each of these characters is homoplastic with many tetanurine theropods, so further work is needed to defend the monophyly of Abelisauroidea. ABELISAURIDAE (Bonaparte, 1991) is generally taken to be a stem-based name that refers to all taxa closer to Carnotaurus than to Elaphrosaurus. Currently, it includes Abelisaurus, Carnotaurus, Xenotarsosaurus, Indosuchus, Indosaurus, Majungasaurus (Sampson et al., 1996), and possibly some other fragmentary taxa such as Noasaurus. Only Carnotaurus is known from a relatively complete skeleton. To date, only phenetic resemblances have been offered in support of the recognition of Abelisauridae. Geographic (Gondwanan) and stratigraphic distribution of its members are consistent with abelisaurid monophyly, but the record is so incomplete as to be consistent with many other scenarios of theropod phylogeny. Because stembased names such as Abelisauridae are predicated on a more general node, in this case the weakly supported Abelisauroidea, and because no rigorous phylogenetic analysis of these poorly known taxa has been conducted, the monophyly of Abelisauridae should be viewed with caution. Carnotaurus (Figs. 2 and 3) exhibits unusual morphology in which the frontal bones bear blunt, laterally facing bony outgrowths over the orbits. The premaxilla, and the snout in general, is blunt and deep, as has been observed in other abelisaurid taxa (Bonaparte et al., 1990; Sampson et al., 1996). The forelimbs of Carnotaurus are profoundly reduced, like the famous condition in Tyrannosaurus, but the pectoral

110 girdle remains well developed. The scapula is straplike, as is the case in tetanurine theropods (Gauthier, 1984). The hand has four stubby fingers, including an incongruously elongated fourth metacarpal. Both the pubis and the ischium are distally expanded. Skin impressions from the neck, shoulder, torso, and tail were found with the original skeleton. The texture of the skin in these regions was rough and pebbly, with larger, intermittently spaced conical thickenings of the hide. The skulls of several ceratosaurs display bizarre ornamentation. Ceratosaurus has a nasal horn and brow hornlets. Carnotaurus sastrei lacks a nasal horn, but instead sports massive brow horns and a rugose surface on top of its snout. Majungasaurus crenatissimus (Sampson et al., 1996) appears to have a large knob of bone on its skull roof and rugose nasal bones. These structures may have functioned in visual displays, although the robust construction has led to speculation about their use in intraspecific bouts. The lack of a rigorous phylogenetic analysis, plus discordance in the distribution of purported synapomorphies, leaves doubt regarding the monophyly of Abelisauridae, Abelisauroidea, and Neoceratosauria. Nevertheless, all taxa currently referred to these names seem clearly to lie outside Tetanurae, and all are extinct. The relationships among these taxa, and among basal theropods generally, remain vexing problems (Fig. 1b).

See also the following related entries: TETANURAE ● THEROPODA

References Bakker, R. T., Williams, M., and Currie, P. (1988). Nanotyrannus, a new genus of pygmy tyrannosaur, from the latest Cretaceous of Montana. Hunteria 1, 1–30. Bonaparte, J. F. (1985). A horned Cretaceous dinosaur from Patagonia. Natl. Geogr. Res. 1, 149–151. Bonaparte, J. F. (1991). The Gondwanan theropod families Abelisauridae and Noasauridae. Historical Biol. 5, 1–25. Bonaparte, J. F., Novas, F. E., and Coria, R. A. (1990). Carnotaurus sastrei Bonaparte, the horned, lightly built carnosaur from the middle Cretaceous of Patagonia. Contrib. Sci. Nat. History Museum Los Angeles County 416, 1–42. Colbert, N. E. (1995). The Little Dinosaurs of Ghost Ranch, pp. 250. Columbia Univ. Press, New York. Cope, E. D. (1889). On a new genus of Triassic Dinosauria. Am. Nat. 23, 626.

Ceratosauria Currie, P. J. (1995). Phylogeny and systematics of theropods (Dinosauria). J. Vertebr. Paleontol. 15(Suppl. to No. 3), 25A. Gauthier, J. A. (1984). A cladistic analysis of the higher systematic categories of the Diapsida. Ph.D. dissertation, University of California, Berkeley. Gilmore, C. W. (1920). Osteology of the carnivorous Dinosauria in the United States National Museum, with special reference to the genera Antrodemus (Allosaurus) and Ceratosaurus. Bull. U.S. Natl. Museum 110, 1–154. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 86, 1100–1117. Marsh, O. C. (1884). Principal characters of American Jurassic dinosaurs. Part VIII: The order Theropoda. Am. J. Sci. 27(38), 329–341. Novas, F. E. (1991). Relaciones filogeneticas de los dinosaurios teropodos ceratosaurios. Ameghiniana 28, 401–414. Novas, F. E. (1996). Dinosaur monophyly. J. Vertebr. Paleontol. 16, 723–741. Raath, M. A. (1969). A new coelurosaurian dinosaur from the Forest Sandstone of Rhodesia. Arnoldia 4, 1–25. Rowe, T. (1989). A new species of the theropod dinosaur Syntarsus from the Early Jurassic Kayenta Formation of Arizona. J. Vertebr. Paleontol. 9, 125–136. Rowe, T., and Gauthier, J. A. (1990). Ceratosauria. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 151–168. Univ. of California Press, Berkeley. Sampson, S. D., Krause, D. W., Dodson, P., and Forster, C. A. (1996). The premaxilla of Majungasaurus (Dinosauria: Theropoda), with implications for Gondwanan paleobiogeography. J. Vertebr. Paleontol. 16, 601–605. Schwartz, H. L., and Gillette, D. D. (1994). Geology and taphonomy of the Coelophysis quarry, Upper Triassic Chinle Formation, Ghost Ranch, New Mexico. J. Paleontol. 68, 1118–1130. Sereno, P. C., Wilson, J. A., Larsson, H. C. E., Dutheil, D. B., and Sues, H.-D. (1994). Early Cretaceous dinosaurs from the Sahara. Science 266, 267–271. Sereno, P. C., Dutheil, D. B., Iarochene, M., Larsson, H. C. E. et al. (1996). Predatory dinosaurs from the Sahara and Late Cretaceous faunal differentiation. Science 272, 986–991. Welles, S. P. (1954). New Jurassic dinosaur from the Kayenta Formation of Arizona. Bull. Geol. Soc. Am. 65, 591–598. Welles, S. P. (1970). Dilophosaurus (Reptilia: Saurischia), a new name for a dinosaur. J. Paleontol. 44, 989. Welles, S. P. (1984). Dilophosaurus wetherilli (Dinosauria, Theropoda), osteology and comparisons. Palaeontographica A 185, 85–180.

Chemical Composition of Dinosaur Fossils HERVE´ BOCHERENS Universite´ Pierre et Marie Curie Paris, France


he chemical composition of fossilized vertebrate hard tissues is the result of the uptake, exchange, and loss of chemical elements, in two different sets of circumstances. First, during the life of the animal, chemical elements are taken from the surrounding environment through food and drinking water, and they are incorporated into the living tissues under physiological conditions. Second, during the diagenetic evolution of the mineralized tissues (i.e., fossilization), this original organization of the chemical elements is altered under the physical, chemical, and microbiological conditions existing when the dead tissues are buried in the sediment. The organization of the chemical elements can be linked to some aspects of the biology of the living animal, such as its genetic pool, its physiology, its diet, and the climatic conditions of the ecosystem. Knowing these relationships in modern animals, it is possible to retrieve some information about the paleobiology of extinct animals from the chemical composition of their fossilized remains, but only if the modifications occurring after death are minimal and do not totally overprint the biological signals. Bones, teeth, and eggshells are the only preserved fragments of dinosaurs that could eventually provide paleobiological information from their chemical composition. Whenever remnants of dinosaur soft tissues are exceptionally preserved, they are only permineralized pseudomorphs or prints in the sediment, and thus they do not bear any chemical relationship to the original living tissue. In contrast, most fossilized bones, teeth, and eggshells present the same global chemical composition as the living original tissue. The inorganic part is still formed of calcium phosphate in fossil bones and teeth, and of calcium carbonate in fossil eggshells. Although the organic part of these tissues is extensively degraded, some recognizable fragments may still be recovered. The question for the ‘‘chemical’’ paleontologist is whether these minerals

and these fragments of organic matter still bear any useful information about the paleobiology of the animal when it was alive. This possibility has been explored sporadically for dinosaurs and other extinct vertebrate groups since the beginning of this century, but this field of research became more firmly based during the past decades (for a historical perspective on this research, see Fig. 1). In ‘‘classical’’ paleontology, information is retrieved at the morphological and at the histological levels of the fossil hard tissues. In chemical paleontology, information is tentatively retrieved at lower levels of organization—the molecular, chemical, and isotopic levels. We will examine for each level of organization the kind of information recorded in the living tissue and then the possibility of preservation of this information in the fossilized dinosaur remains.

Molecules Two major groups of molecules can be distinguished in vertebrate mineralized tissues: mineral and organic molecules. The mineral molecules are crystals— calcium phosphate (apatite), with hydroxyl (OH⫺) and numerous ionic substitutions including carbonate (CO2⫺ 3 ) ions (this mineral is thus called carbonate hydroxylapatite; CHA), in bones and teeth; and calcium carbonate in eggshells. The crystallographic properties of CHA, such as its size and crystallinity (i.e., the perfection of the crystals), are mostly determined by the type of tissue (bone and dentine have very small crystals, whereas the crystals are much bigger in enamel) and the genetics (the size of the bioapatite crystals in dentine differs according to the species). Although the great majority of dinosaur fossil bones are still formed of calcium phosphate, the crystallographic properties of the CHA crystals of bone are clearly changed during diagenesis, with an increase in the perfection of the crystals, as shown by X-ray diffractometry of fossil bone powder (Pflug



Chemical Composition of Dinosaur Fossils

FIGURE 1 Historical perspective on the studies of dinosaur fossils’ chemical composition. The arrows indicate the date of the first publication about a topic, and references to relevant work on the same topic are also indicated.

and Stru¨bel, 1967; Person et al., 1995). However, these changes do not alter the histological features of the fossil bones in most cases. The enamel crystals seem less affected by the diagenetic changes than the bone and dentine crystals. Eggshell calcium carbonate crystals have a specific mineralogy (aragonite for turtles, and calcite for other reptiles and birds). The occurrence and composition of the organic molecules in mineralized tissues are mostly determined by the genetic program. They do not depend on environmental parameters. If organic molecules are retrieved intact from fossil tissues, they may provide phylogenetic information by comparison of homologous molecules in modern species or other fossil species. They can also be used as an uncontaminated support of biogenic chemical or isotopic signals bearing paleoenvironmental information.

Collagen and Other Proteins Fresh bone and dentine contain about 20% organic matter, 90% of which is collagen. This fibrillar protein is easily recognizable by its specific amino acid composition, with 30% glycine, about 10% proline and 10% hydroxyproline, and a few percent hydroxyly-

sine. These last two amino acids are found almost exclusively in collagen. This protein is not ideal for phylogenetic studies because it does not change very much from one group to another. Collagen is rather insoluble when intact, but it is quickly degraded after death by hydrolysis and generates smaller peptides as degradation products. These peptides are soluble and are likely to be leached out of the bone or tooth during fossilization (Hare, 1980). Amino acid patterns very similar to those of collagen have been reported for organic matter extracted from dinosaur bones (Wyckoff, 1969; Nowicki et al., 1972; Davidson et al., 1978; Bocherens et al., 1991b), and even for Devonian material 300 million years old (Davidson et al., 1978). However, actual peptide remnants of collagen, that would indicate the preservation of this protein at the molecular level have been reported only once, and even so, the amino acid composition of the peptides was not identical to collagen, suggesting a significant amount of contamination (Gurley et al., 1991). The non-collagenic proteinic material present in bone and dentine, as well as in enamel and eggshells, may have a better potential for preservation at the molecular level than collagen. In bone and dentine,

Chemical Composition of Dinosaur Fossils this material is composed of sialoproteins, osteocalcin, phosphoproteins, and serum-derived proteins. Some non-collagenous proteins have been detected immunologically in dinosaur bones (Muyzer et al., 1992), but they have not been extracted and purified yet. Phosphorylated amino acids suggesting the preservation of phosphoproteins have been extracted from Lower Cretaceous crocodile enamel from Niger (Glimcher et al., 1990), but no investigation has been performed on the contemporaneous dinosaur material. Proteinic material has also been reported from dinosaur eggshells (Voss-Foucard, 1968; Kolesnikov and Sochava, 1972; Krampitz et al., 1977; Marin and Dauphin, 1991).

DNA DNA has been successfully retrieved from mammal fossil bones as old as 150,000 years (Hagelberg et al., 1994). This ancient DNA allowed phylogenetic studies of extinct species such as moa, mammoth, saber-toothed cat, and cave bear (Cooper et al., 1992; Janczewski et al., 1992; Hagelberg et al., 1994; Ha¨nni et al., 1994). The discovery of ancient DNA in insects preserved in amber of the same age as dinosaurs (Cano et al., 1993) raised the hope to extract DNA from dinosaur bones as well. No DNA has yet been retrieved from dinosaur bones, but the research is under way (Morell, 1993).

113 elements than just calcium and phosphate. These atoms or ions substitute into the CHA structure to form a very complex chemical formula (Fig. 2). For instance, there is more fluorine (F) in bones and teeth of marine vertebrates than in those of non-marine vertebrates (Klement, 1938; Schmitz et al., 1991). Strontium (Sr) and barium (Ba) may replace calcium in the bioapatite, but because the organisms discriminate against Sr and Ba during digestion, there is a decrease in the amount of these elements from herbivores to carnivores (Elias et al., 1982; Ezzo, 1994). Similar phenomena have been reported for eggshells, but the details are less well known (Dauphin, 1988). The amount of such trace elements in archaeological bones has been tentatively used to investigate the diet of ancient people and of extinct animals. However, the problem of diagenetic alteration of trace element amounts in permineralized fossil bones precludes their use in dinosaur bones. The deposition and formation of authigenic minerals in the bone cavities, such as calcite, quartz, sulfides, iron, and manganese oxides, dramatically changes the chemical composition of fossil bone relative to fresh bone. Careful study of these authigenic minerals can yield valuable information about the diagenetic evolution of a fossil bone (Clarke and Barker, 1993) and thus

Other Organic Molecules Carbohydrates such as mucopolysaccharides and lipids are also contained in fresh bones. Lipids (Everts et al., 1968; Pawlicki, 1977a) and mucopolysaccharides (Pawlicki, 1977b) have been identified in dinosaur bones but with no clear paleobiological implication. However, a reassessment of these results is necessary with improved technologies.

Chemical Elements The amount of the major chemical elements in bone, such as carbon (C), hydrogen (H), nitrogen (N), oxygen (O), and sulfur (S) in organic matter and calcium (Ca), phosphorus (P), oxygen, and carbon in CHA, is determined mostly genetically and partly by the health status. However, for some elements present in very small amounts (less than 1%) in CHA, also called trace elements, there is a control through the external environment. Indeed, the calcium phosphate of bone, dentine, and enamel includes many other chemical

FIGURE 2 Summary of the possible substitution in bioapatite, their bearing for biological studies, and the evolution of this chemical composition during fossilization. Arrows represent the possible substitutions. The underlined elements have a chemical composition that depends on the environmental conditions; elements in relief are not present in living bioapatite but incorporated during diagenesis. Authigenic minerals are formed during fossilization in the porosity of bone.

114 indirectly help to solve some alteration problems. Microprobe techniques allow the analysis of welldefined spots in a fossil tissue and thus focus on the apatite part of the bone, which is the only one in which the preservation of a biogenic signature is possible. However, these techniques are not as accurate as the techniques for analyzing the bulk of the fossil bone. Microprobe studies have demonstrated the diagenetic alteration of phosphate in dentine (Bocherens et al., 1994). A few studies attempted to link the amounts of different chemical elements in bones, such as calcium, phosphorus, lead, iron, and magnesium, to the health status of a Late Cretaceous ornithomimid from Mongolia (Pawlicki and Bolechala, 1987, 1991). Unfortunately, the conclusions are impaired by the diagenetic alteration of these chemical compositions, and they remain very dubious. The chemical composition of enamel may be less altered than in bone and dentine, due to the greater compactness and the larger size of the enamel crystals (Dauphin, 1989; Bocherens et al., 1994). Applications to the paleobiology of dinosaurs have not yet been achieved. An additional problem when dealing with ancient reptiles is the poor knowledge of the variability of the amount of these trace elements in modern reptile bones. Thus, the interpretation of the values measured in the fossil specimens is difficult (Bocherens et al., 1994). Chemical contents of dinosaur eggshells have also been used (Iatzoura et al., 1991), but, in that case, the control for diagenetic alteration was not satisfactory.

Isotopes Chemical elements usually exist in the natural environment under different forms called isotopes. The different isotopes of a given chemical element exhibit identical gross chemical properties, but their atomic masses are slightly different. For each element, one isotope is in much greater abundance than the other(s), and very accurate measurements can distinguish between substances of different origins. The isotopes relevant for dinosaur studies are carbon (13C/ 12C), nitrogen (15N/ 14N), and oxygen (18O/ 16O). For these light elements, the isotopic ratios in the mineralized tissues reflect the ratios in the source food or drinking water, and the fractionations in the organism during metabolism, which depends on the environmental conditions and the physiology of the animal.

Chemical Composition of Dinosaur Fossils Carbon is the major constituent of organic matter and is also present in the mineral phase of bone and tooth as carbonate (CO2⫺ 3 ) substituted for phosphate (Fig. 3), and in the carbonate of eggshells. The source of carbon is the diet, and the isotopic abundance of carbon varies according to the diet. For instance, in modern environments, the 13C amount is lower in terrestrial plants than in marine plants, and many tropical grasses have higher 13C amounts than other terrestrial plants and marine plants. Nitrogen is a major constituent of proteins. The amount of 15N is higher in an animal’s tissues than in its diet. Thus, a carnivore has a higher 15N amount than its herbivorous prey. Moreover, 15N amount increases in a given animal when the food becomes scarcer, which is usually linked to arid conditions of the environment. Oxygen is present in the phosphate and carbonate of bones, teeth, and eggshells. Its isotopic composition depends primarily on those of its drinking water and food and also on the temperature of crystallization of the minerals. The amount of 18O decreases when drinking and food water have a lower temperature and also when crystallization occurs at a lower temperature. In an animal with constant body temperature, the oxygen isotopic abundances provide information about the oxygen isotopic abundances of water in the environment and thus indirectly about the temperature of the environment. In an animal whose temperature changes according to that of the environment, the oxygen isotopic abundances also reflect the temperature of formation of the crystals, which may differ among different locations within the body—for example, if the organ is close to the source of body heat (trunk) or at the periphery of the body (limbs and tail: Barrick and Showers, 1994). Stable isotopes have great potential for providing information about the paleobiology of fossil animals. Diet, aridity, temperature, and possibly thermal physiology are reflected in carbon, nitrogen, or oxygen isotopic ratios of bones, teeth, and eggshells of a living animal. However, the biggest problem in using stable isotopes in fossil bones is the possible alteration of these values due to fossilization. One must find a stable support for isotopic values and show that the biogenic isotopic values are preserved. Organic molecules such as collagen can be used as a support for stable isotopic information. It has been shown on Pleistocene bones and teeth up to around 60,000 years

Chemical Composition of Dinosaur Fossils


FIGURE 3 Preservation of biomolecules and chemical and isotopic compositions in fossil bones and teeth. The time-scale is logarithmic from the present to 65 million years and is linear for the Mesozoic period (65–220 million years). Solid bold lines indicate the range of collagenic amino acid pattern in a dinosaur bone from the Maastrichtian of Lan˜o, Spain (Bocherens et al., 1991b). (1) Collagen and collagenic amino acid pattern: a, b, collagenic amino acid pattern in a dinosaur toe bone from the Campanian Nemegt Beds, Mongolia (Nowicki et al., 1972); c, collagenic amino acid pattern in an atlantosaurid bone from the Lower Cretaceous, Gadoufaoua, Niger (Davidson et al., 1978); d, fragments of proteins with some collagenic affinity in a Seismosaurus bone from the Upper Jurassic Morrison Formation, New Mexico (Gurley et al., 1991); e, collagenic amino acid pattern in a dinosaur bone from the Upper Jurassic Morrison Formation, Wyoming (Wyckoff, 1969); f, collagenic amino acid pattern in a fish dermal plate from the upper Devonian of Niger (Davidson et al., 1978). (2) Non-collagenic proteins: a, immunological evidence for osteocalcin in a Lambeosaurus bone from the Campanian Judith River Formation, Alberta, Canada (Muyzer et al., 1992); b, immunological evidence for osteocalcin in a sauropod bone from the Upper Jurassic Morrison Formation (Muyzer et al., 1992). (3) Lipids and fatty acids: a, lipids (staining thin sections with Sudan B) in a Tarbosaurus bone from the Campanian, Gobi Desert, Mongolia (Pawlicki, 1977b). (4) Mucopolysaccharids: a, mucopolysaccharids (histochemical reaction) in a Tarbosaurus bone from the Campanian, Gobi Desert, Mongolia (Pawlicki, 1977a). (5) DNA: preservation of chemical compositions. (6) Trace elements in enamel: a, trace element composition in dinosaur enamel from the Judith River Formation, Alberta (Bocherens et al., 1994); preservation of isotopic compositions. (7) Carbon and nitrogen isotopic abundances in collagen: a, isotopic abundances in ‘‘collagen’’ extracted from bones and teeth of dinosaur and other reptiles from the Campanian Judith River Formation, Alberta, are claimed to reflect the trophic structure (Ostrom et al., 1993). These values are likely to be diagenetically altered. (8) Carbon in enamel carbonate hydroxylapatite: a, carbon isotopic abundances in enamel of synapsids from the Late Permian from South Africa are claimed to reflect the isotopic changes in the contemporaneous marine carbonates (Thackeray et al., 1993). (9) Oxygen (from the phosphatic group) in bone hydroxylapatite: a, oxygen isotopic abundances in phosphate of Tyrannosaurus bone are claimed to reflect thermal physiology (Barrick and Showers, 1994). These values are probably diagenetically altered (Morell, 1993).

old that when collagen with its specific amino acid composition can be extracted from a bone, the carbon and nitrogen isotopic values are preserved. Checking the preservation of isotopic values in collagen-like organic matter extracted from fossil bones is performed by measuring isotopic abundances in bones from species of known dietary habits found in the

fossil locality. For instance, horse, reindeer, wolf, and hyena are found in abundance in European Pleistocene localities, and their isotopic abundances can be predicted according to their respective diets (Bocherens et al., 1991a, 1995). A similar attempt was made by Ostrom et al. (1990, 1993) on organic matter extracted from dinosaur and other vertebrate bones and teeth

116 from the Late Cretaceous Dinosaur Park Formation in Alberta, Canada. In this study, the amino acid composition of the extracted organic matter was not identical to the amino acid composition of collagen. Moreover, the carbon isotopic values were different from what is expected for such an ecosystem, and the nitrogen isotopic values, although showing some tendencies that could be related to the trophic level of the specimens, presented a far too large range of variation within a given species when compared to modern ecosystems. It is thus likely that these values have been at least partially overprinted by diagenesis and cannot be used for paleobiological reconstructions. Other organic molecules could be better preserved than collagen in dinosaur fossils (Muyzer et al., 1992) and could be a good support for stable isotopic signature (Ajie et al., 1991). However, the quantities that could be recovered would be very low. Another possibility to short-cut the contamination problem of proteinic residues in dinosaur bones would be to use new techniques of analysis, such as a gas chromatograph system, to separate the different amino acids, connected to a combustion oven generating CO2 from each separated amino acid separately, and finally an isotopic mass spectrometer to analyze the isotopic abundances of carbon in the different amino acids. By carefully choosing the relevant amino acids, it is conceivable to track back the isotopic value of the diet, and the quantities required for such an analysis are very small. The carbon isotopic abundances of a dinosaur diet have also been recorded in the carbonate of its CHA and eggshells. In bones, the isotopic values of CHA are very quickly altered (Koch et al., 1990) but they can be retained in enamel for millions of years (LeeThorp and van der Merwe, 1987). Carbon isotopic abundances are possibly preserved in Paleocene mammal tooth enamel (Koch et al., 1992) and in Permian therapsid tooth enamel (Thackeray et al., 1990). These specimens are just younger and older, respectively, than typical (Mesozoic) dinosaurs. The problem with dinosaurs is the thinness of enamel, less than .5 mm thick, and no attempt has been made to date to measure carbon isotopic abundances on dinosaur enamel CHA. Carbon isotopic abundances have been measured on dinosaur eggshell carbonates (Folinsbee et al., 1970; Erben et al., 1979; Iatzoura et al., 1991; Sarkar et al., 1991), but due to diagenetic

Chemical Composition of Dinosaur Fossils alterations these values could well be those of recrystallized carbonates and thus be very different from the biogenic values. Oxygen isotopic abundances are also likely to be diagenetically altered in eggshell carbonates. On the other hand, oxygen isotopic abundances in bone phosphate have been claimed to be preserved, in some cases at least, and to bring new light on dinosaur thermal physiology (Barrick and Showers, 1994). In this study, oxygen isotopic abundances in different bones from one Tyrannosaurus rex skeleton were compared, assuming that a rather uniform distribution of these values would indicate a homeothermic physiology. Unfortunately, the test for diagenetic alteration, i.e.. trying to find the expected isotopic variations on an ectotherm vertebrate, such as a lizard or a crocodile, has not been made. Also, preliminary data from Kolodny et al. (1997) indicate that the oxygen isotopic values in Late Cretaceous vertebrate bones from North America are related to latitude and not to the thermal physiology of the specimens within one locality, suggesting an isotopic equilibration of the oxygen phosphate with the circulating groundwater. Due to the possibility of diagenetic alteration, the conclusions about Tyrannosaurus thermal physiology cannot be taken for granted until the control for diagenetical alteration is made.

Conclusions Biogeochemistry of fossil biomineralizations is an expanding new field of vertebrate paleontology. This approach has already brought valuable new lines of information for relatively young specimens, dozens of thousands of years old. Until now, biogeochemical investigations of dinosaur fossils have provided no unambiguous paleobiological information, but many clues indicate the preservation of biogenic material in some dinosaur specimens—bones, teeth, and possibly eggshells. The continuous improvement of technology, using lower quantities of a purer material, and the search for different kinds of preservations, will most probably lead in the near future to spectacular results from this approach in the case of dinosaurs. More complete studies of the fossilization process itself, at the chemical level, are also required to understand the preservation of geochemical signals in fossil material.

Chemical Composition of Dinosaur Fossils See also the following related entries: BIOMINERALIZATION ● GENETICS ● JURASSIC PARK ● PERMINERALIZATION

References Barrick, R. E., and Showers, W. J. (1994). Thermophysiology of Tyrannosaurus rex: Evidence from oxygen isotopes. Science 265, 222–224. Bocherens, H., Fizet, M., Mariotti, A., Lange-Badre´, B., Vandermeersch, B., Borel, J. P., and Bellon, G. (1991). Isotopic biogeochemistry (13C, 15N) of fossil vertebrate collagen: Implications for the study of fossil food web including Neandertal Man. J. Hum. Evol. 20, 481–492. Bocherens, H., Brinkman, D. B., Dauphin, Y., and Mariotti, A.. (1994). Microstructural and geochemical investigations on Late Cretaceous archosaur teeth from Alberta (Canada). Can. J. Earth Sci. 31, 783–792. Clarke, J. B., and Barker, M. J. (1993). Diagenesis in Iguanodon bones from the Wealden Group, Isle of Wright, Southern England. Kaupia, Darmsta¨dter Beitrage zur Naturgeschichte 2, 57–65. Dauphin, Y. (1991). Chemical composition of reptilian teeth. 2. Implications for paleodiets. Palaeontographica, A 219(4–6), 97–105. Davidson, F. D., Lehman, J.-P., Taquet, P., and Wyckoff, R. W. G. (1978). Analyse des prote´ines de Verte´bre´s fossiles de´voniens et cre´tace´s du Sahara. Comptes Rendus l’Acade´mie Sci. Paris 287, 919–922.

117 Muyzer, G., Sandberg, P., Knapen, M. H. J., Vermeer, C., Collins, M., and Westbroek, P. (1992). Preservation of the bone protein osteocalcin in dinosaurs. Geology 20, 811–814. Ostrom, P. H., Macko, S. A., Engel, M. H., Silfer, J. A., and Russell, D. A. (1990). Geochemical characterization of high molecular weight material isolated from Late Cretaceous fossils. Organic Geochem. 16, 1139–1144. Ostrom, P. H., Macko, S. A., Engel, M. H., and Russell, D. A. (1993). Assessment of trophic structure of Cretaceous communities based on stable nitrogen isotope analyses. Geology 21, 491–494. Person, A., Bocherens, H., Salie`ge, J.-Fr., Paris, F., Zeitoun, V., and Ge´rard, M. (1995). Early diagenetic evolution of bone phosphate: A X-ray diffractometry analysis. J. Archaeol. Sci. 22, 211–221. Wyckoff, R. W. G. (1972). The Biochemistry of Animal Fossils, pp. 152. Scientechnica, Baltimore, MD.

Chengdu College of Geology Museum, People’s Republic of China see MUSEUMS



Erben, H. K., Hoefs, J., and Wedepohl, K. H. (1979). Paleobiological and isotopic studies of eggshells from a declining dinosaur species. Paleobiology 5, 380–414. Folinsbee, R. E., Fritz, P., Krouse, H. R., and Robblee, A. R. (1970). Carbon-13 and oxygen-18 in dinosaur, crocodile, and bird eggshells indicate environmental conditions. Science 168, 1353–1356. Gurley, L. R., Valdez, J. G., Spall, W. D., Smith, B. F., and Gillette, D. D. (1991). Proteins in the fossil bone of the dinosaur, Seismosaurus. J. Prot. Chem. 10, 75–90. Marin, F., and Dauphin, Y. (1991). Composition de la phase prote´ique soluble des coquilles d’œufs de dinosaures du Rognacien (Cre´tace´) du Sud-Est de la France. Neues Jahrbuch Geol. Pala¨ontol. 4, 243–255. Morell, V. (1993). Dino DNA: The hunt and the hype. Science 261, 160–162.

Children’s Museum of Indianapolis, USA see MUSEUMS




Chinese Dinosaurs DONG ZHIMING Institute of Vertebrate Paleontology and Paleoanthropology, Academia Sinica Beijing, People’s Republic of China


n the past two decades, great progress has been made in the knowledge of Chinese dinosaurs. A huge number of dinosaur skeletons, eggs, and footprints have been collected, studied, and displayed, with more than 100 species of dinosaurs, including eggs and footprints, described and named so far (Dong, 1992). The study of dinosaurs in China not only has greatly extended the list of dinosaurs over the world but also has made great contributions to the understanding of general theoretical problems concerning these fascinating animals (Russell, 1993).

A Brief History of the Study of Dinosaurs in China The Chinese people call themselves ‘‘descendants of the dragon.’’ The fossilized vertebrate bones are known as ‘‘dragon bones.’’ For instance, as early as the Jin Dynasty (265–317 A.D.), a book titled Hua Yang Guo Zhi already recorded the discovery of dragon bones in Wuchen, which covered the present Santain County, Sichuan Province. Because most of the exposed strata belong to Jurassic deposits, it is highly probable that the bones discovered were actually dinosaur bones. The earliest scientific discoveries of dinosaurs in China were made in the 1910s by Russians in the southern banks of the Heilongjiang (Amur) River. The finds were referred to a hadrosaur, Mandschurosaurus. It is the first named dinosaur in China (Riabinin, 1925, 1930). In 1913–1915 an American geologist, George Louderback of the University of California at Berkeley, reported the first dinosaur fossils to be found in the Sichuan Basin (Louderback, 1935; Camp, 1935). A German mining engineer, Berhagel, found several fossils of dinosaurs in Mengyin, Shandong, in 1916. This site was excavated by an Austrian paleontologist, Zdansky, in 1922 and 1923. This was the wellknown Euhelopus (Wiman, 1929). Reports on these early discoveries of dinosaur fos-

sils aroused the attention and interest of many Western paleontologists. A series of multinational expeditions came to China, such as the CENTRAL ASIATIC EXPEDITIONS of the AMERICAN MUSEUM OF NATURAL HISTORY (1921–1930), the SINO –SWEDISH EXPEDITIONS (1927–1935), and Sino–French Scientific Expeditions (1930). The Central Asiatic Expedition was the largest of the three and proved to be the most fruitful (Osborn, 1924; Andrews, 1932). The years 1933 to 1949 represent the initial stage in the study of dinosaur fossils in China. The progress accomplished during this period was mainly undertaken by Dr. C. C. Young. After pursuing his studies in Germany, he returned to China in 1928 and devoted himself to the study of paleovertebrates. From 1933 onwards, he began to focus his attention on the study of reptiles and conducted a series of excavations for dinosaur fossils in Sichuan, Yunnan, Xinjiang, and Gansu. He undertook the famous excavation at Lufeng Basin, Yunnan, in 1938 (Young, 1951). In 1951 Dr. Young led an excavation of Late Cretaceous dinosaurs at Laiyang, Shandong, and collected a complete skeleton of Tsintaosaurus. A large sauropod skeleton, Mamenchisaurus, was studied by Young in 1954. In 1957, a new nearly complete skeleton of Mamenchisaurus was unearthed from Hechuan, Sichuan. This giant dinosaur was named M. hechuanensis and is the longest dinosaur ever discovered in Asia (Fig. 1). From 1959 to 1960, Sino–Soviet paleontological expeditions conducted large-scale excavations in the Erlen and Alxa Gobi areas of Inner Mongolia and obtained a sizable quantity of Early Cretaceous dinosaur fossils, which included Probactrosaurus (Iguanodontidae), Chilantaisaurus (Theropoda), and an ankylosaurian dinosaur. For political reasons, the expeditions were suspended in 1960. From 1963 to 1966, the Institute of Vertebrate Paleontology and Paleoanthropology of Chinese Academy of Sciences organized expeditions that explored


Chinese Dinosaurs

FIGURE 1 Tsintaosaurus, a hadrosaur from China.

the Junggar and Turpan basins of Xinjiang for 3 years and discovered the pterosaur fauna of Urhe (Dong, 1973). Nearly simultaneously, a large hadrosaur skeleton was found in Shandong. It was named Shantungosaurus, measuring 15 m in length. In 1974, 106 crates of dinosaur fossils weighing more than 10 tons were collected by the Chongqing Museum from the Late Jurassic of Zigong area. These finds include two skeletons of Omeisaurus, one skeleton of Szechuanosaurus, and Tuojiangosaurus (Dong et al., 1983). In an attempt to fill in the gaps in the evolution of dinosaurs, Chinese paleontologists engaged in the study of dinosaurs have focused on the Early and Middle Jurassic strata. They began by exploring Yunnan, Guizhou, Sichuan, and the eastern part of Xizang (Tibet) from 1976 to 1979. They found Middle Jurassic sauropods, theropods, and stegosaurs (Dong, 1992). In 1976, a large, incomplete skeleton of a sauropod was collected from the Xiashaximiao (Lower Shaximiao) Formation of the Middle Jurassic in Dashanpu of Zigong City. This specimen was named Shuno-

119 saurus lii (Dong et al., 1983). Now this site has become the well-known Dashanpu Dinosaur Quarry. From May 1979 to July 1981, the author led the dinosaur excavations in the Dashanpu dinosaur quarry. More than 40 tons of dinosaur fossils were uncovered by the Institute of Vertebrate Paleontology and Paleoanthropology, the Chongqing Municipal Museum, and the Zigong Salt Industry History Museum. Approximately 8000 bones have been excavated, many of them large and some articulated as skeletons. The fossils include complete skeletons of sauropods (Shunosaurus, Omeisaurus, and Datousaurus), carnosaurs (Gasosaurus), stegosaurs (Huayangosaurus), ornithopods (Agilisaurus and Xiaosaurus), pterosaurs, a plesiosaur, amphibians, and fishes. They represent more than 100 individual animals and include at least 12 reptiles, including 6 different kinds of dinosaurs. From July 1981 to May 1982, this quarry was worked by a Sichuan expedition. Dashanpu Quarry has proved to be one of the richest and most rewarding localities for Middle Jurassic dinosaurs in the world (see ZIGONG DINOSAUR MUSEUM). The Zigong Dinosaur Museum was built on the site and opened in the spring of 1987. An outstanding feature of this magnificent museum is that it was erected over the stratum containing the dinosaur skeletons. The newly reformed, open politics of China welcome foreign scholars to cooperate with Chinese colleagues. Some projects were in cooperation with the British Museum of Natural History (1982), Texas Tech University (1985), and Canada (1986–1990). The most exciting finds were made by the China–Canadian Dinosaur Project (CCDP) from 1986 to 1990. In 1986, a dinosaur project, the CCDP was organized. This is the first joint paleontological expedition into the northwestern interior of China since the 1940s (Dong et al., 1988, 1989; Currie, 1991). The main aim of CCDP is to study Mesozoic continental strata in the north part of China, to find superb dinosaur fossils, and to trace the relationships among various groups of dinosaurs of Asia and North America. Fieldwork has been carried out in the Junggar Basin of Xinjiang, the Gobi Desert of Inner Mongolia, and the badlands of Alberta and the Arctic islands of Canada. Two special issues of the Canadian Journal of Earth Sciences were published as the scientific results of CCDP. Eight new genera and 11 new species of turtles and dinosaurs were described in these special

Chinese Dinosaurs

120 issues. A traveling exhibition, ‘‘Dinosaur World Tour,’’ was displayed in several countries. From 1992 to 1993, an expedition from the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP) explored for dinosaurs along an ancient silk road. A new dinosaur site from the Early Cretaceous was found in the Mazunshan area, Gansu. Many dinosaur fossils, including iguanodontids, sauropods, and primitive protoceratopsids, were collected.

Distribution and Biostratigraphy of Chinese Dinosaurs China has excellent outcrops yielding dinosaur remains. Rocks deposited during the lifetime of the dinosaurs were laid down on low plains and in basins by rivers and lakes. This was important for the preservation of the remains of land vertebrates. The current climate and topography is such that the rocks at many of the best dinosaur sites are not deeply covered with soil or concealed under thick vegetation. This allows the discovery and excavation of dinosaur bones. Chinese dinosaur fossils can be divided into five dinosaur faunas and provide an almost unbroken record from the Late Triassic to the Late Cretaceous periods (Dong, 1980, 1992) (Fig. 2): 1. Early Jurassic, Prosauropod–Lufengosaurus fauna 2. Middle Jurassic (Bathonian–Callovian), Sauropod–Shunosaurus fauna 3. Late Jurassic, Sauropod–Mamenchisaurus fauna 4. Early Cretaceous, Psittacosaur–Pterosaur fauna 5. Late Cretaceous, Hadrosaurid–Titanosaurid fauna

The Early Jurassic Lufengosaurus Fauna This period is a significant time in dinosaur evolution. Dinosaurs were extensively distributed throughout all the continents. This cosmopolitan fauna could be named the Circum–Tethys Dinosaur Fauna (Dong, 1983). In China, this dinosaur assemblage comes mainly from southern China, such as the Lufeng and Yimen basins of Yunnan; the Weiyuan Basin of Sichuan, and the Dafang Basin of Guinzhou. So far, 23 genera of vertebrates have been recorded from the Lower Lufeng Formation. It represents what is here termed the prosauropod–Lufengosaurus faunal complex (Dong, 1980, 1992; Zhen et al., 1985). Representative dinosaurian taxa include three genera of prosauropods, Lufengosaurus, Anchisaurus (Gyposaurus), and Yunnanosaurus; theropods (Lukosaurus, Sinosaurus, and Dilophosaurus); a stegosaur (Tatisaurus); and small ornithopods (Dianchungosaurus) (Young, 1941, 1942, 1951). Kunmingosaurus wudingensis was a primitive sauropod from the Early Jurassic of Wuding, Yunnan. Its head and lower jaw are rather deep, with spoonshaped teeth; the sacrum consists of six vertebrae. Its pelvic girdle shows sauropod-like features, modified in the form of a plate. This dinosaur is 7.5 m long and has a relatively short neck. Its heavy body is supported by four massive and straight legs (Dong, 1992). The discovery of Dilophosaurus in the Lower Lufeng Formation was an important find in China made by the Kunming Museum in 1986 (Wu, 1992). It provided evidence that the age of the Lower Lufeng Formation was Early Jurassic (see GLEN CANYON GROUP). Young (1951) argued for a Rhaetic age of the prosauropod–Lufengosaurus fauna of the Lower Lufeng Formation. It is now considered Early Jurassic in age by most paleontologists (Chen et al., 1982; Wu, 1991; Wu et al., 1993) (Fig. 3).

The Middle Jurassic Shunosaurus Fauna

FIGURE 2 Dinosaur localities in China.

The Middle Jurassic is a period when many major dinosaur taxa seemed to make their first appearances. They became the dominant members through a rapid radiation. Middle Jurassic dinosaurs in China come from two major areas: the Sichuan and the Junggar basins. In Sichuan, the Dashanpu is a well-known Middle Jurassic (Bathonian–Callovian) dinosaur site. It produced a primitive sauropod–Shunosaurus fauna

Chinese Dinosaurs


FIGURE 3 Lufengosaurus, appropriately on display in its namesake, the Lufeng Dinosaur Museum.

from the Lower Shaximiao (Xiashaximiao) Formation. The Shunosaurus fauna contained sauropods (Protognasaurus, Shunosaurus, Datousaurus), theropods (Gasosaurus and Szechuanosaurus), a stegosaur (Huayangosaurus), and ornithopods (Xiaosaurus and Agilisaurus). Shunosaurus, a short-necked sauropod, is the best known dinosaur, with 12 or more complete skeletons and three well-preserved skulls discovered. Omeisaurus is a large sauropod. Omeisaurus tianfuensis has a bony club at the end of the tail for defense. Agilisaurus louderbacki was a small fabrosaurid dinosaur (Peng, 1990). The material includes a nearly complete skeleton, with a complete skull. Many remains ranging from juveniles to adults were found at the Dashanpu site. Huayangosaurus was a rather primitive stegosaur, with six or seven small teeth on the premaxilla. The bony plates are variable in shape and are arranged symmetrically along its back from the neck to the end of the tail. A pair of large bony plates also lie on the shoulders. Twelve individuals were found at Dashanpu Quarry. Middle Jurassic deposits are also distributed extensively in the Junggar Basin, Xinjiang. They are called the Wuciawan Formation and are composed of light gray, fine to medium-grained feldspathic quartzitic sandstones, sandy mudstones, and siltstones. The environments of deposits were fluvial to deltaic. Bellu-

saurus sui was a small sauropod (4.8 m). Seventeen individuals have been found from the single quarry of Konglonggou (Dinosaur ravine), Kelamaili region. Evidently a herd of these animals had been overwhelmed in a flash flood. Morphological features suggested that they could be a group of juveniles. Monolophosaurus is an allosaurid with a well-developed ridge on the top of the head. The material is a nearly complete skeleton with a complete skull and was collected from the Wucaiwan Formation in the Jiangjunmiao site in 1984 (Zhao and Currie, 1993).

The Late Jurassic Mamenchisaurus Fauna The Late Jurassic represents a golden age of dinosaurs. Sauropods were flourishing and became the most abundant taxon, evolving giantism during the Late Jurassic. They reached their maximum size and greatest diversity, with a nearly global distribution. Chinese records of the Late Jurassic dinosaur-bearing strata to date are mainly from the Shishugou Formation of the Junggar Basin, the Xiangtang Formation of Gansu Province, the Mengyin Formation of Shandong Province, and the Upper Shaximiao (Shangshaximiao) Formation of the Sichuan Basin. The main dinosaur fossils yielded by the Upper Shaximiao Formation include sauropods (Mamenchisaurus and Omeisaurus), theropods (Szechuanosaurus, Yangchuanosaurus, and Sinraptor), ornithopods


FIGURE 4 Tuojiangosaurus, a stegosaur from the Late Jurassic of Sichuan Basin.

(Gongbusaurus and Yandusaurus), and stegosaurs (Chialingosaurus, Tuojiangosaurus, and Chungkingosaurus) (Fig. 4). Mamenchisaurus hechuanensis is the most famous sauropod of China, known from a fairly complete skeleton lacking only the skull and forelimbs. Mamenchisaurus was placed in a family of its own. C. C. Young pointed out that the Mamenchisauridae is similar to the Diplodocidae. Recently, however, an incomplete skeleton with a nearly complete skull of Mamenchisaurus was found in Raxian County by the Municipal Museum of Chongqing and a complete lower jaw was collected by CCDP. The skull of Mamenchisaurus is tiny when compared with the enormous size of the animal. It has the spatula-like teeth instead of the pencil-like teeth of the Diplodocidae, so Mamenchisauridae is valid. Yangchuanosaurus was an allosauroid. It is known from an almost complete skeleton that lacks only forelimbs and some caudal vertebrae. This genus has three species, Y. shangyouensis, Y. magnus, and Y. hepingensis. The latter is a large form and was collected from the same beds as the former. The material consists of a complete skull, vertebrae, pelvic girdle, and hindlimbs. Recently it was referred to Sinraptor as a new species, S. hepingensis (Currie and Zhao, 1993), in a new taxon, Sinraptoridae. Stegosaurs are the most bizarre dinosaurs found in the Upper Shaximiao Formation. Since the first skeleton of stegosaurs was found, the bony back

Chinese Dinosaurs plates of these dinosaurs have vexed paleontologists; both their function and their arrangement are still being argued. Tuojiangosaurus’s bony plates are symmetrically arranged in pairs. This is similar to those of Kentrosaurus of eastern Africa. A pair of large and symmetrical bony plates lying on the shoulders is preserved in a new specimen collected from Heping, near Zigong, with a piece of skin also found from this specimen. In the summer of 1987, the CCDP worked in the Jiangjunmiao region and found a lower jaw and a series of cervical vertebrae at the Shishugou Formation. It was identified as Mamenchisaurus sinocandorum, the largest sauropod in Asia (Russell and Zheng, 1993). A nearly complete allosauroid, Sinraptor, was collected from the same horizon. It was described and named S. dongi by Currie and Zhao in 1993. This is the most complete skeleton of a theropod from China (Fig. 5).

The Early Cretaceous Psittacosaurus Fauna In Late Mesozoic times, East Asia comprised three major blocks: Siberia, north China, and south China (Lee et al., 1987). Paleomagnetic data appear to point out that all these parts of East Asia occupied the same relative latitude in the Lower Cretaceous. The landscape in the northeastern part of Asia was dominated by a plateau, including north China, Mongolia, and southern Siberia (North China Block). It was covered with alluvial plains or some large lake basins. A river, the ancient Heilongjiang, flowed west to east on this highland, where vegetation flourished (Chen, 1977). This area formed a special ecological province where there lived an endemic dinosaur assemblage (Dong, 1992; Russell, 1993). This unique group of dinosaurs (the Psittacosaurus fauna) had evolved by the early Cretaceous in northeast Asia. Several fossil birds were found in the Early Cretaceous deposits of northeast China (Sereno et al., 1990; Hou et al., 1993). These animals are collectively known as the Psittacosaurus fauna and are found in the Qingshan Formation of Shandong, the Tugulu Group of the Junggar of Xinjiang, the Zhidan Group, the Ejinhoro Formation of Ordos, and so on. This fauna contains Psittacosaurus, theropods (Kelmayisaurus, Phaedrolosaurus, and Tugulusaurus), an iguanodontid (Probactrosaurus), a stegosaur (Wuerhosaurus), a protoceratopsid (Microceratops), and a pterosaur (Dsungaripterus).

Chinese Dinosaurs


FIGURE 5 Mamenchisaurus, in China, is the analog of the sauropod Diplodocus from North America.

During the Early Cretaceous Period, there were two separate dinosaur faunas in the nonmarine deposits of north and south China, which are regarded as a separate district of the biogeographic province. The Psittacosaur–Pterosaur fauna is mainly found in the northern part of China (Dong, 1993). A dinosaur fauna from the Early Cretaceous was reported at Tebch by Bohlin in 1950. Tebch means the ‘‘black plate’’ in Mongolian because the black lava (basalt) that may be of the mid-Aptian age (110 ⫹ 0.52 Ma) lies on top of the hill, with a thickness of 1.5–3 m (Eberth et al., 1993). The dinosaur fauna consists of Prodeinodon sp. and Psittacosaurus mongoliensis. In the summer of 1990, this locality was reexamined by the CCDP. Nanshiungosaurus is a most interesting dinosaur from the Nanxiong Basin and is known from an incomplete skeleton. This animal has a special pelvis that identifies it as a therizinosaur. The ilium is low, and the anterior apophysis is well developed and extends outward. The pubis is straight, and the exterior edge is thick. The ischium is thin and plate-like and the distal end expands and is fused. The age of the Nanxiong Formation was suggested as Maastrichtian. Wannanosaurus is a small dome-headed dinosaur (pachycephalosaur), with a large supratemporal fenestra and a completely flattened cranial roof. The frontoparietal region is thick, and the external surface of the skull roof cranial bone is ornamented by small and densely distributed bony processes; the ornamentation on the temporal region is well developed. Protoceratops was the most common dinosaur discovered from Bayan Mandahu, with 66 specimens of all sizes ranging from skull lengths of 2 cm to more than 1 m. The most important and interesting discov-

ery at this locality was two mass graves of the ankylosaurid Pinacosaurus including 12 individual juveniles the size of small sheep. They were found in a nest, lying in their original positions covered by presumably windblown sand. Small theropods, including Velociraptor and Oviraptor, were collected, and many turtles, lizards, numerous nests of dinosaur eggs, and mammals were unearthed. In sedimentological features the beds are apparently similar to the Djadokhta Formation of Mongolia.

Dinosaur Eggs Dinosaur eggs from the Late Cretaceous are abundant in China. At first, only simple descriptions were published based on the outer structure of the eggshell. Recent studies, mainly by Zhao and collaborators, are based on observations of the microstructure of the eggshell. Microscopic analysis of eggshells from the Late Cretaceous of China indicate that at least 12–15

FIGURE 6 Elongatoolithus eggs from the Late Cretaceous of South China.

124 dinosaur species are represented by the eggs. We have observed eggshells from the Nanxiong Basin that are unusual, indicating eggs from the end of Cretaceous that did not hatch. The dinosaurian eggs from the Wangshi Formation of the Late Cretaceous were studied by Chow in 1951. Thereafter, Young (1953) and Zhao (1979) restudied and reclassified two groups: Spheroolithid and Elongatoolithid. These eggs were distributed in the red clays of the middle-upper part of the Wangshi Formation (Fig. 6).


References Andrews, R. C. (1932). The new conquest of Central Asia. The Amer. Mus. Nat. Hist. of Central Asia. American Museum of Nat. History, New York. pp. 1–678.

Chinese Dinosaurs Eberth, D. A., Russell, D. A., Braman, D. R., and Deino, A. L. (1993). The age of the dinosaur-bearing sediments at Tebch, Inner Mongolia, People’s Republic of China. Canadian. J. Earth Sci. 30, 2101–2106. Hou, L.-H., Zhou, Z.-H., Martin, L. D., and Feduccia, A. (1995). A beaked bird from the Jurassic of China. Nature 377, 616–618. Lee, Gidong, et al. (1987). Eastern Asia in Cretaceous, new paleomagnetic data from south Korea and a new look at Chinese and Japanese data. J. Geophys. Res. 29, 3580–3596. Louderback, G. D. (1935). The stratigraphic relations of the Junghsien fossil dinosaur in Sichuan Red Bed of China. Bull. Dept. Geol. Univ. Calif. 23, 459–466. Osborn, H. F. (1924). Three new theropoda, Protoceratops Zone, central Mongolia. Am. Mus. Nat. Hist. Novitates 144, 12 pp. Peng, G. H. (1990). A new small ornithopod (Agilisaurus louderbacki gen. et sp. nov.) from Zigong, Sichuan, China. Zigong Dinosaur Museum Newletter 2, 19–27 (in Chinese). Riabinin, A. N. (1925). A restored skeleton of a huge Trachodon amurensis nov. sp. Izvestia. Geol. Com. XLIV (1) p. 1–12.

Camp, C. C. (1935). Dinosaur remains from the Province of Szechuan. Bull. Dept. Geol. Univ. Calif. 23, 467–471.

Riabinin, A. N., (1930). Mandschurosaurus amurensis nov. gen. sp., a hadrosaurian dinosaur from the Upper Cretaceous of Amur River. Societe Paleon. de Russie, Mem. S.2 p. 1–36. Leningrad.

Currie, P. J. (1991). The Sino-Canadian Dinosaur Expeditions, 1986–1990. Geotimes 36 (4), 18–21.

Russell, D. A. (1993). The role of central Asia in dinosaurian biogeography. Canadian J. Earth Sci. 30, 2002–2012.

Currie, P. J., and Zhao, X. J. (1993). A new large theropod (Dinosauria, Theropoda) from the Jurassic of Xinjiang, People’s Republic of China. Canadian J. Earth Sci. 30, 2037–2081.

Russell D. A., and Zheng, Z. (1993). A large mamenchisaurid from the Junggar Basin, Xinjiang, People’s Republic of China. Canadian J. Earth Sci. 30, 2082–2095.

Dong, Z. M. (1973a). Cretaceous stratigraphy of Wuerho District, Dsungar (Zunggar) Basin. Memoirs of the Institute of vertebrate Paleontology and Paleoanthropology Academia Sinica 11, 1–7. Dong, Zhiming (1980). On the dinosaurian fauna and their stratigraphical distribution in China. J. Stratigraphy 4, 24–38. Dong, Z. M. (1992). Dinosaurian Faunas of China. 188 pp. Springer-Verlag, Berlin. Dong, Z. M., Russell, D. A., and Currie, P. J. (1988). The Dinosaur Project, an international cooperation program on dinosaurs. Vertebrata PalAsiatica 26, 235–240. Dong, Z. M., Currie, P. J., and Russell, D. A. (1989). the 1988 field program of the Dinosaur Project. Vertebrata PalAsiatica 27 (3), 293–295. Dong, Z. M., Zhou, S. W., and Zhang, Y. H. (1983). The dinosaurian remains from Sichuan Basin, China. Palaeontologia Sinica, No. 162 (New Series C, No. 23), pp. 1– 145., 43 plates. (In Chinese.)

Sereno, P. C., and Rao, C. G. (1992). Early evolution of avian flight and perching: new evidence from the Lower Cretaceous of China. Science 255, 845–848. Wiman, C. (1929). Die Kreide-dinosaurier aus Shantung. Pal. Sin. C 1, 1–67. Young, C. C. (1941). Gyposaurus sinenses (sp. nov.), a new Prosauropoda from the Upper Triassic Beds at Lufeng, Yunnan. Bull. Geol. Soc. China. 21, 205–252. Young, C. C. (1942). Yunnanosaurus huangi (gen. et sp. nov.), a new Prosauropoda from the red Beds at Lufeng, Yunnan. Bull. Geol. Soc. China. 22, 63–104. Young, C. C. (1951). The Lufeng Saurischian Fauna in China. Pal. Sin. C 13, 1–96. Zhao, X. J., and Currie, P. J. (1993). A large crested theropod from the Jurassic of Xinjiang, People’s Republic of China. Canadian J. Earth Sci. 30, 2027–2036. Zhen, S., Mateer, N., and Lucas, S. G. (1985). The Mesozoic reptiles of China. Studies of Chinese fossil vertebrates. Bull. Geol. Inst. Univ. Uppsala N.S. 11, 133–150.

Chinle Formation

Chinle Formation J. MICHAEL PARRISH Northern Illinois University DeKalb, Illinois, USA

The Chinle Formation is an extensive sequence of sedimentary rocks that was deposited over much of what today is the Colorado Plateau during the Late Triassic (Carnian–Norian). Consisting primarily of a series of fluvial/floodplain deposits, the rivers that deposited Chinle sediments appear to have drained into a large lake in the middle of the basin, a location that corresponds to present-day southeastern Utah (Blakey and Gubitosa, 1983). Paleoclimatological studies suggest that Chinle climates were relatively humid and marked by a monsoonal weather pattern that produced striking seasonality (Dubiel et al., 1991). The Chinle Formation is abundantly fossiliferous, with extensive records of plants, vertebrates, and invertebrates. Plants include vast deposits of petrified wood referred to the form genus Araucaryoxylon, including logs tens of meters in length. Other plant records include the horsetail Neocalamites and abundant leaf records of ferns, seed ferns, and conifers (Ash, 1986). The invertebrate record primarily consists of nonmarine molluscs, conchostrachans (clamshrimp), and rare insect fossils. Floodplain deposits are dominated by amphibious vertebrates, notably the large-headed metoposaurid amphibians and the long-snouted phytosaurian archosaurs. In more upland environments, common taxa include the armored aetosaurs and (in local concentrations) the dicynodont synapsid Placerias, large carnivorous archosaurs such as Postosuchus, and early crocodylomorphs (Long and Padian, 1986). Dinosaurs known from the Chinle include the ornithischian tooth form genus Revueltosaurus (Padian, 1990), the ceratosaurian theropod Coelophysis (Padian, 1986; Colbert, 1989), and the herrerasaurid Chindesaurus (Long and Murry, 1995). The basal units of the Chinle Formation include the Shinarump, a coarse conglomerate that is the major source for uranium ore in southeastern Utah, and the fluviolacustrine Monitor Butte Member. The main fossil-bearing parts of the Chinle Formation are con-

125 tained in a section usually referred to as the Petrified Forest Member. This member is commonly subdivided into upper and lower units, separated by an extensive sheet sandstone, the Sonsela. The pollen and fossil vertebrate records are consistent with the lower unit being Carnian in age and the upper unit Norian (Litwin et al, 1991). Overlying the Petrified Forest Member are the Owl Rock Member, a thick series of lacustrine facies, and the fluviolacustrine Church Rock Member. Lucas and Lucas (1989) have advocated merging the Chinle and DOCKUM FORMATIONS within the Chinle Group and proposed elevating many of the members to formation status. The details of this proposal have yet to be fully published, and as yet it has not been widely accepted by other workers. Others, such as Blakey and Gubitosa (1983) and Dubiel et al. (1991), maintain that the Chinle and Dockum formations were deposited in adjacent sedimentary basins at approximately the same time.

See also the following related entries: ARCHOSAURIA ● PETRIFIED FOREST ● TRIASSIC PERIOD

References Ash, S. L. (1986). Fossil plants and the Triassic Jurassic Boundary. In The Beginning of the Age of Dinosaurs (K. Padian, (Ed.), pp. 21–30. Columbia Univ. Press, New York. Blakey, R. C., and Gubitosa, R. (1983). Late Triassic paleogeography and depositional history of the Chinle Formation, southern Utah and northern Arizona. In Mesozoic Paleogeography of West Central United States (M. W. Reynolds and E. D. Dolly, Eds.), Rocky Mountain Paleogeography Symposium 2, pp. 57–76. Rocky Mountain Section, Society of Economic Paleontologists and Minerologists. Colbert, E. H. (1989). The Triassic dinosaur Coelophysis. Bull. Museum. N. Arizona 57, 1–174. Dubiel, R. F, Parrish, J. T., Parrish, J. M., and Good, S. C. (1991). The Pangean megamonsoon—Evidence from the Upper Triassic Chinle Formation, Colorado Plateau. Palaios 6, 347–370. Litwin, R. L., Traverse, A., and Ash, S. R. (1991). Preliminary palynological zonation of the Chinle Formation, Southwestern U.S.A., and its correlation to the Newark Supergroup (eastern U.S.A.). Rev. Palynol. Paleobot. 68, 269–287.

Cleveland–Lloyd Dinosaur Quarry

126 Long, R. A., and Murry, P. A. (1995). Late Triassic (Carnian and Norian) tetrapods from the Southwestern United States. New Mexico Museum Nat. History Sci. Bull. 4, 1–254. Long, R. A., and Padian, K. (1986). Vertebrate biostratigraphy of the Late Triassic Chinle Formation, Petrified Forest Natural Park, Arizona: Preliminary results. In The Beijing of the Age of Dinosaurs (K. Padian, Ed.), pp. 161–170. Cambridge Univ. Press, New York. Lucas, S. G., and Lucas, A. P. (1989). Revised Triassic stratigraphy in the Tucumcari Basin, east-central New Mexico. In Dawn of the Age of Dinosaurs in the American Southwest. (S. G. Lucas and A. P. Hunt, Eds.), pp. 150– 170. New Mexico Museum of Natural History, Albuquerque. Padian, K. (1986). On the type material of Coelophysis Cope (Saurischia: Theropoda) and a new specimen from the Petrified Forest of Arizona. The Beginning of the Age of Dinosaurs (K. Padian, Ed.), pp. 45–60. Cambridge Univ. Press, New York. Padian, K. (1990). The ornithischian form genus Revueltosaurus from the Petrified Forest of Arizona (Late Triassic: Norian; Chinle Formation). J. Vertebr. Paleontol. 10, 268–269.



Cleveland–Lloyd Dinosaur Quarry JOSHUA B. SMITH University of Pennsylvania Philadelphia, Pennsylvania, USA

Located in the Brushy Basin Member of the Morrison Formation (Tithonian), the Cleveland–Lloyd Dinosaur Quarry represents one of the most impressive dinosaur mass accumulations known. The site, designated a national landmark in 1967, is located in Emery County in east-central Utah (Fig. 1). The date of the quarry’s discovery is not known but the first elaborate

excavations occurred from 1937 to 1939 under the direction of William Stokes of Princeton University (Stokes, 1986). The quarry sits within a gray calcareous claystone, floodplain deposit capped by a micritic limestone and a volcanic ash (Bilbey, 1992, 1993). The preliminary interpretation of the paleoenvironment is a groundwater-fed wetland (Bilbey, 1992, 1993). The Cleveland–Lloyd Quarry differs from many other mass assemblages (e.g., Derstler, 1995) in that the animals found do not appear to be allochthonous, water-transported carcasses. The remains found at the quarry are mostly disarticulated, and are commonly broken. Evidence of predation or scavenging is common at Cleveland–Lloyd, and it is one of the few places where carnosaur bones are marked (Molnar and Farlow, 1990). The bones are randomly oriented, and there is no evidence of a prevalent current direction. One of the most striking features about the Cleveland–Lloyd fauna is its almost exclusively dinosaurian composition. Of these remains, 75% have been identified as the large theropod Allosaurus fragilis. Indeed, Madsen (1976) reported that A. fragilis material from at least 44 individuals had been recovered from the quarry by 1975. This is an uncommon occurrence because dinosaur predator/prey ratios are usually weighted towards the herbivores. Also, Dodson (1971) and Dodson et al. (1980) observed that in many dinosaur mass accumulations, theropods are greatly outnumbered (but see GHOST RANCH QUARRY). The stratigraphy of the quarry has been well studied (e.g., Bilbey, 1992). The common belief is that it was a predator trap, and that it may indicate gregarious behavior in Allosaurus (Richmond and Morris, 1996). Invertebrate and floral remains from the Cleveland–Lloyd Quarry are extremely rare. Madsen (1976) reported rare, disassociated charophytes (Plantae), several gyrogonites (from genera of charophyte reproductive structures), and as many as four genera of gastropods (Mollusca). The dinosaur faunal list compiled by Madsen (1976) includes the theropods A. fragilis), Ceratosaurus sp. (elements from ?one individual), Stokesaurus clevelandi (two individuals), Marshosaurus bicentesimus (two individuals); the sauropods Camarasaurus cf. lentus (two or three individuals) and elements belonging to at least one and per-

Cleveland–Lloyd Dinosaur Quarry


FIGURE 1 Locus map of the Cleveland–Lloyd Dinosaur Quarry in Emery County, Utah. (modified from Madsen, 1976).

haps several non-camarasaurid sauropods; the ornithischians Camptosaurus cf. browni (at least five individuals), Stegosaurus cf. stenops (at least two individuals), and Stegosaurus sp. (?-two or three individuals), and material related to one possible ankylosaur.


References Bilbey, S. A. (1992). Stratigraphy and sedimentary petrology of the Upper Jurassic–Lower Cretaceous rocks at Cleveland–Lloyd dinosaur quarry with a comparison to the Dinosaur National Monument quarry, Utah, pp. 313. PhD dissertation, University of Utah, Provo. Bilbey, S. A. (1993). Depositional environments in the Morrison Formation at the Cleveland–Lloyd Dinosaur Quarry, Utah. J. Vertebr. Paleontol. 13(3), 26A. Derstler, K. (1995). The Dragon’s Grave—An Edmontosaurus bonebed containing theropod egg shells and juve-

Cloverly Formation

128 niles, Lance Formation (Uppermost Cretaceous), Niobrara County, Wyoming. J. Vertebr. Paleontol. 15(3), 26A. Dodson, P. (1971). Sedimentology and taphonomy of the Oldman Formation (Campanian), Dinosaur Provincial Park, Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol. 10, 21–74. Dodson, P., Behrensmeyer, A. K., Bakker, R. T., and McIntosh, J. S. (1980). Taphonomy and paleoecology of the dinosaur beds of the Jurassic Morrison Formation. Paleobiology 6(2), 208–232. Madsen, J. H., Jr. (1976). Allosaurus fragilis: A revised osteology. Utah Geol. Survey Bull. 109, 163. Miller, W. E., Horrocks, R. D., and Madsen, Jr., J. H. (1996). The Cleveland–Lloyd Dinosaur Quarry, Emery County, Utah: A U.S. Natural Landmark. Brigham Young University Geological Studies 41, 3–24, four maps. Molnar, R. E., and Farlow, J. O. (1990). Carnosaur paleobiology. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 210–224. Univ. of California Press, Berkeley. Richmond, D. R., and Morris, T. H. (1996). The dinosaur death-trap of the Cleveland–Lloyd Quarry, Emery County, Utah. Mus. N. Ariz. Bull. 60, 533–545. Stokes, W. L. (1986). The geology of Utah. Utah Museum of Natural History Occasional Papers, No. 6.

Cleveland Museum of Natural History, Ohio, USA see MUSEUMS



Cloverly Formation W. DESMOND MAXWELL New York College of Osteopathic Medicine Old Westbury, New York, USA

The Lower Cretaceous Cloverly Formation (Aptian– Albian) is exposed in the Bighorn Basin of southcentral Montana and northern Wyoming. It has an average thickness of approximately 90 m and consists of predominantly variegated claystones with channel-filling sandstones and conglomeratic sandstones (Moberly, 1962). The claystones represent soils resulting from fluvial overbank deposits and pyroclastic ash weathered in hot climates in seasonal swamps and lakes. The top of the Cloverly is marked by a transgressive marine facies, deposited by a south-

ward-encroaching inland sea (Moberly, 1962; Young, 1970). The terrestrial deposits of the Cloverly have yielded relatively few, but highly significant, dinosaurian remains. Most significantly perhaps are the remains of the raptorial, pack-hunting dromaeosaurid Deinonychus antirrhopus (Ostrom, 1969, 1990; Maxwell and Ostrom, 1995). Other important members of the dinosaurian fauna include the basal ornithopod Tenontosaurus tilletti, the nodosaurid Sauropelta edwardsi, and the small theropod Microvenator celer (Ostrom, 1970). Recent discoveries at a microsite include several varieties of dinosaurian eggshell and the only neonate dinosaurian remains known from any Lower Cretaceous deposit (Maxwell and Horner, 1994). Continued collecting at the site has produced new varieties of dinosaurian eggshell and additional remains of neonates, including the teeth of neonate dromaeosaurid and sauropod individuals. The Cloverly fauna is a transitional one, mediating the sauropod- and stegosaur-dominated faunas of the Late Jurassic and the hadrosaur- and ceratopsiandominated faunas of the Late Cretaceous. Further discoveries in the Cloverly and other Lower Cretaceous formations may help us to better understand the evolutionary mechanics of the Jurassic– Cretaceous transition in western North America.

See also the following related entries: CEDAR MOUNTAIN FORMATION ● CRETACEOUS PERIOD

References Maxwell, W. D., and Horner, J. R. (1994). Neonate dinosaurian remains and dinosaurian eggshell from the Lower Cretaceous Cloverly Formation of south-central Montana. J. Vertebr. Paleontol. 14(1), 143–146. Maxwell, W. D., and Ostrom, J. H. (1995). Taphonomy and paleobiological implications of Tenontosaurus– Deinonychus associations. J. Vertebr. Paleontol. 15(4), 707–712. Moberly, R., Jr. (1962). Lower Cretaceous history of the Bighorn Basin, Wyoming. In Wyoming Geological Association Guidebook, 17th Annual Field Conference (R. L. Enyert and W. H. Curry, III, Eds.), pp. 94–101. Wyoming Geological Association. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Bull. Peabody Museum Nat. History 30, pp. 165. Ostrom, J. H. (1970). Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Big-

Coelurosauria horn Basin area, Wyoming and Montana. Bull. Peabody Museum Nat. History 35, pp. 234. Ostrom, J. H. (1990). The Dromaeosauridae. The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 269–279. Univ. of California Press, Berkeley. Young, R. G. (1970). Lower Cretaceous of Wyoming and the southern Rockies. In Wyoming Geological Association Guidebook, 22nd Annual Field Conference. (R. L. Enyert, Ed.), pp. 147–160. Wyoming Geological Association.

Coelurosauria JOHN R. HUTCHINSON KEVIN PADIAN University of California Berkeley, California, USA

Coelurosauria (‘‘hollow-tailed reptiles’’) (Fig. 1) was a name coined by F. von Huene (1914, 1920, 1926), based on the genus Coelurus, to include a variety of

129 small, lightly built theropod dinosaurs. It was distinguished from CARNOSAURIA, larger forms with great skulls and enormous teeth. Since that time both terms have been used in variously formal and informal senses, but it has often been thought that the large forms comprised a more or less natural group, the lineage beginning with Early Jurassic forms such as Dilophosaurus and extending through the Late Jurassic Ceratosaurus, Megalosaurus, and Allosaurus and through the Late Cretaceous tyrannosaurs (see THEROPODA). Despite occasional misgivings to the contrary, such as Ostrom’s (1969) description of the maniraptoran Deinonychus, which hinted at the possible artificiality of the small coelurosaur—large carnosaur dichotomy, coelurosaurs have generally been considered an informal taxon of small theropods ranging from the Late Triassic through the Late Cretaceous. However, Gauthier (1986) redefined Coelurosauria in cladistic terms as a stem-based taxon comprising birds and all theropods closer to birds than to Carnosauria. Hence, Coelurosauria and Carnosauria were sister

FIGURE 1 Phylogeny of the Coelurosauria and various outgroups; after Holtz (1994, 1995, 1996) and other sources, courtesy T. Rowe and L. Dingus.

130 taxa within what Gauthier (1986) called TETANURAE, a stem-based taxon. Holtz (1994) formalized AVETHEROPODA as the node-based taxon within Tetanurae comprising Coelurosauria and Carnosauria. Gauthier (1986) listed ORNITHOMIMIDAE, Compsognathus, Ornitholestes, Coelurus, Microvenator, Saurornitholestes, Hulsanpes, ELMISAURIDAE, Caenagnathidae, DEINONYCHOSAURIA, and AVIALAE (Archaeopteryx ⫹ Aves) as component taxa of Coelurosauria. To these can be added THERIZINOSAUROIDEA, now generally considered close to OVIRAPTORIDAE (Russell and Dong, 1993; Holtz, 1994, 1996a,b; Clark et al., 1995). Novas (1992) realized that tyrannosaurs did not belong in Carnosauria, and an analysis by Holtz (1994) indicated that they were the sister group to BULLATOSAURIA (TROODONTIDAE ⫹ ORNITHOMIMIDAE) within Coelurosauria. Hence tyrannosaurs, despite their great size, are not carnosaurs but rather coelurosaurs, as von Huene first recognized. Gauthier’s (1986) analysis did not recognize specific subgroups within Coelurosauria, except to link DEINONYCHOSAURIA with AVIALAE (Archaeopteryx ⫹ AVES in his formulation; Aves of most other workers). Gauthier’s definition of Deinonychosauria (Colbert and Russell, 1969) included Troodontidae and Dromaeosauridae, but troodontids are now regarded as the sister taxon to ornithomimids; therefore, Deinonychosauria could be retained as the node corresponding to Holtz’s (1994) unnamed ‘‘node 11.’’ However, Holtz (1996b), in reanalyzing the nomenclature, called this node MANIRAPTORIFORMES. Holtz’s (1994, amended 1996b) analysis of the phylogenetic relationships among the major subclades of Theropoda found that within Coelurosauria, Dromaeosauridae and Avialae form a monophyletic group called MANIRAPTORA [following Gauthier’s (1986) usage; Holtz 1996b non 1994]; a second monophyletic group (ARCTOMETATARSALIA) is formed by Ornithomimosauria ⫹ Troodontidae (Bullatosauria) ⫹ TYRANNOSAURIDAE; and OVIRAPTORIDAE, Caenagnathidae, and THERIZINOSAUROIDEA may form a third monophyletic group (designated OVIRAPTOROSAURIA by Russell and Dong, 1993). Ornitholestes and Compsognathus appear to be outgroups to all these taxa within Coelurosauria (Fig. 1). Recent analyses have also placed the enigmatic Late Cretaceous ‘‘carnosaur’’ Dryptosaurus as a basal coelurosaur (Denton, 1990; Holtz, 1996a), although this is strange given its late appearance. Other poorly known taxa, such as

Coelurosauria the recently discovered Bagaraatan (Osmo´lska, 1996) and Deltadromeus (Sereno et al., 1996) and the long known but poorly understood Coelurus, may be other basal (i.e., nonmaniraptoriform) coelurosaurs, members of other less inclusive clades within Coelurosauria, or noncoelurosaurian tetanurines (see also Norman, 1990). Coelurosauria is supported as a node by synapomorphies including an ischium reduced to two-thirds or less the length of the pubis, the loss of the ischial foot, an expanded circular orbit, a triangular obturator process on the ischium, an ascending process of the astragalus that is greater than one-fourth the length of the epipodium, and 15 or fewer caudal vertebrae bearing transverse processes (Holtz, 1994). Apart from Compsognathus, the other coelurosaurians share an ulna that is bowed posteriorly, a long and slender metacarpal III, a posterodorsal margin of the ilium that curves ventrally in lateral view, flexed cervical zygapophyses, a jugal expressed on the rim of the antorbital fenestra, and a first metacarpal that is one-third or less as long as the third (Holtz, 1994). Coelurosaurians apart from Compsognathus (Fig. 2) and Ornitholestes (Fig. 3), which Holtz (1996b) termed Maniraptoriformes, share a distally placed obturator process on the ischium, a third antorbital fenestra, and elongated anterior cervical zygapophyses (Holtz, 1994). Dromaeosauridae ⫹ Avialae form Maniraptora, and Troodontidae ⫹ Ornithomimosauria form Bullatosauria. Avimimus is removed from the analysis because it may be a chimera (Holtz, 1996a,b). Phylogenetic analysis of the coelurosaurian groups must be regarded as substantially in flux. This is the result of several factors, including considerable homoplasy, the uncertain identity or association of some taxa (e.g., Avimimus), incomplete knowledge of some groups (e.g., Elmisauridae), differences in coding and polarizing characters and character states, and the programs and options used in computerized phylogenetic analysis. Nevertheless, considerable strides have been made by recent explorations and systematic analyses. The hypothesis that birds are coelurosaurs most closely related to dromaeosaurs has been sustained, ornithomimids and oviraptors have been shown to be not especially closely related within Coelurosauria, therizinosauroids appear to be coelurosaurs and not aberrant ornithischians or basal sauropodomorphs, and tyrannosaurs are closely related to ornithomimids and troodontids. It can be expected



FIGURE 2 Compsognathus, after Ostrom (1978) with permission.

that future discoveries and analyses will change some details of the phylogeny as presented here. Coelurosaurs, as small theropods, have traditionally been considered active, agile bipeds, feeding primarily on small tetrapods, sometimes including those of their own species. These dinosaurs show virtually all the features classically considered to demonstrate cursoriality (Coombs, 1978). The femur is shorter than the tibia and, as in all dinosaurs, is set off from the pelvic girdle by a rounded head placed at 90⬚ to the femoral shaft. The knee is a hinge, as is the ankle, and the lower extremities (lower leg bones, metatarsals, and phalanges) are particularly elongated. The fibula is reduced and strap-like and does not move against the tibia; nor is there substantial movement between the astragalus and calcaneum nor between these proximal tarsals and the leg. The proximal tar-

sals and distal tarsals (which form a cap to the metatarsals) articulate to create a mesotarsal hinge joint. These features are characteristic not only of basal coelurosaurs but also of basal theropods, dinosaurs, and even ornithodirans (see BIPEDALITY; FUNCTIONAL MORPHOLOGY; ORNITHODIRA; ORNITHOSUCHIA). In this way, the femur’s rotation is generally around a subhorizontal orientation, and during locomotion it is elevated and depressed through an angle of perhaps 30–45⬚. The tibia, meanwhile, swings in a wide anteroposterior arc, as do the bones of the foot, and it is distal to the knee that most of the excursion of the stride takes place. This is why, for example, ornithomimids are considered to have been ostrich-like, rapid runners, because their lower leg bones and metatarsals are so long. Conversely, Deinonychus has an almost surprisingly robust hindlimb with a low

FIGURE 3 Ornitholestes, after Norman (1990).

132 tibia:femur ratio; Ostrom (1969) associated this anomaly with the function of the reverse-jointed second toe and hypertrophied claw in attacking prey rather than in maximizing running ability. Among ornithischian dinosaurs, only some basal forms, such as Lesothosaurus and Heterodontosaurus, are comparable to theropods in the cursorial ratios of their hindlimb elements. ORNITHISCHIANS took refuge, as did sauropodomorphs, not in their running ability but in large size, which made them more formidable prey and probably also facilitated the digestive processing of the plants that they ate. Several ornithischian clades also show evidence of group behavior, which would have made individual members more difficult targets (see BEHAVIOR). There is some evidence that some coelurosaurs, such as Deinonychus, foraged in packs (Ostrom and Maxwell, 1996), which may have given them some advantage over large prey. Coelurosaurs are generally considered small theropods, and certainly most known Mesozoic coelurosaurs were small (under 2 m in length); most living coelurosaurs—i.e., birds (see BIRD ORIGINS)—are even smaller, seldom exceeding 1 m in length, and mostly 10–15 cm long. However, as shown previously, not all coelurosaurs were small. Whether one accepts Eoraptor and herrerasaurids as basal theropods, as dinosauriformes close to the origin of dinosaurs, or at some phylogenetic level in between (see HERRERASAURIDAE), it is clear that these forms are not large representatives of carnivorous dinosaurs (Eoraptor has a skull just over 10 cm long, and that of Herrerasaurus is about three times as long). Ceratosaurs, as currently understood, are small when they first appear in the fossil record (e.g., Coelophysis and Syntarsus), and the first well-known large forms are the Early Jurassic Dilophosaurus and the Late Jurassic Ceratosaurus (which, enigmatically, appears to be the sister taxon of the Cretaceous ABELISAURIDS as well as phylogenetically the most basal member of Ceratosauria; see Rowe and Gauthier, 1990). A great number of small problematic theropod remains, usually grouped in Coelurosauria but actually too indeterminate or basal to qualify, are known from the Late Triassic (Norman, 1990), but no very large forms are known. The true coelurosaurs, then, began at small size, as exemplified by their basal members Compsognathus, Coelurus, and Ornitholestes, all from the Late Jurassic.

Coelurosauria However, large forms evolved several times within the lineage. Ornithomimids, such as Harpymimus, Gallimimus, and Struthiomimus, reached lengths of 3–5 m, with skulls up to 30 cm long (see ORNITHOMIMIDAE). The giant Deinocheirus, represented only by a shoulder girdle and forelimb, has a variety of apparently general theropod characters that make it difficult to classify but has the three subequal metacarpals characteristic of ornithomimids (see Norman, 1990, 292); if it is indeed a member of this group, or of Coelurosauria, it is certainly on the large end of the size range. Tyrannosaurids, now reclassified as coelurosaurs, are also among the largest, if not the largest known theropods. The skull of Deinonychus is approximately 30 cm long, so it must be regarded as at least a mediumsized carnivore, and Utahraptor must have been much larger. Dryptosaurus, whose precise systematic position is uncertain, is now regarded as a basal coelurosaur (Denton, 1990); it is very large, but it occurs in the latest Cretaceous, so it would probably be unwise to accord its size much influence in considering the state of basal coelurosaurs in the Late Jurassic, when the group is first known. Given the outgroup comparisons previously noted, Dryptosaurus and all the other large coelurosaurs are almost certainly secondarily large. Finally, as noted previously, coelurosaurs are the only dinosaurian clade that did not become extinct by the end of the Cretaceous. Its surviving subclade, the birds, first appear in the Late Jurassic (Archaeopteryx) and so are known as far back in the fossil record as any group of coelurosaurs, as Gauthier (1986) noted. Intimations of their avian relationships have been provided dramatically with the discovery that at least some obviously nonavian coelurosaurs apparently had a type of feathered covering; this further suggests that the origin of feathers did not evolve directly for the purpose of flight, and whether thermoregulation, display, or a related function was the selective force for the evolution of feathers, the behavioral implication is clear that coelurosaurs were and are different from other, nonornithodiran reptiles, then and now (see AVES; BIRD ORIGINS; FEATHERED DINOSAURS).



References Colbert, E. H., and Russell, D. A. (1969). The small Cretaceous dinosaur Dromaeosaurus. Am. Museum Novitates 2380, 1–49. Coombs, W. P., Jr. (1978). Theoretical aspects of cursorial adaptations in dinosaurs. Q. Rev. Biol. 53, 393–418. Denton, R. K., Jr. (1990). A revision of the theropod Dryptosaurus (Laelaps) aquilunguis (Cope 1869). J. Vertebr. Paleontol. 9(Suppl. to No. 3), 20A. Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117.

133 Sereno, P. C., Duthiel, D. B., Iarochene, M., Larsson, H. C. E., Lyon, G. H., Magwene, P. W., Sidor, C. A., Varricchio, D. J., and Wilson, J. A. (1996). Predatory dinosaurs from the Sahara and Late Cretaceous faunal differentiation. Science 272, 986–991. von Huene, F. (1914). Das naturliche System der Saurischia. Zentralblatt Mineral. Geol. Pala¨ontol. B 1914, 154–158. von Huene, F. (1920). Bemerkungen zur Systematik und Stammesgeschichte einiger Reptilien. Zeitschrift Indukt. Abstammungslehre Vererbungslehre 24, 162–166. von Huene, F. (1926). The carnivorous Saurischia in the Jura and Cretaceous formations, principally in Europe. Revista Museo de La Plata 29, 35–167.

Holtz, T. R., Jr. (1995). A new phylogeny of the Theropoda. J. Vertebr. Paleontol. 15(Suppl. to No. 3), 35A. Holtz, T. R., Jr. (1996a). Phylogenetic analysis of the nonavian Tetanurine dinosaurs (Saurischia, Theropoda). J. Vertebr. Paleontol. 16(Suppl. to No. 3), 42A. Holtz, T. R., Jr. (1996b). Phylogenetic taxonomy of the Coelurosauria (Dinosauria: Theropoda). J. Paleontol. 70, 536–538. Norman, D. B. (1990). Problematic Theropods: ‘‘Coelurosaurs’’. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 280–305. Univ. of California Press, Berkeley. Novas, F. E. (1992). La evolucio´n de los dinosaurios carnivo´ros. In Los Dinosaurios y su Entorno Biotico. Actas II, Curso de Paleontolo´gia en Cuenca (J. L. Sanz and A. Buscalione, Eds.), pp. 123–163. Instituto ‘‘Juan de Valdes,’’ Cuenca, Spain. Osmo´lska, H. (1996). An unusual theropod dinosaur from the Late Cretaceous Nemegt Formation of Mongolia. Acta Palaeontol. Polonica 41, 1–38. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Bull. Peabody Museum Nat. History 30, 1–165. Ostrom, J. H. (1978). The osteology of Compsognathus longipes. Zitteliana 4, 73–118.

Coevolution see PLANTS



College of Eastern Utah Prehistoric Museum, Utah, USA see MUSEUMS AND DISPLAYS

Colombian Dinosaurs Unidentified theropod dinosaurs have been excavated from the ‘‘Ortega’’ formation in Colombia.


Ostrom, J. H. (1990). Dromaeosauridae. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 269–279. Univ. of California Press, Berkeley. Ostrom, J. H., and Maxwell, W. D. (1996). J. Vertebr. Paleontol. 16. Rowe, T. R., and Gauthier, J. A. (1990). Ceratosauria. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 151–168. Univ. of California Press, Berkeley. Russell, D. A., and Dong, Z. (1993). The affinities of a new theropod from the Alxa Desert, Inner Mongolia, People’s Republic of China. Can. J. Earth Sci. 30, 2107– 2127.

Coloniality There is some evidence that some dinosaurs were at least periodically, and perhaps usually, colonial (i.e., that they congregated in large, cooperative groups). See BEHAVIOR entry for a fuller treatment of this issue and related inferences.



Color MICHAEL J. RYAN Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

ANTHONY P. RUSSELL University of Calgary Calgary, Alberta, Canada

The fossil record of life on earth consists of copious amounts of preserved hard tissues and relatively few records of soft tissues. For vertebrates, bones and teeth comprise by far the greatest component of the record, with soft tissues, including the integument, playing a minor role. In vertebrates the skeleton, whether it is relatively superficial (such as the bony scales of many fish) or deep (such as the long bones of the limbs and the ribs), is overlain by soft tissues and therefore is normally unpigmented. For vertebrates, then, we rely on superficial tissues, especially the SKIN, to reveal color and color patterns. This, in itself, poses two major problems because soft tissues, including the skin, typically do not preserve well, and pigments that occur in biological tissues are generally not stable and thus do not normally withstand the rigours of fossilization. The geochemistry of fossilized pigments and their derivatives is poorly known and the material relatively rare. Almost all the information that we have on pigments of extinct organisms comes from mollusks, although evidence is also available for brachiopods, trilobites, crinoid echinoderms, and insects. Most animal pigments are soluble in water and thus are prone to rapid disappearance after death (Parsons and Brett, 1991). Melanins (pigments responsible for brown and black colors) are generally much less soluble, however, and thus have a better record of preservation. The more recent the fossil the better the chances of the preservation of any pigments. Although our evidence of pigments for mollusks is better than that for other groups and dates back to the Permian, most examples are known from remains no older than the Tertiary. From what we do know of the preservation of pigmentation, it is dependent upon the original pigment composition (i.e., stability—most are unstable),

Color the location in the body (hard parts vs soft parts), the mineralogy of the hard parts if the pigments are preserved there, the rapidity of burial, and the nature of the subsequent fossilization process. If alteration of pigments occurs, color patterns (as opposed to true color preservation) may be evident. Thus, light/dark patterns may be evident, as in a specimen of the trilobite Anomocare vittata (Raymond, 1922) from the Ordovician, which was preserved with three transverse, alternating light/dark bands on its pygidium. Extensive evidence of color pattern preservation in fossil mollusks is documented by Hollingsworth and Barker (1991). The ink sacs of dibranchiate cephalopods are well known for Jurassic seriopods from the Holzmaden and Solhofen deposits of Germany (Boucot, 1990). Preserved color patterns are known for a variety of terrestrial fossil insects, including elaborate wing patterns with eye spots (Jarzembowski, 1984). For fossil vertebrates, a cryptic color pattern is preserved from an amphibian of the Upper Carboniferous (Lund, 1978), with chromatophores known from frogs from the Jurassic (Hecht, 1970) and the Eocene (Voight, 1935). Grande (1982) and others have described chromatophores for Eocene fish. The radius and ulna of an ornithomimid from the late Campanian of Alberta possesses short, black markings that are regularly spaced parallel and sub-parallel to the long axes of these bones (M. Ryan and A. Russell, personal observation). Their origin and association to the dinosaur are unknown. A small, feathered compsognathid dinosaur from China (Late Jurassic–Early Cretaceous) exhibits a black discoloration on both sides of the preserved sclerotic ring. A bird feather from the Eocene Green River Shale of Colorado preserves a color pattern (Lambrecht, 1933). For the reasons outlined previously, color preservation in vertebrates is extremely rare. Holman and Sullivan (1981) described dark ocular markings on the turtle Chrysemys from the Miocene. Sullivan et al. (1988) report the preservation of the original color pattern of black spots against a reddish-brown matrix in the preserved keratinous epidermal scales of the early Paleocene turtle Neurankylus. One of the most frequently asked questions by students of dinosaur paleontology is ‘‘what color were dinosaurs?’’ The simple answer is ‘‘ we don’t know’’ and probably never will because of the chemical composition of such pigments and their solubility,

Como Bluff and because of the extreme unlikelihood that skin will be preserved at all, let alone with the colors intact. Such eventualities mean that speculation, based on what we know of pigmentation in living vertebrates, is our chief vehicle for understanding color and color patterns in dinosaurs. Dinosaur artists from the first half of the century (e.g., Charles Knight and, to a lesser degree, Zdenek Burian and Rudolph Zallinger) typically painted dinosaurs in drab green, browns, or grays. These artists would have based their interpretations on modernday analogs such as large mammals (elephants and hippopotamus) and reptiles [alligators, crocodiles, and large varanid lizards (e.g., the Komodo Dragon, Varanus komodensis)]. These animals usually have subdued colors over the majority of their bodies, although they often exhibit countershading and some spots of other colors. Recent artists (Mark Hallett, Greg Paul, William Stout, and many others) have expanded their pallet and presented us with brightly colored and patterned dinosaurs reflecting the relatively recent rethinking of dinosaurs as energetic, sociable animals. Modern vertebrates display a wide variety of color patterns in their integument (skin, scales, feathers, and hair). Based on preserved fossil skin impressions for dinosaurs (primarily large dinosaurs) of the Mesozoic, we know their skin was usually covered in non-bony scales of a variety of sizes, which abutted each other in a mosaic-like pattern. Modern reptiles, especially lizards and snakes, possess a wide variety of colors and color patterns. There is nothing, in theory, precluding dinosaurs from being similarly endowed. Colors and color patterns can serve a number of functions in modern reptiles including use in social interactions, thermoregulation, aposematism (color patterns that signal a noxious or otherwise dangerous quality of a prey to a potential predator, such as that found in the coral snake), and defense (cryptic or mimicking coloration or for suddenly revealed bright colors for startling potential predators) (see Pough, 1988, and references therein). Dinosaurs probably used coloration patterns in these ways as well, although which species possessed which qualities will probably never be known with certainty.

See also the following related entry: SKIN


References Boucot, A. J. (1990). Evolutionary Paleobiology of Behaviour and Coevolution, pp. 725. Elsevier, Amsterdam. Carpenter, F. M. (1971). Adaptations among Paleozoic insects. Proc. North Am. Paleontol. Congr. II, 1236–1251. Grande, L. (1982). A revision of the fossil genus Knightia, with a description of a new genus from the Green River Formation. Am. Museum Novitiates, 2731. Hecht, M. K. (1970). The morphology of Eodiscoglossus, a complete Jurassic frog. Am. Museum Novitiates, 2227. Hollingsworth, N. T. J., and Barker, M. J. (1991). Color pattern preservation in the fossil record: Taphonomy and diagenetic significance. In The Processes of Fossilization (S. K. Donovan, Ed.) pp. 105–119. Columbia Univ. Press, New York. Holman, J. A., and Sullivan, R. M. (1981). A small herpetofauna from the type section of the Valentine Formation (Miocene: Barstovian), Cherry County, Nebraska. J. Paleontol. 55, 138–144. Jarzembowski, E. A. (1984). Early Cretaceous insects from southern England. Mod. Geol. 9, 71–94. Lambrecht, K. (1933). Handbuch der Pala¨oornithologie, pp. 1024. Borntraeger, Berlin. Lund, R. (1978). Anatomy and relationships of the Family Phlegethontiidae (Amphibia, Aistopoda). Ann. Carnegie Museum 47, 53–79. Pough, F. H. (1988). Antipredator mechanisms in reptiles. Biol. Reptilia 16, 1–152. Raymond, P. E. (1922). A trilobite retaining color markings. American Journal of Science 4, 461–464. Sullivan, R. M., Lucas, S. G., Hunt, A. P., and Fritts, T. H. (1988). Color pattern on the selmacryptodiran turtle Neurankylus from the early Paleocene (Puercan) of the San Juan Basin, New Mexico. Contrib. Sci., Nat. History Museum Los Angles County 401. Voight, E. (1935). Die Erhaltung von Epithelzellen mit Zellkeren, von Chromatophoren und Corium in fossiler Froschhaut aus der mittel eoza¨nen Braunkohle des Geiseltales. Nova Acta Leopoldina, 3, 339–360.

Como Bluff BRENT H. BREITHAUPT University of Wyoming Laramie, Wyoming, USA


ne of the most renowned dinosaur collecting areas in the world is Como Bluff in the northern part of the Laramie Basin, in southeastern Wyoming. Como

136 Bluff is a roughly east–west trending anticline plunging to the west, with a gently dipping southern limb and a steeply dipping northern limb. Como Bluff is approximately 15 km long and 1.5 km wide. The southern limb (Como Ridge) of this breached anticline is capped by a highly resistant sandstone of the Cretaceous Cloverly Formation. Beautiful exposures of the Upper Jurassic Morrison Formation crop out beneath the Cloverly Formation cap rock on the north face of Como Ridge. The marine Jurassic Sundance Formation and some of the Triassic red beds are also exposed. The south face is a slightly vegetated, gentle slope of Cretaceous strata (e.g., Cloverly Formation, Mowry Shale, Muddy Sandstone, and Thermopolis Shale). Como Bluff derives its name from the spring-fed Lake Como at the western end of Como Bluff. The lake was given this name by early surveyors who thought that this 1.5-km-long body of water resembled the famous Lake Como (Lago de Como) in the Italian Alps (Urbanek, 1988). In 1877, Professor Othniel Charles Marsh at Yale University received a letter from Messrs. Harlow and Edwards, identified as Union Pacific Railroad workers in the Wyoming Territory. They reported finding numerous large bones of a ‘‘Megatherium’’ from ‘‘Tertiary Period’’ units near Laramie City, Wyoming Territory. They claimed to have found a 142-cm-long shoulder blade and a 25-cm-long vertebra. A few months after their first letter, Marsh received the bones in Connecticut. Marsh sent his field assistant, Samuel W. Williston, to check the site. Williston wrote back to Marsh of a long exposure containing dinosaur bones in an amount and diversity far exceeding any site that had yet been found. As Marsh had suspected, the bones were from the same fossiliferous formation on which he had crews working in Colorado. Marsh transferred his crews to Wyoming Territory to begin collecting later that same year. Williston discovered that the two men who wrote to Marsh were the station agent and section foreman of the Union Pacific Railroad Como Station, approximately 90 km north of Laramie. Their full names, which they had not used to keep their find secret, were William Harlow Reed and William Edward Carlin. The value of prehistoric animal remains and the importance of keeping sites secret were the result of Marsh’s already legendary confrontations with archrival fossil collector Edward Drinker Cope.

Como Bluff Although Marsh’s crews from Yale University did most of the collecting at Como Bluff from 1877 to 1889, Cope’s field crews also opened quarries there at that time. The Marsh–Cope rivalry extended into the field operations of the two rival camps (Ostrom and McIntosh, 1966; Colbert, 1968; Wilford, 1986). Reed and Carlin continued to work for the railroad but also helped Marsh’s crews excavate fossil bones. Although Carlin eventually would work for Cope and then quit the fossil collecting business altogether, Reed was a devoted and loyal employee to Marsh. In fact, Reed became so enamored with discovering dinosaurs that he would eventually work full time as a fossil collector (Breithaupt, 1990). Some of the first published reconstructions of a sauropod (the skeleton of Brontosaurus, now known as Apatosaurus) were based on a fairly complete skeleton found by Reed at Como Bluff ’s Quarry 10 (Marsh, 1883, 1891). These famous reconstructions, in which Camarasaurus elements were used to replace various missing parts, including the head, dramatically affected views of this dinosaur for almost a century. William Harlow Reed was born in Hartford, Connecticut, in 1848. In 1877, while working as a section foreman for the Union Pacific Railroad at Como Station, Wyoming Territory, he accidentally discovered large bones on the nearby ridge. These specimens launched him in a career in vertebrate paleontology that he would pursue for the next 38 years. Although frustrated by certain aspects of field-work and by lack of recognition as a field paleontologist, he was a diligent and loyal collector for Marsh. He gave this same dedication in later years to W. C. Knight at the University of Wyoming, W. J. Holland at the Carnegie Museum, and W. Granger at the American Museum of Natural History. Although not formally educated in the sciences, Reed’s desire to learn, interest in natural phenomena, and association with the notable paleontologists of his time allowed him to gain a background in geology and paleontology. After more than 25 years of significant discoveries of dinosaurs, ichthyosaurs, plesiosaurs, pterosaurs, mammals, and cycads in Wyoming, Reed was given the position as curator of the museum and instructor in geology at the University of Wyoming in 1904. He held this position until his death in 1915. In 1897, the American Museum of Natural History sent collecting crews to Como Bluff after Marsh and

Computers and Related Technology Cope’s crews had left the area. The American Museum crews, under the direction of Jacob Wortman and later Walter Granger (Professor Henry Fairfield Osborn’s chief collectors) found some new dinosaur quarries, but most of the localities at Como Bluff were barren of fossils. Major collections were made for the American Museum from Como Bluff and the region around this famous site between 1897 and 1905 (Colbert, 1968; McIntosh, 1990). By the turn of the century, museum crews from the Carnegie Museum and the University of Wyoming were also making collections in the area. Although by the early 1900s most collecting in the Como Bluff area had ceased, museum crews continue to routinely excavate dinosaur bones and small, fossil vertebrate material from the region (Bakker, 1985, 1990). Since the discovery of Como Bluff in 1877, thousands of vertebrate fossils have been recovered by institutions throughout the country. This extensive collecting has resulted in one of the best known terrestrial faunas in the world, ranging from bivalves to pterosaurs. The fossil vertebrate fauna from Como Bluff is dominated by large herbivorous dinosaurs. Five types of sauropods (e.g., Apatosaurus, Diplodocus, Camarasaurus, Pleurocoelus, and Barosaurus) have been identified from Como Bluff (Ostrom and McIntosh, 1966). Other herbivorous dinosaurs that inhabited the region during the Late Jurassic were the armored Stegosaurus and the bipedal Camptosaurus, Laosaurus, Othnielia, Drinker, and Dryosaurus. The carnivores Allosaurus, Ceratosaurus, Ornitholestes, and Coelurus were also present. Many smaller animals have been found fossilized at Como Bluff as well. These include lungfishes, frogs, salamanders, turtles, lizards, rhynchocephalians, crocodiles, pterosaurs, and a diverse group of early mammals. Although early studies interpreting the depositional environment of the Morrison Formation varied, current research suggests that the formation represents a complex fluviatile–lacustrine floodplain deposited in a seasonally wet–dry, warm-temperate environment (Dodson et al., 1980).

See also the following related entries: BONE CABIN QUARRY ● MORRISON FORMATION

References Bakker, R. T. (1985). Dinosaur Heresies, pp. 481. Morrow, New York.

137 Bakker, R. T. (1990). A new latest Jurassic vertebrate fauna from the highest levels of the Morrison Formation at Como Bluff, Wyoming. Hunteria 2, 1–19. Breithaupt, B. H. (1990). Biography of William Harlow Reed: The story of a frontier fossil collector. Earth Sci. History 9, 6–13. Colbert, E. H. (1968). Men and Dinosaurs, pp. 283. Dutton, New York. Dodson, P., Behrensmeyer, A. K., Bakker, R. T., and McIntosh, J. S. (1980). Taphonomy and paleoecology of the dinosaur beds of the Jurassic Morrison Formation. Paleobiology 6, 208–232. Marsh, O. C. (1883). Principal characters of American Jurassic dinosaurs. Part VI: Restoration of Brontosaurus. Am. J. Sci. 26, 81–85. Marsh, O. C. (1891). Restoration of Triceratops and Brontosaurus. Am. J. Sci. 41, 339–342. McIntosh, J. S. (1990). The second Jurassic dinosaur rush. Earth Sci. History 9, 22–27. Ostrom, J. H., and McIntosh, J. S. (1966). Marsh’s Dinosaurs, pp. 388. Yale Univ. Press, New Haven, CT. Urbanek, M. (1988). Wyoming Place Names, pp. 233. Mountain Press, Missoula, MT. Wilford, J. N. (1986). The Riddle of the Dinosaur, pp. 304. Knopf, New York.

Computers and Related Technology RALPH E. CHAPMAN National Museum of Natural History Smithsonian Institution Washington, DC, USA

DAVID B. WEISHAMPEL Johns Hopkins University School of Medicine Baltimore, Maryland, USA

Computers are in the process of revolutionizing the way we look at dinosaurs. Early uses of computers with dinosaurs date to the 1960’s, concentrated in research applications (e.g., data analysis) and early attempts by some museums (e.g., the National Museum of Natural History of the United States) to store their collections data electronically. These applications became more and more prevalent, and progressively more sophisticated through the 1970’s and

138 1980’s. It is during the 1990s, however, that computers finally have expanded to start assisting paleontologists in the field, during the preparation process, and in making a great difference in how dinosaurs are being presented to the general public. The application of computers and related technology to dinosaurs can be divided into five major areas: field work and specimen collection, specimen preparation, collections and data management, research, and exhibition. We will touch on these subjects in order. Finding dinosaurs in the field still continues to be done mostly in the same way it has been done for more than a century: finding areas with the right rock from the right time and the right paleoenvironments, and walking the outcrops looking for exposed bones. For the most part this will continue to be the approach taken, but new technology promises to change much of the process of getting into the field, determining where you are once there, and even analyzing what has been found and what is still in the ground. Formerly, it took much skill to use the available maps and keep track of where you were in the field; many mistakes were made and much of the old location data are inaccurate. This is changing because global positioning systems (GPS) using satellite technology can locate a position within a 100 m or less anywhere on earth and are being used by paleontologists during exploration. For example, a GPS system was used by the American Museum of Natural History in their new Gobi Desert expeditions (McKenna, 1992) to keep track of where they were in unmapped or badly mapped areas. These locality data have yet to be used extensively with geographic information systems (GIS), which combine geographic data with computer databases. In the future this will allow predictive mapping of potential outcrop areas, and the GIS will suggest where new prospecting should be done. Small-scale geographic data (e.g., quarry maps of bone orientations and positions) were once taken using compasses and were plotted in field notebooks, but now are being taken automatically using electronic distance measurement devices with millimeter accuracy (see Jorstad and Clark, 1995, for work on paleohominid applications) and, in even smaller scale and with higher accuracy, using three-dimensional digitizers (e.g., see Jefferson, 1989, on Pleistocene

Computers and Related Technology Rancho La Brea material). Finally, technology is being developed to allow paleontologists to determine, in some cases, the nature of fossils buried in an area. One example is the application of geophysical diffraction tomography by Witten, et al. (1992), who tried to determine the extent of the material buried for a specimen of the sauropod Seismosaurus. Computers and related technology have had only limited effect on specimen preparation but changes are well on the way. Standard X-rays have been used for years during specimen preparation (see Zangerl and Schultze, 1989), but more advanced imaging methods such as computed tomography (CT) are allowing, in some cases, significantly better indications of what fossil material is present in an unprepared specimen (see Clark and Morrison, 1994). In fact, the CT scanning of dinosaur eggs has become a standard operating procedure (e.g., Hirsch et al., 1989), and many fetal dinosaur fossils are now being found. It may be common in the near future for preparators to have three-dimensional models of the specimens they are preparing as an aid to the process. Specimen casting also will be changed by threedimensional computer modeling. Fossils can be digitized using various scanning or related technologies and three-dimensional reconstructions made in wax, plastic, or some other medium for exhibition or study using methods of automated casting (also known as prototyping or, in some cases, stereolithography; see Burns, 1993) thus avoiding more destructive ways of casting specimens. These casts can also be varied in scale to enable them to be viewed at a more manageable scale; very large specimens can be reduced to allow more easy manipulation and very small specimens enlarged to allow them to be viewed without a microscope. The use and development of collections materials has changed dramatically and will continue to change in the coming years due to computers. The initial transfer of collections data to an electronic format was done in the 1960’s (e.g., the National Museum of Natural History of the United States), with many now using their third, fourth, or later generation of database software, and nearly all having some electronic storage. Where data used to be kept in card catalogs that took long search times to extract simple information, data are now available in large computer databases that can be searched for very complex infor-

Computers and Related Technology mation almost instantaneously. These databases are also being stored on-line and, in many cases, are available for searching on the Internet. Another big change due to computer technology is the nature of data being made available; computer databases are not restricted to just text anymore. Image scanners are making it possible to store pictures of specimens as well as the text information that goes along with it. Furthermore, with the scanning technology being developed, three-dimensional computer images can now be stored and viewed from a variety of angles. A first attempt at storing such images was made by Rowe et al. (1993) for high-resolution CT scans of the skull of the cynodont Thrinaxodon. The CD-ROM released contains both the important descriptive literature and CT scan data that can be viewed from front-to-back, side-to-side, or top-tobottom. The use and storage of bibliographic data on dinosaurs is changing rapidly. The Bibliography of Fossil Vertebrates is being made available in electronic format and the next release of A Bibliography of the Dinosauria (Chure and McIntosh, 1989) will be made available in electronic format as well. Many dinosaur paleontologists have developed and maintained their own electronic bibliographic databases using their personal computers. Computers have had a major impact on the types of research being done in the natural sciences. The relatively unquantitative approaches taken by scientists in the early days mostly have been replaced by quantitative ones as methods have become more and more rigorous. Computer-based studies of dinosaurs are still in their infancy, however, because most of the research done on them still proceeds mostly in a qualitative fashion. This is due, to a large part, to the difficulties involved in doing research on a group represented by relative few individuals that are often incomplete and fragmentary. However, there are a number of important exceptions and we will discuss them within four major areas of research: morphometrics, mathematical and/or statistical studies of variation used to solve taxonomic or evolutionary problems; phylogenetic analyses, the analysis of the relationships among taxa; functional morphology, studies of the biomechanics and locomotion of dinosaurs; and distributional analyses, studies of the distribution of dinosaurs through time and space.

139 Morphometrics is the quantitative analysis of shape. Before the availability of computers, paleontologists were limited to analyses of two or three variables, usually within the context of allometry, the study of size and its consequences. One classic study of this type is the analysis of bivariate (two-variable) allometry in groups of ceratopsian dinosaurs by Gray (1946). Another morphometric approach, the application of D’Arcy Thompson’s (1942) transformation grids, could be done without computers and grids were generated for a number of dinosaur groups (e.g., Lull and Gray, 1949, for ceratopsians). Computers allowed calculations to be done much faster than ever before, which opened the door for multivariate analyses, those using many variables simultaneously (e.g., Dodson, 1975, 1976; Chapman, et al., 1981), as well as very sophisticated geometric methods of shape analysis (e.g., Chapman, 1990). These methods have led to a much better understanding of growth in dinosaurs, have allowed sexual dimorphs to be described in some cases (e.g., see Chapman, et al., 1981, for the pachycephalosaurid Stegoceras), and have started to be used more within studies of phylogeny and functional morphology. The next step will be to do even more sophisticated analyses of shape in three-dimensions and use highlevel computer graphics to show the results. Phylogenetic Analyses try to determine relationships among the dinosaur taxa being studied. Here, we will concentrate on numerical cladistic analyses, which attempt to reconstruct these relationships on trees, called cladograms, using the principle of parsimony; looking for the shortest trees by minimizing the number of evolutionary steps needed to generate the tree and minimizing instances of convergence or parallelism. Originally done by hand, most cladistic analyses of dinosaurs, as well as all other groups of organisms, are carried out on computers using programs such as Phylogenetic Analysis Using Parsimony (PAUP) (Swofford and Begle, 1993) and MacClade (Maddison and Maddison, 1992). These packages search for the shortest trees (cladograms) that account for the characters forming the database supplied by the researcher, while reducing the number of reversals and convergences in these characters. Computer programs are necessary because the number of possible trees increases exponentially as more taxa are studied.

140 Cladistic research on dinosaurs is burgeoning, and nearly every major group has been analyzed to some degree. Much of this work was begun in the 1980’s, spearheaded by the research of Gauthier (1984, 1986), and it has become the standard for reconstructing phylogenetic relationships. As more numerical cladistic analyses are done, we have begun to get a better understanding of how the different dinosaur groups are related and a much better understanding of the anatomy of dinosaurs and why they look the way they do. An additional way to use phylogenetic analysis in research is to follow the evolution of a single character (an anatomical feature) on a tree to see how it varies across a dinosaur taxon. Other ways are to superimpose geographic locations, ecological characteristics, or time on trees to see how they have influenced the history of dinosaurs. As computers get stronger, phylogenetic analyses will provide better information. Functional morphology is the study of how organisms work. These studies are still relatively rare for dinosaurs but such analyses are now becoming more common. Functional analyses typically make use of architectural (e.g., Weishampel, 1993) and/or machine analogies to understand the evolution and operation of a particular anatomical structure. They commonly use physical models, graphical representations, mathematical computations, computer simulations, and thought experiences to analyze this anatomy. Examples include Alexander’s (1989) analyses of dinosaur locomotion and Weishampel’s (1981) study of the nasal systems of lambeosaurine hadrosaurids. Computers can help increase the level of sophistication possible in such studies, especially through the use of high-level computer graphics, and should provide a strong impetus for a great increase in the number of functional studies on dinosaurs. To date, computer applications in dinosaur biomechanics have been limited to studies of feeding mechanisms and locomotion. For example, Weishampel (1984) has used a three-dimensional kinematics computer program (developed by engineers) to analyze a series of ornithopod skulls as chewing machines. Heinrich et al. (1993) studied locomotion in the Late Jurassic iguanodontian Dryosaurus lettowvorbecki by modeling the femur as a beam. Bone cross-sections provided indications of both strength and patterns of loading on the living dinosaurs. Studying juvenile and adult specimens allowed them to postulate

Computers and Related Technology changes in locomotory patterns with age for that species. Distributional studies analyze the distribution of organisms through time and space. Dinosaur studies of this kind have been very limited so far because of the nature of the fossil record for dinosaurs, but a number of paleontologists are now actively studying distributional problems with some success. To date, studies using computer databases have been able to track dinosaur diversity through space and time (Weishampel and Norman, 1989; Dodson, 1990) as well as the rate of study of dinosaurs by paleontologists (Dodson and Dawson, 1991). The large compilation by Weishampel (1990) of dinosaur localities is making it possible to analyze dinosaur paleobiogeography quantitatively for the first time (R. Chapman and D. Weishampel, work in progress). Such studies, however, can only be done using computers because they involve the mathematical and statistical analysis of large matrices of data. Clearly, this is one area of research that will be expanding greatly because of computers and related technology. The interface between the general public and dinosaurs is one area of great change because of computers. Computers make available a wide range of educational approaches for teaching people about dinosaurs, especially using CD-ROM and multi-media technology. More people are also gaining access to data about dinosaurs through the Internet using online computer databases. One of the biggest effects will be in changing the ability of the public to visualize what dinosaurs looked like. Conventional approaches of reconstructing dinosaurs (e.g., Paul, 1987) are being supplemented by sophisticated three-dimensional computer graphics that use computer visualization technology (e.g., Nielson and Shriver, 1990) to help generate lifelike dinosaurs such as those seen in the film Jurassic Park (Shay and Duncan, 1993) and can support the development of more life-like robotic dinosaurs (e.g., Poor, 1991). The development and distribution of better systems for virtual reality will allow researchers and the public alike to tour a dinosaur’s morphology and even view it from the inside [(see Fro¨hlich, et al., (1995) and Stevens (1995) for discussions in the field of biology and medicine). Once these approaches are developed, they will be used more in conjunction with the original fossil material within exhibitions. Most modern exhibits on

Computers and Related Technology


dinosaurs include some computer technology and this will increase more with time. Clearly, computers can vastly improve how the general public is introduced to dinosaurs and increase their general knowledge.

Gauthier, J. (1986). Saurischian monophyly and the origin of birds. In The Origin of Birds and the Evolution of Flight (K. Padian, Ed.), Memoirs, No. 8, pp. 51–55. California Academy of Sciences.


Heinrich, R. E., Ruff, C. B., and Weishampel, D. B. (1993). Femoral ontogeny and locomotor biomechanics of Dryosaurus lettowvorbecki (Dinosauria, Iguanodontia). Zool. J. Linnean Soc. 108, 179–196.


Gray, S. W. (1946). Relative growth in a phylogenetic series and in an ontogenetic series of one of its members. Am. J. Sci. 244, 792–807.

Hirsch, K. F., Stadtman, K. L., Miller, W. F., and Madsen, J. H., Jr. (1989). Upper Jurassic dinosaur egg from Utah. Science, 243, 1711–1713.

Alexander, R. M. (1989). Dynamics of Dinosaurs and Other Extinct Giants, pp. 167. Columbia Univ. Press, New York.

Jefferson, G. T. (1989). Digitized sonic location and computer imaging of Rancho La Brea specimens from the Page Museum salvage. Curr. Res. Pleistocene, 6, 45–47.

Burns, M. (1993). Automated Fabrication. Improving Productivity in Manufacturing, pp. 369. Prentice-Hall, Englewood Cliffs, NJ.

Jorstad, T., and Clark, J. (1995). Mapping human origins on an ancient African landscape. Prof. Surveyor, 15(4), 10–12.

Chapman, R. E. (1990). Shape analysis in the study of dinosaur morphology. In Dinosaur Systematics, Approaches and Perspectives (K. Carpenter and P. J. Currie, Eds.), pp. 21–42. Cambridge Univ. Press, New York.

Lull, R. S., and Gray, S. W. (1949). Growth patterns in the Ceratopsia. Am. J. Sci., 247, 492–503.

Chapman, R. E., Galton, P. M., Sepkoski, J. J., Jr., and Wall, W. P. (1981). A morphometric study of the cranium of the pachycephalosaurid dinosaur Stegoceras. J. Paleontol. 55(3), 608–618. Chure, D. J., and McIntosh, J. S. (1989). A Bibliography of the Dinosauria (Exclusive of the Aves) 1677–1986, Paleontology Series No. 1, pp. 226. Museum of Western Colorado. Clark, S., and Morrison, I. (1994). CT scan of fossils. In Vertebrate Paleontological Techniques, Volume 1 (P. Leiggi and P. May, Eds.), pp. 323–329. Cambridge Univ. Press, New York. Dodson, P. (1975). Taxonomic implications of relative growth in lambeosaurid dinosaurs. Systematic Zool. 24(1), 37–54. Dodson, P. (1976). Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. J. Paleontol. 50(5), 929–940. Dodson, P. (1990). Counting dinosaurs: How many kinds were there? Proc. Natl. Acad. Sci. (USA), 87, 7608–7612. Dodson, P., and Dawson, S. D. (1991). Making the fossil record of dinosaurs. Mod. Geol. 16(1/2), 3–15. Fro¨hlich, B., Grunst, G., Kru¨ger, W., and Wesche, G. (1995). The responsive workbench: A virtual working environment for physicians. Comp. Biol. Med. 25(2), 301–308. Gauthier, J. (1984). A cladistic analysis of the higher systematic categories of the Diapsida, pp. 564. PhD dissertation, Univ. of California, Berkeley.

Maddison, W. P. and Maddison, D. R. (1992). MacClade. Analysis of Phylogeny and Character Evolution, Version 3, pp. 398. Sinauer, Sunderland, MA. McKenna, P. C. (1992). GPS in the Gobi: Dinosaurs among the dunes. GPS World, 3(6), 20–26. Nielson, G. M., and Shriver, B. (Eds.). (1990). Visualization in Scientific Computing, pp. 282. IEEE Computer Society Press, Los Alamitos, CA. Paul, G. S. (1987). The science and art of restoring the life appearance of dinosaurs and their relatives: a rigorous how-to guide. In Dinosaurs Past and Present. Volume II (S. J. Czerkas and E. E. Olson, Eds.), pp. 4–49. Los Angeles County Museum of Natural History/University of Washington Press, Seattle. Poor, G. W. (1991). The Illusion of Life: Lifelike Robotics, pp. 96. Educational Learning Systems, San Diego, CA. Rowe, T., Carlson, W., and Bottorf, W. (1993). Thrinaxodon: Digital Atlas of the Skull. Univ. of Texas Press, Austin. Shay, D., and Duncan, J. (1993). The Making of Jurassic Park, An Adventure 65 Million Years in the Making, pp. 196. Ballantine Books, New York. Stevens, J. E. (1995). The growing reality of virtual reality. BioScience 45(7), 435–439. Swofford, D. L., and Begle, D. P. (1993, March). User’s Manual for PAUP: Phylogenetic Analysis Using Parsimony. Version 3.1, pp. 257. Laboratory of Molecular Systematics, Smithsonian Institution, Washington, DC.

142 Thompson, D’A. W. (1942). On Growth and Form, The Complete Revised Edition, 1992 ed., pp. 1116. Dover, New York. Weishampel, D. B. (1981). Acoustic analysis of potential vocalization in lambeosaurine dinosaurs (Reptilia: Ornithischia). Paleobiology 7, 252–261. Weishampel, D. B. (1984). Evolution of jaw mechanisms in ornithopod dinosaurs. Adv. Anat. Embryol. Cellular Biol. 87, 1–116. Weishampel, D. B. (1990). Dinosaurian distribution. The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 63–139. Univ. of California Press, Berkeley. Weishampel, D. B. (1993). Beams and machines: Model-

Computers and Related Technology ing approaches to the analysis of skull form and function. The Skull, Volume 3 ( J. Hanken and B. K. Hall, Eds.), pp. 303–343. Univ. of Chicago Press, Chicago. Weishampel, D. B., and Norman, D. B. (1989). Vertebrate herbivory in the Mesozoic: jaws, plants, and evolutionary metrics [Special paper]. Geol. Soc. of Am. 238, 87–100. Witten, A., Gillette, D. D, Sypniewski, J., and King, W. C. (1992). Geophysical diffraction tomography at a dinosaur site. Geophysics 57(1), 187–195. Zangerl, R., and Schultze, H.-P. (1989). X-radiographic techniques and applications. Paleotechniques (R. M. Feldmann, R. E. Chapman, and J. T. Hannibal, Eds.), The Paleontological Society, Spec. Publ. No. 4, pp. 165–178.

Connecticut River Valley JOANNA WRIGHT University of Bristol Bristol, United Kingdom


inosaur tracks from the Connecticut Valley have been known for more than 150 years. They were first scientifically studied and described by Edward Hitchcock (1836, 1858). Hitchcock’s fossil trackway collections are now held in the YALE PEABODY MUSEUM, Connecticut, and the Pratt Museum of Amherst College, Massachusetts. Most of the dinosaur tracks that have been found in the Connecticut Valley were collected in the 19th century. This is because at that time conditions in the area were ideal for fossil collection. Much of the economy was based on farming and the population was sparse by today’s standards. The only suitable building stone in the area was the Turners Falls Sandstone and the Portland Formation: These are the sediments in which the footprints occur and many of the footprints were discovered in quarries. In the middle of the 1800’s the Great Plains were discovered and exploited and many farms were abandoned as it became uneconomical to farm in New England. The abandoned farms were left to be reclaimed by the forest so that a large number of the quarry and farm footprint sites are now completely overgrown. In addition, dams were built at several points along the river, which had the effect of raising the water level along much of the river with the result that many of Hitchcock’s footprint sites are now either submerged or only accessible by boat.

Previous Work After Hitchcock, the only other person to review the whole of the track collection at Amherst College was Lull (1904, 1915), who attempted to sort out the morass of ichnogenera and ichnospecies names that was Hitchcock’s legacy. From when he first started to study the fossil tracks Hitchcock had assigned them genus and species names in the Linnaean binomial system and he would change these as his perceptions of the trackmakers changed and he rarely acknowledged previous names in subsequent publications. Lull also tried to work out what (kinds of ) animals

made the different types of prints. Unfortunately, much of Lull’s nomenclatural work served only to confuse the issue further because he tended to resurrect old and abandoned names. In recent years the main workers on these prints have been Olsen and co-workers (Olsen and Baird, 1986; Olsen et al., 1992), who also attempted to revise some of the tangled nomenclature. They demonstrated that if the lengths of specimens of the ichnogenera Grallator, Anchisauripus, and Eubrontes are plotted against their widths, all three types show a complete gradation in size and proportions and their species frequently overlap.

Geological Background The track-bearing strata of the Connecticut Valley are contained in the Hartford and Deerfield Basins of Connecticut and Massachusetts, two of a series of rift basins along the east coast of North America that opened in the early Mesozoic in response to the NE–SW extension associated with the breakup of Pangaea. The sediments occur in a north–south elongate half graben extending over more than 160 km from the northern border of Massachusetts to Long Island Sound (Fig. 1). More than 4000 m of predominantly red, gray, and black clastic sediments and tholeiitic basalt was deposited in the basin during its approximately 35 million year existence during the Late Triassic and Early Jurassic (Fig. 2). Strata of the Hartford and Deerfield basins constitute two major genetic sequences, a lower Late Triassic age fluvial and alluvial arkose and an upper Early Jurassic age lacustrine and alluvial siltstone to conglomerate with interbedded basalts very low in the sequence. The basal strata, the New Haven and Sugarloaf Arkoses, are dominantly alluvial fan and braided stream redbeds laid down by rivers that flowed from the crystalline highlands in the east. Abundant calcretes suggest that the palaeoclimate was tropical and semi-arid with perhaps 100–500 mm of seasonal rain


Connecticut River Valley


FIGURE 1 Simplified geology of the Hartford and Deerfield Basins.

and a long dry season. Only rare reptile remains have been recovered from these formations. In the Deerfield Basin, the Mount Toby Conglomerate and Turners Falls Sandstone overlie the Sugarloaf Arkose and Deerfield Basalt; they largely reflect

alluvial fan, allluvial plain floodplain, and local lacustrine depositional environments. In the Hartford Basin, the red and gray, fine-grained lacustrine deposits of the Shuttle Meadow and the East Berlin Formations are sandwiched between three tholeiitic basalt flows,

Connecticut River Valley


FIGURE 2 Correlation of the formations between the Hartford and Deerfield Basins.

the topmost of which is overlain by the Lower and then the Upper Portland Formation, which represent similar environments to that of the Turners Falls Sandstone and the Mount Toby Conglomerate, respectively. Olsen et al. (1992) have equated the Holyoke and Deerfield Basalts of the Hartford and Deerfield Basins, but the Mount Toby Conglomerate and Turners Falls Sandstone cannot currently be directly correlated with the sedimentary formations of the Hartford Basin. These sedimentary units, with the exception of the Mount Toby Conglomerate and the Upper Portland Formation, are the rocks in which most of the trackways occur. It is likely that the climate throughout this period was subtropical, monsoonal characterized by alternating episodes of high precipitation and aridity.

morph. Otozoum has recently been identified as the tracks of a basal thyreophoran. Some tracks seem to have been produced by juvenile theropods. There are some indeterminate tridactyl footprints, most of which are probably theropod prints, and there are some indeterminate quadrupedal prints, perhaps made by an animal such as a sphenodontid reptile. If all these kinds of prints, including invertebrate traces, are plotted as numbers of trackways (in the tracks collection of Amherst College) in a pie chart (Fig. 4), it can be seen that there is a very strange

Faunal Diversity Although the taxonomy of the footprint fauna has not recently been revised and can therefore not be taken as a reliable diversity indicator, the tracks can nevertheless be divided into several groups that represent the main types of animals present in the Connecticut Valley in the Early Jurassic (Fig. 3). Grallator, Anchisauripus, and Eubrontes represent small, medium, and large theropods, respectively. Anomoepus represents a small ornithopod (probably fabrosaurid). Batrachopus represents a small crocodilo-

FIGURE 3 Some of the more common vertebrate tracks and their likely producers.

Connecticut River Valley


FIGURE 4 Pie chart showing the proportions of the different types of trackways found in the Valley.

faunal distribution pattern. One-third of the fauna is composed of theropods, 15% consists of invertebrates, and only 11% is made up of herbivores. This is obviously an unsustainable ecological situation. The maker of Batrachopus was probably an insectivore/ carnivore and most of the indeterminate tridactyl prints were probably made by theropods—this would bring the proportion of carnivores to more than 50%. This is obviously not an accurate picture of the Early Jurassic fauna of the Connecticut Valley; therefore, what could be the explanation for the strong bias in the footprint fauna? There are several possible reasons. First, higher activity of the theropods—they moved around more so they made more footprints. Second, the theropods may have waited at strategic places, such as a watering hole, for their prey and thus spent a longer time in areas where their footprints were more likely to be preserved. Also, there is a possibility that our way of recognizing the makers of footprints is flawed and not all the footprints that are assigned to theropods were made by theropods. However, no successful method has been developed to test this hypothesis. Invertebrates An unusual feature of the Connecticut Valley fauna is that it also contains very good invertebrate trackways; the sediments that preserve

the invertebrate traces also preserve plant fossils. Many of the invertebrate trackmakers were hexapods; again the ichnotaxonomy is in great need of revision. The only person to have worked on the invertebrate traces is Edward Hitchcock; they have largely if not totally been ignored by subsequent workers, who preferred to concentrate on the larger, more spectacular dinosaur tracks and trackways [in Lull’s (1953) book of 285 pages he devotes only 17 to the invertebrate traces].

Conclusions The extensive trace fossil collections from the Connecticut Valley are a valuable resource. Both vertebrate and invertebrate traces are preserved and, in addition, plant fossils also occur in the same sediments; therefore, quite a complete picture of the fauna can be obtained. Much work still remains to be done, however, especially on the invertebrate traces, but there is, in addition, a great deal of scope for statistical analysis on the dinosaur footprints.




References Hitchcock, E. (1836). Ornithichichnology. Description of the footmarks of birds (ornithichnites) on New Red Sandstone in Massachusetts. Am. J. Sci. 29, 307–340. Hitchcock, E. (1858). Ichnology of New England: A Report of the Sandstone of the Connecticut Valley, Especially Its Fossil Footmarks, pp. 232. White, Boston. Lull, R. S. (1904). Fossil footprints of the Jura–Trias of North America. Mem. Boston Soc. Nat. History 5, 461–557. Lull, R. S. (1915). Triassic Life of the Connecticut Valley. Connecticut Geol. Nat. History Surv. Bull. 24. Olsen, P. E., and Baird, D. (1986). The ichnogenus Atreipus and its significance for Triassic biostratigraphy. In The Beginning of the Age of Dinosaurs: Faunal Change across the Triassic–Jurassic Boundary (K. Padian, Ed), pp. 61–87. Cambridge Univ. Press, Cambridge, UK. Olsen, P. E., McDonald, N. G., Huber, P., and Cornet, B. (1992). Stratigraphy and paleoecology of the Deerfield Rift Basin (Triassic–Jurassic, Newark Supergroup), Massachusetts. In Guidebook for Field Trips in the Connecticut Valley Region of Massachusetts and Adjacent States (P. Robinson and J. B. Brady, Eds.), Vol. 2, pp. 488–535.

Connecticut State Museum of History, Connecticut, USA see MUSEUMS



Constructional Morphology DAVID B. WEISHAMPEL The Johns Hopkins University School of Medicine Baltimore, Maryland, USA

gether, constructional morphology adds a third factor in biological explanations of form—a bautechnische (roughly translated as architectural or fabricational) factor. Thus, biological form is understood only as the result of three interacting factors: ecological– adaptive, historic—phylogenetic, and bautechnische. Traditional biology views form as the result of adaptation and phylogeny. Bautechnik recognizes that geometry, natural materials, and growth processes also regulate morphologic patterns. As a consequence, bautechnische factors are ahistorical elements that express biological possibilities and limits on evolutionary change that stem from the physical and chemical properties of available materials (strength and failure, elasticity, adhesion, and viscosity); cybernetic controls on development, maintenance, and repairs; and, finally, the geometry of pattern formation and self-organization. The particular materials, growth programs, and regulatory systems are naturally acquired as specific historical (evolutionary) events; hence the transfer of ahistorical factors to phylogenetic clades and lineages. Likewise, the ways in which these ahistorical factors affect reproductive/ evolutionary fitness also shift such influences to the unique historical nexus of adaptation.


References Reif, W.-E., Thomas, R. D. K., and Fischer, M. S. (1985). Constructional morphology. Acta Biotheor. 34, 233–248. Seilacher, A. (1970). Arbeitskonzept zur Konstruktions– Morphologie. Lethaia 3, 393–396.

The term ‘‘constructional morphology’’ as currently understood comes from an important evolutionary biological research program in Germany called Konstruktionsmorphologie. Founded by A. Seilacher at the University of Tu¨bingen, constructional morphology treats biological form not just as a consequence of function and phylogeny, although these are admittedly necessary parts of the explanation of such form. However, because neither alone in isolation provides a sufficient understanding of form, nor do both to-

Coprolites KAREN CHIN University of California Santa Barbara, California, USA

Under exceptionally favorable conditions, ancient feces have been preserved as fossils called coprolites.

148 Because feces are largely composed of soft material, coprolites are usually much rarer than skeletal fossils. They can be locally abundant, however, and such concentrations contributed to early interest in these unusual formations. The first published report of fossil feces actually predates the earliest descriptions of dinosaur bones and was made without benefit of precedent when William Buckland (1823) compared some enigmatic white fossil lumps with fresh hyena feces and deduced a fecal origin. He later (Buckland, 1835) applied the term ‘‘coprolite’’ to fossilized feces. Coprolites have been found on every continent, with the oldest known vertebrate specimens dating back to the Silurian (Gilmore, 1992). Although some Quaternary feces have been preserved through desiccation and resemble modern dried dung (e.g., Mead and Agenbroad, 1992), most coprolites have been substantially altered during fossilization. A small number of specimens have been preserved as carbonaceous compressions (e.g., Hill, 1976), but the overwhelming majority of coprolites are lithified. Most fossil feces have been recognized by their familiar fecal shapes, but coprolites are highly variable fossils. This variation reflects differences in the animals that produced the feces, fluctuations in diet, and disparate diagenetic conditions. Numerous coprolite morphologies have been described, including spherical, cylindrical, fusiform, spiral, blocky, pancake-like, and amorphous forms. This range of shapes is similar to that found in modern animal droppings. The colors of coprolites also vary considerably: Different diagenetic regimes have resulted in brown, white, cream, orange, black, gray, and even bluish specimens. As trace fossils, coprolites provide a record of animal activity and have the potential to supplement information obtained from skeletal fossils. Although coprolitic interpretations are complicated by diagenetic variability, well-preserved specimens can contain recognizable dietary inclusions that provide information on trophic interactions in ancient ecosystems. This information can be enhanced by some knowledge of the animal of origin.

Coprolite Provenance The assignment of a coprolite to its source animal remains one of the more difficult problems of coprolite analysis. Unless a formed but unextruded fecal mass is found in the body cavity of an articulated

Coprolites specimen, its origin remains speculative. Certain factors, however, can help constrain a list of likely producers. A spiral configuration is the only distinctive coprolite morphology that can be reliably associated with a type of source animal. Because the spiral valve that affects the egestion of coiled feces is absent in teleosts and tetrapods (Romer and Parsons, 1986), spiral coprolites are attributable to other taxa such as sharks, gars, or lungfish (Gilmore, 1992). Spiral coprolites have been recovered from many Paleozoic and Mesozoic localities (Hantzschel et al., 1968). The widespread distribution of these specimens indicates that feces deposited in aquatic environments have a high preservation potential. This suggests that many non-spiraled coprolites may have also been produced by aquatic animals or by terrestrial vertebrates that defecated in or near bodies of water. Although non-spiraled coprolites can have distinctive shapes and/or striations, such morphologies are found in feces produced by many different taxa (Thulborn, 1991; Hunt et al., 1994). This necessitates the evaluation of non-morphological characters. Of prime consideration is the fact that the stratigraphic distribution of potential source animals must be consistent with the age, locality, and depositional environment of the coprolite itself. The co-occurrence of skeletal elements in the same sediments may pose strong arguments for associations between coprolites and source animals, but such associations remain speculative without additional evidence. Coprolite size can provide information on possible animal producers, but interpretations of the significance of size must be made carefully. Although the quantity of egested feces is proportional to body size, direct correlations can be misleading. A small coprolite, for example, may have broken off a larger fecal mass. In addition, because some large extant animals produce quantities of small pelletoid feces, an isolated pelletoid coprolite could have been produced by a relatively large animal. Small animals, on the other hand, cannot produce large fecal masses. These considerations suggest that coprolite size should be primarily used to infer minimum sizes of possible producers. This criterion is particularly useful for identifying possible dinosaur coprolites. Coprolite composition can also help constrain the number of likely producers by providing clues to the feeding strategies of source animals. Inclusions of

Coprolites bone fragments, teeth, fish scales, or mollusc shells, for example, provide evidence of carnivory. Unfortunately, recognizable dietary residues have often been destroyed by digestion and diagenesis. In such cases, carnivory can be implicated by a predominance of calcium phosphate. Bradley (1946) noted that most coprolites are phosphatic and that carnivore feces contain relatively high percentages of phosphorus. He suggested that carnivore feces are preferentially fossilized because of the availability of dietary calcium phosphate. This is consistent with the observation that permineralized coprolites containing substantial plant material are relatively rare and are almost invariably calcareous or siliceous. Although it is difficult, if not impossible, to ascertain the origin of a coprolite, the concomitant analysis of stratigraphic occurrence, morphology, size, and composition can help characterize likely source animals. The informative value of these factors is variable, however. The identities of some coprolite producers may remain poorly resolved if coprolite specimens have very common features, whereas unusual or distinctive attributes might provide significant clues to coprolite provenance.

Dinosaur Coprolites Many coprolites have been found in Mesozoic sediments, but few described specimens have been confidently attributed to dinosaurs. The identification of possible dinosaur coprolites is complicated by the fact that dinosaurs shared Mesozoic ecosystems with many other animals. Non-spiraled small or mid-sized phosphatic coprolites might be particularly difficult to identify because they could have been produced by a number of carnivorous vertebrates, including teleosts, turtles, crocodilians, or dinosaurs. Very large coprolites may be more reasonably ascribed to dinosaurs, but sizable bona fide coprolites are rare. The paucity of large specimens probably reflects preservational biases: Large fecal masses would have been highly susceptible to mechanical disruption, and large animals may have rarely defecated in depositional environments that were conducive to the fossilization of feces. These considerations help account for the poor record of dinosaur coprolites. Moreover, some specimens previously interpreted as dinosaur feces must be re-evaluated, such as those from the Bernissart Iguanodon Quarry in Belgium. Bertrand (1903) ini-

149 tially attributed those coprolites to theropods, but a subsequent discussion (Abel, 1935) suggested that the feces could have been produced by crocodiles. In another report, Matley (1939) used size criteria to assign large Cretaceous coprolites from India to titanosaurs whose bones were found in the same sediments. The coprolites were probably not produced by herbivores, however, because the specimens are phosphatic and contain no traces of plant tissue. The large droppings are up to 10 cm wide and 17 cm long, and certainly implicate hefty source animals, but it is not clear if fecal masses produced by large carnivorous dinosaurs can be distinguished from those left by large crocodilians. Still other purported dinosaur feces may not be coprolites at all. Large, bulbous, siliceous nodules commonly found in Jurassic deposits have sometimes been interpreted as dinosaur coprolites (Spendlove, 1979). These specimens, however, lack organic inclusions or other positive evidence that supports a fecal origin and may simply be inorganic concretions. The origin of large Mesozoic plant-filled coprolites is less ambiguous, because few large herbivores coexisted with dinosaurs. One unusual grouping of more than 250 compressed pellets containing plant cuticle was found in Jurassic sediments in England (Hill, 1976). The individual pellets are small (8- to 18mm diameter), but the total assemblage represents a sizable fecal mass that could have been produced by a herbivorous dinosaur. Much larger herbivore coprolites found in Montana are undoubtedly dinosaurian. These large (up to 24 ⫻ 33 ⫻ 34 cm) blocky Cretaceous specimens contain conifer stem fragments. They lack a familiar coprolite shape, but a fecal origin has been corroborated by the presence of backfilled dung beetle burrows in the specimens (Chin and Gill, 1997). The recognition of atypical coprolitic masses suggests that additional dinosaur droppings may be identified with more careful examination of Mesozoic sediments. Continued analyses of dinosaur feces will increase our understanding of dinosaur diets and their interactions with other organisms because coprolites can provide paleobiological information that is unavailable from skeletal fossils.

See also the following related entries: DIET ● TRACE FOSSILS




Romer, A. S., and Parsons, T. S. (1986). The Vertebrate Body, pp. 679. Saunders College, Philadelphia.

Abel, O. (1935). Vorzeitliche Lebensspuren, pp. 644. Fischer, Jena, Germany.

Spendlove, E. (1979). Henry Mountain coprolites. Rock Gem 9, 60–64.

Bertrand, C. E. (1903). Les Coprolithes de Bernissart. I. partie: Les Coprolithes qui ont e´te´ attribue´s aux Iguanodons. Mem. Musee Royal d’histoire Nat. Belgique 1(4), 1–54.

Thulborn, R. A. (1991). Morphology, preservation and palaeobiological significance of dinosaur coprolites. Palaeogeogr. Palaeoclimatol. Palaeoecol. 83, 341–366.

Bradley, W. H. (1946). Coprolites from the Bridger Formation of Wyoming: Their composition and microorganisms. Am. J. Sci. 244, 215–239. Buckland, W. (1823). Reliquiae Diluvianae, pp. 303. Arno Press, New York [Reprint of the 1823 ed. published by Murray, London, 1978] Buckland, W. (1835). On the discovery of coprolites, or fossil faeces, in the Lias at Lyme Regis, and in other formations. Trans. Geol. Soc. London Ser. 2, 3(1), 223–236. Chin, K., and Gill, B. D. (1997). Dinosaurs, dung beetles, and conifers: Participants in a Cretaceous food web. Palaios, in press.

Corpus Christi Museum of Science and History, Texas, USA see MUSEUMS



Gilmore, B. G. (1992). Scroll coprolites from the Silurian of Ireland and the feeding of early vertebrates. Palaeontology 35(2), 319–333. Hantzschel, W., El-Baz, F., and Amstutz, G. C. (1968). Coprolites, an Annotated Bibliography. Memoir 108, pp. 132. Geological Society of America, CO. Hill, C. R. (1976). Coprolites of Ptiliophyllum cuticles from the Middle Jurassic of North Yorkshire. Bull. Br. Museum Nat. History Geol. 27, 289–294. Hunt, A. P., Chin, K., and Lockley, M. G. (1994). The palaeobiology of vertebrate coprolites. In The Palaeobiology of Trace Fossils (S. K. Donovan, Ed.), pp. 221–240. John Wiley, Chichester, UK. Matley, C. A. (1939). The coprolites of Pijdura, Central Provinces. Rec. Geol. Surv. India 74(4), 530–534. Mead, J. I., and Agenbroad, L. D. (1992). Isotope dating of Pleistocene dung deposits from the Colorado Plateau, Arizona and Utah. Radiocarbon 34(1), 1–19.

Cranbrook Institute of Science, Michigan, USA see MUSEUMS



Cranial Comparative Anatomy see SKULLS, COMPARATIVE ANATOMY

Craniofacial Air Sinus Systems LAWRENCE M. WITMER Ohio University Athens, Ohio, USA


n unusual anatomical system pervaded the heads of dinosaurs. Insinuated among such conventional soft tissues as muscles, nerves, blood vessels, and sense organs was a complicated system of air-filled sinuses. These sinuses formed as thinwalled, epithelial outgrowths (diverticula) of other air-filled cavities, often invading and resorbing surrounding bone and producing foramina and cavities within these bones. This process is called pneumatization and the resulting state of having air-filled bones is known as pneumaticity. Many dinosaurs are highly pneumatic animals indeed, with most of the bony skull literally riddled with foramina, channels, and cavities. More technical treatments of this topic have been published by Witmer (1990, 1995, 1997) and Currie and Zhao (1993a,b) and the articles cited therein.

Pneumatic Systems in Dinosaurs There are two well-known pneumatic systems in dinosaurs, one arising as outgrowths of the nasal cavity and the other as outgrowths of the tympanic (middle ear) cavity. Not just dinosaurs, but all archosaurs— living and extinct—have at least one main paranasal air sinus, known as the antorbital sinus, that forms as an outgrowth of the main nasal cavity (Fig. 1). The antorbital sinus produces a large cavity and opening in the side of the face known respectively as the antorbital cavity and antorbital fenestra. In many archosaurs, the antorbital sinus itself has subsidiary outgrowths that may pneumatize surrounding bones, producing so-called accessory cavities. In addition to the nearly ubiquitous antorbital sinus, a few kinds of dinosaurs have air sacs deriving from a different part of the nasal cavity, namely, the front-most portion known as the nasal vestibule. Such vestibular sinuses tend to pneumatize the bones surrounding the bony nostril (i.e., premaxilla and nasal). Humans and most other mammals have similar (but not homologous) paranasal sinuses; these are the sinuses that become


congested when we have colds and that are involved in our ‘‘sinus headaches.’’ Paratympanic air sinuses are less common in archosaurs. In those non-dinosaurian taxa with paratympanic pneumaticity (such as crocodylomorphs and pterosaurs), the bones of the braincase are the ones that are usually invaded by air sacs. Among perhaps all archosaurs, tympanic recesses are best developed in theropod dinosaurs. As with paranasal sinuses, humans and most other mammals have a set of sinuses associated with the tympanic cavity, and particularly bad middle-ear infections may spread to our paratympanic air sinuses. In addition to paranasal and paratympanic sinuses, there are other, more poorly known, systems pneumatizing the head skeleton. The first may simply represent diverticula from the cervical system of pulmonary air sacs that extend beyond the neck vertebrae into the occipital region of the skull. Some of the pneumatic cavities of certain theropod dinosaurs may result from these pulmonary diverticula. The second is the median pharyngeal system that forms as a midline outgrowth from the roof of the throat and invades the base of the skull in the region of the basisphenoid and basioccipital bones in many archosaurs. It is not always demonstrably of pneumatic origin in many archosaurs but is almost certainly so in theropod dinosaurs. Although the resulting ‘‘basisphenoid sinus’’ is often regarded as a derivative of a ‘‘median Eustachian tube,’’ it is clearly distinct from the definitive auditory (Eustachian) tubes that, along with the tympanic cavities, have their embryological origins from the paired first pharyngeal (or branchial) pouches; whereas in some archosaurs the median system eventually connects up with the tympanic cavity, it does not always do so. Some have suggested that the median pharyngeal pneumatic system results from aeration of the embryonic hypophysial pouch (of Rathke)—a precursor of

Craniofacial Air Sinus Systems


FIGURE 1 Paranasal pneumaticity in theropod dinosaurs. (Left) skull of Allosaurus fragilis in oblique view, showing the antorbital paranasal air sinus and some of its epithelial diverticula (modified from Witmer, 1997, with permission). (Right) skull of Sinraptor dongi in left lateral view with the major paranasal pneumatic accessory cavities labeled (modified from Currie and Zhao, 1994a, with permission). The accessory cavities result from pneumatization by the subsidiary diverticula of the antorbital sinus.

part of the pituitary gland—and this is an idea worthy of further investigation.

Ornithischian Dinosaurs Paratympanic air sinuses are very uncommon in ornithischians. The middle ear sac was certainly present (as evidenced by, among other things, the discovery of columellae—the slender ear bones), but apparently it did not typically send out diverticula that invaded surrounding bones. A few ornithischians, however, such as the basal thyreophoran Scelidosaurus and the ornithopod Hypsilophodon, do seem to have some excavations of portions of the braincase that are best interpreted as being of pneumatic origin. In these forms, there is a fairly extensive cavity associated with the canal for the major artery supplying the brain, the cerebral (or internal) carotid artery. This cavity is directly behind and medially undercuts a curving ridge of bone—nearly ubiquitous in archosaurs—known as the otosphenoidal crest, which runs from the basipterygoid process to the paroccipital process and segregates the orbital contents in front from the middle ear contents behind. Although such a rostral tympanic recess is, as we will see, fairly common in theropod dinosaurs (as well as in a variety of other archosaurs), it is rather rare in ornithischians.

Paranasal pneumaticity, on the other hand, is present in probably all ornithischians in that, like all other archosaurs, they possessed an antorbital sinus. This antorbital air sinus was lodged in a cavity, the antorbital cavity, located in front of the orbit and bounded by primarily the maxilla and lacrimal, and sometimes also the jugal, palatine, and nasal. In fact, this description holds for most dinosaurs (indeed, most archosaurs). Pneumatic bony accessory cavities, produced by subsidiary diverticula of the antorbital sinus, are relatively rare in ornithischians, although they are present in a few taxa, such as the intramaxillary sinuses of Protoceratops and its relatives and the maxillary recesses of some basal thyreophorans. Higher thyreophorans, in particular ankylosaurid ankylosaurians, deserve special mention here in that within their highly armored skulls is a maze of pneumatic sinuses. The precise pattern and arrangement of ankylosaurid paranasal sinuses remain poorly known, and it is not entirely clear if some are accessory cavities of the main antorbital sinus, novel paranasal sinuses, or even diverticula of the nasal vestibule. Pneumatic evaginations of the nasal vestibule, however, are clearly expressed in lambeosaurine hadrosaurids, such as Corythosaurus. In lambeosaurines, the narial region is greatly enlarged, and the bones enclosing

Craniofacial Air Sinus Systems

FIGURE 2 The evolving antorbital cavity of ornithischian dinosaurs, especially Ornithopoda. Most clades of ornithischians, such as Ornithopoda, show marked trends for reduction and enclosure of the antorbital cavity by laminae of the maxilla and lacrimal. Modified from Witmer (1997) and references cited therein with permission.

the nasal vestibule (the premaxilla and nasal) are folded into a complicated collection of passages and chambers, all of which are perched atop the remainder of the skull. Perhaps the most remarkable aspect of skull pneumaticity in ornithischians is its recurrent trend for reduction. In other words, ornithischians tend to become relatively less pneumatic when comparing more advanced members of a clade with more basal members. For example (Fig. 2), the basal ornithischian Lesothosaurus has a more or less primitive—and hence, fairly large—antorbital cavity; it even has a small pneumatic accessory cavity associated with its palatine bone. In ornithopodan ornithischians, however, there is a general trend for reduction of the antorbital cavity and enclosure by lateral sheets of the maxilla and lacrimal bones such that the antorbital cavity becomes a relatively small and completely internalized space. The broad outlines of this trend can be viewed by making a phylogenetic march from a basal ornithopod such as Heterodontosaurus (which retains a relatively large antorbital cavity but also shows the beginnings of the lateral enclosure), through Hypsilophodon (which shows further lateral enclosure) and iguanodontians such as Camptosaurus and Iguanodon (which show further reduction and enclosure), to hadrosaurids (in which the antorbital

153 cavity is relatively tiny and completely closed laterally). A similar phylogenetic trend can be identified in thyreophorans. These trends almost certainly relate to (especially in ornithopods) the expansion and elaboration of the feeding apparatus (in particular, the dentition and its bony supports). Thus, as the feeding apparatus expanded phylogenetically, the antorbital sinus and its bony cavity contracted. Before concluding the discussion of ornithischian craniofacial pneumaticity, the supracranial cavities of ceratopsids such as Triceratops need to be considered. These cavities, formed by the folding of the frontal bones, are often referred to as ‘‘frontal sinuses’’ and are commonly thought to be of pneumatic origin. They are often compared to the frontal air sinuses of modern bovid mammals (cattle, sheep, etc.) because in both ceratopsids and bovids the sinuses form strutted chambers that extend up into the base of the horn cores. It is possible that the supracranial cavities of ceratopsids are indeed pneumatic, but the source of the air-filled diverticulum remains obscure. It is not yet clear how an outgrowth from either the nasal cavity or the tympanic cavity could reach the skull roof. Its mode of development is also unusual for a pneumatic recess. Thus, pending further research, the supracranial cavities of ceratopsids will remain functionally enigmatic.

Saurischian Dinosaurs Saurischia includes neornithine birds, the most pneumatic of all known vertebrates, but not all saurischians display extensive craniofacial pneumaticity. In fact, other than the antorbital cavity itself, sauropodomorphs do not display many pneumatic features in their skulls, which is ironic because their axial skeletons are otherwise often marvels of pneumatization. Basal sauropodomorphs (i.e., prosauropods) have relatively primitive and simple antorbital cavities, although a subsidiary diverticulum of the antorbital sinus excavates a pneumatic accessory cavity in the nasal of Plateosaurus. The antorbital cavity of most sauropods is relatively reduced, being telescoped between the orbit behind and the greatly expanded nasal vestibule in front. As in ornithopods, the antorbital sinus appears to be, in a sense, ‘‘crowded out’’ by other structures. Paratympanic pneumaticity is also relatively poorly developed in sauropodomorphs. A few basal taxa, such as Anchisaurus and Plateosaurus,

154 have, much as does Hypsilophodon, moderate development of a rostral tympanic recess (i.e., the cavity associated with the cerebral carotid artery and bounded by the otosphenoidal crest). Other paratympanic recesses appear to be virtually absent, although a few sauropods (e.g., Camarasaurus) have a deep excavation in the back surface of the quadrate bone that could be interpreted as pneumatic in nature. In contrast to sauropodomorphs, theropod dinosaurs display the most extensive craniofacial pneumaticity of all archosaurs with perhaps the exception of some pterosaurs. Not only are the paranasal and paratympanic systems well developed in theropods but also the median pharyngeal system—for the first time in dinosaurs—takes on clearly pneumatic attributes. In virtually all theropods, the antorbital cavity is huge, occupying in some cases more than half of the total skull length; thus, the enclosed antorbital paranasal air sinus must have been very voluminous (Fig. 1). As in other archosaurs, the maxilla and lacrimal were the major bones housing the air sac, although the nasal, jugal, and palatine were also commonly involved. With the exception of the bizarre oviraptorosaurs, which have dramatically complicated pneumatic skulls, there is no evidence that the facial skeleton of theropods was pneumatized by any air sacs other than the antorbital sinus. Nevertheless, one of the remarkable aspects of paranasal pneumaticity in theropods was the tendency for the antorbital sinus to send out subsidiary diverticula that penetrated into the adjacent facial bones, often producing large pneumatic accessory cavities (Fig. 1). Although many other archosaurs have more or less shallow pneumatic depressions on various facial bones, it is theropods alone that routinely exhibit a facial skeleton that is produced into an open series of hollowed struts and chambers. The most commonly observed accessory cavities are those in the maxilla, with the two most consistent maxillary sinuses being the promaxillary recess and the maxillary antrum. Both these accessory cavities are widely (but not universally) distributed among neotetanurans, and many (but again not all) ceratosaurians have a single maxillary sinus, which is probably homologous to the promaxillary recess of higher theropods. The lacrimal bone is also commonly pneumatized by a subsidiary diverticulum of the antorbital sinus. It is this lacrimal sinus that invades and hol-

Craniofacial Air Sinus Systems lows out the ‘‘horns’’ of theropods such as Allosaurus and Ceratosaurus. The nasal, jugal, and palatine bones are less frequently pneumatized, but in some cases these recesses can result in spectacular structures, such as the inflated nasal crest of Monolophosaurus or the puffed-up palatine bones of Tyrannosaurus. The phylogenetic distribution of these paranasal pneumatic accessory cavities can be rather confusing. Some groups, such as tyrannosaurids, tend to be fairly consistent. Other groups, however, can be more variable. For example, among dromaeosaurids, Deinonychus has all the accessory cavities noted previously, but the closely-related form, Velociraptor, has only a few of them. Moreover, a few taxa, such as the basal tetanuran Torvosaurus and the aberrant maniraptoran Erlikosaurus, lack many or even all of the accessory cavities that phylogenetics indicates they ‘‘should’’ have. Nevertheless, despite these problems, one point that emerges is that the facial skeleton of theropods, as a group, is highly pneumatic; in fact, pneumaticity has been recorded in every bone of the facial skeleton except two (the prefrontal and vomer). A final aspect of the paranasal air sinus system of theropods involves a diverticulum of the antorbital sinus that only very rarely pneumatizes bone. This air sac, the suborbital diverticulum, is almost always present in modern theropods (i.e., birds). It forms as an outgrowth of the back wall of the antorbital sinus and expands into the orbit where it often encircles the eyeball and interleaves with the jaw musculature. In at least a few nonavian theropods, there are good reasons to believe that a bird-like suborbital diverticulum was present (Fig. 1). The significance of this suborbital sac is that it provides a mechanism for actively ventilating the antorbital sinus (i.e., pumping air in and out). Movements of the lower jaw—such as in closing and opening the mouth—will set up positive and negative pressures in the suborbital sac because of its intimate relationship to the jaw muscles. These pressure changes are transferred to the antorbital sinus and, thus, like a bellows pump, air passes to and fro between the nasal cavity and antorbital sinus. This situation is unique: In other animals with paranasal sinuses, such as crocodilians and mammals, the sinuses are never actively ventilated but rather are stagnant, dead-air spaces. What role such a paranasal bellows pump plays in the physiology of birds and probably other theropods is still unknown.

Craniofacial Air Sinus Systems As with the paranasal system, the paratympanic air sinuses are generally very diverse and extensive in theropods. Most of the sinuses invade the bones of the braincase and otic (ear) region, such as the prootic, opisthotic, basisphenoid, and basioccipital. Some of these recesses are obviously associated with the middle ear sac, but others, as mentioned previously, seem to result from median pneumatic outgrowths of the roof of the pharynx and perhaps even from extensions from the pulmonary air sacs in the neck. The pneumatic sinuses of the braincase will be briefly discussed using a hypothetical form (Fig. 3) because no known species has all the air cavities. There are three fairly consistent pneumatic cavities that clearly derive from diverticula of the middle ear sac, and they all aptly bear the name ‘‘tympanic re-

155 cess.’’ Perhaps the most commonly encountered tympanic recess is the rostral or lateral tympanic recess noted previously in some other dinosaurs. This recess is located just behind the otosphenoidal crest in the area of the cerebral carotid artery foramen. In some theropods, such as Syntarsus or Dilophosaurus, this recess resembles that of other archosaurs in being a more or less simple but expanded cavity, whereas in others, such as some coelurosaurs, it becomes complicated and multichambered. For example, taxa such as Deinonychus and Struthiomimus have a discrete prootic recess within the rostral tympanic recess just ventral to the facial nerve foramen, and a varied group of tetanurans have a more ventral cavity, the subotic recess, that excavates the basal tubera. In some forms, such as ornithomimids, troodontids, and birds, the

FIGURE 3 Braincase of a hypothetical theropod dinosaur showing the diversity of pneumatic recesses. Pneumatization from the middle ear sac produces the dorsal, caudal, and rostral tympanic recesses and the recesses within the quadrate and articular bones (not shown). The subcondylar recesses derive from either the middle ear sac or extensions from the pulmonary air sacs. Pneumatization from a median pharyngeal system produces the basisphenoid recess. The basipterygoid and subsellar recesses are not clearly pneumatic structures, and the source of the diverticulum, if present, is also uncertain. fo, fenestra ovalis (or vestibularis) [note: not foramen ovale, which is the maxillary n. foramen in mammals); fpr, fenestra pseudorotundum (or cochlearis); osc, otosphenoidal crest.

156 rostral tympanic recess and part of the main middle ear sac are covered laterally by a thin sheet of bone, the parasphenoid. The caudal tympanic recess is typically found in Coelurosauria (Troodon being an important exception) and involves a large air space within the paroccipital process that opens into the tympanic cavity via an oval foramen on the front of the base of the paroccipital process. The dorsal tympanic recess, a moderate to very deep depression on the dorsolateral surface of primarily the prootic bone, has a very patchy distribution but is found in ornithomimids, velociraptorine dromaeosaurids, and all known birds. In a few groups of nonavian theropods, such as birds, tyrannosaurids, and at least some ornithomimids and troodontids, the quadrate and/or articular bones are also invaded by outgrowths of the middle ear sac. In some respects, such pneumaticity makes good sense in that the quadrate and articular bones have their embryological origins as parts of the first pharyngeal (⫽ mandibular) arch; recall that the middle ear sac itself derives from the same pharyngeal arch. In fact, it is a mystery as to why more groups of theropods—especially bird-like forms such as Deinonychus that have extensive braincase pneumaticity—lack mandibular arch pneumaticity. Interestingly, crocodyliforms, many pterosaurs, and some other non-dinosaurian archosaurs have pneumatic quadrates and articulars. The median pharyngeal system, a rudimentary blind pit or small foramen between the basioccipital and basisphenoid in other archosaurs, takes on the unmistakable appearance of an invasive air-filled cavity in theropod dinosaurs, in which it is usually referred to as a ‘‘basisphenoid sinus’’ (Fig. 3). Typically, the sinus is roughly pyramidal or conical, with its base being a large opening in the midline of the basicranium between the paired basal tubera and basipterygoid processes and its apex being directed dorsally toward the pituitary fossa. Even in basal forms such as Coelophysis and Dilophosaurus, the basisphenoid sinus is already somewhat expanded. However, in most tetanurans the basisphenoid sinus becomes a large expansive cavity, invading back into the basioccipital bone, sometimes even into the occipital condyle. Sometimes a median septum is preserved within the sinus, partially dividing it into left and right sides.

Craniofacial Air Sinus Systems Paired openings on the occipital surface of the cranial base below the occipital condyle, known as the subcondylar recesses (Fig. 3), are not widely distributed in theropods but are well developed in tyrannosaurids and ornithomimids. These cavities are obviously of pneumatic origin in that they expand within the bone and are multichambered. What is uncertain is the source of the air-filled diverticulum. The diverticulum could indeed derive from the middle ear sac—that is, it is another tympanic recess. Certainly, the subcondylar recess is close enough to the tympanic cavity that it would not be a ‘‘far reach.’’ Another idea, however, cannot be ruled out: The subcondylar recesses derive from diverticula of the cervical division of the lung air-sac system. The cervical air sacs pneumatize the backbone up to the second cervical vertebra in tyrannosaurids and ornithomimids and thus are very close to the occiput. Furthermore, they presumably interleaved with neck muscles that attached to the occiput in the vicinity of the subcondylar recesses. Choosing between a tympanic or pulmonary source is currently very difficult. A couple of other cavities may be of pneumatic origin but are much more uncertain (Fig. 3). The basipterygoid recess is a depression within the lateral surface of the basipterygoid process within the orbit of theropods such as Allosaurus. If it is pneumatic, the source of the diverticulum is unclear. Instead, it may be a site of muscular attachment, perhaps for a palatal protractor (if indeed such muscles were present in nonavian theropods). The subsellar recess is a median ventral cavity located just in front of the basisphenoid recess at the base of the parasphenoid rostrum. As its name implies, it resembles the basisphenoid recess in being directed toward the pituitary fossa, but obviously both recesses could not result from a diverticulum tracking along the single embryonic hypophysial stalk (if in fact either did). Clearly, both the basipterygoid and subsellar recesses are structures that need a good deal more research. Two more ‘‘problem sinuses’’ need to be considered. Large cavities within both the ectopterygoid and squamosal bones are clearly pneumatic. The source of the diverticula, however, remains unclear. The ectopterygoid recess is an almost ubiquitous feature of theropods. It usually takes the form of a simple, smooth-walled ventromedial ‘‘pocket’’ in the bone, although in some tyrannosaurids it becomes a

Craniofacial Air Sinus Systems multichambered affair. However, did the diverticulum come from the middle ear sac, the antorbital sinus, or from some unknown source (perhaps yet another pharyngeal outgrowth!)? All these ideas are possible but have problems. Likewise, the large recess within the squamosal bones of tyrannosaurids and ornithomimids could have been produced by subsidiary diverticula of either the antorbital sinus or the middle ear sac, and we currently have little strong evidence that allows us to make a reasoned choice. In summary, theropod dinosaurs obviously exhibit an extraordinary—often bewildering—diversity of air-filled sinuses. Without question, pneumaticity is the single most important anatomical system shaping theropod skull morphology. Interestingly, however, it seems fraught with high levels of homoplasy (the evolutionary loss and/or convergent acquisition of a feature). Certainly, there are some very consistent, phylogenetically informative pneumatic characters, such as the presence of the two main maxillary sinuses in neotetanurans or the caudal tympanic recess in coelurosaurs. However, many of the pneumatic characters seem to have been evolved or been lost repeatedly—often yielding a morphology so similar that the only hint of homoplasy comes through phylogenetic analysis. The dorsal tympanic recess of coelurosaurs is a good, but by no means the only, example. This recess is found in Mesozoic birds, such as Hesperornis and Archaeopteryx, as well as the dromaeosaurids Deinonychus and Velociraptor. So far, so good, but it is absent in the dromaeosaurid Dromaeosaurus. It is also absent in the bullatosaur Troodon but is clearly present in the bullatosaur Struthiomimus. The addition of other taxa only complicates the picture further. When the workings of pneumatic systems—that is, the soft tissue systems that literally produces the bony recesses—are understood, morphogenetic mechanisms for phylogenetic reversal and convergence become rather easy to envision. For example, in the diagram in (Fig. 4) showing a hypothetical ancestor–descendant sequence of species, the ancestral form has a relatively simple middle ear sac with no dorsal tympanic diverticulum. A descendant species might evolve some anatomical change in the conformation of the general region (in this case, a higher and shorter braincase) that, for some reason, permits evagination of a diverticulum, and this diverticulum excavates a pneumatic cavity on the prootic. A further


FIGURE 4 A transgression–regression model for homoplasy in pneumatic systems. Hypothetical evolutionary transformation series of theropod braincases depicting the changing status of a dorsal tympanic diverticulum of the middle ear sac and, hence, the variable presence or absence of the bony recess. As the anatomical conformation of the whole region changes, the possibility for a diverticulum to evaginate changes. Thus, the mechanism for pneumatic homoplasy is easily envisioned, with an apt analogy being the fluctuating sea levels associated with transgression and regression.

conformational change might prevent evagination of this diverticulum or cause it to shift its position, which would appear to us as a reversal. Subsequent changes could permit the diverticulum to evaginate once again, perhaps resulting in a bony recess that is identical to the second species, which then would represent convergence. The point is that these epithelial air sacs are highly labile structures, the position (or even the presence) of which is easily affected by surrounding anatomical structures. Thus, given this almost capricious ebb and flow of pneumatic diverticula, perhaps it is unreasonable even to expect these bony cavities to fall out neatly on a cladogram.

Functions of Sinuses It is somewhat ironic that, despite air-filled sinuses being very prominent components of head anatomy in not just birds and other dinosaurs but also in mammals, the function of pneumaticity has remained obscure. Numerous ideas have been proposed over the years, including sinuses acting as shock absorbers, flotation devices, vocal resonators, thermal insula-

158 tors, weight-reducing air bubbles, and the list goes on. Many of these ideas at first may seem absurd but still may apply to some animals in some cases. For example, the sinuses within the crests of lambeosaurine hadrosaurids are thought to have functioned as resonating chambers. However, these functions obviously cannot apply broadly across all animals that possess the sinus. In fact, because almost all the ideas that have been proposed fail because of limited applicability, the problem of the general function of pneumatic sinuses has been regarded as one of the great mysteries of craniofacial functional morphology. It is significant that virtually all the suggestions hinge on the empty space within the sinuses being that which provides the suggested functional benefits. With regard to the lambeosaurine example discussed previously, the empty space would provide the chamber in which sound resonates. It now seems that it is this focus on the enclosed volume of air that has led us astray. Sinuses are not truly ‘‘empty,’’ but rather they contain the air sac itself—that is, the thin epithelial balloon that lines the bony recess. It turns out that the epithelium—not the empty space—may be the key. Pneumatic epithelium (and its associated tissues) appear to have the intrinsic capacity to expand and to pneumatize bone. Therefore, these air sacs may simply be opportunistic pneumatizing machines, expanding as much as possible—but within certain limits. These limits are provided by the local biomechanical loading regimes (i.e., the stresses and strains within the bone substance). Bone is sensitive to and responsive to its local stress environment, such as the forces involved in chewing or biting or laying down new bone to maintain sufficiently strong and stiff structures. Thus, there are two competing tendencies, one involving expanding air sacs and the other involving mechanically mediated bone deposition. A compromise is struck ensuring that pneumatization does not actually jeopardize the strength of the whole structure. An interesting consequence of this ‘‘battle’’ is that mechanically ‘‘optimal’’ structures will result as a secondary and completely incidental by-product—in other words, there is no reason to invoke natural selection to act directly to produce structures of maximal strength but minimal materials. Put simply, architecturally ‘‘elegant’’ structures are automatic outcomes of these two intrinsic, but opposite, processes. Thus, what is the function of the antorbital sinus

Craniofacial Air Sinus Systems or the diverticula of the middle ear sac? The answer is probably, at its base, no particular function whatsoever. That is not to say that any of these air sacs could not be secondarily pressed into service for some positive function, subsequently honed by natural selection. Such is likely the case for the resonating chambers of hadrosaurs and, additionally, aspects of the expanded tympanic cavities of theropod dinosaurs, which enhance hearing due to certain acoustic properties. However, at the same time, positive functions need not be sought for every pneumatic recess in every species that has them. Again, they are simply intrinsic properties of pneumatic systems. This new view on the function of pneumaticity helps explain certain trends in various dinosaur groups. For example, the reduction of the antorbital cavity in ornithopod ornithischians that was discussed previously (Fig. 2) clearly reflects a situation in which bone deposition ‘‘won out’’ over pneumatically-induced resorption of bone. The evolution and refinement of chewing in ornithopods required extensive bony buttressing and reinforcement, and so expansion of the antorbital sinus was severely limited. On the other hand, skull pneumatization in theropods appears to have progressed with few constraints and was clearly opportunistic. Although carnivory may involve fairly large bite stresses (force per unit area), the stresses and strains that a theropod skull as a whole underwent were minimal in comparison with the repetitive masticatory forces to which chewing animals such as ornithopods were subject. Thus, in theropods, sinuses were generally free to expand. It is significant, however, that the remaining bony struts in theropod skulls are positioned in mechanically advantageous locations, giving the appearance of exquisite design. It is also worth mentioning that Tyrannosaurus rex, a theropod that secondarily became adapted for particularly hard biting, increased the dimensions of some bony buttresses and bars to resist the stresses, yet it maintained pneumaticity—and even expanded some sinuses—by pneumatizing the bones internally, yielding a skull composed of a series of hollow, often truly tubular bars and plates.


Cretaceous Period

References Currie, P. J., and Zhao, X.-J. (1993a). A new carnosaur (Dinosauria, Theropoda) from the Jurassic of Xinjiang, People’s Republic of China. Can. J. Earth Sci. 30, 2037– 2081. Currie, P. J., and Zhao, X.-J. (1994b). A new troodontid (Dinosauria, Theropoda) braincase from the Dinosaur Park Formation (Campanian) of Alberta. Can. J. Earth Sci. 30, 2231–2247. Witmer, L. M. (1990). The craniofacial air sac system of Mesozoic birds (Aves). Zool. J. Linnean Soc. 100, 327–378. Witmer, L. M. (1995). Homology of facial structures in extant archosaurs (birds and crocodilians), with special reference to paranasal pneumaticity and nasal conchae. J. Morphol. 225, 269–327. Witmer, L. M. (1997). The evolution of the antorbital cavity of archosaurs: A study in soft-tissue reconstruction in the fossil record with an analysis of the function of pneumaticity. Mem. Soc. Vertebr. Paleontol (J. Vertebr. Paleontol.) 17(Suppl.), 1–75.


Cretaceous Extinction see EXTINCTION, CRETACEOUS

Cretaceous Period EVA B. KOPPELHUS Geological Survey of Denmark and Greenland Copenhagen, Denmark

The Cretaceous is the last period of the Mesozoic. It lasted for approximately 80 million years, ending 65 million years ago. The name is derived from the Latin word ‘‘creta,’’ which means ‘‘chalk,’’ and refers to the thick beds of Cretaceous chalk that are characteristic of parts of Europe.

159 Once divided into three epochs, the Cretaceous of Europe is now divided into the Early and Late Cretaceous, which are further subdivided into 12 stages. These were all defined on the basis of strata in the Anglo–Paris–Belgian area. In North America, the period is also divided into Early (Comanchean) and Late (Gulfian) Cretaceous. The earth’s climate was generally warm in the Cretaceous. Early in the period, conditions were becoming more humid and seasonal, which favored explosive diversification of the floras and the animals that fed on them. By the end of the Cretaceous, global cooling led to a drop in diversity of plants and animals in the higher latitudes. Throughout the Cretaceous, the continents drifted apart to approach their current positions. The Atlantic Ocean opened up, India reached Asia, and extensive inland seas subdivided the continental masses. Cretaceous rocks are widely distributed and exposed, and it is not surprising that almost half of the known dinosaurs have been found at this level. Hadrosaurs and ceratopsians in particular are common in the Late Cretaceous of the Northern Hemisphere. It was a time of great diversification, and dinosaurs shared their world with many other vertebrate groups that have changed little since the Cretaceous. Placental mammals, birds, snakes, and many other animals established the bauplans that are familiar to us today. Insects and other invertebrates were taking on a more modern appearance. It was a time when the terrestrial floras changed from being entirely dominated by pteridophytes and gymnosperms in the early part of the Cretaceous to angiosperms in the Late Cretaceous. The floral assemblages from the Late Jurassic persisted into the Early Cretaceous in the lower latitudes, whereas angiosperms became more prominent in the higher latitudes. By the early Late Cretaceous, there was a dramatic change into assemblages dominated by angiosperms. Four floral provinces based on pollen have been recognized in the Cenomanian: northern Laurasia, southern Laurasia, northern Gondwana, and southern Gondwana (Brenner, 1976). By Santonian–Campanian times, angiosperm pollen assemblages have been used to divide Laurasia into a southern Normapolles province and a northern Aquilapollenites province (Batten, 1984). The Cretaceous is a most interesting period from an evolutionary point of view. Toward the end of the period, however, things


160 changed dramatically. Ammonites and the great marine reptiles disappeared from the seas, pterosaurs disappeared from the air, and nonavian dinosaurs suffered a major extinction event.


References Batten, D. J. (1984). Palynology, climate and the development of floral provinces in the northern hemisphere: A review. In Fossils and Climate (P. Brenchley, Ed.), pp. 127–164. Wiley, London. Brenner, G. J. (1976). Middle Cretaceous floral provinces and early migration of angiosperms. In Origin and Early Evolution of Angiosperms (C. B. Beck, Ed.), pp. 23–47. Columbia Univ. Press, New York.

Crocodylia JAMES M. CLARK George Washington University Washington, DC, USA

The term Crocodylia was originally coined to encompass living crocodiles, but as fossil forms were discovered, they were also included. Many of these, however, were outside the group formed by living crocodiles (crown group Crocodylia), although obviously related to them. T. H. Huxley separated the fossil and recent Crocodylia into three groups: Protosuchia, Mesosuchia, and Eusuchia, distinguished by the posterior extent of their secondary palate and other features. These categories are now recognized not as monophyletic units but as grades of organization, the first two being paraphyletic. Most investigators use the term Crocodylomorpha to include these taxa plus the closely related Sphenosuchidae, which themselves may be paraphyletic relatives of crocodiles but share with crocodiles an elongated radial– ulnar and other features. Some workers use the term

Crocodylia to include ‘‘Protosuchia,’’ ‘‘Mesosuchia,’’ and ‘‘Eusuchia’’; others prefer to confine the term to crown group Crocodylia (a subset of Eusuchia). Because crocodilians and birds are each other’s closest living relatives, the evolutionary history of each group’s stem began at the same time, when they initially diverged from a common ancestor. The lineage leading to birds included dinosaurs and pterosaurs, whereas the lineage culminating in living crocodilians included a variety of extinct groups, such as phytosaurs, aetosaurs, rauisuchids, and a host of forms more or less similar to living crocodilians (see ARCHOSAURIA). Two names are currently applied to the lineage leading to crocodilians; the inaptly named PSEUDOSUCHIA (stem based) is defined as crocodilians and all extinct taxa more closely related to them than to birds, whereas the CRUROTARSI (node based) refers to the same group of taxa but is not based on the concept of representing the evolutionary limb leading to crocodilians (see PHYLOGENETIC SYSTEM). The crocodilian lineage first appears in the fossil record in the Middle Triassic, at about the same time the first fossils of the bird lineage (stem-based ORNITHOSUCHIA or node-based ORNITHODIRA) occur. Although the 25 or so living species of crocodilians are all semiaquatic and have a somewhat sprawling gait, their distant ancestors held their hindlimbs erect and were probably terrestrial. At the other extreme, one extinct group, thalattosuchians (Early Jurassic–Early Cretaceous), included forms that were apparently committed to an aquatic lifestyle. Many terrestrial groups were small, even smaller than the living dwarf crocodile (Osteolaemus), and two (Candidodon and Chimaerasuchus) developed mammal-like (cusped) teeth and may have chewed their food. The largest Pseudosuchian was Deinosuchus, a close relative of living crocodilians from the Late Cretaceous that may have reached 13 m (50 ft) in length. The first crown group crocodilians (i.e., those belonging to a living group of crocodilian) appear in the Late Cretaceous, and crocodilians apparently were little affected by the events surrounding the extinction of nonavian dinosaurs at the end of the Mesozoic.

See also the following related entry: ARCHOSAURIA

Crystal Palace

References Brochu, C., and Poe, S. Crocodylia section of the Tree of Life web site. http:/ /phylogeny.arizona.edu/tree/phylogeny. html. Clark, J. M. (1994). Patterns of evolution in Mesozoic Crocodyliformes. In In the Shadow of the Dinosaurs (N. C. Fraser and H.-D. Sues, Eds.), pp. 84–97. Cambridge Univ. Press, Cambridge, UK. Parrish, J. M. (1993). Phylogeny of the Crocodylotarsi, with reference to archosaurian and crurotarsan monophyly. J. Vertebr. Paleontol. 13(3), 287–308. Poe, S. (1997). Data set incongruence and the phylogeny of the Crocodylia. Systematic Biol., in press. Ross, C. A. (Ed.) (1989). Crocodiles and Alligators. Facts on File, New York. Sereno, P. C. (1991). Basal archosaurs: Phylogenetic relationships and functional implications. Soc. Vertebr. Paleontol. Mem. 2. Wu, X.-C., and Sues, H.-D. (1996). Anatomy and phylogenetic relationships of Chimaerasuchus paradoxus, an unusual crocodyliform reptile from the Lower Cretaceous of Hubei, China. J. Vertebr. Paleontol. 16(4), 688–702. Wu, X.-C., Brinkman, D. B., and Lu, J.-C. (1994). A new species of Shangungosuchus from the Lower Cretaceous of Inner Mongolia (China), with comments on S. chuhsienensis Young, 1961 and the phylogenetic position of the genus. J. Vertebr. Paleontol. 14(2), 210–229

Crystal Palace WILLIAM A. S. SARJEANT University of Saskatchewan Saskatoon, Saskatchewan, Canada

The creation of the world’s earliest three-dimensional restorations of dinosaurs was an event of the mid-19th century and the consequence of a suggestion by Prince Albert, consort of Queen Victoria. The prince had been much involved in developing the Great Exhibition, held in London in 1851. Its principal feature was a highly innovative prefabricated structure of glass and iron, the Crystal Palace. When the exhibition closed, it was decided that the Crystal Palace should be dismantled and re-erected in park-like grounds at Sydenham, on the south side of London. The prince had been greatly intrigued by Richard Owen’s accounts of antediluvian creatures, in partic-

161 ular the giant ground-sloth Megatherium; he suggested that life-sized models of these creatures might be displayed. Sir Joseph Paxton, who had designed the Crystal Palace, was also given charge of planning the new park. He decided to include in its southwest corner a ‘‘geological illustration’’ of the British Isles in the form of structures that would represent all major stratigraphical horizons, together with the principal mineral reserves. David T. Ansted (1814–1880), a leading economic geologist, served as geological adviser, while Richard Owen supervised the construction of the models of the extinct creatures that would be superposed on this geological landscape. Because Owen had himself named the dinosaurs a decade earlier, he made sure that they would be featured. As Doyle (1994) reports, The ‘‘Geological Illustrations’’ were logically arranged in three separate but interconnected parts of the park, surrounding a tidal lake which acted as a reservoir for the great central fountains of the park. The first part of the exhibit was a cliffline ‘‘exposing’’ a series of natural strata, representing the older rocks of northern England, Wales and Scotland. The exhibit was three-dimensional and dynamic; a water course issued from a spring line at the base of the Carboniferous (Mountain) Limestone. The limestone cliff in turn had a three-quarter scale reconstruction of a lead mine through which observers could pass. There was an accurate representation of the Coal Measures of Clay Cross [Derbyshire], complete with a dipping and fracture coal seam and associated beds of ironstone, and an unconformity, a discordance between two rock sequences indicating tectonic upheaval and erosion. None of these features represented a compromise of scientific accuracy, and all were dynamic in their impact. The lake itself contained the younger formations, set in a series of small islands. Scientifically the exhibits make sense; a fault can be ‘‘mapped-in’’ to explain the relationship of the younger strata in the lake with the older strata in the cliffline. On the islands were represented the great reptiles of the Secondary (Mesozoic) Era and the mammals of the Tertiary (Cenozoic) Era, each sited on the geologically most appropriate strata getting successively younger to the northeast. (pp. 7–8)

The task of restoring the extinct beasts was given to Benjamin Waterhouse Hawkins (1807–1889), who had been Paxton’s assistant superintendent. Hawkins was a skilled artist and sculptor; under Owen’s supervision, first he made small-scale models, then clay models at presumed life-size. From these, the final

Crystal Palace


FIGURE 1 ‘‘The Secondary Island’’ in Crystal Palace Park, London (from Anonymous, 1893).

restorations were made, but the task was not simple. As Hawkins reported (1854), Some of these models contained 30 tons of clay, which had to be supported on four legs. . . . In the instance of the Iguanodon, it is not less than building a house upon four columns, as the quantities of material of which the standing Iguanodon is composed, consist of 4 iron columns 9 feet long by 7 inches diameter, 600 bricks, 650 5-inch half-round draintiles, 900 plain tiles, 38 casks of cement, 90 casks of broken stone, making a total of 640 bushels of artificial stone. (p. 448)

By present-day standards, the restorations fail in accuracy. The Carboniferous Labyrinthodon was, in consequence of a misattribution and misreading of New Red Sandstone Chirotherium footprints, reconstructed as a toad-like creature, while the mammallike reptile Dicynodon was quite wrongly depicted— presumably on the basis of its almost toothless jaws—as a turtle-like creature with a carapace. The dinosaurs especially suffered from Owen’s misinterpretations. One must remember that the reconstructions were based only on partial skeletons and a presumed analogy with living reptiles. Nevertheless, to modern eyes, it is startling to see both Megalosaurus and Iguanodon depicted as extremely massive quadrupeds and the latter genus with a sharp, nasal horn. These errors were not unreasonable; rhinocerine lizards were well known to Owen

(and indeed, several genera of rhinocerine dinosaurs were discovered subsequently), whereas there were no living parallels to what was, in truth, Iguanodon’s spike-like thumb. As for the concept of dinosaur bipedality, that had not even arisen: Wealden (Early Cretacous) dinosaur tracks, though already reported by Tagart (1846), were still thought to be those of birds. The third dinosaur to be reconstructed, Hylaeosaurus, was correctly depicted as a quadruped. It is still poorly known, but is now considered to be an ankylosaur. The row of spikes, which Owen placed upon its back, is no longer acceptable; Hylaeosaurus seems instead to have had spikes directed laterally outward from its flanks. Debus believes Owen had a particular motive in shaping these models (see Desmond, 1979), considering (Debus, 1993) that Owen enlisted dinosaurs in his crusade to defend paleontology from the progressive evolutionists, who claimed species naturally transmuted into more advanced forms. If dinosaurs were, as Hawkins constructed them, grandiose Mesozoic overlords, this would imply that there had been no progression. (p. 12)

Before the park was opened to the public, a New Year’s dinner party was held on December 31, 1853. Its setting was unique because 12 of the 22 guests

Crystal Palace


FIGURE 2 The savants dining inside the unfinished Iguanodon (from The Illustrated London News, 1854).

actually dined inside the mould of the still-uncompleted Iguanodon, with the others at a table alongside. Spalding (1993) reports, Owen presided at the head of the table, which was appropriately located inside the head of the animal. Edward Forbes (1815–54; Professor of Natural History at the Museum of Practical Geology) wrote a song for the occasion, which the company sang with gusto. Thus the eminent cavorted, joked, and sang, and newspapers reported that their noise could be heard across the park. The humorous magazine Punch commented solemnly, in an article called ‘‘Fun in a Fossil,’’ that ‘‘if it had been an earlier geological period they might perhaps have occupied the Iguanodon’s inside without having any dinner there.’’ (pp. 66–67)

It is likely that the inspiration for this event came from Hawkins, whose father had attended a banquet hosted in 1802 by Charles Willson Peale inside the skeleton of a Mastodon discovered in New York State (Debus, 1993, p. 12). The park was officially opened

on June 10, 1854, to an estimated 40,000 visitors. The dinosaurs—set on islands in the lake—excited much attention because they afforded the first demonstration to the public at large of how awesomely huge those creatures were. This may well have been the true beginning of ‘‘dinomania’’ (see Torrens, 1993, p. 277). The exhibit established Hawkins’s reputation and brought many further opportunities for illustrating extinct creatures. Soon his work was becoming even better known through his illustrations in books and on educational wall-charts. Eventually he was invited to the United States, where he studied vertebrate remains in several major museums, making casts of dinosaur bones in Philadelphia and being invited to develop a series of dinosaur restorations for a proposed ‘‘Palaeozoic Museum’’ in Central Park, New York. Unfortunately, this project fell victim to the machinations of politicians (Colbert, 1959; Desmond, 1974); also, none of the casts of American vertebrates prepared by Hawkins are known to survive (Debus,

164 1993, pp. 11, 18). However, 15 (of an original 17) mural paintings by Hawkins of the life of past geological epochs are still to be seen in Guyot Hall, Princeton University. Although Sir Joseph Paxton’s project for the geological illustrations was never quite carried to completion (Doyle and Robinson, 1993) and the Crystal Palace itself burned down in 1936, Waterhouse Hawkins’s restorations still survive in the park in Sydenham. Somewhat oddly, they have been scheduled by the National Trust as ‘‘grade II listed buildings’’ and are protected against vandalism (McCarthy and Gilbert, 1994). They serve as a visible reminder of the earliest attempts at three-dimensional restoration of dinosaurs.

See also the following related entry: HISTORY OF DINOSAUR DISCOVERIES

References Anonymous. (1893). Crystal Palace: Illustrated Guide to the Palace and Park. Dickens and Evans, London. Colbert, E. H. (1959). The Palaeozoic Museum in Central Park, Or the Museum that Never Was. Curator 2, 137–150. Debus, A. A. (1993). The first great paleo-artist: Benjamin Waterhouse Hawkins. Earth Sci. News Bull. Earth Sci.

Crystal Palace Club Northern Illinois 44(September), 11–15; (October), 17–21. Desmond, A. J. (1974). Central Park’s fragile dinosaurs. Nat. History 63, 64–71. Desmond, A. J. (1979). Designing the dinosaur: Richard Owen’s response to Robert Edmond Grant. Isis 70, 224–234. Doyle, P. (1994). Crystal Palace revisited. Ethical Rec. 99, 7–8. Doyle, P., and Robinson, E. (1993). The Victorian ‘Geological Illustrations’ of Crystal Palace Park. Proc. Geol. Assoc. 104, 181–194. Hawkins, B. W. (1854). On visual education as applied to geology. J. Soc. Arts 2, 444–449. Illustrated London News (1854). January 7, p. 22. McCarthy, S. and Gilbert, M. (1994). The Crystal Palace Dinosaurs . . . The Story of the World’s First Prehistoric Sculptures. (London: Crystal Palace Foundation), 99 p. Spalding, D. A. E. (1993). Dinosaur Hunters, pp. ix, 310. Key Porter, Toronto. [Republ. by Prima, Rocklin, CA, 1995]. Tagart, E. (1846). On markings in the Hastings Sands near Hastings, supposed to be the footprints of birds. Q. J. Geol. Soc. London 2, 262.

Cultural Impact of Dinosaurs see POPULAR CULTURE, LITERATURE

D Dakota Dinosaur Museum, North Dakota, USA see MUSEUMS



Dallas Museum of Natural History, Texas, USA see MUSEUMS



Dalton Wells Quarry BROOKS B. BRITT Museum of Western Colorado Grand Junction, Colorado, USA

KENNETH L. STADTMAN Brigham Young University Provo, Utah, USA

The Dalton Wells dinosaur quarry is an extraordinarily rich deposit of fossil bone—rich not only in the number of bones but also in the number of dinosaurian taxa represented. Most of the fauna is undescribed but is currently being studied by the authors. Stratigraphically, the quarry is at the base of the Cedar Mountain Formation and is located in east-central Utah, near the town of Moab. The site is particularly important to paleontologists because it provides a window into the little known world of Early Cretaceous dinosaurs in North America. Casual collectors have known about this site for more than 50 years, and probably at least since the

1930s when a Civilian Conservation Corps camp was constructed less than 0.5 km from the bone-bearing layer. Mr. J. Leroy ‘‘Pop’’ Kay showed the site to James A. Jensen of Brigham Young University in the early 1960s when the site was thought to be within the Late Jurassic Morrison Formation. It was not until Lyn Ottinger discovered a partial, tooth-bearing iguanodontid maxilla that the significance of the site was recognized. Galton and Jensen (1979) made the maxilla the type of Iguanodon ottingeri and recognized that the fauna was Early Cretaceous in age. The deposit is a bone bed approximately 0.3 km long, consisting mainly of disarticulated elements of many animals, including juveniles and adults. The bones of several individuals, however, occur in clusters. Articulated elements are rare but include, significantly, a sauropod cranium with three cervical vertebrae. The bone bed varies from the thickness of a single bone to about 0.75 m. The matrix is a silty mudstone with occasional fine-grained chert pebbles, and the bone-bearing horizon is virtually devoid of internal sedimentary structures. The bone bed is tentatively interpreted to have been deposited in a volcanic ash-choked stream. The rate of flow was such that even large sauropod vertebrae were often severely broken. Medium-bedded fluvial sandstones overlie the quarry unit, which in turn are capped by mudcracked, limy mudstone, preserving sauropod and ornithopod footprints made along the shore of a small, Early Cretaceous lake. An analysis of the Dalton Wells fauna is still in the early stages, but with nearly 1000 prepared elements on hand six dinosaurian genera are recognized. Two genera from the locality have been described, the dromaeosaurid theropod Utahraptor (Kirkland et al., 1993) and Iguanodon ottingeri (Galton and Jensen, 1975, 1979). Bones of an ornithomimid theropod have been recovered and are currently being described. Two sauropod genera are present, a titanosaurid and a possible camarasaurid. These represent, respec-


Deccan Basalt

166 tively, the earliest and latest occurrences of these clades in North America. The titanosaurid is recognized on the basis of strongly procoeleous caudal vertebrae, dorsal vertebrae with reclined neural spines, and ulnae with short but robust olecrenon processes. Iguanodon ottingeri is regarded as a nomen dubium by Norman and Weishampel (1991), but at least one iguanodontid genus is present in the fauna. It is characterized by tall neural spines similar to those of Ouranosaurus. The Dalton Wells ornithopod is large, with an estimated length of 8 meters. A small nodosaur is also present and appears to be the same genus as the undescribed nodosaur found at the nearby Gaston Quarry. The only nondinosaurian taxon is a turtle, represented by a single carapace fragment.

Dayton Public Library Museum, Ohio, USA see MUSEUMS



Deccan Basalt ASHOK SAHNI Punjab University Chandigarh, India

The Deccan Traps constitute one of the most exten-

Kirkland, J. I., Burge, D., and Gaston, R. (1993). A large dromaeosaur (Theropoda) from the Lower Cretaceous of eastern Utah. Hunteria 2(10), 1–16.

sive continental flood basalt provinces of the Phanerozoic and have now been radiometrically shown to lie at the Cretaceous–Tertiary boundary. The basaltic flows are intercalated at the base with thin, highly fossiliferous sedimentary horizons that have yielded a diverse biota including mammals, dinosaurs, lower microvertebrates, and freshwater flora. Along the western coast of India, the exposed thickness of the Deccan Traps exceeds 2 km, but it gradually thins to the east. The Deccan Traps have been subdivided into three subgroups composed of compound and sheet flows. Magnetostratigraphy suggests a N-R-N sequence for the composite flow, with the majority of flows showing reverse polarity, probably corresponding to the 29R Chron. Deccan volcanism is considered to be one possible cause of the mass extinctions at the end of the Cretaceous.

Norman, D. B., and Weishampel, D. B. (1991). Iguanodontidae and related ornithopods. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 510–533. Univ. of California Press, Berkeley.


See also the following related entries: CEDAR MOUNTAIN FORMATION ● CRETACEOUS PERIOD

References Galton, P. M., and Jensen, J. A. (1975). Hypsilophodon and Iguanodon from the Lower Cretaceous of North America. Nature. Galton, P. M., and Jensen, J. A. (1979). Remains of ornithopod dinosaurs from the Lower Cretaceous of Colorado. Brigham Young Univ. Geol. Stud., 25, 1–10.

Gauthier (1986) cladistically defined and diagnosed Dashanpu see ZIGONG MUSEUM


Colbert and Russell’s (1969) term Deinonychosauria, including in it DROMAEOSAURIDAE and TROODONTIDAE. Gauthier’s definition was taxon based and so requires a stem- or node-based definition in the PHYLOGENETIC SYSTEM. Because Troodontidae has been determined to be closer to Ornithomimidae than to Dromaeosauridae (Holtz, 1994), the latter taxon carries approximately the same information as Deinonychosauria. Accordingly, it is appropriate to redefine Deinonychosauria as a stem-based taxon comprising all maniraptorans closer to Deinonychus than to birds; its sister taxon is AVIALAE.

Denver Museum of Natural History

References Colbert, E. H., and Russell, D. A. (1969). The small Cretaceous dinosaur Dromaeosaurus. Am. Museum Novitates 2380, 1–49. Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55. Holtz, T. R., Jr. (1994). The phylogenetic position of the Tyrannosauridae: Implications for theropod systematics. J. Paleontol. 68, 1100–1117.

Denver Museum of Natural History KENNETH CARPENTER Denver Museum of Natural History Denver, Colorado, USA

167 dinosaur specimens, mostly from the Morrison and the Hell Creek formations. Dinosaurs on exhibit include articulated skeletons of the small dinosaur Coelophysis, Stegosaurus, and Allosaurus (Fig. 1), Diplodocus, the skull of Brachiosaurus from Colorado, five juvenile skeletons of the ornithopod Othnielia, and a skeleton of the hadrosaur Edmontosaurus that shows a partially healed injury from the attack of a Tyrannosaurus. In addition, there is a walk-through diorama showing two life-sized male Stygimoloch pachycephalosaurs fighting for a female. Current research by the museum is concentrated on the vertebrate fauna of the Morrison Formation, especially the dinosaurs, at Can˜on City.

See also the following related entries: CAN˜ ON CITY ● HELL CREEK FORMATION ● MORRISON FORMATION

Located in Denver, Colorado, the Denver Museum of Natural History has had a long but sporadic history of collecting dinosaurs. The museum’s research collection is modest, holding only approximately 300

Department of Geology, University of Buffalo, New York, USA see MUSEUMS



Department of Geology, University of Texas at El Paso, Texas, USA see MUSEUMS

FIGURE 1 The skeletons of Allosaurus and Stegosaurus on exhibition at the Denver Museum of Natural History. Photo by Rick Wicker, courtesy of the Denver Museum of Natural History.



Devil’s Canyon scovery Center, Fruita, Colorado, USA see MUSEUMS




Devil’s Coulee Dinosaur Egg Historic Site CLIVE COY Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Devil’s Coulee is an unimposing pocket of Badlands that derives its name from a trident-shaped drainage system that ultimately feeds the Milk River in the southwestern corner of Alberta, Canada. In 1987, dinosaur embryos were discovered in eggs at Devil’s Coulee by staff of the Royal Tyrrell Museum of Palaeontology. The embryonic skeletons found together in the nest were described as a new type of hadrosaurian dinosaur called Hypacrosaurus stebingeri (Horner and Currie, 1994). Devil’s Coulee was the first dinosaur nesting site to be discovered in Canada and only the second in North America. Eggs and beautifully preserved embryonic skeletal material have now been found at numerous sites in Devil’s Coulee. Six types of eggshell have been identified, although only two of these are represented by nests of eggs. Although many dinosaur skeletons have been collected from Alberta during the past 100 years, they provide little evidence about the birth and development of dinosaurs. Study of the sediments of Devil’s Coulee is assisting in our understanding of the environment in which these animals nested and providing clues about the season during which nesting occurred. Examination of the nests and nesting behavior of these dinosaurs may eventually provide insight into the yearly routines of these animals. Speculations

Devil’s Coulee Dinosaur Egg Historic Site can be made regarding the animals living in groups, possibly moving as herds. As modern birds do today, dinosaurs appeared to have nested in great numbers in close proximity to each other in an attempt to simply overwhelm any predators. Three-dimensional preservation of the eggs guides our understanding of dinosaur egg physiology and indicates that not all the eggs were fertile. The excellent skeletal remains at Devil’s Coulee show a range of skeletons from embryos to nestlings. This sample of growth in a single species may indicate how often the dinosaurs reproduced, how long it took the young to develop, and what kinds of stresses (disease and predation) affected populations of duck-billed dinosaurs. Devil’s Coulee was purchased from the landowner by the government of Alberta and has been designated a protected historical resource, ensuring its safety for future generations. The Royal Tyrrell Museum has a continuing field excavation project at Devil’s Coulee and has collected hundreds of specimens there, including teeth of half a dozen dinosaur taxa, two new mammals, a new bird, turtles, and amphibians.

See also the following related entries: EGGS, EGGSHELLS, AND NESTS ● JUDITH RIVER WEDGE

References Horner, J. R., and Currie, P. J. (1994). Embryonic and neonatal morphology and ontogeny of a new species of Hypacrosaurus (Ornithischia, Lambeosauridae) from Montana and Alberta. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 312–336. Cambridge Univ. Press, Cambridge, UK.

Diet MICHAEL J. RYAN MATTHEW K. VICKARYOUS Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada


inosaurs can be defined as being primarily carnivorous (THEROPODA) or herbivorous (SAUROPODOMORPHA and ORNITHISCHIA) based on tooth morphology, tooth wear facets, inferred jaw mechanics, as well as general body morphology (see sections in this volume). Additionally, isotopic biogeochemistry, the preserved ‘‘stomach contents’’ of some fossils, coprolites, and the depositional environment in which dinosaurs are found, in association with the local plant mega- and microfossils, can help us to infer what they may have been eating. Dinosaur diets have long been closely tied to assumptions concerning their thermal physiology in terms of what types, quantities, and qualities of food they would have needed, as well as to what their internal food processing organs would have consisted of and how they would have functioned. This article attempts to synthesize the current understanding of dinosaur diets. For reviews of dietary habits, associated biomechanical functioning of the skulls, and interactions with food items see Galton (1986), Farlow (1976), Weishampel (1984), and Weishampel and Norman (1989). Living carnivores and herbivores have selected a wide variety of food items to meet their dietary needs which occasionally seem inconsistent with their skeletal characteristics (e.g., herbivorous pandas). In addition to the demands of specific physiologies, diets can be at least partially determined by age, social organization, habitat preference, and the availability of food, all of which are not easily determined from the fossil record. Palaeontologists parsimoniously assume that the most common food item associated with a dinosaur will be its most probable dietary food. However, we know from living animals that some taxa can be very selective of their source of food and may even go for long periods of time without eating, even in the presence of edible items (e.g., many constrictor snakes and whales). Thus, all speculations about dinosaurian diets should be taken with caution. Dinosaurs evolved from a carnivorous ornitho-

suchid ancestor sometime during the Early Triassic. By the time the first dinosaurs are recognized in the fossil record the two major dietary types appear to have been well established. Eoraptor lunensis, from the Ischigualasto Formation (Upper Triassic–middle Carnian), is perhaps the oldest known dinosaur and the basal member of the Theropoda. Its unique heterodont dentition has serrated, recurved crowns typical of theropods in the maxillae, whereas the lower teeth are leaf-shaped resembling those seen in basal sauropodomorphs (Sereno et al., 1993). Pisanosaurus mertii (Bonaparte, 1976), from the same formation, is the oldest known ornithischian. Its closely packed teeth suggest herbivorous habits.

Theropoda The generalized dinosaurian carnivore is a bipedal theropod with powerful legs designed for running, long forearms (reduced in tyrannosaurids) designed for some type of prey manipulation, and jaws with a wide gape and a large number of laterally compressed teeth. The teeth typically have serrations (denticles) on mesial and distal carinae. All carnosaurs appear to have evolved as active predators that could locate, track, and capture the appropriate Hunting styles and the size of prey items that theropods utilized have long been debated in the literature and will continue to be, probably without resolution. Late Cretaceous tyrannosaurids were either the same size or larger than their prey items, whereas allosaurs from the Jurassic were in some cases 10 times smaller than adult sauropods. If they were limited to prey items of equal size or smaller, the tyrannosaurs could have made a healthy living off of hadrosaurs and ceratopsians, but the allosaurs either fed on immature sauropods, stegosaurs, or campotosaurs or developed a method to bring down the massive sauropods. Auffenberg (1981) has suggested that the jaws of Allosaurus functioned like the canines of sabretooth cats and that these dinosaurs used a ‘‘hit-and-


170 run,’’ possibly pack-style, attack that allowed them to inflict deep wounds and then quickly withdraw and wait for their prey to succumb to its wounds. Trackway evidence suggests that some theropods hunted prey (sauropods) in packs. Large carnivores, such as Giganotosaurus or Tyrannosaurus rex, would have been well suited for active predation but they may also have intimidated other carnivores out of their actively hunted prey and subsisted by scavenging. In addition to ‘‘hit-and-run’’ attacks, some theropods may have suffocated their prey by clamping their jaws onto their necks or practiced ambushstyle attacks. Most large theropods would have had powerful mandibular adductors providing for a strong bite. Erickson et al. (1996) reported on a Triceratops pelvis with multiple bite and puncture marks attributed to T. rex. Their research concluded that T. rex had the greatest bite force of any animal measured to date (up to 13,400 N) and suggested that T. rex had strong, impact-resistant teeth that could regularly puncture the bones of prey. Broken teeth in dinosaurs would not have caused long-term problems because dinosaurs demonstrate a pattern of continual replacement. Tooth wear on large theropod teeth and tooth marks on a variety of dinosaur bones (see Jacobson, this volume) indicates that these dinosaurs regularly tore the flesh off carcasses. Shed theropod teeth associated with skeletons and bone beds suggests that carnivorous dinosaurs were feeding at these sites. Tooth shape in Carcharodontosaurus and tooth and jaw shape in Baryonyx and Spinosaurus suggests that these dinosaurs may have been piscivorous. Almost all theropods have hind limbs that are similar in shape and function. The femur is usually shorter than the tibia and designed for rapid running. The feet are functionally three-toed (except in therizinosaurids) and equipped with terminal claws that could have assisted in the killing and/or dismembering of prey. DROMAEOSAURS were small maniraptorans with a highly derived pedal digit II designed for hypertension and that terminated in an enlarged trenchant ungual. This claw seems to be designed to disembowel prey when the dromaeosaurid was either balanced on one foot or possibly leaping through the air. The forelimbs in CERATOSAURS, COELUROSAURS, and MANIRAPTORANS are strong and relatively long in order to assist in grappling prey. Maniraptorans probably possessed limited pronation/supination when the wrist was flexed or extended, allowing them at least

Diet to manipulate their food. One can visualize these theropods leaping at a large prey item, grasping hold with both hands, and slashing away with their raptorial claws. If these animals hunted in packs, as has been suggested by some authors (Ostrom, 1969), then no prey item may have been too large for them. Russell and Se´guin (1982) have suggested that Troodon formosus may have been crepuscular, preying on small, nocturnal mammals or lizards. Juvenile theropods probably preyed on different animals than they would as adults and may also have utilized different hunting strategies (e.g., pack hunting in juveniles vs solitary hunting in adults). The juveniles were more gracile than adults, and probably were fast, energetic hunters regardless of the hunting styles they may have adopted as adults. Invertebrates and small vertebrates may have constituted a large part of their diet (Farlow, 1976). Some theropods (Ornithomimosauria and Oviraptorosauria) were secondarily edentulous; the jaws were covered in a horny beak (now known from an ornithomimid; P. Currie, personal communication, 1996). The jaws of ornithomimids are considered to have been relatively weak, suggesting that their diet consisted of soft items such as insects or small mammals. Oviraptorids have jaws that are powerfully built and connected to an akinetic skull, suggesting an extremely strong bite force. They probably fed on small lizards and mammals as well as insects and eggs (Fig. 1).

FIGURE 1 Oviraptor mongoliensis (© S. R. Bissette, 1996).



FIGURE 2 Therizinosaurus, a therizinosaurid from Asia (© S. R. Bissette, 1996).

Stomach contents are rarely preserved in theropods but are known for Compsognathus (Ostrom, 1978), containing the small lepidosaur Bavarisaurus, and small fragments of bone mesial to the gastralia of Syntarus rhodesiensis (Raath, 1969). A new, feathered compsognathid dinosaur from China also has a small lizard preserved in its stomach cavity (P. Currie, personal communication, 1996). Cannibalism has been reported for Coelophysis from the bone beds of Ghost Ranch, New Mexico. The Therizinosauridae (⫽ Segnosauridae) (Fig. 2) are a unique group of theropods with long arms and greatly elongated manual unguals. Various authors have suggested that these dinosaurs were piscivorous or even herbivorous.

Sauropodomorphs and Ornithischians All sauropodomorphs and all ornithischians are believed to have been primarily or exclusively herbivorous. These dinosaurs can be divided broadly into two groups, gut processors and mouth processors. Gut processors are characterized by simple dentitions and are assumed to have had modified guts for the digestion of high-fiber, low-nutrition plants. Some appear to have had gastroliths to assist in the breakdown of plant matter. Fermentation in the gut would have further broken down the fibrous plant material and produced nutrients that subsequently were absorbed by the host. Mouth processors have modified dentitions and/or jaw structure/mechanics that allowed for active grinding or pulping of low-fiber, high-nutrition foods. Prosauropods and most ornithischians are believed to have possessed cheeks that allowed them to contain food in the mouth while they processed it.

When dinosaurs first appeared in the Middle to Late Triassic, the flora still contained some Late Paleozoic members (e.g., herbaceous lycopods, large arborescent and small herbaceous horsetails, and possibly some conifers such as Brachyphyllum) but was starting to take on the more gymnosperm-dominated aspect typical of late Mesozoic. These plants included a variety of ferns, true cycads, cycadeoids, ginkgos, conifers, and the seed fern order Caytoniales. These groups dominated the Jurassic and Early Cretaceous, varying in diversity and numbers depending on latitude and the local environment. By the Middle Jurassic such modern conifer families as Araucariaceae, Pinaceae, and Taxodiaxeae had appeared. Throughout the Jurassic and the Early Cretaceous, the sauropods were the dominant herbivores. Large groups of sauropods would have consumed immense quantities of plant matter and substantially impacted local environments. Some authors (e.g., Bakker, 1978) have suggested that by clearing large amounts of upperstory foliage, the underbrush was opened up to be exploited by the fast-growing, ‘‘weedy’’ angiosperms, which then quickly evolved and out-competed their gymnosperm relatives. By the Early Cretaceous (Valanginian) angiosperm pollen is present in the fossil record. The dramatic increase in angiosperm floras from the Barremian onward came at the expense of the cycads, cycadeoids, and the seed ferns. The evolution of the Late Cretaceous angiospermdominated forests at least parallels the evolution of the primary Late Cretaceous herbivores, and the Cerapoda (specifically the ornithopods and the ceratopsians) effectively replaced the sauropods in the Northern Hemisphere (see Basinger, this volume, for a complete review of Mesozoic floras).

172 The Prosauropoda first appeared in the Carnian and represent the first radiation of herbivorous dinosaurs. Prosauropods and the giant sauropods that replaced them in the Late Jurassic are considered to have been primarily gut processors, and gastroliths are known from Massospondylus, Sellosaurus, and Seismosaurus. The skulls are lightly built with a relatively weak jaw musculature. Prosauropod teeth are narrow, subconical, and closely spaced, showing little in the way of wear facets, indicating that the teeth did not regularly occlude. This suggests that they lived on relatively soft plants. A variety of sauropod families have teeth with distinctive wear patterns (see Fiorillo and Weishampel, TOOTH WEAR, this volume) indicating that different plant groups may have been utilized by different sauropods. Prosauropods show a modest lengthening of the cervical vertebrae that would have allowed them to reach plants as high as 3 m, making them the dominant high-browsing animals of the Triassic. They may have fed on lycopsid fructifications (Weishampel, 1984). Likewise, the later sauropods exhibit extreme elongation of the neck and were designed to be high browsers. Some sauropods, such as the diplodocids Apatosaurus and Diplodocus, may have been able to achieve a tripodal stance using the long tail as a counterbalance to reach plant material more than 18 m above the ground (Bakker, 1971). Sauropod food groups may have included ferns, ginkgophytes, conifers, nilssonian fructifications, and possibly Czekanowskiales and Caytoniales. Late Cretaceous forms may have also fed on angiosperm fructifications (Weishampel, 1984). The Ornithischia show a progressive increase in the complexity of the dentition and associated jaw mechanics from their first appearance in the Late Triassic until the end of the Cretaceous. All ornithischians appear to utilize some degree of oral processing. Basal ornithischians such as Pisanosaurus have closely packed teeth showing continuous wear, foreshadowing the dental batteries of more derived ornithischians. These early herbivores would have foraged on low-level ground cover. Their apparent lack of gastroliths suggests that these herbivores may have utilized food with lower fiber than the gut processors. The Thyreophora includes ankylosaurs, stegosaurs, and their basal relatives Scelidosaurus, Scutellosaurus, and Emausaurus. Like the prosauropods, which they functionally replaced, and the sauropods,

Diet thyreophorans were mostly obligatory quadrupeds with simple spatulate dentition and were probably primarily gut processors. Other than in the basal Scutellosaurus, most thyreophorans show evidence of tooth wear from imprecise occlusion. Oral processing was likely limited to slicing and/or puncturing and crushing. Diet was probably restricted to relatively low-lying plants (e.g., 1 or 2 m above the ground). Most authors believe that these dinosaurs fed on nonabrasive, ‘‘soft’’ plants that may have included the fleshy components of bennettitalian, nilsonnialian (Jurassic), caytonialian inflorescences (Jurassic to Cretaceous), and angiosperm fructifications (Cretaceous) (Weishampel, 1984). Among the Marginocephalia, the pachycephalosaurs and, to a lesser degree, the psittacosaurs had dentitions very similar to those of thyreophorans and probably also relied on gut processing to a high degree. However, their bipedal lifestyle would have allowed them the ability to move more rapidly across the landscape, not unlike modern deer, perhaps in search of different, higher quality foods. Ornithopods were the first group of herbivores to develop a transverse chewing stroke (side-to-side grinding), either by the slight rotation of the lower jaw as in heterodontosaurids or through the rotation of the upper jaw via pleurokinesis as in hypsilophodontids, iguanodontids, and hadrosaurids (Weishampel and Norman, 1989). This progression from simple isognathy (bilateral occlusion), seen in basal forms, to an increasingly more complex series of jaw mechanics and tooth/tooth row structures in the more advanced Iguanodontia allowed for more extensive processing of high-quality, low-fiber food. The basal ornithopods (Heterodontosauridae, Hypsilophodontidae, Tenontosaurus, and Dryosauridae) tended to be relatively small and were probably browsers of low (ⱕ2 m and under) ground cover. Typically, the rostral portion of their premaxilla was narrow, edentulous, and covered by a cornified rhamphotheca suggesting that these dinosaurs selectively cropped their food. The more derived Euornithopoda (Hadrosauridae and Iguanodontidae) show a more advanced adaptation to herbivory. Their teeth are generally narrow and buccally lanceolate (maxillary) or leaf-shaped (dentary). Their diet was probably a mix of gymnosperms and the proliferating late Mesozoic angiosperms. All members of the Hadrosauridae are character-

Diet ized by having broad, edentulous, ‘‘duck-billed beaks’’ and a complex maxillary/dentary dentition organized into batteries of up to several hundred closely packed teeth. Dental batteries formed a single wear facet at the point of occlusion. Hadrosaur teeth differ from those of other ornithopods: they are taller than wide and have lancolate crowns bearing heavy enamel, required for efficiently grinding plant matter. Hadrosaurs appear to have been well suited to living off low-quality, high-fiber vegetation (Weishampel and Norman, 1989). Some authors have argued that some hadrosaurs invested in parental care of their young and may have brought food or provided regurgitate for nest-bound young. Several hadrosaurs have been collected that have putative stomach contents (see Currie et al., 1995) located in the thoracic cavity. The best studied consists of 1- to 4-cm-long sections of 5- to 7-year-old twigs from angiosperms and gymnosperms as well as seeds and seed pods. Although this material would be consistent with food processed through the jaws of a large hadrosaur, the authors could not rule out the possibility that this material was washed in after death. All members of the Ceratopsia (Psittacosauridae, Protoceratopsidae, and Ceratopsidae) have a mouth that terminates in an edentulous parrot-like beak formed from the rostral and predentary bones. The surface of each would have been cornified and formed a sharp cutting surface. All ceratopsians have leaf-shaped teeth in both maxillae and dentaries. Psittacosaurs have a single tooth row of low, leaf-shaped teeth characterized by broad planar wear surfaces with self-sharpening cutting edges. Polished gastroliths have been associated with some skeletons, suggesting that gastric mills may have played some role in processing food. Although the protoceratopsids have only a single replacement tooth present per position, the Ceratopsidae have a dental battery similar to, but less extensive than, that of hadrosaurs. Both protoceratopsids and ceratopsids have vertically inclined wear facets on the teeth, indicating that during chewing the power stroke was restricted to orthal slicing movements. Ceratopsians may have browsed at a level of approximately 2 m and under. Their sharp beaks appear to have been well adapted to cropping off small trees and even processing them whole. Their large guts may have housed large fermentary en-

173 gines. Palynological evidence suggests that Triceratops may have subsisted on the small herbivorous plant, Gunnera (Rich, 1996). Isotopic (C13 and N15) examinations of high-molecular-weight material isolated from fossils have recently been used to infer possible diets for some dinosaurs. Isotopic work by Ostrom et al. (1990) suggests that some ceratopsids may have been omnivorous. Coprolites have been attributed to a variety of dinosaurs, but their use in determining dinosaur dietary behaviour has been limited to date. Coprolites attributed to herbivores can be used to infer dietary fiber content and food quality based on the presence or absence of various plant tissues. The presence of bone fragments, scales, or teeth can indicate carnivorous excrement. A coprolite from the Maastrichtian Frenchman Formation of Saskatchewan with bone fragment inclusions from what appears to be a sub-adult ornithischian has been attributed to T. rex based in part on its large size.


References Auffenberg, W. (1981). The Behavioral Ecology of the Komodo Monitor, pp. 406. Univ. Presses of Florida, Gainesville. Bakker, R. T. (1978). Dinosaur feeding behavior and the origin of flowering plants. Nature 274, 661–663. Bonaparte, J. F. (1976). Pisanosaurus mertii Casamiquela and the origin of the Ornithischia. J. Paleontol. 50, 808–820. Currie, P. J., Koppelhus, E. B., and Muhammad, A. F. (1995). ‘‘Stomach’’ contents of a hadrosaur from the Dinosaur Park Formation (Campanian, Upper Cretaceous) of Alberta, Canada. In Sixth Symposium on Mesozoic Terrestrial Ecosystems and Biota, Short Papers (A. Sun and Y. Wang, Ed.), pp. 111–114. Beijing, China. Erickson, G. M., Van Kirk, S. D., Su, J., Levenston, M. C., Caler, W. E., and Carter, D. R. (1996). Bite-force estimation for Tyrannosaurus rex from tooth-marked bones. Nature 382, 706–708. Farlow, J. O. (1976). Speculations about the diet and foraging behavior of large carnivorous dinosaurs. Am. Midland Nat. 95, 186–191.


174 Galton, P. M. (1986). Herbivorous adaptations of Late Triassic and Early Jurassic dinosaurs. In The Beginning of the Age of Dinosaurs (K. Padian, Ed.), pp. 203–221. Cambridge Univ. Press, New York. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Bull. Peabody Nat. History Museum 30, 1–165.

Dinosaur Discovery Center, Can˜on City, Colorado, USA see MUSEUMS



Ostrom, J. H. (1978). The osteology of Compsognathus longipes Wagner. Zitteliana 4, 73–118. Ostrom, P. H., Macko, S. A., Engel, M. H., Silfer, J. A., and Russell, D. A. (1990). Geochemical characterization of high molecular weight material isolated from Late Cretaceous fossils. Org. Geochem. 16, 1139–1144. Raath, M. A. (1969). A new coelosaurian dinosaur form the Forest Sandstone of Rhodesia. Arnoldia 4, 1–25.

Dinosaur Eggs see EGGS, EGGSHELLS,



Rich, F. J. (1996). Palynological interpretation of matrix associated with a Triceratops burial site, Lance Creek Area, Wyoming. Geol. Soc. Am. 28(4), 36. [Abstracts with programs] Russell, D. A., and Se´guin, R. (1982). Reconstruction of the small Cretaceous theropod Stenonychosaurus inequalis and a hypothetical dinosauroid. Syllogeus 37, 1–43. Sereno, P. C., Forster, C. A., Rogers, R. R., and Monetta, A. M. (1993). Primitive dinosaur skeleton from Argentina and the early evolution of Dinosauria. Nature 361, 64–66. Weishampel, D. B. (1984). Interactions between Mesozoic plants and vertebrates: Fructifications and seed predation. Neues Jahrbuch Geol. Pala¨ontol. Abhandlungen 167, 224–250. Weishampel, D. B., and Norman, D. B. (1989). Vertebrate herbivory in the Mesozoic: Jaws, plants, and evolutionary metrics [Special paper]. Geol. Soc. Am. 238, 87–100.





Dinosauria: Definition KEVIN PADIAN University of California Berkeley, California, USA


he first use of the word ‘‘dinosaur’’ was in 1842, when the great British comparative anatomist and paleontologist, Richard Owen (Fig. 1), applied it to three partially known but impressively large fossil reptiles from the English countryside (see HISTORY: EARLY DISCOVERIES). The Dinosauria were inaugurated in a published version of a lecture on British fossil reptiles that Owen had given to the British Association for the Advancement of Science in Plymouth in August 1841. However, in the lecture, which went on for more than 2 hr, Owen evidently did not use the word Dinosauria because it was not reported in any account of the lecture (Torrens, 1992); its origin must be traced to the updated version that appeared in the report of the meeting, published in April 1842. The exact date of the first use of the term Dinosauria is perhaps not as important as what it meant at the time and why Owen erected it. He based the Dinosauria on three previously named taxa: the large carnivore Megalosaurus, which Buckland had de-

FIGURE 1 Portrait of Richard Owen, ca. 1858.

scribed in 1824; the ornithopod Iguanodon, the teeth of which were described by Mantell in 1822 and first recognized from a giant reptile in 1825; and the armored Hylaeosaurus, which Mantell had described in 1833 (Fig. 2). He did not include other material that could reasonably have been considered, such as Palaeosaurus, a tooth of uncertain origin (probably pseudosuchian), and Thecodontosaurus, known from a toothed lower jaw and perhaps some fragmentary material (now called ‘‘Palaeosauriscus’’), both described by Riley and Stutchbury in 1836, as well as Cetiosaurus, which was described by Owen himself in 1842 (and at that time considered a kind of giant crocodile-like aquatic form with clawed, webbed feet and a swimming tail). He also did not consider any continental material, such as the nomenclatorially problematic taxon Streptospondylus, which at the time was based on both British and continental material (and which Owen considered a crocodile because the referred material included both crocodilian and dinosaurian remains); Plateosaurus, known from fragmentary remains described by Meyer in 1837; or Poikilopleuron, considered a megalosaurid by EudesDeslongchamps in 1838. Owen erected the Dinosauria (‘‘fearfully great lizards’’ as he translated the Greek) to receive these taxa because he recognized that they were completely distinct from other reptiles. They were large, but other fossil reptiles, including mosasaurs, plesiosaurs, ichthyosaurs, and some crocodiles, were also large; the dinosaurs, however, were terrestrial, not aquatic. Owen pointed to the five fused sacral vertebrae and the hips structured so that the animals demonstrably walked upright (Fig. 3; see PELVIS). These were not like living reptiles, and not only because of their size: They could not have sprawled. He also pointed to the height, breadth, and sculpturing of the dorsal neural arches; the two-headed ribs; the ‘‘broad and sometimes complicated coracoids and long and slender clavicles’’; and the proportionally large but thinwalled limb bones that indicated terrestrial habits.



Dinosauria: Definition

FIGURE 3 Owen’s 1854 reconstruction of Megalosaurus as a quadruped. The lower jaw and other available skeletal bones are indicated.

FIGURE 2 The jaw of Megalosaurus, part of the type specimen; Mantell’s specimen of the Maidstone Iguanodon, on which he based his restoration; and the type specimen of Hylaeosaurus described by Mantell.

Part of Owen’s motivation in constructing this new taxon may not have been entirely taxonomic. In those early Victorian days, the concept of evolution had many meanings, and both materialistic and idealistic theories were proposed to explain it. The pattern of evolution, to many, meant continuous progress

through time, from the earliest humble beginnings of life through the rise of the vertebrates to the ascent of mammals and their pinnacle, Homo sapiens. Owen did not accept progressivism because he knew from the fossil record that new forms continued to be produced, and life was not a ladder of ascending complexity. However, by showing that some extinct reptiles were more ‘‘advanced’’ structurally than living reptiles—that is, that they approached the mammalian and avian grades of organization—he could deny the validity of progressivism and its easy connection to materialistic transmutation, advocated by such scientists as Owen’s rival at University College London, Robert Edmond Grant (Desmond, 1979). Regardless of motivation, Owen’s concept of the Dinosauria took hold, and the unveiling of Waterhouse Hawkins’ statues of these dinosaurs at the CRYSTAL PALACE EXPOSITION in 1854 sealed their fixation in the minds of the public. Although almost no other new dinosaurs would be discovered in Britain until the 1970s, it will always be regarded as the birthplace of the dinosaurs. In the late 19th and early 20th centuries, more dinosaurs were discovered on the continent, including complete skeletons of Iguanodon in Belgium and of Plateosaurus in Germany. However, notably in the 1850s, the first dinosaurs in North America were discovered in New Jersey and soon after in Montana. By the 1880s, dinosaurs were well known from the western United States, and from both theropod and ornithopod remains it became clear that Owen was more correct than he had supposed: Many dinosaurs not only walked upright but also their short forelimbs proved that they walked bipedally—habits that had never been seen in a reptile (Desmond, 1975; see BIPEDALITY).

Dinosauria: Definition The concept of Dinosauria altered radically once again in 1887 and 1888 when Owen, now in his eighties and retired from the Natural History Museum that he had founded, was only a few years from his death. Harry Govier Seeley, a former student of Owen and an expert on both PTEROSAURS and anomodont dicynodont therapsids, surveyed the known skeletal material of dinosaurs and concluded that there were two consistent types of pelvic structure in this group. One type, which characterized a group that he named SAURISCHIA, was much like those of other reptiles, with a pubis directed mostly anteriorly and an ischium directed mostly posteriorly. The other type, which characterized a group that he named ORNITHISCHIA, superficially resembled the pelves of birds in having the shaft of the pubis retroverted to lie posteriorly next to the ischium; in most forms a new prong of the pubis developed anteriorly and outwardly, unlike the original anteroventrally and medially directed pubic shaft (see PELVIS). Seeley (1887, 1888) concluded that the Dinosauria was really composed of two separate groups, and so he erected Saurischia and Ornithischia as orders of the Reptilia. In retrospect, all that Seeley had done was to recognize the distinctness of Ornithischia and Saurischia based on the structure of the pelvis, vertebrae, braincase, and armor. In deconstructing the Dinosauria, he was neglecting the similarities in the vertebrae, pelvis, and hindlimb, among other structures, that Owen had noted in the animals on which he had originally based the taxon. However, Seeley could have been expected to treat those similarities as simply convergences, exigencies of large size and terrestrial living. He had used the same argument several years before in disputing T. H. Huxley’s contention that birds were allied to dinosaurs (specifically theropods such as Megalosaurus and Allosaurus), asking why the similarities could not be merely convergences associated with bipedality, without indicating particularly close relationship (Desmond, 1982). Seeley’s question, which as Desmond (1982) notes probably ‘‘stemmed more from his love of morphological tabulation than any evolutionary imperative,’’ nonetheless influenced paleontology for most of the ensuing century. The Dinosauria was no longer regarded universally as a natural group, and often the term ‘‘dinosaurs’’ was only informally used by paleontologists. Even within Saurischia, there were frequent doubts that Sauropoda and Theropoda had a

177 particularly close relationship. Often, Saurischia was broken into three component groups, listed as if of equal rank: PROSAUROPODA, SAUROPODA, and THEROPODA. In all three editions of A. S. Romer’s influential textbook Vertebrate Paleontology (1933, 1945, 1966), Dinosauria was not listed as a taxon, and Saurischia and Ornithischia were simply listed as orders of Reptilia. Until the 1970s, the question became largely a matter of individual judgment hinging largely on the taxonomic weight placed on various evolutionary similarities and differences among the various ‘‘dinosaurian’’ subgroups. The 1970s brought the first stirrings of cladistics (phylogenetic systematics; see PHYLOGENETIC SYSTEM; SYSTEMATICS), a methodology that has since fundamentally changed not only the practice of taxonomy but also the approach to comparative biology as a whole. However, the initial paper that reconsidered the monophyly of Dinosauria was not primarily cladistic in its thrust but rather paleophysiological. R. T. Bakker and P. M. Galton (1974) argued that Dinosauria, including the birds, should be elevated to a new class of vertebrates, united by a suite of morphological features related to upright stance, bipedality, and a metabolic level elevated above those of typical reptiles. These arguments were based on the revival and general acceptance of the hypothesis that birds descended from theropod dinosaurs; on renewed comparative studies of the origins of dinosaurs augmented by the new discoveries of small Late Triassic archosaurs (Lagosuchus and Lagerpeton) with obvious cursorial capabilities and many similarities to early dinosaurs (see DINOSAUROMORPHA); and on considerable circumstantial evidence from bone histology, zoogeography, functional morphology, and inferences about behavior that suggested strongly that if dinosaurs were not exactly like living birds and mammals, they were more like them than like living reptiles (see PHYSIOLOGY). The debate raged for another decade, particularly regarding the last conclusion, and has flared episodically in various new forms since then (Thomas and Olson, 1980). However, if the taxonomic premise of Bakker and Galton’s argument, that Dinosauria (including birds) should be considered a new class of vertebrates, was not generally accepted, the phylogenetic conclusion, that Dinosauria was monophyletic and included birds, almost universally was. For most workers, the question of dinosaurian pa-

Dinosauria: Definition

178 leobiology must be separated from the question of their systematic identity. Phylogenetic systematics groups organisms into hierarchically arranged taxa, based on the distribution of shared derived characters, or synapomorphies (see SYSTEMATICS). In the 1980s this approach was first extensively applied to the groups hypothesized to comprise Dinosauria. The most thorough analysis, and most seminal from the standpoint of later work, was by Jacques Gauthier (1986), who, in the course of trying to decide this question, brought the levels of analysis below dinosaurs to Archosauria (the crown group formed by birds and crocodiles) and down to the level of Diapsida and its immediate outgroups (Gauthier, 1984). Gauthier determined that Dinosauria, including the monophyletic groups Ornithischia and Saurischia, was itself monophyletic, united by a suite of nine synapomorphies of the skull, shoulder, hand, hip, and hindlimb. Later workers have reanalyzed and modified this original listing, but the node has remained robust. The question that ultimately arises from these historical considerations is the following: What, finally, are dinosaurs, and can this taxon have any stability? In this work, we follow the basic principles of the phylogenetic system developed in a series of papers by de Queiroz and Gauthier (1990, 1992, 1994)—as opposed to the traditional Linnean system and the attendant rules of nomenclature followed by the International Commission on Zoological Nomenclature. We omit traditional categories of hierarchical rank, such as order and family, and focus on the monophyly of taxa and their relationships to other taxa. The monophyly of a taxon depends on an adequate definition and an adequate diagnosis (see PHYLOGENETIC SYSTEM). The diagnosis of a taxon is a matter of determining synapomorphies that apply to it; synapomorphies are hypotheses of homologous characters that obtain at particular hierarchical levels, but with increasing knowledge these may be shown to apply to more or less general hierarchical levels or they may be found to be homoplasies (convergences). Hence the stability of a taxon rests more on its definition than on its diagnosis. Definitions may be stem-based, nodebased, taxon-based, or apomorphy-based (see PHYLOGENETIC SYSTEM); as the phylogenetic system has progressed, it turns out that the former two are far preferable to the latter two in the interests of stability. However, many taxa have been based cladistically

on lists of taxa or on one or more presumed synapomorphies, and although their priority should be respected when at all possible, in some cases they need to be adjusted for uniformity and ease of use. Gauthier (1986, p. 44) appears to have defined Dinosauria in a taxon-based sense as ‘‘Herrerasauridae*, Ornithischia, Saurischia, Sauropodomorpha, and Theropoda—including birds’’ (the asterisk denoted a metataxon, or a taxon with no synapomorphies of its own). However, from other contexts in the same work, including the diagnosis he gave of Dinosauria (1986, p. 45), it is clear that he meant to exclude Herrerasauridae per se. The diagnosis of Dinosauria begins by recognizing the Ornithischia, Sauropodomorpha, and Theropoda as monophyletic separately and as a group, and it ends by recognizing herrerasaurs and other ornithodirans as successively more remote outgroups (see also Gauthier, 1986, pp. 14–15). In either case, however, this is a taxon-based diagnosis. Recognizing this, Padian and May (1993) proposed ‘‘to define Ornithischia as all those dinosaurs closer to Triceratops than to birds, and Saurischia as all dinosaurs closer to birds than to Ornithischia. Dinosauria is defined as all descendants of the most recent common ancestor of birds and Triceratops.’’ This made the taxa stem based and node based, respectively. Regrettably, it was only later that T. R. Holtz (personal communication) suggested that the first two described dinosaurs, Megalosaurus and Iguanodon, included in Owen’s original Dinosauria, would have been more fitting end members than birds and Triceratops! As dinosaurian phylogeny is currently understood, this would have made no difference to the membership of the group, and it would have paid homage to Owen’s foresight. To be a dinosaur, then, according to current definition within the phylogenetic system, a given animal must be a member of the group descended from the most recent common ancestor of birds and Triceratops. The diagnosis of this group, and its membership, will change as we learn more about the included taxa and modify the distributions of synapomorphies accordingly. However, what cannot change in the phylogenetic system is the valid definition of Dinosauria.




References Bakker, R. T., and Galton, P. M. (1974). Dinosaur monophyly and a new class of vertebrates. Nature 248, 168–172. de Queiroz, K., and Gauthier, J. (1990). Phylogeny as a central principle in taxonomy: Phylogenetic definitions of taxon names. Systematic Zool. 39, 307–322. de Queiroz, K., and Gauthier, J. (1992). Phylogenetic taxonomy. Annu. Rev. Ecol. Systematics 23, 449–480. de Queiroz, K., and Gauthier, J. (1994). Toward a phylogenetic system of biological nomenclature. Trends Ecol. Evol. 9, 27–31. Desmond, A. J. (1975). The Hot-Blooded Dinosaurs: A Revolution in Palaeontology. Blond & Briggs, London. Desmond, A. J. (1979). Designing the dinosaur: Richard Owen’s response to Robert Edmond Grant. Isis 70, 224–234. Desmond, A. J. (1982). Archetypes and Ancestors: Palaeontology in Victorian London, 1850–1875. Muller, London. Gauthier, J. A. (1984). A cladistic analysis of the higher systematic categories of Diapsida. Unpublished Ph.D. dissertation, Department of Paleontology, University of California, Berkeley. Gauthier, J. A. (1986). Saurischian monophyly and the origin of birds. Mem. California Acad. Sci. 8, 1–55. Owen, R. (1842). Report on British fossil reptiles, Part II. Rep. Br. Assoc. Adv. Sci. 1841 60–294. Padian, K., and May, C. L. (1993). The earliest dinosaurs. New Mexico Museum Nat. History Sci. Bull. 3, 379–381. Romer, A. S. (1933). Vertebrate Paleontology, 1st ed. Univ. of Chicago Press, Chicago.

Dinosaur National Monument DANIEL J. CHURE Dinosaur National Monument Jensen, Utah, USA

Dinosaur National Monument is a unit of the U.S. National Park Service that straddles the Utah– Colorado border. It was established in 1915 to protect Upper Jurassic dinosaur fossils of the Morrison Formation, then being excavated by the Carnegie Museum. Some 350 tons of fossils were shipped back to the Carnegie from the discovery of the quarry in 1909 to the cessation of excavations in 1924. The remains of several hundred dinosaurs belonging to 10 genera have been found in the quarry, making it the most diverse Upper Jurassic dinosaur site in the world. Part of the quarry, with some 1500 dinosaur bones prepared in high relief and left in situ, is enclosed within the Quarry Visitor Center and can be viewed by the public. Recent backcountry excavations have uncovered a diverse Morrison flora and fauna including mammals, lizards, sphenodontians, turtles, and some of the earliest known frogs and salamanders.


Romer, A. S. (1945). Vertebrate Paleontology, 2nd ed. Univ. of Chicago Press, Chicago. Romer, A. S. (1966). Vertebrate Paleontology, 3rd ed. Univ. of Chicago Press, Chicago. Seeley, H. G. (1887). On the classification of the fossil animals commonly named Dinosauria. Proc. R. Soc. London 43(206), 165–171. Seeley, H. G. (1888). The classification of the Dinosauria. Rep. Br. Assoc. Adv. Sci. 1887, 698–699. Thomas, R. D. K., and Olson, E. C. (Eds.) (1980). A Cold Look at the Warm-Blooded Dinosaurs. Westview Press, Boulder, CO. Torrens, H. (1992, April 4). When did the dinosaur get its name? New Scientist, 40–44.

Dinosaur Museum, Dorchester, United Kingdom see MUSEUMS



Dinosauromorpha ANDREA B. ARCUCCI Universidad Nacional de La Rioja La Rioja, Argentina

Sereno (1991) defined Dinosauromorpha as ‘‘ornithodirans more closely related to the dinosaur-avian clade than to pterosaurs.’’ Apart from dinosaurs themselves, this includes taxa informally known as ‘‘lagosuchids,’’ small, long-limbed carnivorous archosaurs from the Middle Triassic of Argentina. They were originally described by Romer (1971) based on fragmentary specimens. Several other specimens recovered later allowed a more detailed description (Bonaparte, 1975; Arcucci, 1986). Recent reviews of



FIGURE 1 Location map showing the distribution of sites in the Triassic Ischigualasto–Villa Union basin in the north west of Argentina.

FIGURE 2 Reconstruction of Lagerpeton pelvis and hindlimb in lateral view. From Sereno and Arcucci (1994 a).

archosaurian phylogeny have found that these animals constitute the sister group of Dinosauria, showing several derived anatomical features that support their monophyly (Sereno and Arcucci, 1994a,b). Marasuchus (Lagosuchus) and Lagerpeton are preserved in the same levels and were probably contemporary, taking part in one of the richest and most diverse paleofaunas recorded in the Middle Triassic worldwide. Although they are too fragmentary to evaluate their affinities, Lewisuchus and Pseudolagosuchus, which come from the same levels, probably belong to the same clade. The beds of the Chan˜ares Formation are extensively distributed in the Ischigualasto–Villa Union Basin in the southwest of the La Rioja Province, Argentina, near the Andes Range (Fig. 1). They consist of a relatively thin rock unit (about 60 m thick) divided into two members; the lower one yields the vertebrate fossils.

The paleoenvironment recorded from the Chan˜ares Formation is an extensive floodplain, with thin layers of paleosols. The preservation of the bones is often very good, showing the smallest anatomical details. The skeletons are usually enclosed in strongly cemented nodules, without internal structure, that probably caused selective preservation reflected in the small and medium sizes of the animals recorded. There is a bias toward the preservation of certain body parts, such as the hindlimbs and vertebrae, apparently because they are lighter in structure than the rest of the skeleton, and not as a consequence of collection bias. The skull is only partially known in Marasuchus and Lewisuchus, and they share an overall profile of an elongated snout with numerous small serrated teeth. These dinosauromorph archosaurs consist of approximately 10 specimens and were apparently scarce in a large faunal sample of approximately 300 specimens of therapsids and other archo-

Dinosauromorpha saurs recovered from the site. This proportion could correspond to the predator–prey ratio in the faunal assemblage. Several genera of non-dinosaurian dinosauromorphs are known from these deposits. They are described here in order of their increasing proximity to dinosaurs. The position of Eoraptor and herrerasaurs is currently debated. They may be basal theropods, outside Dinosauria, or basal saurischians, depending on the analysis (see HERRERASAURIDAE; PHYLOGENY OF DINOSAURS). Lagerpeton Romer, 1971 (Fig. 2) Type species: Lagerpeton chanarensis Romer 1971 Diagnosis: Small archosaur with posterior dorsal vertebrae with anterodorsally inclined neural spines, iliac blade with sinuous dorsal margin, ischium with broad convex ventromedial flange and ventrally deep puboischiadic suture, proximal end of femur with flat anteromedial surface, astragalus with tongue-shaped posterior ascending process, pedal digit IV and metatarsal IV longer than pedal digit III and metatarsal III, respectively (in part from Sereno and Arcucci, 1994a) (Fig. 2). Novas (1992) defined Dinosauriformes to include the common ancestor of Lagosuchus and dinosaurs and all its descendants.

181 Lagosuchus Romer, 1971 Type species: Lagosuchus talampayensis Romer 1971 Marasuchus Sereno and Arcucci, 1994 Type species: Marasuchus lilloensis (Romer 1971) Diagnosis: Small archosaur with anterodorsally projected cervical neural spines, marked fossa ventral to the transverse process in the last cervical vertebrae and the first dorsals, mid-caudal vertebrae twice the length of the anterior caudals, and broad scapular blade (Fig. 3) (in part from Sereno and Arcucci, 1994b). Pseudolagosuchus Arcucci, 1987 Type species: Pseudolagosuchus major Arcucci 1987 Diagnosis: Medium-sized archosaur with pubis longer than femur, elongated proximal caudal vertebrae?, and rounded process projected from the posterior face of the astragalus (in part from Arcucci, 1987). Lewisuchus Romer, 1972 Type species: Lewisuchus admixtus Romer 1972 Diagnosis: Small archosaur with elongated cervical vertebrae, long and narrow scapular blade, small oval dermal scutes on the cervical, and dorsal neural spines (Fig. 4) (in part from Romer, 1972a). These last two dinosauriforms are in revision at the moment, and although not much of the available

FIGURE 3 Reconstruction of Marasuchus (Lagosuchus); skull roof, hands, and gastralia restored. From Sereno and Arcucci (1994b).



FIGURE 4 Reconstruction of Lewisuchus; skull roof, hands, and gastralia restored. Not to scale. Modified from Paul (1988).

material overlaps, some of it suggests that these could represent a single taxon. The phylogenetic relationships of these animals are generally clear, but details are sketchy. Lagerpeton chanarensis is clearly associated with the dinosaurian radiation, but it is not possible to evaluate its precise relation to the dinosauriforms mentioned previously. Its particular sacral and pedal specializations distinguish it from the rest of the archosaurian fauna from the Chan˜ares Formation, but it shares a set of derived characters with Dinosauriformes and Dinosauria—for example, the transverse extension of the calcaneum, the acute corner of the astragalus, and the reduction of the articular facet for the fifth metatarsal (Fig. 5). The Dinosauriformes (Marasuchus, Lagosuchus, Pseudolagosuchus, and probably Lewisuchus) share with dinosaurs several characteristic features such as the proportions of the forelimbs, the partly open acetabulae, the trochanteric shelf on the posterior side of the proximal part of the femur, and the parallelogram shape of the cervical vertebrae. These features strongly suggest that these Middle Triassic archosaurs are more closely related to Dinosauria than to Lagerpeton or the pterosaurs (Fig. 5). The fragmentary preservation of the available material of all these taxa keeps some phylogenetic relationships obscure, even after detailed analysis of the characters.

Functional Morphology These reptiles are the first ones preserved in the fossil record that developed obligatory BIPEDALITY. They probably explored new ecological roles using locomotor capabilities not previously recorded in other tetrapods. Although it is difficult to assign a specific locomotor gait to extinct taxa, despite the completeness of the available material, there is general consensus that Lagerpeton and Marasuchus were undoubtedly bipeds.

FIGURE 5 Cladogram depicting the phylogenetic relationships among the clades of basal archosaurs. From Sereno and Arcucci (1994a).


183 plete set of hindlimb functional features are the same as those in the other Dinosauriformes, including primitive dinosaurs, and appear for the first time in these reptiles (Fig. 6). Obviously, they represent evolutionary novelties that involved extensive changes in muscle attachments and the function of the hindlimb, but their meaning is difficult to reveal in the current state of our knowledge.


● ●

References Arcucci, A. B. (1986). Nuevos materiales y reinterpretacion de Lagerpeton chan˜arensis Romer (Thecodontia, Lagerpetonidae nov.) del Triasico medio de La Rioja, Argentina. Ameghiniana 23, 233–242.

FIGURE 6 Dorsal view of the pes of (A) Marasuchus and (B) Lagerpeton, showing the differences among the tridactyl and didactyl pes. From Sereno and Arcucci (1994a).

Even though they show similar enlargement of the hindlimbs relative to the body, they present a completely different set of specializations in their pelvis and hindlimb. Lagerpeton has a very wide and short pelvis, relative to the limbs, and the pubis and ischium developed flat transverse surfaces, very much like the ones present in primitive archosaurs (Fig. 2). The astragalus and calcaneum are co-ossified in adults, unlike most known related archosaurs (except pterosaurs). The astragalus itself is unusual, presenting a posterior ascending process and lacking the anterior foramen. The last dorsal neural spines are inclined anteriorly, as in some saltatorial mammals (Arcucci, 1987). Finally, the functionally didactyl pres is also consistent with a saltatory gait, which could be similar to the living small ricochetal springhare Pedetes (Sereno and Arcucci, 1994a). Marasuchus, on the other hand, has an elongated, rod-like pubis and ischium and a narrow pelvis in dorsal view. The dorsal neural spines incline posteriorly (Fig. 3). The pes is basically tridactyl, and the astragalus has an anterior ascending process, like those of dinosaurs, but not very developed. Its com-

Arcucci, A. B. (1987). Un nuevo Lagosuchidae (Thecodontia–Pseudosuchia) de la fauna de Los Chanares (Edad Reptil Chan˜arense, Triasico medio), La Rioja, Argentina. Ameghiniana 24, 89–94. Bonaparte, J. F.(1975). Nuevos materiales de Lagosuchus talampayensis Romer (Thecodontia–Pseudosuchia) y su significado en el origen de los Saurischia. Chan˜arense inferior. Triasico medio de Argentina. Acta Geol. Lilloana 13, 1–90. Bonaparte, J. F. (1982). Faunal replacement in the Triassic of South America. J. Vertebr. Paleontol. 2, 362–371. Paul, G. (1988). Predatory Dinosaurs of the World, pp. 464. Simon & Schuster, New York. Romer, A. S. (1971). The Chan˜ares (Argentina) Triassic reptile fauna X. Two new but incompletely known long-limbed pseudosuchians. Breviora 378, 1–10. Romer, A. S. (1972a). The Chan˜ares (Argentina) Triassic reptile fauna. XIV. Lewisuchus admixtus gen. et sp. nov. a further thecodont from the Chan˜ares beds. Breviora 390, 1–13. Romer, A. S. (1972b). The Chan˜ares (Argentina) Triassic reptile fauna. XV. Further remains of the thecodonts Lagerpeton and Lagosuchus. Breviora 394, 1–7. Sereno, P. C. (1991). Basal archosaurs: Phylogenetic relationships and functional morphology. J. Vertebr. Paleontol. II (Suppl.), 1–65. Sereno, P. C., and Arcucci, A. B. (1994a). Dinosaur precursors from the Middle Triassic of Argentina: Lagerpeton chan˜arensis. J. Vertebr. Paleontol. 13, 385–399. Sereno, P. C., and Arcucci, A. B. (1994b). Dinosaur precursors from the Middle Triassic of Argentina: Marasuchus lilloensis gen. nov. J. Vertebr. Paleontol. 14, 53–73.

Dinosaur Provincial Park


Dinosaur Provincial Park CLIVE COY Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Established in 1955, Dinosaur Provincial Park occupies an area of 73 square km along the Red Deer River near the center of southern Alberta, Canada. No single place of equal size on earth has produced as many individual skeletons of different dinosaurs or attracted so much research. In 1979, the park was designated a World Heritage Site by UNESCO in recognition of the exceptional abundance and diversity of dinosaur and other vertebrate fossils, the largest and most spectacular area of badlands found in Canada, and the endangered riparian habitat of plains cottonwood trees. Meltwaters of retreating glaciers 12,000–14,000 years ago exposed 120 vertical meters of Upper Cretaceous sediments in the park. These sediments are divided into three distinct periods within the Judith River Group. In 1909, a rancher from Alberta reported to Barnum Brown of the American Museum of Natural History that fossil bones like those on display at the museum were common on his ranch. Brown started working on the Red Deer River the following year, and by the summer of 1912 he had set up camp within current park boundaries. He continued working in the area until 1915. Brown’s impressive collection included complete skeletons (with skulls) of Centrosaurus, Corythosaurus, Gorgosaurus, Prosaurolophus, and Struthiomimus. American fossil collector Charles H. Sternberg, and his three sons Charles M., Levi, and George, were hired by the Canadian Geological Survey (1911) to collect dinosaurs that would stay in Canada. The Sternbergs’ first summer yielded the type specimen of Gorgosaurus libratus, two fine skulls of hadrosaurs, the horned dinosaurs Chasmosaurus (with skin impressions), Centrosaurus, and the type of the spikefrilled Styracosaurus albertensis. This friendly competition between Brown and the Sternbergs began a period of intense collection (1911–1917) that set the stage for more than 80 years of successful fossil collecting in the park. This has produced more than 250 articulated large skeletons representing 36 different species of dinosaur and another 84 species of vertebrates includ-

ing fish, salamanders, frogs, turtles, lizards, crocodiles, pterosaurs, birds, and mammals. Dinosaur Provincial Park’s rich bounty is housed in more than 30 institutions around the world. The Royal Tyrrell Museum of Palaeontology has been conducting regular collection and research projects within the park since 1978. In 1987, the government of Alberta acknowledged the overwhelming significance of the park’s fossiliferous deposits by building the permanent research station and field laboratory of the Royal Tyrrell Museum of Palaeontology in Dinosaur Provincial Park for use by researchers from around the world. The field station has a small but nice display of dinosaurs, including original specimens of Centrosaurus and Daspletosaurus.


Dinosaur Ridge MARTIN LOCKLEY University of Colorado at Denver Denver, Colorado, USA

Dinosaur Ridge, also known as the ‘‘Dakota Hogback,’’ is a national natural landmark, that forms part of the elongate north–south ridge that comprises the easternmost range of the Rocky Mountain foothills, just west of Denver, Colorado. The segment named Dinosaur Ridge is situated between the town of Morrison in the south and the southern outskirts of Golden in the north and consists of eastward-dipping Jurassic Morrison Formation and Cretaceous (Dakota Group) strata. Dinosaur Ridge is also the type section for the famous Late Jurassic Morrison Formation, which in 1877 began producing many well-known dinosaurs including Allosaurus, Apatosaurus, Diplodocus, and Stegosaurus (the Colorado State fossil). These discoveries sparked off a ‘‘dinosaur gold rush’’ and figured prominently in the famous ‘‘bone wars’’ between Edward D. Cope and Othniel C. Marsh. Cretaceous Dakota Group strata at Dinosaur Ridge are replete with dinosaur tracks. Approximately 500

Dinosaur Valley tracks, at as many as 10 different levels, represent the activity of more than 70 individuals. Most tracks are attributable to ornithopod dinosaurs, probably iguanodontids, that walked on all fours. These tracks have been assigned to the track type Caririchnium because of their similarity to tracks of the same name from Cretaceous strata in the Carir basin region of Brazil. There are also many unnamed footprints of slendertoed, theropod dinosaurs (probably coelurosaurs), and a few tracks of crocodiles. Dinosaur Ridge is also considered a sister institution to Dinosaur Valley in western Colorado. Dinosaur Ridge is open to the public year-round and is furnished with interpretive signs and a visitors’ center. Guidebooks and other documentation of the site are available (Lockley, 1990, 1991; Lockley and Hunt, 1995). Collecting is not permitted, but track replicas can be obtained from the visitors’ center.

See also the following related entries: FOOTPRINTS AND TRACKWAYS ● MORRISON FORMATION

References Lockley, M. G. (1990). A Field Guide to Dinosaur Ridge, pp. 29. Friends of Dinosaur Ridge and Univ. Colorado. at Denver, Denver. Lockley, M. G. (1991). Tracking Dinosaurs: A New Look at an Ancient World, pp. 238. Cambridge Univ. Press, Cambridge, UK. Lockley, M. G., and Hunt, A. P. (1995). Dinosaur Tracks and Other Fossil Footprints of the Western United States, pp. 338. Columbia Univ. Press, New York.

185 education about dinosaur science and fund scientific research. Overseen by scientists on a voluntary basis, the Dinosaur Society is now a leading international funder of dinosaur research. Since its inception, the society has published the monthly children’s newspaper, Dino Times, which provides current and scientifically vetted findings on dinosaur science to a young audience. An adult quarterly, The Dinosaur Report, is also published. A major source of financial support and educational value to the society since 1993 has been its traveling ‘‘Dinosaurs of Jurassic Park’’ exhibit, created by Research Casting International in cooperation with Amblin Entertainment and Universal Studios. Two editions of this exhibit have toured North America, Europe, and South America. Revenues from the exhibit have become the major source of income for the Dinosaur Society. Grants are awarded to applicants who submit proposals to a scientist review panel on a quarterly basis. Paleontological research, from the Arctic to Australia, and from Brazil to Montana, has been supported through the program. A special artists’ grant program has also been established to encourage artists to assist scientists in the preparation of scientific illustrations. The society is involved in many other programs to support dinosaur science, art, and education. It has created artists’ displays and worked with artists to secure employment and gallery opportunities and has published their work in the society’s own calendar. It has developed a teacher’s kit and a list of recommended children’s and adult dinosaur books, updated annually. The society hopes to remain a major force in furthering cooperative support for science and education among professional and amateur dinosaur workers and enthusiasts, young and old.

Dinosaur Society DON LESSEM

Dinosaur Valley

Dinosaur Society Waban, Massachusetts, USA


Distressed by the inadequacy of funding for dino-

University of Colorado at Denver Denver, Colorado, USA

saur science and the poor quality of dinosaur-related education, author Don Lessem and paleontologists Dr. David Weishampel of Johns Hopkins University and Dr. Peter Dodson of the University of Pennsylvania founded a nonprofit organization, the Dinosaur Society, in 1991. Its purposes are to promote public

Dinosaur Valley, as used in Colorado, is an alternate name for the Grand Valley region in the greater Grand Junction–Fruita area. The term also refers more specifically to the main paleontological branch of the Museum of Western Colorado, situated on 4th


186 and Main streets in Grand Junction and open to visitors year-round. Within the greater Dinosaur Valley region there are a number of important dinosaur sites including Dinosaur Hill, the FRUITA PALEONTOLOGICAL AREA, RABBIT VALLEY, and Riggs Hill. Dinosaur Valley, Colorado, should not be confused with Dinosaur Valley State Park (in Texas) or other sites with the same or similar names in other parts of the world. DINOSAUR RIDGE, in eastern Colorado, is considered a sister institution to Dinosaur Valley.

See also the following related entry: MORRISON FORMATION

extensive trampling. A ‘‘dinoturbation index’’ has been proposed (Lockley, 1991), based on the percentage of the surface covered by footprints, to define areas of light, moderate, and heavy trampling (respectively disturbing 0–33, 34–66, and 67–100% of the surface). Heavy dinoturbation can result from the biological activity of many animals, from the geological effects of long periods of exposure of the substrate to trackmakers, or from a combination of both factors. The term dinoturbation should not be used to describe trampling caused by vertebrates that are not dinosaurs.

See also the following related entries: FOOTPRINTS AND TRACKWAYS ● TAPHONOMY




Lockley, M. G. (1991). Tracking Dinosaurs: A New Look at an Ancient World, pp. 238. Cambridge Univ. Press, Cambridge, UK.

University of Colorado at Denver Denver, Colorado, USA

Dinoturbation is a term for trampling of soil or sedimentary substrates by dinosaurs. The word derives from the more general term ‘‘bioturbation,’’ which refers to all manner of burrowing and disturbance of sedimentary substrates by plant roots, invertebrates, and vertebrates. A good example of bioturbation, Charles Darwin noted, is the constant recycling of soil through the gut of earthworms. The term dinoturbation was first coined in 1980, in the same year that the term ‘‘Megabioturbation’’ was used to describe trampling caused by mammoths in Ice Age sediments. The term dinoturbation has caught on and been widely used in recent years, whereas the latter term has not. Dinoturbation is particularly common in late Mesozoic sediments in which the tracks of large gregarious herbivores (mainly sauropods and ornithopods) are abundant. At this time in earth history, trampling reached a peak (Lockley, 1991; Lockley and Hunt, 1995). It has even been suggested that this trampling stimulated the evolution of flowering plants, which are capable of rapid growth and regeneration. Certainly there is evidence that dinosaurs were destructive in trampling flora and fauna underfoot. In theory, any dinosaur track is an example of dinoturbation, but the term is often taken to indicate

Lockley, M. G., and Hunt, A. P. (1995). Dinosaur Tracks and Other Fossil Footprints of the Western United States, pp. 338. Columbia Univ. Press, New York.



Distribution and Diversity PETER DODSON University of Pennsylvania Philadelphia, Pennsylvania, USA

We now know that dinosaurs lived on all seven continents. Antarctica was the last continent to produce dinosaur fossils, and the Early Jurassic theropod Cryolophosaurus was only described in 1994, whereas

Distribution and Diversity Late Cretaceous ornithopod and ankylosaurian remains have been reported but not yet named. The distribution of dinosaurs throughout the Mesozoic is to an extent determined by the positions of the wandering continents. The earliest dinosaurs or dinosaur relatives currently known appear to be those of the Late TRIASSIC (Carnian) Ischigualasto Formation of Argentina and the Santa Maria Formation of Brazil. At this time, Pangea was substantially intact, and no major barriers impeded intercontinental dispersal. By the succeeding Norian stage of the Late Triassic, prosauropods had appeared in Germany, Greenland, and South Africa. Early Jurassic prosauropods reached China and the southwestern United States. By the end of the Cretaceous, approximately 163 million years after the first appearance of dinosaurs, the continents were close to their current positions. It is probable that some degree of faunal exchange was possible among South America, Antarctica, and Australia. For essentially the entire Mesozoic, Europe was an archipelago, and for significant lengths of time during the Jurassic and especially the Cretaceous, North America was flooded by epicontinental seas that tended to separate the eastern and western parts of the continent. As a rule of thumb, dinosaurs that appeared early have a cosmopolitan distribution, whereas those appearing later are more restricted. PROSAUROPODS (Late Triassic) and SAUROPODS (Early Jurassic) are rather cosmopolitan, whereas CERATOPSIANS (late Early Cretaceous) are found only in eastern Asia and western North America (claims of solitary bone specimens from South America and Australia are unconvincing to some; but see POLAR DINOSAURS). HYPSILOPHODONTIDS (Middle Jurassic) are cosmopolitan, but LAMBEOSAURINE HADROSAURIDS are known only from western North America and eastern Asia. HADROSAURINE HADROSAURIDS (early Late Cretaceous) are predominantly of Laurasian distribution. Hadrosaurs reached South America (Secernosaurus), possibly by sweepstakes dispersal from North America. Many specimens have been collected, but few have been studied or described. STEGOSAURS (Middle Jurassic) are cosmopolitan, but show greater diversity in China than anywhere else. The sauropod family Titanosauridae has a broad distribution in the Late Cretaceous on southern continents. Titanosaurids also entered Europe (Magyarosaurus and Hypselosaurus) and North America (Alamosaurus), a continent from which sau-

187 ropods had previously become extinct. The abelisaurids may be a Gondwanan family of THEROPODS of broad distribution (South America, India, or Europe?) perhaps related to Ceratosauria. It has been argued that rather than being typical of worldwide dinosaurs of the Late Cretaceous, dinosaurs of western North America and eastern Asia were isolated from the world dinosaur fauna and exhibited instead a high degree of endemism (e.g., tyrannosaurids, lambeosaurines, ceratopsians, and ankylosaurids). Assessments of dinosaur diversity and of diversity trends through time are difficult matters. In the simplest form, a tabulation of described genera represents one estimation of dinosaur diversity. More than 600 genera of dinosaurs have been named, but of these only about 325 are currently (as of 1995) regarded as valid. However, the number has doubled since 1970 and may double again in another 25 years. Two further approaches have been taken to estimate dinosaur diversity. Dodson (1990) estimated the mean generic longevity of dinosaurs to be 7.7 million years (range: 5–10.5 million years). This suggests nearly 100% faunal turnover per geological stage. He estimated the maximum number of dinosaur genera living at one time (Campanian–Maastrichtian, latest Cretaceous) to be 100. Depending on several different models of diversity change through time, and integrating across intervals for which no fossils are known, this method yields estimates of dinosaur diversity on the order of 900–1200 genera. Russell (1994) estimated dinosaur diversity on the basis of the relationship between area of continental landmass and diversity. This method yields an estimate of 3300 genera. The two methods frame reasonable endmember estimates. One approach suggests that our knowledge of the fossil record of dinosaurs is currently nearly 30% complete; the other a more conservative 10% complete. Did dinosaur diversity increase through time? If it is assumed that certain time intervals are relatively well sampled (e.g., Norian–Sinemurian, Late Triassic–Early Jurassic; Kimmeridgian–Tithonian, Late Jurassic; and Campanian–Maastrichtian, Late Cretaceous), relevant observations may be made. In a typical early dinosaur assemblage, there were one or two sauropodomorphs, one or two theropods, and possibly one ornithischian. In a typical Late Jurassic assemblage, there were three or more sauropods, two or more theropods, one or two stegosaurs, and one or

188 two ornithopods. In a Late Cretaceous North American assemblage, there were three or more hadrosaurs, a basal ornithopod, a pachycephalosaur, two or more ankylosaurs, two or three ceratopsians, and three or more theropods. It thus appears that diversity increased through the Mesozoic. Two factors may be mentioned as contributing to increased diversity in the Late Cretaceous. One is the fractionation of continental landmasses compared with the earlier Mesozoic, leading to increased endemism. The other is the smaller body sizes of ornithopods compared to sauropods, with the attendant greater possibilities for niche specialization and consequent greater diversity at the community level.

See also the following related entries: BIOGEOGRAPHY ● MIGRATION ● PLATE TECTONICS

References Dodson, P. (1990). Counting dinosaurs: How many kinds were there? Proc. National. Acad. Sci. 87, 7608–7612. Dodson, P. and Dawson, S. D. (1991). Making the fossil record of dinosaurs. Mod. Geol. 16, 3–15. Russell, D. A. (1994). China and the lost worlds of the dinosaurian era. Historical Biol. 10, 3–12. Smith, A. G., Smith, D. G., and Funnell, B. M. (1994). Atlas of Mesozoic and Cenozoic Coastlines, pp. 99. Cambridge Univ. Press, Cambridge, UK. Weishampel, D. B. (1990). Dinosaurian distributions. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 63–139. Univ. California Press, Berkeley.

Djadokhta Formation TOM JERZYKIEWICZ Geological Survey of Canada Calgary, Alberta, Canada

The Djadokhta Formation of Campanian (Late Cretaceous) age is famous for yielding the first unquestionable finds of dinosaur eggs (discovered in 1922 by the third ASIATIC EXPEDITION led by Roy Chapman Andrews) and numerous exceptionally well-preserved dinosaur skeletons, most notably Protoceratops andrewsi, Pinacosaurus grangeri, and Velociraptor mongoliensis (Figs. 1–3). The type localities of the forma-

Djadokhta Formation tion at BAYN DZAK, UKHAA TOLGOD, and TUGRIG of pre-Altai Gobi (Mongolia) and correlative strata at BAYAN MANDAHU of the southern Gobi (Inner Mongolia, China) have been the subject of many later successful paleontological expeditions (most notably from the United States, Russia, Mongolia, Poland, and Canada). The dominant lithology of the formation is poorly cemented reddish brown, fine-grained eolian sandstone. Beds of water-deposited sandstone, mudstone, and conglomerate are subordinate. Sedimentary facies indicate that the Djadokhta redbeds were deposited in semiarid, alluvial-to-eolian settings. The presence of mature caliche paleosols and wind-blown sediments suggests overall semiarid conditions of accumulation for the formation. Stratal geometry of wind-blown sediments indicates the presence of large straight-crested dune forms and smaller barchan and parabolic dunes. Some layers within the dunes are rich in trace fossils, suggesting that the dunes were organically rich and seasonally moist. An assemblage of fossil vertebrates found in the Djadokhta Formation consists of ankylosaurs, ceratopsians, theropods, turtles, crocodiles, lizards, and mammals. The fossil vertebrates occur largely in association with fine-grained eolian sandstones and paleosols, and these have been interpreted as the remains of autochthonous ‘faunal’ components, adapted to living in semiarid environments. In contrast, rare and fragmentary specimens of large dinosaurs that occur in coarse-grained alluvial deposits have been interpreted as the remains of allochthonous faunal components. These rare components include the tyrannosaurid Tarbosaurus, another large, nontyrannosaurid THEROPOD, a SAUROPOD(s), and ORNITHOPODS. The low diversity of the Djadokhta fossil assemblage and the overall small to medium size of its constituents indicate a relatively stressed paleoenvironment. Furthermore, the absence of fishes or any other undoubted aquatic organisms, together with the localized abundance of dermatemydid (terrestrial) turtles, further suggests a fully terrestrial vertebrate assemblage. The environment of the Djadokhta dinosaurs developed during Late Cretaceous time as a result of block-faulting tectonic movements that affected central Asia and transformed large, Early Cretaceous perennial lakes into semiarid steppes located far from

Djadokhta Formation


FIGURE 1 Partly unearthed skeleton of ‘‘standing’’ Protoceratops in the Djadokhta Formation at Toogreeg. (a) Oblique view; (b) side view (photo from T. Jerzykiewicz, Polish–Mongolian Expedition of 1971).

the seashore. These semiarid steppes were partly covered by dune fields and drained by seasonal streams into intermittent ponds whose distribution and occurrence may have been controlled by plugged caliche horizons. Such a landscape, shaped to a large degree by eolian processes (including dust and sand storms), must have made living conditions rather difficult and supposedly contributed to an increased mortality of vertebrates during droughts. The poses of some fossil dinosaurs in the eolian sandstones indicate that the animals died in situ and were not transported. A number of articulated Protoceratops, for instance, ‘‘stand’’ on their hindlimbs, with snouts pointing upward and forelimbs tucked in at their sides (Figs. 1a and 1b). Such a pose clearly indicates that the animals were encased within the surrounding sediment at the time of their deaths. It is likely that some of these creatures died while attempting to free themselves from a sandstorm deposit during or shortly after the storm event. In the

exceptional co-occurrence of Protoceratops and Velociraptor—the so called ‘‘fighting dinosaurs’’ (Fig. 2), the theropod’s right forelimb is ‘‘clenched’’ between the closed jaws of the prostrate CERATOPSIAN indicating rapid burial by eolian sands shortly after the animals came in contact with one another. Monospecific death assemblages of both Protoceratops and Pinacosaurus were also noted. In one case, a group of five mediumsized adult Protoceratops were found lying parallel aligned, side by side, and inclined about 20⬚ from the horizontal, probably on a slope of a dune. All the specimens were lying on their bellies with their heads facing upslope. Yet another example consists of a group of juvenile Pinacosaurus that died in situ, probably as a result of burial during sandstorms (Fig. 3). The occurrence of large numbers of fossil vertebrate eggs in association with the eolian sandstone and semiarid paleosols provides additional evidence of a semiarid climate during deposition of the Djadokhta Formation.

FIGURE 2 Partly unearthed skeletons of Protoceratops and Velociraptor at Toogreeg (pre-Altai Gobi). Note the claws (arrow) of the left hindfoot of the Velociraptor imbedded within the thorax of the Protoceratops and the claw (arrow) of the hand clasp the top of the skull of the presumed prey (photo from T. Jerzykiewicz, Polish–Mongolian Expedition of 1971).

FIGURE 3 Excavating the mass grave of juvenile Pinacosaurus at Bayan Mandahu. Inset shows one of the skulls recovered from the grave.




Dockum Group PHILLIP A. MURRY Tarleton State University Stephenville, Texas, USA

Barsbold, R., and Perle, A. (1983). On taphonomy of a joint burial of juvenile dinosaurs and some aspects of their ecology. Sovmestnaya Sovetsko–Mongolskaya Paleontologicheskaya Ekspeditsia. Trudy 24, 121–125. [In Russian].


Berkey, C. P., and Morris, F. K. (1927). Geology of Mongolia, Natural History of Central Asia, Volume II, pp. 475. American Museum of Natural History, New York.

Upper Triassic-age redbeds, representing fluvial, lacustrine, and paleosol facies. Four Dockum units are recognized in Texas; in ascending order, these include an informal ‘‘Pre-Tecovas Horizon,’’ Tecovas Formation, Trujillo Formation, and Cooper Canyon Formation. Dinosaur remains have been reported from all Dockum units in Texas, except the Trujillo Formation. The poorly defined, mudstone-rich Pre-Tecovas Horizon is regarded as Middle Carnian age; localities of this age are in Howard, Borden, and Scurry counties within the southern portion of the Dockum depositional basin. Fossiliferous localities within the mudstone-dominated Tecovas Formation are found within the central Dockum basin of Crosby and Dickens counties, and in the lower mudstone units in the Palo Duro Canyon area (Randall County) and Canadian River Valley (Oldham and Potter counties) within the northern Dockum of Texas. The sandstonerich Trujillo Formation is best exposed in the Palo Duro Canyon area, and fossiliferous units in the interbedded coarse and fine clastics of the Cooper Canyon Formation are known from Garza County, in the south-central portion of the Dockum Basin (Long and Murry, 1995). Although thousands of reptilian individuals are represented in the Dockum collections, only a few are dinosaurian. Furthermore, most Dockum material reported as dinosaurs by previous workers has proven to be that of non-dinosaurian archosauromorphs. Cope (1893) named Palaeoctonus dumblianus and P. orthodon from isolated teeth from Palo Duro Canyon, and referred them to theropod dinosaurs. These teeth are actually phytosaurian. Case (1922) referred a left femur (UMMP 3396) to the Dinosauria that in reality represents a robust aetosaurian, probably Desmatosuchus. He also re-

Currie, P. J., and Peng, J. H. (1993). A juvenile specimen of Saurornithoides mongoliensis from the Djadokhta Formation (Upper Cretaceous) of northern China. Can. J. Earth Sci. 30, 2224–2230. Gradzinski, R., Kielan-Jaworowska, Z., and Maryanska, T. (1977). Upper Cretaceous Djadokhta, Barun Goyot and Nemegt formations of Mongolia, including remarks on previous subdivisions. Acta Geol. Polonica 27, 281–318. Granger, W., and Gregory, W. K. (1923). Protoceratops andrewsi, a preceratopsian dinosaur from Mongolia. Am. Museum Novitates 42, 1–9. Jerzykiewicz, T., Currie, P. J., Eberth, D. A, Johnston, P. A., Koster, E. H., and Zheng, J. J. (1993). Djadokhta Formation correlative strata in Chinese Inner Mongolia: An overview of the stratigraphy, sedimentary geology, and paleontology and comparisons with the type locality in the pre-Altai Gobi. Can. J. Earth Sci. 30, 2180–2195. Lefeld, J. (1971). Geology of the Djadokhta Formation at Bayn Dzak (Mongolia). Palaeontol. Polonica 25, 101–127. Tverdokhlebov, V. P., and Tsybin, Yu. I. (1974). Genezis verkhnemelovykh mestorozhdeniy dinozavrov Tugrikin-Us i Alagteg. Sovmestnaya Sovetsko– Mongolskaya Paleontologicheskaya Ekspeditsia. Trudy 1, 314–319. [In Russian].


Pleasanton, California, USA

The Dockum Group of western Texas consists of

192 ferred a series of cervical and anterior dorsal vertebrae (UMMP 7507) found in Crosby County to Coelophysis (Case, 1922, 1927). In 1932, von Huene designated this specimen the holotype of a supposed saurischian dinosaur, Spinosuchus caseanus, and referred the braincase described by Case to this taxon. The braincase belongs to a rauisuchian (probably Postosuchus) and the vertebral column does not appear to be Saurischian. Relationships have been proposed among Spinosuchus caseanus and Lotosaurus adentus from the Middle Triassic of China (Fa-Kui, 1975) and Ctenosauriscus koeneni from the Middle Buntsandstein of Germany (Krebs, 1969). These three longspined Triassic archosauromorphs are probably not closely related but are merely convergent in the development of dorsal sails. There is no evidence that these taxa are closely related to dinosaurs. Caudal vertebrae (UMMP 7277 and UMMP 9805) and compressed recurved teeth from Crosby County, identified as theropod by Case (1927), cannot be referred with confidence to the Theropoda, although the teeth do appear to represent non-phytosaurian carnivorous archosauromorphs. Case (1932) also described and illustrated a large series of associated caudal vertebrae from Potter County that he referred to Coelophysis sp.; these are not referable to Coelophysis and appear to represent a rauisuchian. Elder (1987) reported ‘‘coelurosaur’’ material from Quarries 1, 3, and 3a at Otis Chalk, and Gregory (1972) listed Coelophysis from Howard, Borden, Crosby, Randall, and Potter counties. Although there are a number of lightly constructed bones in the TMM collections, there is no irrefutable evidence of dinosaurs or herrerasaurs in any of these collections other than the herrerasaurid Chindesaurus. Colbert (1961) referred a fragment of a right ilium (UMMP 11748) from Howard County to Poposaurus and concluded that Poposaurus represented a primitive theropod. However, subsequent workers have placed the poposaurs within the Rauisuchia. Shuvosaurus inexpectatus is based on the remains of at least three individuals from the Post Quarry, including a well-preserved skull with lower jaw (TTU P9280). Chatterjee (1993) regarded Shuvosaurus as a new family of ORNITHOMIMOSAUR, based primarily on a comparison of skull characters including a pointed, hooked premaxillary beak, short preorbital region, participation of the nasal in the development of the

Dockum Group maxillary fenestra, mandibular articulation of the quadrate by the lateral and medial condyle, and the presence of a secondary dental shelf at the jaw symphysis. We have not studied this skull in detail, and our conclusions are tentative. We believe that Chatterjee’s diagnosis of Shuvosaurus fails to include it specifically within the Dinosauria. Probable remains of S. inexpectatus recovered in both New Mexico and Texas are closely associated with elements of Chatterjeea elegans, a rauisuchian based on postcrania (Long and Murry, 1995). These taxa closely match in size and preservation. The morphology of Chatterjeea has a number of convergent features with those of ornithomimosaurs, including attenuation of the vertebral column and limbs, development of a synsacrum, and the presence of an ornithomimosaur-like deltopectoral crest on the humerus. It is possible that all these specimens represent one taxon, in which case Shuvosaurus would have priority. Based on postcranial characters, it could not be referred to the Dinosauria. In this case, the range of Shuvosaurus would be at least from the Middle Carnian to Lower Norian. Although many reports of dinosaurs in the Dockum Group are false, there are evidently both saurischians and ornithischians within the Upper Triassic deposits of Texas. Chatterjee (1986, p. 145) reported a juvenile specimen referable to Coelophysis, and examination of his collection verifies the presence of a small theropod dinosaur femur. Chindesaurus bryansmalli is a 2- to 4-m-long HERRERASAURID dinosauromorph or dinosaur relative that is primarily based on a single specimen from the Lower Norian of Petrified Forest National Park. It is the most primitive North American dinosauromorph known. It is also among the oldest North American dinosauromorphs, as the proximal portion of a femur (TMM 31100-523) from Howard County is identical to that of the type specimen, and is believed to be of Middle Carnian age (Long and Murry, 1995). The Petrified Forest specimen is also the youngest known record of Herrerasauridae. Another Texas specimen of Chindesaurus is a right ilium (UMMP 8870) from Crosby County that Case (1927) tentatively referred to Coelophysis. The presence of a prominent groove along the ventral articular surface of the astragalus, the glutealform shape of its distal surface, and the apparent absence of a fibular facet on the astragalus

Dockum Group differentiates Chindesaurus from Herrerasaurus (Long and Murry, 1995). Chatterjee (1984) described Technosaurus smalli from the Cooper Canyon Formation near Post, Texas. According to Sereno (1991), at least a portion of the holotype (TTU P9021) displays features that are consistent with the ‘‘Prosauropoda,’’ (basal sauropodomorpha) although exhibiting no clear sauropodomorph synapomorphies. However, Hunt and Lucas (1994) agree with Chatterjee that Technosaurus is an ornithischian, although the material could not be more specifically assigned. Hunt and Lucas (1994) described isolated dentary and premaxillary teeth from Crosby County as Tecovasaurus murryi that they believe represents undoubted ornithischians. Other material probably referable to this species was collected by Murry (1982, 1986) from Crosby and Potter counties. Asymmetrically crowned teeth with large, compound denticles of this morphotype evidently show marked heterodonty. The range of dental variation in Triassic dinosaur teeth, convergence between non-dinosaurian archosauromorphs (especially rauisuchians) and dinosaurs, and convergence among dinosaurian taxa present problems in naming dinosaurs on the basis of isolated teeth. However, the presence of several dinosaur-like tooth morphotypes in the Texas collections may indicate that more, as yet undescribed, dinosaurs were present within the Upper Triassic Dockum Group of Texas.

See also the following related entries: CHINLE FORMATION ● TRIASSIC PERIOD

References Case, E. C. (1922). New reptiles and stegocephalians from the Upper Triassic of Western Texas. Carnegie Institution of Washington Publ. 321, 1–84. Case, E. C. (1927). The vertebral column of Coelophysis Cope. Univ. Michigan Museum Paleontol. Contrib. 2(10), 209–222. Case, E. C. (1932). On the caudal region of Coelophysis sp. and on some new or little known forms from the Upper Triassic of Western Texas. Univ. Michigan Museum Paleontol. Contrib. 4(3), 81–91.

193 Chatterjee, S. (1986). The Late Triassic Dockum vertebrates: their stratigraphic and paleobiogeographic significance. In The Beginning of the Age of Dinosaurs. Faunal Change across the Triassic–Jurassic Boundary (K. Padian, Ed.), pp. 139–150. Cambridge Univ. Press, Cambridge, UK. Chatterjee, S. (1993). Shuvosaurus, a new theropod. Nat. Geogr. Res. Exploration 9(3), 274–285. Colbert, E. H. (1961). The Triassic reptile Poposaurus. Fieldiana, Geol. 14, 59–78. Cope, E. D. (1893). A preliminary report on the vertebrate paleontology of the Llano Estacado. Fourth Annual Report of the Geological Survey of Texas. pp. 11–87. Elder, R. L. (1987). Taphonomy and paleoecology of the Dockum Group, Howard County, Texas. J. Arizona– Nevada Acad. Sci. 22, 85–94. Fa-Kui, Z. (1975). A new thecodont Lotosaurus, from Middle Triassic of Hunan (in Chinese). Vertebr. Palasiatica 13(3), 144–147. Gregory, J. T. (1972). Vertebrate faunas of the Dockum Group, Triassic, eastern New Mexico and west Texas. In New Mexico Geological Society, Annual Field Conference Guidebook, Vol. 23, pp. 120–123. New Mexico Geological Society. Hunt, A. P., and Lucas, S. G. (1994). Ornithischian dinosaurs from the Upper Triassic of the United States. In In the Shadow of the Dinosaurs. Early Mesozoic Tetrapods (N. C. Fraser and H.-D. Sues, Eds.), pp. 227–241. Cambridge Univ. Press, Cambridge, UK. Krebs, B. (1969). Ctenosauriscus koeneni (v. Huene), die Pseudosuchia und die Buntsandstein-Reptilien. Eclogae Geol. Helvetiae 62(2), 697–714. Long, R. A., and Murry, P. A. (1995). Late Triassic (Carnian and Norian) tetrapods from the southwestern United States. New Mexico Museum Nat. History Bull. 4, 1–254. Murry, P. A. (1982). Biostratigraphy and paleoecology of the Dockum Group, Triassic, of Texas, pp. 459. PhD dissertation, Southern Methodist University, Dallas, Texas. Murry, P. A. (1986). Vertebrate paleontology of the Dockum Group, western Texas and eastern New Mexico. In The Beginning of the Age of Dinosaurs. Faunal Change across the Triassic–Jurassic Boundary (K. Padian, Ed.), pp. 109–137. Cambridge Univ. Press, Cambridge, UK. Sereno, P. C. (1991). Lesothosaurus, ‘‘fabrosaurids,’’ and the early evolution of Ornithischia. J. Vertebr. Paleontol. 11(2), 168–197.



Dromaeosauridae PHILIP J. CURRIE Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

The first dromaeosaurids described were Dromaeosaurus albertensis (Matthew and Brown, 1922) and Velociraptor mongoliensis (Osborn, 1924), both from the Upper Cretaceous strata of the Northern Hemisphere. Because of the rarity of small theropod fossils, however, the significance of these animals was not fully understood until the discovery of Deinonychus in 1964. Since that time, dromaeosaurids have been a focal point for research on the interrelationships of theropods, the origin of birds, dinosaur physiology, dinosaur brain size, and dinosaur behavior. Two subtaxa of dromaeosaurids are currently recognized. Velociraptorine dromaeosaurids include Deinonychus (Ostrom, 1969), Saurornitholestes, and Velociraptor. Dromaeosaurus is the only unquestionable dromaeosaurine dromaeosaurid, but the poorly known Adasaurus from Mongolia has also been re-

ferred to this genus. Giant dromaeosaurids from the Lower Cretaceous of the United States (Utahraptor), Japan, and Mongolia are poorly known and cannot be assigned to either subtaxon with confidence at this time. Most dromaeosaurids were between 2 and 3 m in length. Their most conspicuous character is found in the second toe of the hindfoot, which bears a large raptorial claw. This claw is strongly recurved and was more than twice as long as the other claws on the foot. Because of its sharp point and knife-like lower edge, it was held off the ground in normal situations. Although dromaeosaurid footprints are unknown at present, the raised position of this claw can be seen in several articulated skeletons. Many other dromaeosaurid apomorphies make this one of the better diagnosed theropod taxa. These include a pubis that is more bird-like than those of any other known dinosaur. The pubis faces down and backward, parallel to the ischium. The tail is unusual because there are long but delicate rods in the tail that extend anteriorly from the prezygopophyses and the haemal arches. They form a cable-like network that would have stiffened the tail without making it completely rigid. As in other theropods, inter-

FIGURE 1 Reconstruction of the skull of Dromaeosaurus albertensis.

Dromaeosauridae dental plates seem to have been present on the inside of the tooth rows, but they are fused together so that individual plates can no longer be distinguished. The jaw articulation is peculiar in that there is a tall, slender, vertical process behind the joint (Currie, 1995). Velociraptorines and dromaeosaurines are easily distinguished from each other on the basis of differences in their teeth. The serrations on the front of a dromaeosaurine tooth are about the same size as the serrations on the back of the same tooth. In velociraptorines, the posterior denticles are much larger than the anterior ones. The teeth in the premaxilla of Dromaeosaurus are all about the same size, whereas the second tooth of this bone is the largest in velociraptorines. Dromaeosaurine skulls seem to have been more heavily constructed (Fig. 1). Dromaeosaurids are often considered to be the most bird-like of the small theropods. The brain is relatively large, the lightly built skulls are often pneumatic, they have clavicles and ossified sternals (breastbones), their arms are relatively long, and the pubis is retroverted. Unlike most other Cretaceous theropods, the metatarsal bones (found in the flat of the foot of humans) are relatively short and unspecialized and are more similar to the metatarsals of early birds. Even the specialized raptorial claw has now been found in Cretaceous birds from Madagascar and Argentina. There are so many similarities between dromaeosaurids and birds that some have even speculated that dromaeosaurids were birds that lost their ability to fly. However, in some characters, such as the stiffened tail, dromaeosaurids are too specialized to have been good ancestors for birds. All known dromaeosaurids are Cretaceous in age, although they no doubt originated sometime in the Jurassic, and other maniraptoran taxa are known from the Late Jurassic (see BIRD ORIGINS). The dromaeosaurid Velociraptor has become a wellknown dinosaur thanks to its role in the book and movie called Jurassic Park. Depicted as relatively intelligent, vicious, warm-blooded, pack-hunting animals, dromaeosaurids have done much to change the public perception of dinosaurs as slow-witted, solitary, cold-blooded creatures. Like troodontids, dromaeosaurids had relatively large brains. Although this does not indicate that they were as intelligent as living birds and mammals, it does suggest that they had the same capabilities as some birds and mammals. The light, agile bodies, long fingers, and raptorial

195 claws show that they were probably effective predators. This view is supported by a remarkable pair of skeletons from Mongolia. A Velociraptor and a Protoceratops were discovered together in Upper Cretaceous sediments, apparently locked in mortal combat. It appears that the predator attacked the herbivore and killed it. However, before the protoceratopsian died, it locked its jaws on the arm of the dromaeosaurid. It might have escaped if a sandstorm had not been in progress, and both animals were completely sealed in sand for 75 million years. Determining conclusively whether or not dinosaurs such as Velociraptor were warm-blooded has been elusive (see PHYSIOLOGY). However, the warm-blooded proponents gained some support in 1996 by the discovery in China of a pair of small ‘feathered’ theropods.

See also the following related entries: COELUROSAURIA ● MANIRAPTORA

References Currie, P. J. (1995). New information on the anatomy and relationships of Dromaeosaurus albertensis (Dinosauria: Theropoda). J. Vertebr. Paleontol. 15, 576–591. Matthew, W. D., and Brown, B. (1922). The family Deinodontidae, with notice of a new genus from the Cretaceous of Alberta. Am. Museum Nat. History Bull. 46, 367–385. Osborn, H. F. (1924). Three new theropoda, Protoceratops Zone, central Mongolia. Am. Museum Nat. History Novitates 144, 12. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Peabody Museum Nat. History Bull. 30, 165.

Dromaeosaurinae see DROMAEOSAURIDAE

Drumheller Dinosaur and Fossil Museum see MUSEUMS



Dry Mesa Quarry


Dry Mesa Quarry BROOKS B. BRITT Museum of Western Colorado Grand Junction, Colorado, USA

BRIAN D. CURTICE Brigham Young University Provo, Utah, USA

The Dry Mesa Quarry (Uncompahgre) has yielded the most diverse dinosaurian fauna known from any single quarry in the Morrison Formation (Britt, 1991). All the common Morrison taxa such as Camarasaurus, Diplodocus, Allosaurus, Stegosaurus, and others are present, as are rare taxa such as Ceratosaurus, Marshosaurus, Stokesosaurus, and an unnamed nodosaurid (based on a scapulocoracoid). At least two small theropod taxa remain to be identified to the generic level. The quarry is the type locality of the robust megalosaurid-grade theropod Torvosaurus and the sauropods Dystylosaurus, Supersaurus, and Ultrasauros (⫽ Ultrasaurus Jensen 1985). The validity of Dystylosaurus (Jensen, 1985) remains to be determined. Through the years, there has been much confusion over Supersaurus and Ultrasaurus. Supersaurus, an enormous diplodocid, is valid (B. Curtice et al., manuscript in preparation) and still ranks as one of the largest animals known. It is now recognized that the 2.5-mlong scapulocoracoid that Jensen (1985) referred to Ultrasauros, once thought to be the largest dinosaur ever found, is referable to Brachiosaurus. Furthermore, the Brachiosaurus scapulocoracoid is slightly shorter than the largest one known from Tendaguru. The type specimen of Ultrasauros was not the scapulocoracoid but a large diplodocid dorsal vertebra, now referred to Supersaurus. Thus, Ultrasauros is a junior synonym of Supersaurus (Curtice et al., manuscript in preparation). The Dry Mesa site was discovered by Eddie and Vivian Jones, who reported their find to James A. Jensen of Brigham Young University (BYU). Jensen opened the quarry in the spring of 1972 and the quarry has been operated by BYU in nearly every field season since that time. More than 4000 elements have been recovered and the quarry face is more than 120 m in length. The quarry is positioned near the

base of the Brushy Basin Member of the Morrison Formation. The bone-bearing lithosome is a fluvial deposit consisting of a broad, conglomeratic sandstone channel incised into overbank mudstone deposits. Bones were deposited at the base of trough crossbeds in the bottom of the river channel. Portions of bones not immediately buried in the sands were weathered before being completely buried by subsequent depositional events. Also, many of the bones exhibit varying degrees of pitting attributed to an as yet unknown biologic agent. Most of the skeletons were completely disarticulated, making it difficult to associate bones with particular taxa. In 1985, a partially articulated juvenile diplodocid was discovered. Before its discovery the few articulated bones consisted mainly of vertebral column segments. Richmond and Morris (1997) attributed the concentration of bones to a drought followed by a flash flood with a maximum flow velocity approaching 2 m/sec. Although the quarry is most renowned for its dinosaurian fauna, its nondinosaurian composition is surprisingly diverse and remains to be studied in detail. The nondinosaurian fauna consists of fishes, including the lungfish Ceratodus, an amphibian, two turtle genera, a crocodilian, and a prototherian mammal (Prothero and Jensen, 1983). Many small vertebrates have been recovered from the matrix encasing the large dinosaur bones, including several amphibians, various reptiles, maniraptoran dinosaurs that may be either small dromaeosaurs or birds, and the type material of the pterodactyloid pterosaur Mesadactylus ornithosphyos (Jensen and Padian, 1989). As is the case with most Morrison Formation quarries, no plant fossils were preserved.

See also the following related entry: MORRISON FORMATION

References Britt, B. B., (1991). Theropods of Dry Mesa Quarry (Morrison Formation, late Jurassic), Colorado, with emphasis on the osteology of Torvosaurus tanneri. Brigham Young Univ. Geol. Stud. 37, 1–72. Jensen, J. A., (1985). Three new sauropod dinosaurs from the Upper Jurassic of Colorado. Great Basin Nat. 45(4), 697–709. Jensen, J. A., and Padian, K. (1989). Small pterosaurs and dinosaurs from the Uncompahgre fauna (Brushy Basin



FIGURE 1 Dryosaurus with young (© Mark Schultz, 1996). Member, Morrison Formation: ?Tithonian), Late Jurassic, western Colorado. J. Paleontol. 63(3), 364–374. Prothero, D. R., and Jensen, J. A. (1983). A mammalian humerus from the Upper Jurassic of Colorado. Great Basin Nat. 43(4), 551–553.

Dryosauridae MICHAEL J. RYAN Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Dryosaurids (Fig. 1) were small, primitive, bipedal herbivorous ornithopods from the Upper Jurassic and Lower Cretaceous (Sues and Norman, 1990). Dryosaurids reached 앑6.5 m in length, more than half of which was tail, and weighed 앑70 kg. They closely resemble members of the Hypsilophodontidae and were originally placed within this family. Subsequent work recognized a number of unique features (see Sues and Norman, 1990) that placed dryosaurids in their own monophyletic family and made them a basal taxon within the Iguanodontia. The Dryosauridae comprise four species with a wide distribution across both hemispheres. Dryosaurus altus is known from the MORRISON FORMATION (late Kimmeridgian–early Tithonian) of Colorado, Wyoming, and Utah of the United States. Dryosaurus lettowvorbecki is known from the Tendaguru beds (late Kimmeridgian) of Mtwara, Tanzania. Despite the wide geographic separation, these two species appear

to be very closely related and show only very minor differences (i.e., longer palpebrals in D. altus and a more pronounced olecranon process in D. lettowvorbecki ). This close resemblance has been used to support the close ties of Laurasia and Gondwana into the latter part of the Jurassic (Galton, 1977). Valdosaurus (Galton, 1977) is a poorly known descendant of Dryosaurus from the Lower Cretaceous. Valdosaurus canaliculatus (Galton, 1975) is known primarily from post-cranial material from the Wealden of the Isle of Wright and West Sussex, England, and the Bauxite of Cornet, Romania (Berriasian–Barremian). It is distinguished from Dryosaurus by the deeper, more prominent extensor groove on the distal end of the femur. Valdosaurus nigeriensis (Galton and Taquet, 1982) has been named from a single femur from the El Rhaz Formation of Agadez, Nigeria (late Aptian). Hooker et al (1991) have reported on ornithopod material from the Late Cretaceous of West Antarctica (Vega Island) with dryosaurid affinities. Dryosaurids are relatively small with a short snout giving the skull a triangular appearance in lateral view. The premaxillae are edentulous and smooth and would have born a bony beak in life. The premaxillae do not enclose the external narial openings dorsally. A long dorsocaudal process of the premaxilla contacts the prefrontal and lacrimal separating the nasal from contact with the lacrimal and the maxilla (Sues and Norman, 1990). There is a small gap between the rostral margin of the maxilla and the premaxilla in both embryos and adults. In D. altus the palpebral is a long tapering element that extends com-


198 pletely across the orbit. In embryos, the palpebral extends to three-quarters of the orbit as in the adult D. lettowvorbecki. Of note is a rostral embayment (not present in the embryonic D. altus) on the quadrate similar to the foramen in the quadratojugal of Hypsilophodon. The dentary is narrow and straight margined with an elevated coronoid process. The predentary has a bilobed ventral process similar to that seen in ceratopsids. The thickly enameled buccal surfaces of the maxillary teeth and the corresponding lingual surfaces of the dentary teeth bear one prominent and a number of secondary ridges, the latter of which usually terminate at well-developed denticles along the tooth crown margin. These denticles are seen even in embryonic material (Scheetz, 1991). Dryosaurs would have been efficient processors of vegetable matter. The joint between the premaxilla and the maxilla functioned as a diagonal hinge allowing the skull to rotate slightly relative to the muzzle when the dinosaur chewed. This meant that as the jaws closed on food, with the teeth of the upper jaw rotated slightly outward increasing the shear force between food and teeth. Postcranially, dryosaurs closely resemble hypsilophodontids possessing at least nine cervicals, 15(?) dorsals, six sacrals, and 29⫹ caudal vertebrae (complete tails are not known). The hind limbs are long and slender. The femur has anterior bowing seen in all small ornithopods. It possesses a distinct anterior intercondylar extensor groove on the distal head, a deep separation of the greater and lesser trochanters, and a very well-developed fourth trochanter with a pit at its base for insertion of the M. caudifemoralis longus. The manus is poorly known from incomplete material. The pedal formula is 2-3-4-5-0 with the presence of a vestigial metatarsal I. The structure of the legs would have made dryosaurs very fast runners— something they would have needed because they lacked any other defense against predators. Thulborn (1982) estimated their top speed to be in excess of 40 km/hr, making Dryosaurus one of the fastest ornithischian dinosaurs to be studied to date. Dryosaurus is well known from bone bed material from both Africa and the United States. One site near Uravan, Colorado, has produced more than 2000 Dryosaurus bone fragments with a minimum of eight individuals ranging from 25 cm (hatchling) to 165 cm long. Carpenter (1994) has described an articulated baby D. altus skull from Dinosaur National Monu-

ment in Utah. Heinrich et al. (1993) examined the biomechanical strength of hatchling D. lettowvorbecki femora and determined that young dryosaurs were quadrupedal until at least 2.5 months of age, becoming functionally bipeds when they reached their next size class. Chinsamy (1995) examined the bone histology of D. lettowvorbecki and concluded that this dinosaur had a rapid, uninterrupted pattern of bone growth. Such growth is typical of mammals and birds that maintain a high, constant body temperature by endogenous means. However, like most reptiles, dryosaurs also appear to have had a pattern of indeterminate growth.

See also the following related entry: ORNITHOPODA

References Carpenter, K. (1994). Baby Dryosaurus from the Upper Jurassic Morrison Formation of Dinosaur National Monument. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 288–297. Cambridge Univ. Press, Cambridge, UK. Chinsamy, A. (1995). Ontogenetic changes in the bone histology of the Late Jurassic Ornithopod Dryosaurus lettowvorbecki. J. Vertebr. Paleontol. 15, 96–104. Galton, P. M. (1977). The ornithopod dinosaur Dryosaurus and a Laurasia–Gondwanaland connection in the Upper Jurassic. Nature 268, 230–232. Galton, P. M., and Taquet, P. (1982). Valdosaurus, a hypsilophodontid dinosaur from the Lower Cretaceous of Europe and Africa. Ge´obios 15, 147–159. Heinrich, R. E., Ruff, C. B., and Weishampel, D. B. (1993). Femoral ontogeny and locomotor biomechanics of Dryosaurus lettowvorbecki (Dinosauria, Iguanodontia). Zool. J. Linnean Soc. 108, 179–196. Hooker, J. J., Milner, A. C., and Sequeira, S. E. K. (1991). An ornithopod dinosaur from the Late Cretaceous of West Antarctica. Antarctic Res. 3, 331–332. Scheetz, R. D. (1991). Progress report of juvenile and embryonic Dryosaurus remains from the Upper Jurassic Morrison Formation of Colorado. In Guidebook for Dinosaur Quarries and Trackway Tour, Western Colorado and Eastern Utah (W. R. Averett, Ed.), pp. 27–29. Grand Junction Geological Society, Grand Junction, CO. Sues, H.-D., and Norman, D. (1990). Hypsilophodontidae, Tenontosaurus, Dryosauridae. In The Dinosauria (D. B. Weishampel, D. Dodson, and H. Osmolska, (Eds.), pp. 498–509. Univ. of California Press, Berkeley. Thulborn, R. A. (1982). Speeds and gaits of dinosaurs. Palaeogeogr. Palaeoclimatol. Palaeoecol. 38, 227–256.

E Edmonton Group DAVID A. EBERTH Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada


he Edmonton Group is an important dinosaurand coal-bearing unit in the south-central portion of the Alberta Basin that ranges in age from latest Campanian to early Danian (Early Paleocene) and thus spans the Cretaceous–Paleogene boundary. It comprises a southeastward-thinning, largely nonmarine to shallow marine clastic wedge that is exposed in modern drainages throughout south-central Alberta. It conformably overlies and interfingers with marine shales of the late Campanian Bearpaw Formation and is overlain unconformably by sandstones of the Late Paleocene Paskapoo Formation. It has been a source of important dinosaur and other fossil vertebrate, invertebrate, and plant discoveries since the early part of the 20th century and, recently, has figured importantly in studies of terminal Cretaceous extinctions.

Stratigraphy Gibson (1977) included the Horseshoe Canyon, Whitemud, Battle, and Scollard formations (ascending order) together in the Edmonton Group. This stratigraphic arrangement is widely accepted and is used here. The base of the group (Horseshoe Canyon Formation) is conformable on the underlying marine Bearpaw Formation. Whereas the Horseshoe Canyon–Whitemud and Whitemud–Battle formational contacts are regarded as conformable, a significant unconformity has been postulated for the Battle– Scollard contact (Russell, 1983) and Scollard– Paskapoo contact (Lerbekmo et al., 1992).


Following Cant and Stockmal (1989), the wedge resulted from a major tectonic and basin-response event related to the accretion of the Pacific Rim– Chugach Terrane. In the terminology of Kaufmann (1977), the Edmonton Group records the R9 and T/ R 10 cycles as well as the transgression that coincides with the onset of the Danian. In the sequence stratigraphic terminology of Haq et al. (1988), the Edmonton Group includes the uppermost portion of the Upper Zuni A supercycle (cycles 4.4 and 4.5) as well as the lower portion of the Tejas A supercycle (cycles 1.1 and 1.2). Although the base of the group is strongly diachronous west–east, paleomagnetic analyses of the dinosaur-bearing outcrops in the Red Deer River valley (Lerbekmo and Coulter, 1985) show that the lowest 40 m of these exposures is in the uppermost portion of the 33n magnetochron (latest Campanian), and that the uppermost 5 m of the group extends up into the 28r magnetochron. Radiometric data from bentonites reveal an age of approximately 72 or 73 Ma for the base of the exposed group in the Red Deer River valley (estimated from results published by Lerbekmo and Coulter, 1985). The top of the group occurs approximately 40 m above the Cretaceous–Paleogene (K–P) boundary and is estimated here at approximately 63.5 Ma. The K–P boundary has been recently dated at 64.70 [0.09 Ma (McWilliams and Baadsgaard, 1991)], whereas the Kneehills Tuff, a multistoried, silicified to bentonitic volcanic ash horizon in the Battle Formation, has yielded K–Ar dates that suggest an overall age of 66.5–67.0 Ma (inferred from Lerbekmo and Coulter, 1985).

Edmonton Group



Correlations (Fig. 1) North of the North Saskatchewan River (at Edmonton) the Edmonton and Judith River clastic wedges become indistinguishable; there, the combined Campanian–Maastrichtian wedge is referred to as the Wapiti Formation. In the Foothills Belt of western Alberta, the Edmonton Group correlates with the St. Mary River and Willow Creek formations (southern Foothills) and with the middle to upper Brazeau and Coalspur formations (central Foothills). To the east into southern Saskatchewan, the Battle and Whitemud formations are still recognized but the Horseshoe Canyon and Scollard formations are equivalent to the Eastend and Frenchman formations, respectively. South-eastward into Montana, the group is equivalent to the Fox Hills and Hell Creek formations. Farther southward the Scollard is equivalent to the Lance and the lower part of the Tullock Formation.

Sediments and Paleoenvironments Edmonton Group facies reflect a variety of depositional settings as well as tectonic and climatic influences along the Western Interior seaway during latest Campanian, Maastrichtian, and earliest Paleogene time. In general, coarse sediments range from litharenites to volcanic litharenites (Rahmani and Lerbekmo, 1975; Binda et al., 1991). Claystones comprise bentonite and minor kaolinite. The Horseshoe Canyon Formation (up to 230 m thick) forms an overall regressive unit comprising offlapping, stacked parasequences that record rapid changes in relative sea level (Ainsworth, 1994). A complex variety of depositional environments are present including offshore, shoreface, foreshore, bar-

rier, estuarine, back-barrier, lagoonal, tidal flat, peat swamp, salt marsh, deltaic, distributary channel, oyster bank, meandering and straight fluvial channels, and overbank (Rahmani, 1988; Ainsworth, 1994). Incised valleys have been recently identified (Ainsworth and Walker, 1994; Eberth, 1996). The lower onethird of the formation displays a strong, nearshore marine influence and is depauperate in vertebrate fossils. The remaining two-thirds of the formation is dominated by coastal plain facies that yield the vast majority of vertebrate fossils. The Whitemud Formation (up to 20 m thick) consists of light gray to white weathering bentonitic and kaolinitic sandstones, siltstones, and claystones. Its sediments were deposited on an extensive, very lowgradient alluvial plain that stretched across Alberta and western Saskatchewan. Depositional environments include meandering streams, swamps, ponds, and small lakes (Nambudiri and Binda, 1991). The Battle Formation (up to 14 m thick) consists of dark gray to purplish black bentonites and bentonitic shales with thin interbedded, rooted, siliceous paleosols, erroneously referred to as tuffs. Like the Whitemud, its sediments are thought to have been deposited in very low-gradient alluvial plain to paludal environments during a period of frequent volcanic eruptions. The Scollard Formation (up to 80 m thick) can be divided subequally into a lower, non-coaly interval and an upper, coaly interval (Eberth and O’Connell, 1995). The contact between the two units coincides approximately with the K–P boundary. The lower interval (latest Maastrichtian) comprises an alluvial succession of straight, sandy, paleochannel fills; levee

Edmonton Group


and splay deposits; and a variety of rooted and mottled paleosols and floodbasin deposits. The upper (Paleogene) portion of the Scollard is dominated by meandering channel sediments deposited in environments characterized by a very high water table, extensive interchannel lakes, marshes, and peat swamps. Eberth and O’Connell (1995) proposed that the principal cause of the transition from straight to meandering channels upwards across the K–P boundary in southern Alberta was climatic, resulting from increased wetness, locally rising water tables, and changes in discharge characteristics and in-channel sediment load. Along the Red Deer River, north of Drumheller, the K–P boundary has been accurately placed at the base of, or within the lowest few centimeters of, the thin, though locally extensive, subbituminous coal referred to as the Nevis, or No. #13 seam (Lerbekmo

and St. Louis, 1986). This coal occurs approximately in the middle of the formation, at the lowest portion of the upper coaly unit. A kaolinitic clay layer is locally preserved within or at the base of the Nevis seam and has yielded an iridium abundance spike, an anomalous concentration of shocked quartz, and microdiamonds, and is now widely regarded as resulting from a bolide impact event.

Paleontology The Edmonton Group yields an impressive and diverse assemblage of plant, invertebrate, and vertebrate fossils recording important aspects of the paleoecology of western Canada during the past 7 million years of the Cretaceous and the first 1 million years of the Paleogene (Tables I and II). Although the taxonomic literature is extensive, useful reviews can be found in a series of guidebooks published by staff

TABLE I Fossil Vertebrates from the Horseshoe Canyon Formation Chondrichthyes Elasmobranchii Lamna Sclerorhynchidae Myledaphus bipartitus Osteichthyes Acipenseriformes Acipenseridae Acipenser sp. Diphyodus longirostris Polyodontidae Palaeopsephurus sp. Lepisosteiformes Lepisosteidae Atractosteus sp. Amiiformes Amiidae Cyclurus fragosa Aspidorhynchiformes Aspidorhynchidae Belonostomus sp. Amphibia Anura Undetermined gen. and sp. Caudata Undetermined gen. and sp. Reptilia Chelonia Trionychidae Aspideretes sp.

Dermatemydidae Basilemys sp. Sauropterygia Elasmosauridae Leurospondylus ultimus Archosauromorpha Champsosaurus albertensis Crocodylia Crocodylidae Leidyosuchus sp. Stangerochampsa Saurischia Dromaeosauridae cf. Dromaeosaurus sp. Saurornitholestes sp. Troodontidae Troodon formosus cf. Caenagnathidae Chirostenotes pergracilis Ornithomimidae Dromiceiomimus brevitertius Ornithomimus edmontonensis Ornithomimus currelli Ornithomimus ingens Struthiomimus altus Tyrannosauridae Albertosaurus sarcophagus Daspletosaurus cf. D. torosus Aublysodon sp.

Ornithischia Hypsilophodontidae Parksosaurus warreni Hadrosauridae Edmontosaurus annectens Edmontosaurus regalis Edmontosaurus edmontoni Hypacrosaurus altispinus Saurolophus osborni Pachycephalosauridae Stegoceras edmontonense Stegoceras validus Nodosauridae Edmontonia longiceps Panoplosaurus sp. Ankylosauridae Anodontosaurus lambei Euoplocephalus tutus Ceratopsidae Anchiceratops ornatus Anchiceratops longirostris Arrhinoceratops brachyops Pachyrhinosaurus canadensis Mammalia Multituberculata Unidentified gen. and sp. Marsupicarnivora Didelphodon coyi

Edmonton Group

202 TABLE II Fossil Vertebrates from the Scollard Formation Chondrichthyes Heterodontiformes Palaeospinacidae Palaeospinax ejuncidus Elasmobranchii Sclerorhynchidae Myledaphus bipartitus Osteichthyes Acipenseriformes Acipenseridae Acipenser sp. Lepisosteiformes Lepisosteidae Atractosteus sp. Lepisosteus occidentalis Amiiformes Amiidae Cyclurus fragosa Amia sp. Aspidorhyncbiformes Aspidorhynchidae Belonostomus sp. Elopiformes Albulidae Paralbula sp. Unidentified gen. and sp. Semionotiformes Semionotidae Semionotus sp. Holostean sp. A Amphibia Caudata Scapherpetonidae Scapherpeton tectum Urodela sp. Anura Anura sp. Reptilia Chelonia Dermatemydidae Adocus sp. Compsemys victa

Trionychidae Aspideretes sp. Chelyridae Unidentified gen and sp. Archosauromorpha Champsosauridae Champsosaurus sp. Crocodylia Crocodylidae Leidyosuchus sp. Pterosauria Pterodactyloidea Unidentified gen. and sp. Saurischia Troodontidae cf.Troodon sp. Ornithomimidae Struthiomimus sp. Unidentified gen. and sp. Tyrannosauridae Paronychodon sp. Tyrannosaurus rex Unidentified gen. and sp. Ornithischia Hypsilophodontidae cf. Hypsilophodon sp. Thescelosauridae Parksosaurus sp. Thescelosaurus neglectus Thescelosaurus edmontonensis Pachycephalosauridae ?Pachycephalosaurus sp. Stegoceras sp. Protoceratopsidae Leptoceratops gracilis Ceratopsidae Monoclonius sp. cf. Torosaurus sp. Triceratops albertensis Ankylosauridae Ankylosaurus magniventris

at the ROYAL TYRRELL MUSEUM in Drumheller (e.g., Braman et al., 1995) as well as papers by L. Russell (1964), D. Russell (1967), and Fox (1988). It is the vertebrate faunas, especially the dinosaurs, that are the best known fossils from the Horseshoe Canyon Formation (Table I). Some of the earliest finds of significant remains in Alberta were made along the Red Deer River northwest of Drumheller,

Aves Unidentified gen. and sp. Mammalia Multituberculata Ectypodae Cimexomys priscus Mesodma cf. florencae Mesodma hensleighi Mesodma thompsoni Ptilodontidae Cimolodon nitidus Cimolomyidae Cimolomys gracilis Cimolomys trochuus Undetermined order Deltatheridiidae cf. Deltatheroides cretacicus Marsupicarnivora Didelphidae Alphadon marshi Alphadon wilsoni Alphadon rhaister Pediomyidae Pediomys elegans Pediomys cf. florencae Pediomys hatcherii Pediomys krejcii Stagodontidae Didelphodon vorax Eodelphis sp. Insectivora Gypsonictopidae Gypsonictops hypoconus Gypsonictops illuminatus Palaeoryctidae Cimolestes cerberoides Cimolestes magnus Cimolestes propalaeoryctes Batodon tenuis

and since 1910 more than 100 complete or partial skeletons have been excavated. Hadrosaurs are the most numerous dinosaur preserved, comprising roughly 50% of the known articulated–associated remains. Ceratopsians and ornithomimids are present in subequal abundance and together comprise approximately 30% of known articulated–associated specimens. Tyrannosaurids make up only 5% of the

Edmonton Group


TABLE III Stratigraphic Positions of Articulated Dinosaur Specimens below the Cretaceous–Paleogene Boundary [Base of the Nevis (No. 13) Seam]a Specimen

Position (m)

NMC 8861 cf. Triceratops albertensis NMC 9542 cf. Triceratops albertensis UA 9542 uncataloged cf. Torosaurus sp. TMP 81.12.1 Tyrannosaurus rex NMC 8862 Triceratops albertensis NMC 8889 Leptoceratops gracilis Uncollected large ceratopsian (NMC field records) NMC 887, 8888 Leptoceratops gracilis Uncollected large ceratopsian (NMC field records) NMC 8880 Ankylosaurus magniventris AMNH 5213 Ankylosaurus magniventris NMC 8537 Thescelosaurus neglectus

2.3 4.5 8.0 10.5 33.2 33.2 34.8


34.8 42.4 43.9 45.4 48.5

From Braman et al. (1995).

known assemblage. The most common taphonomic modes include mono- and multigeneric bone beds, vertebrate microfossil sites, and articulated to associated skeletons in overbank, channel thalweg, and point-bar deposits. Other common fossil types include macroplants (Bell, 1949, 1965; Aulenbach and Braman, 1991), nonmarine to marine invertebrates and trace fossils (Saunders, 1989), fish, amphibians, nondinosaurian reptiles, and mammals (Braman et al., 1995). Together, the Whitemud and Battle formations yield very little vertebrate fossil bone but are important sources of plants and coprolites (e.g., Broughton et al., 1978; Binda et al., 1991), especially in southern Saskatchewan. Plant microfossil, seed cuticle, and palynomorph assemblages have been examined at sites across Alberta and Saskatchewan and support the interpretation that together, the Whitemud and Battle form a chronostratigraphic datum in this region (Nambudiri and Binda, 1991). Sideritic vertebrate coprolites (attributed to dinosaurs) have been described by Broughton et al. (1978) and interpreted by Binda et al. (1991) as resulting from syndepositional methanic diagenesis, a process that may have been responsible for the dissolution of shells and bone in these same deposits. There are anecdotal

reports of vertebrate microfossils from the Battle Formation in the Drumheller area. The Scollard Formation (Table II) has yielded important dinosaur remains since Barnum Brown made the first collections in 1910. The fauna includes 20 different taxa of dinosaur and a large diversity of mammals (Russell, 1987). Dinosaur specimens are found throughout the lower one-half (latest Maastrichtian) of the Scollard, up to 2.3 m below the K–P boundary at the base of the Nevis seam (Table III). Articulated dinosaurs are rare and the fauna is known largely by vertebrate microfossil remains associated with channel lags. Hadrosaurs are not represented in collections from Alberta, but ankylosaurs are abundant, especially in the lowest 10 m of the exposures. Tyrannosaurids and ceratopsians dominate the vertebrate assemblage in the 20 m of section below the K–P boundary. This dinosaur assemblage constitutes the most northerly known occurrence of the Lancian Triceratops fauna. An overview of the mammal fauna is presented by Fox (1988). In addition to the dinosaurs, a modest invertebrate assemblage comprising bivalves and gastropods is known as well as an extensive palynomorph assemblage (Braman et al., 1995).


● ●

References Ainsworth, R. B. (1994). Marginal marine sedimentology and high resolution sequence analysis; Bearpaw– Horseshoe Canyon transition, Drumheller, Alberta, Canada. Bull. Can. Petroleum Geol. 42, 26–54. Ainsworth, R. B., and Walker, R. G. (1994). Control of estuarine valley-fill deposition by fluctuations of relative sea-level, Cretaceous Bearpaw–Horseshoe Canyon Transition, Drumheller, Alberta, Canada. In Incised-Valley Systems: Origin and Sedimentary Sequences, SEPM Spec. Publ. No. 51, pp. 159–174. Aulenbach, K. R., and Braman, D. B. (1991). A chemical extraction technique for the recovery of silicified plant remains from ironstones. Rev. Paleobot. Palynol. 70, 3–8. Bell, W. A. (1949). Uppermost Cretaceous and Paleocene floras of western Alberta. Geological Survey of Canada, Bull. 11, pp. 231.

204 Bell, W. A. (1965). Upper Cretaceous and Paleocene plants of Western Canada. Geological Survey of Canada, Paper No. 65–35, pp. 46. Binda, P. L., Nambudiri, E. M. V., Srivastava, S. K., Schmitz, M., Longinelli, A., and Iacumin, P. (1991). Stratigraphy, paleontology, and aspects of diagenesis of the Whitemud Formation (Maastrichtian) of Alberta and Saskatchewan. In Sixth International Williston Basin Symposium (J. E. Christopher and F. M. Haidl, (Eds.), Spec. Publ. No. 11, pp. 179–192. Saskatchewan Geological Society. Braman, D. R., Johnston, P. A., and Haglund, W. M. (1995). Upper Cretaceous paleontology, stratigraphy and depositional environments at Dinosaur Provincial Park and Drumheller, Alberta. Canadian Paleontology Conference Field Trip Guidebook No. 4. Fifth Canadian Paleontology Conference, pp. 119. Geological Association of Canada. Broughton, P. L., Simpson, F., and Whitaker, S. H. (1978). Late Cretaceous coprolites from western Canada. J. Geol. 89, 741–749. Cant, D. J., and Stockmal, G. S. (1989). The Alberta foreland basin: Relationship between stratigraphy and Cordilleran terrane-accretion events. Can. J. Earth Sci. 26, 1964–1975. Eberth, D. A. (1996). Origin and significance of mudfilled, incised valleys (Upper Cretaceous) in southern Alberta, Canada. Sedimentology 43, 459–478. Eberth, D. A., and O’Connell, S. C. (1995). Notes on changing paleoenvironments across the Cretaceous– Tertiary boundary (Scollard Formation) in the Red Deer River valley of southern Alberta. Bull. Can. Petroleum Geol. 43, 44–53. Fox, R. C. (1988). Late Cretaceous and Paleocene mammal localities of southern Alberta. Field Trip ‘‘A,’’ Society of Vertebrate Paleontology 48th Annual Meeting. Occasional Paper of the Royal Tyrrell Museum of Palaeontology, No. 6, pp. 38. Gibson, D. W. (1977). Upper Cretaceous and Tertiary coal-bearing strata in the Drumheller–Ardley region, Red Deer River Valley, Alberta. Geological Survey of Canada, Paper No. 76-35, pp. 41. Haq, B. U., Hardenbol, J., and Vail, P. R. (1988). Mesozoic and Cenozoic chronostratigraphy and cycles of sea-level change. In Sea-Level Changes: An Integrated Approach, (C. K. Wilgus, B. S. Hastings, C. G. St. C. Kendall, H. W. Posamentier, C. A. Ross, and J. C. Van Wagoner, Eds.), Spec. Publ. No. 42, pp. 71–108. Society of Economic Paleontologists and Mineralogists. Kauffman, E. G. (1977). Geological and biological overview: Western Interior Cretaceous Basin. Mountain Geologist 14, 75–99.

Edmonton Group Lerbekmo, J. F., Demchuk, T. D., Evans, M. E., and Hoye, G. S. (1992). Magnetostratigraphy, and biostratigraphy of the continental Paleocene of the Red Deer Valley, Alberta, Canada. Bull. Can. Petroleum Geol. 40, 24–35. Lerbekmo, J. F., and Coulter, K. C. (1985). Late Cretaceous to early Tertiary magnetostratigraphy of a continental sequence: Red Deer Valley, Alberta, Canada. Can. J. Earth Sci. 22, 567–583. Lerbekmo, J. F., and St. Louis, R. M. (1986). The terminal Cretaceous iridium anomoly in the Red Deer Valley, Alberta, Canada. Can. J. Earth Sci. 23, 120–124. McWilliams, M., and Baadsgaard, H. (1991). High resolution 40Ar/ 39Ar ages from Cretaceous–Tertiary boundary bentonites in western North America. Event Markers in Earth History, IGCP Project Nos. 216, 293, and 303, pp. 53. [Abstract] Nambudiri, E. M. V., and Binda, P. L. (1991). Paleobotany, palynology and depositional environment of the Maastrichtian Whitemud Formation in Alberta and Saskatchewan, Canada. Cretaceous Res. 12, 579–596. Rahmani, R. A. (1988). Estuarine tidal channel and nearshore sedimentation of a Late Cretaceous epicontinental sea, Drumheller, Alberta, Canada. In Tide-Influenced Sedimentary Environments and Facies (P. L. de Boer, A van Gelder, and S. D. Nio, Eds.), pp. 433–471. Reidel, Dordrecht. Rahmani R. A., and Lerbekmo, J. F. (1975). Heavy mineral analysis of Upper Cretaceous and Paleocene sandstones in Alberta and adjacent areas of Saskatchewan. In The Cretaceous System in the Western Interior of North America (W. G. E. Cladwell, Ed.), Spec. Paper No. 13, pp. 607–632. Geological Association of Canada. Russell, L. S. (1983). Evidence for an unconformity at the Scollard–Battle contact, Upper Cretaceous strata, Alberta. Can. J. Earth Sci. 20, 1219–1231. Russell, D. A. (1967). A census of dinosaur specimens collected in western Canada. Natural History Papers, No. 36, pp. 13. National Museum of Canada. Russell, L. S. (1964). Cretaceous non-marine faunas of northwestern North America. Life Sciences Contrib. No. 61, pp. 24. Royal Ontario Museum. Russell, L. S. (1987). Biostratigraphy and palaeontology of the Scollard Formation, Late Cretaceous and Paleocene of Alberta. Life Sciences Contrib. No. 147, pp. 23. Royal Ontario Museum. Saunders, T. (1989). Trace fossils and sedimentology of a Late Cretaceous progradational barrier island sequence: Bearpaw–Horseshoe Canyon Formation transition, Dorothy, Alberta, pp. 170. M.Sc. thesis, University of Alberta, Edmonton, Alberta, Canada.

Eggs, Eggshells, and Nests

Egg Mountain JOHN R. HORNER Montana State University Bozeman, Montana, USA

Egg Mountain is the locality name for a site in the WILLOW CREEK ANTICLINE of western Montana, within the Upper Cretaceous (Campanian) TWO MEDICINE FORMATION. Egg Mountain was discovered in 1979 by Princeton University undergraduate Fran Tannenbaum, while on an expedition under the direction of John Horner on the ranch of John and James Peebles. Miss Tannenbaum found a complete fossil egg sitting on top of a rock ledge, atop a small grass-covered hill. Excavation of the hill during the following 17 years has produced hundreds of eggs representing two dinosaurian taxa, as well as numerous skeletons of dinosaurs, lizards, and mammals. Egg Mountain is now owned by the Nature Conservancy. When the Egg Mountain locality was first being worked, the skeletal elements were initially identified as belonging to the theropod dinosaur Troodon and thus many of the eggs were attributed to Troodon. During the following years it was discovered that the majority of skeletal elements actually belonged to a hypsilophodontid dinosaur that was eventually named Orodromeus makelai Horner and Weishampel 1988 to honor Bob Makela, who had accomplished the majority of the excavations and had named Egg Mountain. Subsequent discovery of an egg clutch from a nearby site named Egg Island contained the remains of embryonic skeletons initially identified as hypsilophodontid because this seemed to substantiate the identity of the eggs (Horner, 1987; Horner and Weishampel, 1988). In addition to numerous skeletons of Orodromeus, there were also discoveries of juvenile Troodon skeletons, varanid and teid lizards, and primitive mammals (Montellano, 1988). In 1996, further preparation of the eggs containing the embryos revealed that the embryos were not representatives of Orodromeus, but rather of Troodon (Horner and Weishampel, 1996). Additional discoveries on Egg Mountain within the past few years have suggested the hypothesis that Troodon brooded its eggs similar to Oviraptor, constructed a sediment rim around its egg clutches, and nested in colonies (Horner, 1982; Varricchio et al., 1997). The skeletons

205 of Orodromeus might be food items brought to the nesting area by adults or, because a second type of egg is found in clutches on Egg Mountain (Horner, 1997), these may simply represent attritional mortality. There are three separate, but distinct, nesting horizons within the 10-m thick rock unit comprising the Egg Mountain locality. The locality, interpreted as an island near the shores of an alkaline lake, covers less than 10,000 m2 (Horner, 1987).

See also the following related entries: BASTU´ S NESTING SITE ● BEHAVIOR ● DEVIL’S COULEE ● EGGS, EGGSHELLS, AND NESTS

References Horner, J. R. (1982). Evidence of colonial nesting and site fidelity among ornithischian dinosaurs. Nature 297, 675–676. Horner, J. R. (1987). Ecologic and behavioral implications derived from a dinosaur nesting site. In Dinosaurs Past and Present, Vol. II (S. J. Czerkas and E. C. Olson, Eds.), pp. 51–63. Univ. Washington Press, Seattle. Horner, J. R. (1997). Rare preservation of an incompletely ossified embryo. J. Vertebr. Paleontol., in press. Horner, J. R., and Weishampel, D. B. (1988). A comparative embryological study of two ornithischian dinosaurs. Nature 332, 256–257. Horner, J. R., and Weishampel, D. B. (1996). Correction to: A comparative embryological study of two ornithischian dinosaurs (1988). Nature 383, 103. Montellano, M. (1988). Alphadon halleyi (Didelphidae, Marsupialia) from the Two Medicine Formation (Late Cretaceous, Judithian) of Montana. J. Vertebr. Paleontol. 8(4), 378–382. Varricchio, D. J., Jackson, F., Borlowski, J., and Horner, J. R. (1997). Nest and egg clutches for the theropod dinosaurs Troodon formosus and evolution of avian reproductive traits. Nature 385, 247–250.

Eggs, Eggshells, and Nests K. E. MIKHAILOV Palaeontological Institute of the Russian Academy of Sciences Moscow, Russia

Egg-laying is characteristic of all birds and most extant reptiles, with the exception of many second-

206 arily viviparous snakes and lizards. All known species of birds, turtles, and crocodiles lay eggs with mineralized (calcite) shell. These eggs are hardshelled (i.e., their shell is rigid and brittle), except in some chelonian taxa (sea turtles and some emydid turtles) that possess soft or pliable shells because the basic shell units are underdeveloped. The shell units are not well attached to one another, which permits the passage of water in and out of the egg. Some lizards, namely, geckos and dibamids, also lay hardshelled eggs. In many other cases the soft shells of squamat eggs are encrusted with fine calcareous elements (globules, star-like structures, etc.). Softshelled eggs are unlikely to be fossilized for obvious taphonomic reasons. Based on current understanding, many dinosaurs are known to have been oviparous, and some are known to have been viviparous. Correctly assigned dinosaurian eggs are now known from most of the major divisions of ornithischians and saurischians, including ornithopods (hadrosaurs and hypsilophodontids), ceratopsians (protoceratopsians), sauropods, and therizinosaurid (‘‘segnosaur’’), troodontid, oviraptorid, and perhaps dromaeosaurid theropods. No certain eggs are known from the stegosaurs and ankylosaurs. All known dinosaur eggs possessed hard shells with well-organized crystalline matter as in other hard-shelled vertebrate eggs. The hard shells of reptilian and avian eggs consist of basic vertical shell units that start to grow from particular sites on the surface of the fibrous shell membrane. Mineralogically this consists of calcium carbonate, either in an aragonitic crystallographic form (turtles) or in calcitic form (all other reptiles and birds). In many cases the crystalline matter includes a network of organic matrix, and the fine organizations of both organic and mineral phases are in high accordance. The eggshell is pierced by numerous pore canals that enable gas exchange for the embryos. The morphology and ultra-microstructure of the basic shell units and the particular organization of the pore system are important characteristics in the designation and classification of fossil eggs, most of which belong to dinosaurs. The general taxonomic assignment of fossil eggs is based on the stability of structural types of shells. All members of large systematic groups (such as turtles, crocodiles, and different divisions of birds and dinosaurs) exhibit their own unique eggshell structures. The only reliable argument for the final assign-

Eggs, Eggshells, and Nests ment of eggs to particular genera and species is the discovery of associated embryos and hatchlings, although the correlation of eggs with bones in some localities can be helpful for preliminary identification. The large diversity of fossil egg remains, in particular those of dinosaurs, can only be described and systematically ordered using a special egg parataxonomy (as for footprints and other forms and organ taxa). The pertinent parataxonomic nomenclature initiated by the Chinese paleontologists is now generally accepted and coordinated with different structural classifications. The universal system of identifying fossil egg remains is referred to as Veterovata and includes three working categories, namely, oofamily, oogenus, and oospecies. Oofamilies are combined in larger structural groups correlated with the five basic types of eggshell structure known for vertebrates. Dinosaurian oofamilies are distributed amongst three of these types (Fig. 1). The oofamily and oogenus include the root ‘‘oolithus,’’ derived from ‘‘oolithes’’ (meaning ‘‘stone egg’’), which easily allows one to distinguish egg taxa from animal taxa in vertebrate fossil lists and catalogs. Such structural categories as eggshell morphotype, type of pore system, ornamentation type, egg shape type, and general range of eggshell thickness are used in the concise descriptions and diagnoses of ootaxa. Most recently known dino-

FIGURE 1 Shelled eggs are characteristic of amniotes. The microscopic structure of eggshells varies, and dinosaurs possess the lower three structures diagrammed above.

Eggs, Eggshells, and Nests


TABLE I Distribution of Eggshell Types and Oofamilies among Particular Dinosaur Lineages Basic eggshell type


Testudoid Geckoid Crocodiloid


Dinosauroid–prismatic Ornithoid

Spheroolithidae Ovaloolithidae Megaloolithidae Faveoloolithidae Dendroolithidae Dictyoolithidae Prismatoolithidae Elongatoolithidae Oblongoolithidae Laevisoolithidae Gobioolithidae

saurian oofamilies can be positively coordinated with particular larger dinosaurian groups (Table I). Three clear-cut basic types of eggshell structure can be distinguished for dinosaurs (Fig. 1): dinosauroid–spherulitic type (sauropods and therizinosaurs; hadrosaurines and possibly lambeosaurines), dinosauroid–prismatic type (protoceratopsians and hypsilophodonts), and ornithoid type (theropods). In the dinosauroid–spherulitic type, the basic shell unit roughly looks and grows like spherocrystal. Dinosauroid–prismatic shell displays spherocrystalline structure only in its lower (internal) one-half to onefourth part, whereas the upper portion is prismatic in morphology with homogeneous calcite ultrastructure. The ornithoid type most strikingly differs from the others because the basic shell units are expressed as discrete items (mammillae) only in the lower onesixth to one-third of the eggshell’s thickness. Above (external to) this layer is a mass of biocrystalline material with a squamatic (⫽ spongy) ultrastructure that comprises a single continuous layer. This structure of theropod eggshell is essentially similar to that of birds. The three remaining eggshell types are crocodiloid, testudoid, and geckoid, all of which have different states of eggshell ultra-microstructure than those of dinosaurs. The known diversity of dinosaurian egg remains comprises eight oofamilies and 18 oogenera with more than 40 oospecies. Most of these are known from the Cretaceous of central Asia (China, Mongolia,

Taxonomic group Chelonia Gekkota Crocodilia Ornithopoda: hadrosaurs ?Ornithopoda Sauropoda Sauropoda ‘‘Segnosauria’’ ? Ornithopoda: protoceratopsids, hypsilophodontids Theropoda ?Theropoda or ?Aves Aves: Enantiornithine birds Aves: Flying paleognath birds

Kyrgyzstan, eastern Kazakhstan, and Uzbekistan); some from the Cretaceous of southern Europe, India, and South America; and some from the Jurassic and Cretaceous of North America (Fig. 2, 3). Singular specimens have been found in the Cretaceous of

FIGURE 2 Fossil eggs such as this one are found in Ramos Arizpe County, west of the state of Coahuila, Mexico.

Eggs, Eggshells, and Nests


FIGURE 3 Eggshells of different structure and in different oofamilies have been unearthed around the world.

southern Africa. The oofamilies Elongatoolithidae, Prismatoolithidae, Spheroolithidae, Ovaloolithidae, and Megaloolithidae are the most widespread and diverse (6–12 oospecies); the others are restricted to Mongolia and China and are mostly monotypic. Additionally, two families (Laevisoolithidae and Gobioolithidae) of eggs of Mesozoic birds are known from the Upper Cretaceous of Mongolia. Many dinosaurian eggs and their shells still await formal description. Known dinosaurian nests and eggs exhibit a surprising diversity of form. The eggs of sauropod dinosaurs (see parataxonomic correlation in Table I) that had been found in southern Europe and in the Gobi Desert of Mongolia imply that subsurface nest conditions were similar to those of sea turtles, with the eggs developing in a moist substrate not far from water. Those eggs are large and subspherical in shape and have a highly developed pore system (pore orifices occupy about half of the egg surface). In contrast, the eggs of hadrosaurs, hypsilophodonts, protoceratopsians, and theropods were laid in nests composed of soil and vegetation similar to those of extant alliga-

tors, some crocodiles, and even some birds (Australian megapodids). These nests seem to be rather primitive in arrangement in protoceratopsians and more sophisticated in theropods and hadrosaurs. The hadrosaurs and therizinosaurs produced subspherical eggs, whereas protoceratopsians, hypsilophodonts, and theropods laid strikingly different, elongate eggs that were often set in subvertical position and organized in two or three stacked circles. In many cases, the surfaces of dinosaur eggs are characteristically ornamented, although the functional significance of this feature is not yet clear.

See also the following related entries: BASTU´ S NESTING SITE ● BEHAVIOR ● DEVIL’S COULEE ● EGG MOUNTAIN

References Carpenter, K., Hirsch, K. F., and Horner, J. R. (1994). Dinosaur Eggs and Babies, pp. 372. Cambridge Univ. Press, Cambridge, UK. Hirsch, K. F. (1989). Interpretations of Cretaceous and pre-Cretaceous eggs and eggshell fragments. In Dino-

Elmisauridae saur Tracks and Traces (D. D. Gilette and M. G. Lockley, Eds.), pp. 89–97. Cambridge Univ. Press, Cambridge, UK. Hirsch, K. F. (1994). The fossil record of vertebrate eggs. In The Palaeoecology of Trace Fossils (S. K. Donovan, Ed.), pp. 269–294. Wiley, Chichester, UK. Horner, J. R. (1987). Ecologic and behavioral informations derived from a dinosaur nesting site. In Dinosaur Past and Present, Vol. 2 (S. J. Czerkas and E. C. Olson, Eds.). Washington Univ. Press, Seattle. Mikhailov, K. E. (1991). Classification of fossil eggshells of amniotic vertebrates. Acta Palaeontol. Polonica 36, 193–238. Mikhailov, K. E. (1997). Fossil and recent eggshells in amniotic vertebrates: Fine structure, comparative morphology and classification. Spec. Papers Palaeontol., in press. Mikhailov, K. E., Bray, E. S., and Hirsch, K. F. (1997). Parataxonomy of fossil egg remains (Veterovata): Basic principles and applications. J. Vertebr. Paleontol., in press. Zhao, Zi-kui. (1993). Structure, formation and evolutionary trends of dinosaurian eggshell. In Structure, Formation and Evolution of Fossil Hard Tissues (I. Kobayashi, H. Mutvei, and A. Sahni, Eds.), pp. 195–212. Tokai Univ. Press, Tokyo.

Egyptian Dinosaurs see AFRICAN DINOSAURS

Elmisauridae PHILIP J. CURRIE Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Elmisaurids are small, lightly built theropods of the Late Cretaceous of the Northern Hemisphere (Currie, 1990). Unfortunately, no cranial material has been identified for elmisaurids, which are best known from their feet and hands. The first specimens of Elmisaurus were recovered from the NEMEGT FORMATION of Mongolia (Osmo´lska, 1981) and were considered distinctive enough to erect a new family (the Elmisauridae). Currie and Russell (1988) noted similarities in morphological characters with Chirostenotes from North America and concluded that both Elmisaurus and

209 Chirostenotes may be related to Caenagnathus, an oviraptorosaur. Recent preparation of a skeleton of Chirostenotes (Sues, 1994) has shown that it is probably synonymous with Caenagnathus. Elmisaurus is a smaller animal than Chirostenotes, but its tarsometatarsus is co-ossified. This clue led to the discovery of elmisaurid material from Alberta (Currie, 1989). Small, toothless caenagnathid jaws from central Asia (Currie et al., 1993) bearing the name Caenagnathasia may also be from elmisaurids. If better material confirms this suspicion, then Elmisauridae will become a junior synonym of Caenagnathidae. However, there are enough differences to suggest that making such a move at this time would be premature. The first metacarpal of an elmisaurid is a straight, slender bone, intermediate in relative length between those of most theropods and those of ornithomimids. The first metacarpals are also straight and slender in Chirostenotes and Microvenator (Currie and Russell, 1988), whereas those of oviraptorids are shorter and stouter. Digit III is approximately 30% shorter than digit II, whereas the second finger of an oviraptorid is only slightly longer than the third. Whereas phalanx II-2 is the longest of the manual phalanges in Elmisaurus and caenagnathids, phalanx I-1 is longest in oviraptorids. Although elmisaurids, caenagnathids, and oviraptorids all have well-developed extensor ligament attachments on their manual unguals, only elmisaurids and caenagnathids have arctometatarsalian feet. Elmisaurids are unique among these three taxa in that the distal tarsals and the proximal ends of the metatarsals are co-ossified into a tarsometatarsus. The metatarsi of elmisaurids and caenagnathids are elongate. The elmisaurid metatarsus is strongly arched in section at midlength, with the third metatarsal deeply inset from the flexor surfaces of the second and fourth. When more complete specimens are found, it is possible that elmisaurids will turn out to be caenagnathids. In the meantime, there are enough differences to maintain taxonomic separation. Even though it is clear that elmisaurids are closely related to caenagnathids, the lack of cranial material makes it impossible to determine whether they should be included in the Oviraptorosauria.

See also the following related entries: OVIRAPTOROSAURIA ● THEROPODA



References Currie, P. J. (1989). The first records of Elmisaurus (Saurischia Theropoda) from North America. Can. J. Earth Sci. 26, 1319–1324. Currie, P. J. (1990). The Elmisauridae. In The Dinosauria (D. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 245–248. Univ. of California Press, Berkeley. Currie, P. J., and Russell, D. A. (1988). Osteology and relationships of Chirostenotes pergracilis (Saurischia, Theropoda) from the Judith River (Oldman) Formation of Alberta, Canada. Can. J. Earth Sci. 25, 972–986. Currie, P. J., Godfrey, S. J., and Nessov, L. (1993). New caenagnathid (Dinosauria: Theropoda) specimens from the Upper Cretaceous of North America and Asia. Can. J. Earth Sci. 30, 2255–2272. Osmo´lska, H. (1981). Coossified tarsometatarsi in theropod dinosaurs and their bearing on the problem of bird origins. Palaeontol. Polonica 42, 79–95. Sues, H.-D. (1994). New evidence concerning the phylogenetic position of Chirostenotes (Dinosauria: Theropoda). J. Vertebr. Paleontol. 14, 48A.

Embryos Embryos of dinosaurs have been found preserved within eggs in China, Mongolia, Canada, the United States, and South Africa. See GROWTH AND EMBRYOLOGY for a more complete reference.


Australia, and Antarctica, accompanied by the recognition that considerable partially known material already in museum collections around the world pertains to enantiornithine taxa.

See also the following related entry: AVES

Energetics and Thermal Biology see PHYSIOLOGY

Environments of Dinosaur Preservation Dinosaurs are preserved in rocks whose depositional circumstances vary from aolian (⫽ sand dunes) sediments to intercontinental shallow sea sediments. Thus, dinosaurs probably inhabited most environments where large vertebrates now live, barring frigid areas of today’s poles.


Erenhot Dinosaur Museum PHILIP J. CURRIE

Emery County Museum, Utah, USA see MUSEUMS




Enantiornithes is a group of Cretaceous birds distinguished by an unusual shoulder girdle and tarsometatarsal fusion. Not recognized as a group until 1981, they are now the most numerous and widespread group of Cretaceous birds, thanks to new discoveries in South America, China, North America,

Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

Erenhot is a small Chinese city on the border between Inner Mongolia and Mongolia. Because it is on the Beijing–Moscow trunk of the Trans-Siberian Railroad, many international visitors pass through it every year, little realizing its paleontological significance. It lies close to Iren Nor, the locality where the AMERICAN MUSEUM OF NATURAL HISTORY discovered the first central Asian dinosaurs in 1922. The type section of the Iren Dabasu Formation is located here. Believed to be Cenomanian–Campanian in age (Currie and Eberth, 1993), the limited exposures have proven to be amazingly rich in skeletons and bone beds. Large collections have been made by the CENTRAL ASIATIC, SINO –SOVIET, and SINO –CANADIAN expe-



ditions and by smaller parties from the Beijing Natural History Museum, the Inner Mongolia Museum, and the INSTITUTE OF VERTEBRATE PALAEONTOLOGY & PALEOANTHROPOLOGY (Beijing). In 1989, after several years of planning and collection, a local museum opened in Erenhot. The Erenhot Dinosaur Museum displays mostly specimens from the Iren Dabasu Formation, including skeletons of Bactrosaurus and Archaeornithomimus, and isolated bones of hadrosaurs, sauropods, small theropods, and therizinosaurids.

See also the following related entries: CHINESE DINOSAURS ● MONGOLIAN DINOSAURS

Reference Currie, P. J., and Eberth, D. A. (1993). Palaeontology, sedimentology and palaeoecology of the Iren Dabasu Formation (Upper Cretaceous), Inner Mongolia, People’s Republic of China. Cretaceous Res. 14, 127–144.


Euornithopoda was established by Sereno (1986) to include the ornithischian taxa heterodontosaurs, hypsilophodonts, iguanodonts, and hadrosaurs. The node is supported by several synapomorphies, including the ventral offset of the premaxillary tooth row relative to the maxillary tooth row, crescentshaped paroccipital process, jaw joint lower than occlusal plane, and exclusion of the maxillary–nasal contact by the premaxilla. Sereno used the term Euornithopoda as a node-based sister taxon to MARGINOCEPHALIA within ORNITHISCHIA. Most authors have continued to use the term Ornithopoda for this taxon, although in Sereno’s phylogeny the term Ornithopoda was confined to the Euornithopoda excluding heterodontosaurs.

See also the following related entry: ORNITHOPODA


Espe´raza see MUSEE´



Sereno, P.C. (1986). Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). Natl. Geogr. Res. 2, 234–256.

Gryposaurus notabilis is an euornithopodan hadrosaur found in Cretaceous sediments that are exposed at Dinosaur Provincial Park, Alberta, Canada. (Photo by Franc¸ois Gohier.)

European Dinosaurs ERIC BUFFETAUT Laboratoire de Pale´ontologie des Verte´bre´s, Universite´ Paris Paris, France


saur bones first become abundant, however, in higher levels of the Late Triassic usually referred to the late Norian. Sites of that age include the famous German prosauropod localities in the Knollenmergel, such as Trossingen in Wu¨rttemberg and Halberstadt in Sachsen-Anhalt, which have yielded many well-preserved skeletons of Plateosaurus, and various other sites in other parts of Germany, including the Grosse Gleichberg in Thuringia, where the theropod Liliensternus has been found. Outside Germany, Plateosaurus remains have been found in abundance in the upper Norian of Switzerland and in rocks of the same age in the French Jura mountains. Late Triassic dinosaur remains from Britain include the prosauropod Thecodontosaurus from the Magnesian Conglomerate near Bristol, which is Norian to Rhaetian in age. Dinosaur remains from the controversial ‘‘Rhaetian’’ stage, at the top of the Triassic, in various European countries (Britain, France, Germany, and Belgium) are usu-

he first dinosaur remains to be described scientifically were found in Europe at the beginning of the 19th century. Since then, however, European dinosaurs, which are often represented by incomplete specimens, have been somewhat overshadowed by finds of spectacular complete skeletons from other parts of the world, including North America, Africa, and Asia. Nevertheless, the European dinosaur record, which includes skeletal remains, footprints, and eggs, is probably the most complete in the world from a stratigraphic point of view, with relatively few gaps in a series of sites covering the time span from the Late Triassic to the end of the Cretaceous. Many European countries (Fig. 1) have yielded remains belonging to various groups of dinosaurs and a stratigraphic rather than systematic or geographic presentation seems to be the most appropriate way to present the European record in a concise way. Only the main sites and assemblages have been included.

Late Triassic In central and western Europe, the Late Triassic is largely represented by continental deposits, in which dinosaur bones and footprints sometimes occur in great abundance. Although there have been reports of dinosaur footprints in rocks older than the Late Triassic in several parts of Europe, including England and France, they appear to be erroneous or inconclusive. The European dinosaur record actually begins with small three-toed footprints from the Carnian of northern Bavaria and possibly from the Ladinian– Carnian boundary in the Swiss Alps. The earliest skeletal remains are Norian in age. The earliest relatively well-known European dinosaur is the prosauropod Sellosaurus gracilis, from the lower Stubensandstein of Wu¨rttemberg, considered early Norian in age. Slightly more recent, probably middle Norian, levels of the Stubensandstein of the same region have yielded remains of the theropod Halticosaurus. Dino-

FIGURE 1 Map of Europe showing the main dinosaur localities. 䉱, Triassic; 䊉, Jurassic; 夝, Cretaceous.


European Dinosaurs ally very fragmentary elements from bone beds that are not easily identifiable; some can be referred to indeterminate prosauropods. An exception is the melanorosaurid prosauropod Camelotia from Somerset, England. Norian dinosaur footprints, mostly attributable to theropods, are known from several sites in northern Bavaria, and also from northern Italy. Several footprint sites of Rhaetian age are known from France, the most important one being at Le Veillon (Vende´e), on the Atlantic coast. There, hundreds of footprints have been found, most of them referable to theropods (Grallator and Eubrontes), although some have been attributed to ornithischians. ‘‘Rhaeto-Liassic’’ tridactyl footprints have also been reported from Scania in southern Sweden.

Early Jurassic In Europe, the Early Jurassic is represented largely by marine deposits in which dinosaur remains are relatively infrequent. An incomplete theropod skeleton, recently redescribed as Liliensternus airelensis, comes from beds very close to the Triassic–Jurassic boundary, probably basal Hettangian in age, at Airel in Normandy (northwestern France). In England, more or less complete skeletons of the early thyreophoran Scelidosaurus harrisonii have been found since the 19th century in the marine Sinemurian of the Dorset coast. Very incomplete remains of another possible thyreophoran, Lusitanosaurus, were reported from the Sinemurian of Portugal. A third primitive thyreophoran, Emausaurus ernsti, was described on the basis of a skull and some postcranial elements from the Toarcian of Mecklenburg-Vorpommern in northern Germany. Hindlimb elements of a sauropod from the Toarcian Posidonienschiefer of Baden-Wu¨rttemberg (southwestern Germany) have been described as Ohmdenosaurus liasicus. Besides these scattered skeletal remains from marine deposits, which also include a few isolated theropod bones or teeth from England, Scotland, France, and Germany, dinosaur footprints of Early Jurassic age occur in several European countries, often in calcareous rocks deposited in a beach or mudflat environment. Such trackways, mostly from theropods, are known from the Hettangian and Pliensbachian of the southern rim of the French Central Massif. Other Early Jurassic dinosaur footprints, most of them referable to theropods, have been reported from the Hettangian of

213 Hungary and the Holy Cross Mountains of Poland. Among the early Liassic footprints from the Rovereto area in northern Italy, some have been referred to a very early sauropod.

Middle Jurassic The Middle Jurassic dinosaurs of Europe are better known than those from the Lower Jurassic mainly because of a few comparatively rich localities in England and France. Bajocian forms are poorly known, although remains of the theropod Megalosaurus have been described from the Inferior Oolite of Dorset, England. Dinosaur remains have been found in abundance in several Bathonian formations in England, including the famous Stonesfield ‘‘Slate’’ which yielded the remains of Megalosaurus bucklandi, the first dinosaur to receive a proper generic name, in 1824. The Bathonian of England also contains remains of the large sauropod Cetiosaurus oxoniensis and stegosaurs. In France, the Bathonian of the Caen area in Normandy yielded remains of a large theropod, that was described as Poekilopleuron bucklandi, but may in fact belong to Megalosaurus. Theropod and sauropod teeth are also known from the marine Bathonian of Saint-Gaultier in central France. The Callovian dinosaurs of Europe are relatively well known because of fairly numerous finds from the marine Lower Oxford Clay of England and its equivalents in Normandy. The English finds include theropods such as Eustreptospondylus oxoniensis, represented by a nearly complete skeleton, sauropods (Cetiosauriscus and Ornithopsis), ornithopods (Callovosaurus), stegosaurs (Lexovisaurus) and one of the oldest known ankylosaurs (Sarcolestes). The French record from Calvados in Normandy is somewhat less diverse but nevertheless includes theropods (with at least two forms, one of which seems to be referable to Megalosaurus; the status of Piveteausaurus, known from a braincase from Normandy, is uncertain), stegosaurs (with a partial skeleton of Lexovisaurus), and sauropods. Middle Jurassic dinosaur footprints have been reported from several sites in England.

Late Jurassic The Late Jurassic European dinosaur record is good, with important localities in several countries. Oxfordian theropod remains are known from marine deposits in England (with Metriacanthosaurus parkeri

European Dinosaurs

214 from Dorset) and Normandy (with Megalosaurus bones from the Vaches Noires Cliffs of Calvados). Isolated theropod and sauropod teeth have been found in the Oxfordian fluvio-marine Sables de Glos of Normandy. One of the best sauropod skeletons ever found in Europe comes from the Oxfordian of Damparis (Jura, eastern France); it belongs to a brachiosaurid and was found associated with several theropod teeth, indicating in situ scavenging on the sauropod carcass during a phase of emersion of a carbonate platform. In Portugal, the Guimarota lignite mine, of Late Oxfordian or Kimmeridgian age, has yielded hypsilophodontid teeth and, recently, teeth referred to Archaeopteryx. Kimmeridgian dinosaurs are known from the Kimmeridge clay of England. They include poorly known sauropods, stegosaurs (Dacentrurus), and ornithopods (including a fairly complete skeleton of Camptosaurus prestwichii from the vicinity of Oxford). The marine Kimmeridgian of the Normandy coast near Le Havre also contains dinosaur remains, including theropods, sauropods, and stegosaurs. In the Jura mountains of Switzerland, abundant sauropod tracks have been found in Kimmeridgian limestones, as well as an incomplete sauropod skeleton. Other sauropod footprints are known from rocks of the same age in Barkhausen in northern Germany. A few dinosaur bones have been reported from the marine Kimmeridgian of the Boulonnais, in northern France, but most of the remains from that area are from nonmarine Tithonian rocks. They include bones and teeth of a large theropod, a large camarasaurid sauropod (Neosodon), and a small iguanodontid. The small theropod Compsognathus is known from two fairly complete skeletons from Tithonian lithographic limestones, in northern Bavaria and southeastern France. Abundant dinosaur remains have been discovered in the Upper Jurassic (Kimmeridgian and Portlandian) of Portugal. The Kimmeridgian assemblage includes several sauropods, including a probable brachiosaurid, theropods, stegosaurs (Dacentrurus), an early ankylosaur (Dracopelta), and ornithopods. Footprints (of theropods and sauropods) are frequent in the Portlandian. The discovery of numerous dinosaur eggs in the Kimmeridgian of the Lourinha region is especially noteworthy. In Asturias (northern Spain), large footprints referable to sauropods have been reported from Kimmeridgian rocks.

Early Cretaceous The most varied Early Cretaceous dinosaur assemblage in Europe comes from the non-marine Wealden beds of southern England (Sussex and Isle of Wight), which range in age from late Berriasian to early Aptian. The Wealden fauna includes several species of Iguanodon; other ornithopod genera such as Hypsilophodon, Valdosaurus, and Vectisaurus; the nodosaurids Hylaeosaurus and Polacanthus; the early pachycephalosaur Yaverlandia; several sauropods (including a brachiosaurid, a titanosaurid, and a possible diplodocid); and several theropods, among which is the probable spinosaurid Baryonyx. On the continent, the most famous Wealden dinosaur locality is undoubtedly the Bernissart coal mine in Belgium, where about 30 skeletons of Iguanodon, belonging to the species I. bernissartensis and I. atherfieldensis, were discovered in 1878. Dinosaurs also occur in the Wealden of northern Germany, where skeletal remains (including those of a theropod, an ornithopod, and the basal marginocephalian Stenopelix) are usually far less abundant than footprints. The latter include tracks of sauropods, ornithopods, and theropods from the region around Hanover (Bu¨ckeberg and Mu¨nchehagen). An exception is the Nehden locality in Westphalia, a karstic deposit of Aptian age, in which abundant remains of I. bernissartensis and I. atherfieldensis have been discovered. In the eastern Paris Basin, a succession of alternating shallow marine and non-marine beds, ranging in age from Hauterivian to Aptian, has yielded welldated Iguanodon remains that show a succession of species similar to that from the English Wealden. Remains of a brachiosaurid sauropod are also known from the Barremian of that area. In southern France, a few bones of an Allosaurus-like theropod have been described from the marine Valanginian of the department of Gard. Early Cretaceous non-marine rocks in eastern Spain contain dinosaur assemblages. The mainly Barremian beds of the Galve area (Teruel Province) have yielded hypsilophodontids, theropods, and four forms of sauropods, including the camarasaurid Aragosaurus. From the early Aptian of the Morella region (Castellon Province), theropods, sauropods, ankylosaurs, and Iguanodon have been reported. Early Cretaceous ankylosaurs, iguanodontids, and theropods are also known from Burgos Province in north-central

European Dinosaurs Spain. In Cuenca Province in central Spain, the late Hauterivian to early Barremian lacustrine lithographic limestones at Las Hoyas have yielded a few dinosaur remains, including the early multitoothed ornithomimosaur Pelecanimimus. Abundant ornithopod and theropod footprints are known from the Lower Cretaceous of the La Rioja region in northern Spain. A few skeletal remains (of theropods, sauropods, and ornithopods) and tracks are also known from the Lower Cretaceous (Hauterivian and Aptian) of Portugal. Among the few dinosaurs reported from Italy is the skeleton of a very small theropod from the Aptian lithographic limestones of Pietraroia (Benevento Province). In eastern Europe, the most important Early Cretaceous dinosaur assemblage is that from the karstic bauxite deposits (Barremian to Aptian in age) of Cornet, in Transylvania (Romania). It mainly contains ornithopods (Iguanodon, Valdosaurus, and Dryosaurus), as well as a theropod. On the northern outskirts of Europe, ornithopod and theropod footprints have been reported from the Barremian of Spitzbergen. Albian dinosaurs are known from localities in England and France. The English material mainly comes from the Cambridge Greensand, mostly as disarticulated and reworked material. It includes sauropods, ornithopods (including a probable iguanodontid and a hadrosaurid), and ankylosaurs . In France, sauropod remains have been found in the Albian of the Pays de Caux and the Pays de Bray in Normandy, as well as in the ‘‘Gault’’ clay of the eastern Paris Basin. Dinosaurs from the phosphate-bearing Albian of northeastern France include the enigmatic theropod Erectopus. In southeastern France, a sauropod humerus was found in the Albian green sandstones of the Mont Ventoux.

Late Cretaceous The European dinosaurs of the early part of the Late Cretaceous, up to the Campanian, are relatively poorly known. In England, the nodosaurid ankylosaur Acanthopholis horridus comes from the Cenomanian Chalk Marl, and an ornithopod has been reported from the Totternhoe Stone, also Cenomanian in age. A few isolated and fragmentary specimens of sauropods and theropods are known from the Cenoman-

215 ian, Turonian, and Santonian of west-central France. A few dinosaur remains (belonging to a theropod and to the iguanodontid Craspedodon) have been collected from the Santonian of Lonze´e in Belgium. Cenomanian footprints are known from Portugal (theropods and sauropods) and from Croatia (sauropods). The European dinosaur record becomes tolerably good again in the Campanian. A few isolated specimens are known from marine rocks of that age in southern Sweden (theropod and ornithopod) and southwestern France (sauropod). One of the best early Campanian localities was found in lignite-bearing beds at Muthmannsdorf in Austria in the 19th century; it yielded a theropod, the ornithopod Rhabdodon, and the nodosaurid ankylosaur Struthiosaurus. In southern France, the Villeveyrac locality, of the same age, also contains Rhabdodon, an ankylosaur, and a small theropod. In southeastern France, an abelisaurid theropod, Tarascosaurus, has been reported from Campanian beds at Le Beausset (Var). Maastrichtian dinosaurs are known from marine Chalk deposits in the Limburg region of Belgium and the Netherlands, near the city of Maastricht. They include a theropod of uncertain affinities (Betasuchus) and hadrosaurs. Hadrosaur remains have also been found in the marine upper Maastrichtian of the Bavarian Alps, and there is one hadrosaur record from the marine Maastrichtian of the Crimea in Ukraine. However, most of the Maastrichtian dinosaurs from Europe have been found in non-marine rocks in Transylvania, in southern France, in northern Spain, and in Portugal (some of the French-Iberian localities may actually be late Campanian in age). The Transylvanian assemblage is apparently late Maastrichtian in age. It includes Rhabdodon, the primitive hadrosaur Telmatosaurus, the nodosaurid ankylosaur Struthiosaurus, titanosaurid sauropods (Magyarosaurus), and poorly known theropods. Also worth mentioning is the occurrence of eggs associated with remains of hadrosaur embryos. Whereas the Portuguese record is poor (with indeterminate theropods and ornithischians), there are good late Campanian to Maastrichtian dinosaur localities in various parts of Spain. In the Tremp basin of Catalonia, sauropods, hadrosaurs, and Rhabdodon occur in late Maastrichtian deposits. Footprints (including those of sauropods) and eggs have also been found. Other Late Cretaceous dinosaurs are known

European Dinosaurs

216 from various provinces of Spain. One of the richest sites is at Lano, near Vitoria in the Basque Country. It is of late Campanian to early Maastrichtian age and has yielded bones and teeth of abelisaurid theropods, titanosaurid sauropods, Rhabdodon, and the ankylosaur Struthiosaurus. In southern France, dinosaur localities of early Maastrichtian age are known in great number in nonmarine formations extending from Provence in the east to the foothills of the Pyrenees in the west. The best known assemblages are from the Fox-Amphoux area in Provence and the upper valley of the Aude River. Skeletal remains indicate the occurrence of abelisaurid and dromaeosaurid theropods, armored titanosaurid sauropods, the ornithopod Rhabdodon, and an ankylosaur. Eggs are abundant in some areas, especially the Aix-en-Provence basin and the upper Aude valley. Although several types have been distinguished on the basis of shell microstructure, no clear associations with skeletal remains have yet been reported. Less numerous late Maastrichtian localities are also known, mainly from the Corbie`res region and the area of the Garonne valley. They contain a fauna dominated by hadrosaurs, accompanied by theropods and ankylosaurs. The southern French record thus suggests a faunal change, marked by a decline of titanosaurid sauropods and an expansion of hadrosaurs, during the Maastrichtian. From this brief review, it appears that the European record covers most of dinosaur history in a remarkably complete way. In the Triassic and Jurassic, European dinosaur faunas showed clear resemblances to those of other continents, notably North America and Africa. In the Early Cretaceous, resemblances to North America were still marked (with such genera as Iguanodon and Polacanthus in common). In the Late Cretaceous, however, the European assemblages showed characteristics of their own, with taxa of ‘‘Gondwanan’’ affinities such as abelisaurid theropods and titanosaurid sauropods playing an important part together with endemic forms, whereas taxa of ‘‘Asiamerican’’ affinities were few. This of course reflects the changing paleogeographical history of the European continent during the Mesozoic, which alternatingly favored or limited faunal exchanges with North America, Asia, and Africa. The details of this complex biogeographical history still have to be worked out in detail.


References Buffetaut, E. (1995). Dinosaures de France, pp. 144. Editions du BRGM, Orle´ans, France. Buffetaut, E., and Le Loeuff, J. (1991). Late Cretaceous dinosaur faunas of Europe: Some correlation problems. Cretaceous Res. 12, 159–176. Buffetaut, E., Cuny, G., and Le Loeuff, J. (1991). French dinosaurs: The best record in Europe? Mod. Geol. 16, 17–42. Bultynck, P. (1989). Bernissart et les Iguanodons, pp. 115. Institut Royal des Sciences Naturelles de Belgique, Bruxelles, Belgium. Dantas, P. (1991). Dinossaurios de Portugal. Gaia 2, 17–26. Jaeger, M. (1986). Die Dinosaurier der Schweiz und der Bundesrepublik Deutschland, pp. 1–40. Schriften des Bodensee–Naturmuseums, Konstanz, Germany. Jurcsak, T., and Kessler, E. (1991). The Lower Cretaceous paleofauna from Cornet, Bihor County, Romania and its importance. Nymphaea 21, 5–32. Leonardi, G., and Avanzini, M. (1994). Dinosauri in Italia. Sci. Quaderni 76, 69–81. Probst, E., and Windolf, R. (1993). Dinosaurier in Deutschland, pp. 316. C. Bertelsmann, Mu¨nchen, Germany. Sanz, J. L., Buscalioni, A. D., Moratalla, J. J., France´s, V., and Anton, M. (1990). Los reptiles mesozoicos del registro espanol. Monogr. Museo Nacional Ciencias Nat, 1–81. Weishampel, D. B. (1990). Dinosaurian distribution. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmolska, Eds.), pp. 63–139. Univ. of California Press, Berkeley. Weishampel, D. B., Grigorescu, D., and Norman, D. B. (1991). The dinosaurs of Transylvania. Nat. Geogr. Res. 7, 196–215.

Everhart Museum, Pennsylvania, USA see MUSEUMS



Evolution J. DAVID ARCHIBALD San Diego State University San Diego, California, USA


n everyday usage, evolution is used most often in the sense of change, but frequently there is the added connotation of progress. In biology there is no single definition for evolution, which is often qualified as organic or biological evolution. For the population biologist it might be expressed as a change in the frequency of certain gene types in a species or population with concomitant changes in phenotype (visible aspects of the organism). More generally we could define evolution as the change in a species or lineage of populations over some number of generations. One of the best definitions is that of Charles Darwin, who called it ‘‘descent with modification.’’ This is the simplest and most eloquent definition because it embodies both elements that the majority of biologists feel are key in evolution—descent because there is continuity of life from one generation to the next through genetic inheritance, and modification because life demonstrably changes over time with changes in the genes. Note, however, that in none of these definitions or characterizations is there a clear explanation of a cause. Charles Darwin (Fig. 1) is often and incorrectly portrayed as the scientist who ‘‘discovered’’ evolution; rather, he argued persuasively for the principal mechanism that seemed to explain most of what we see in nature. Darwin did not even care for the term evolution, preferring transmutation or descent with modification. The word evolution does not even appear in his famous book On the Origin of Species by Means of Natural Selection or in the Preservation of Favoured Races in the Struggle for Life that appeared in 1859. The word ‘‘evolved’’ does appear, however, but only once as the very last word in the text. It was the English philosopher, Herbert Spencer, who first used evolution in basically the sense we use it today. Previously, the word was usually used by a group of biologists who argued that all organisms began as a preformed individual within the sperm or egg of the parent. To them, evolution was the process of unfolding that occurred as the preformed individual (a

homunculus in the case of humans) grew or unfolded from the sex cell. For Spencer, it became the unfolding of life in general. It was also Spencer who coined the phrase ‘‘survival of the fittest’’ and the now largely discredited sociological theory of ‘‘social Darwinism.’’ Although Darwin began to use the phrase in the fifth edition of The Origin of Species, it has had unfortunate connotations. As philosophers pointed out, it could be construed as a tautology; that is, a statement that is true simply because its terms so define it (all mothers are female). The fittest are by definition those that survive, thus the survival of the fittest becomes survival of the survivors. It might be more accurate to speak of the survival of the adequate because although Darwin later did use Spencer’s phrase, he usually wrote about the variations that organisms possessed that could be advantageous or injurious to their survival. This was the essence of Darwin’s incalculable contribution to evolutionary theory— that nature, just as the breeder of plants and animals, ‘‘selects’’ from among the available variations those that will be passed along to subsequent generations. This is Darwin’s major contribution—natural selection. Natural selection can be defined as the differential contribution of heritable variation to the next generation. Today, natural selection is still regarded as the major mechanism that drives the engine of evolutionary change. We now realize, however, that chance does play a very important role. A cataclysmic ecological event or even the piggybacking of one nonselected gene along with another that is under selective pressure could simply by chance determine which types of genes will survive. From his time as naturalist on the naval surveying vessel, the H. M. S. Beagle, starting in 1831 through the publication of The Origin of Species in 1859, Darwin not only developed and honed his ideas on natural selection but also developed a general explanation of the process of evolution through his keen powers




FIGURE 1 Charles Darwin, 1809–1882.

of observation and deduction. The major points of Darwin’s descent with modification (or evolution) by means of natural selection are as follows: (i) organisms within a species vary, (ii) some of these variations are inherited by offspring, (iii) more offspring are produced than can possibly survive, (iv) usually offspring with variations favored by the environment will survive, (v) surviving offspring will in turn usually leave more offspring with variations favored by environment, and (vi) over time variations favored by the environment will accumulate. The two essential components of evolutionary thought that eluded Darwin are how the variations arise and how they are passed on to the next generation. Darwin’s ideas on the origin of variations and how they are transmitted to the next generation hark back to the older ideas of the French scientist Lamarck, who argued in the early 19th century that influences of the environment can be passed along to the next generation. Much later Lamarck’s ideas on evolution were somewhat incorrectly simplified to the phrase ‘‘inheritance of acquired characteristics.’’

It was Darwin’s contemporary, the Austrian monk Gregor Mendel, who began to reveal the secrets of genetics. It is ironic that as far as can be determined, Darwin did not know of Mendel’s work, although Mendel did know of Darwin. Even if Darwin did know of Mendel, he clearly did not connect the importance of Mendel’s findings to his own work on variation and natural selection. Of the critics of Darwin’s ideas on evolution, the only one of considerable interest here is Richard Owen, the originator in 1842 of the name and concept DINOSAURIA, and the describer in 1862 of the London specimen of the earliest bird, Archaeopteryx lithographica. Both the early ideas of dinosaurs and Archaeopteryx influenced and in turn were influenced by ideas on evolution. Early in their careers, Darwin and Owen were on quite cordial terms. Owen even described in 1840 fossil mammals from South America that Darwin brought back from his voyage on the H. M. S. Beagle. With the publication of The Origin of Species (Darwin, 1859), the relationship between Darwin and Owen became very strained. Often and incorrectly, but not surprisingly, Owen is portrayed as antievolutionist. In fact, it is clear that he thought change had occurred in the organic world. What was lacking in Owen’s limited work on the subject, however, was a clearly articulated idea on the mechanism(s) of evolution. He did not totally reject Darwin’s main theme of natural selection, but he ranked it as a minor factor. The naming and defining of Dinosauria by Owen at a meeting in 1841 (published the following year) has been viewed as antievolutionary. Of course Darwin had not yet published his societally shattering work, but ideas about evolution (or transformation) were becoming more common in the scientific community. Owen’s reconstructions of dinosaurs, later immortalized in the 1850s in statuary by Waterhouse Hawkins, were rather mammal like in that they all were portrayed in a quadrupedal position. This has been used to argue that Owen wished to show that modern reptiles are far less advanced than their dinosaurian precursors, a presumed blow against what we today would call progressive evolution. It is far more likely, however, that Owen was simply applying the best known taxonomic and anatomical techniques of the time to reconstruct his best estimation of what these creatures looked like. Any potential side benefits adding ammunition against the Darwin view of evolution would have been welcomed.

Evolution Turning to Archaeopteryx, Owen secured for the British Museum in 1862 what at the time was the best specimen of A. lithographica. It was hailed by some of Darwin’s supporters as a missing link between reptiles and birds. It was also regarded by Darwin as an exception that proved the rule that the fossil record is very imperfect. Thus, the fossil intermediates that Darwin thought would be required if his theory was true should also be very uncommon. Owen did not regard Archaeopteryx as an intermediate but rather as a true bird that demonstrated his ideas of progress from the general to the more specific. As one of England’s preeminent scientists of the time, jealousy over Darwin’s success seems to have also prevented Owen from accepting natural selection as the single most important evolutionary mechanism. Although evolutionary change as described by Darwin is eloquently simple, there are misconceptions about what evolution by natural selection is and is not. Two commonly held misconceptions about evolution are that it is progressive and that it has or shows purpose. These can be explored by starting with another common but incorrect view of evolution, the scala naturae. The ‘‘scale of nature’’ is traceable back to ancient Greek philosophers such as Aristotle who arranged life from humans, with the highest kind of soul, down (skipping some groups) through mammals, birds, reptiles, fishes, and finally plants at the bottom. This was the great and continuous chain of being that prevailed and continued to grow through the Middle Ages in Europe. This is also why, even today, we still commonly speak of ‘‘higher’’ and ‘‘lower,’’ or ‘‘advanced’’ and ‘‘primitive’’ forms of life. Unconsciously we still view life as a linear hierarchy with humans at the pinnacle. In fact, all species are mosaics of characteristics that have changed little from the ancestor and other characteristics that are greatly changed. For example, both the ‘‘lowly’’ opossum and humans retain the ancestral condition of five digits on hands and feet; humans clearly evolved a much larger, more complex brain than an opossum; but opossums have reduced the number of tooth generations to essentially one whereas humans still must rely on the tooth fairy because we still retain two generations of teeth from our ancient ancestors. Both the study of the pattern of the history of life and the study of the process of evolution emphatically show that life does not form a linear hierarchy. In fact, before evolution became widely accepted in the past century, some scientists who did not even sub-

219 scribe to evolution showed that life is not a ladder, but rather a bush or tree; hence our metaphor of a phylogenetic tree to show the history of life. The great French comparative anatomist, paleontologist, and antievolutionist, Georges Cuvier (Fig. 2), argued that animal life could be arranged in four distinct plans ( gembranchements). The four ‘‘branches’’ that Cuvier recognized were the vertebrates, the molluscs, the articulates (insects and worms), and the radiates ( jellyfish, starfish, etc.). Except for the unnatural radiate grouping, we still recognize the other three groupings along with others within Animalia. With the shift to evolutionary thinking in biology in the mid to late 19th century, it became clear that the different branches of organisms formed a bush of life taking its origin someplace in the deep recesses of time. We now know that many of these major branches of life show up in the fossil record by some 500 million years ago, not much after animals evolved hard supporting structures or skeletons that were

FIGURE 2 Georges Cuvier, 1769–1832. Photo courtesy of the National Library of Medicine.


220 easily preserved. Thus, our own major branch, the vertebrates, and that including insects, the arthropods, appeared within the same geological time frame rather than vertebrates showing up as the last and highest form of life. Humans are not more highly evolved than ants, simply differently evolved. Even though there is clearly no ordering of life from the lowest to the highest species in a scala naturae, why can we not argue that within the branches of life the later appearing forms have progressed compared to older forms? In Darwinian terms, why can we not argue that because the species are still here, they are better adapted to their environment and thus are more progressive? The answer is that species have evolved and adapted (or became extinct) to conditions that prevailed at the time they were alive. The physical conditions, such as climate and topography, and the biological conditions, such as competition within species or food sources, are constantly shifting over geological time. So too must the species shift or become extinct. In fact, extinction is the unquestioned rule rather than the exception: Approximately between 90 and 99.9% of all species that have ever lived are extinct. Progress can certainly be rejected as detectable in evolution. This does not mean that there are no detectable trends or directions, but these are determined only by looking backwards from the present. Also, our predictive powers are of only the broadest kind. Very broad predictions or directions in evolution are obvious if for no other reason than, like all other parts of the known universe, living organisms must obey physical laws. As early plant and animal life evolved, it invaded the land. For animals we know that this invasion was successfully accomplished a bare minimum of five times. Another big jump was to flight, which in vertebrates alone evolved a minimum of three times. Thus, we can detect major directions and even predict what the general physical requirements would be (lungs for breathing and wings for flight). There is another general direction in evolution—a

general increase in complexity. This does not mean that all branches show increased complexity, but rather only that there is increased complexity in some branches. This occurs from the molecular level, with a general increase of genetic information over time in some groups, to the increase in larger structures such as the vertebrate brain, which has reached its greatest complexity in placental mammals. Although it might be tempting to view this increased complexity as progress, we must keep in mind that in only some societies is increased complexity equated with progress; in others an attainment of simplicity is progress. This strongly suggests that societal, value-based concepts such as progress are not applicable when examining questions in science—in this case evolution.


● ●

References Brown, J. (1995). Charles Darwin Voyaging, pp. 605. Knopf, New York. Darwin, C. (1859). On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life, pp. 502. John Murray, London. Dennett, C. C. (1995). Darwin’s Dangerous Idea, pp. 586. Simon & Schuster, New York. Desmond, A., and Moore, J. (1991). Darwin, pp. 808. Warner Books, New York. Futuyma, D. J. (1986). Evolutionary Biology, pp. 600. Sinauer, Sunderland, MA. Gould, S. J. (1996). Full House: The Spread of Excellence from Plato to Darwin, pp. 244. Harmony, New York. Minkoff, E. C. (1983). Evolutionary Biology, pp. 627. Addison-Wesley, Reading, MA. Rupke, N. A. (1994). Richard Owen Victorian Naturalist, pp. 462. Yale Univ. Press, New Haven, CT.


I. Extinction, Cretaceous J. DAVID ARCHIBALD San Diego State University San Diego, California, USA

The phenomenon that we call extinction is arguably one of the most misunderstood biological events. Three common misperceptions contribute to this misunderstanding. First, extinction is thought to be rare, or at least may have been rare before the explosion of the human population. Second, extinction is viewed as a negative process, one that only brings destruction. Third, the simple disappearance of a species is thought to be the same as extinction. Estimates of both the number of living (or extant) SPECIES and the rate at which they are being driven to extinction by humankind vary widely. Estimates for the numbers of species thought to grace the earth today range from a known number of 1.4 million to tens of millions (or more). Rates of present-day extinction are placed from as low as one per year to as high as one per day, with some even higher estimates. Within the constraints of the human timescale, extinction seems to occur at very low levels, even using the highest rates. So what if we lose one species per day if we have millions of species? If such levels continued, however, for only the extent of our Gregorian calendar—which is fast closing in on 2000 years—three-quarters of a million species would have disappeared. This would represent a loss of more than one-half of all known living species and would be very high even if the true number of extant species was closer to 10 million. How high a rate of extinction would this be compared to those in the geological past? It would rank within the top five, if not the top three, of the most significant episodes of extinction in the past 550 million years. The process of extinction, however, was occurring well before the emergence of Homo sapiens, although

we certainly have accelerated it to breakneck rates. Extinction has been ubiquitous during earth history since the origin of life approximately 3.5 billion years ago. All current estimates place the total percentage of species extinction throughout earth history at more than 90%, if not over 99%. This is a surprising figure for the uninitiated, but the surprise soon dissipates when we realize that evolution has constantly added new species throughout the past 3.5 billion years. Although there have been long intervals of increasing species diversity, life remained at a steady state for very long stretches of GEOLOGIC TIME. The only way a steady state of species numbers could be maintained, even while evolution is occurring, is for other species to become extinct. Thus, extinction, rather than being a rare and negative event in human time frames and sensibilities, is actually a very common and positive counterpoint to evolution. Without extinction, the vast majority of extant species, including Homo sapiens, could not and would not have arisen. I have addressed the first two common misperceptions concerning extinction; first, that extinction is thought to be rare, and second, that extinction is viewed as a negative process. However, what of the third—that the simple disappearance of a species is thought to be the same as extinction? The disappearance of species, usually based on disappearance from the fossil record, often occurs because of vagaries of the fossil record. It must also be remembered that many species are only poorly represented in the fossil record, if at all, because they lack preservable hard parts such as a shell or skeleton. If they live in regions that usually do not promote fossilization, such as in the mountains or desert, lack of a fossil record is not by itself a good measure of whether extinction has occurred. Even if we have a good fossil record for a particular species, and it disappears locally, it may continue to survive elsewhere on the globe for a considerable length of time. For example, plants such as dawn redwoods, fish such as coelacanths, and small mammals such as rat opossums (small South Ameri-


222 can marsupials) were known as fossils before living representatives were discovered growing, swimming, and scampering about elsewhere. Finally and probably most important for the disappearance of some species is the process of speciation. During speciation a parent species may give rise to a recognizably new species while the parent species remains relatively unchanged. In other instances the parent species disappears as it splits to form two or possibly even more new daughter species. This is pseudoextinction. Unlike true extinction, which occurs when all the individuals possessing a very similar but variable genome disappear, pseudoextinction occurs when the genomes of individuals in the two daughter species are sufficiently altered so that both are clearly set on new and different evolutionary paths. This most obviously has occurred when the resulting daughter species can no longer interbreed. If we can establish with reasonable certainty that the true extinction of one or more species has occurred, the next step is to place the extinction(s) in the context of other extinctions. Were the extinctions in question normal (or background) or were they part of a mass extinction? The major difference between normal and mass extinctions is one of scale. Because extinctions have been sampled repeatedly throughout the better known past 550 million years of earth’s history, it appears that the rate of extinction is roughly similar. Although it is difficult to provide a specific figure because of vagaries of the fossil record and differences between environments and different kinds of organisms, it can be safely said that normal extinction is below (often well below) 50%. There were intervals, however—five to be precise—during the past 550 million years that witnessed percentages of extinction well over 50%, in one case possibly reaching as high as 95%. This horrendously high level of extinction occurred at the end of the Permian Period approximately 250 million years ago. All five known mass extinctions, from oldest to youngest, were in (or at the end of) the Late Ordovician, Late Devonian, Late Permian, Late Triassic, and Late Cretaceous. There is an unresolved debate as to whether these five mass extinctions represent a separate class of extinction from normal extinctions or instead form a continuum with normal extinctions in both rate and cause. The most recent mass extinction event near the end of the Late Cretaceous includes that of the dino-

Extinction saurs, or more correctly the nonavian dinosaurs. The level of extinction for all species during the Cretaceous–Tertiary (or K–T) transition has been frequently given at approximately 75%, although there are no studies documenting this level of extinctions for species. For backboned animals or vertebrates only, the level of extinction hovers around 50% for species, but this is based on only one area, the Western Interior of North America. Dinosaurs have come to represent not only the K–T mass extinction event for both scientists and the public but also extinction in general. Except for perhaps the hapless Dodo, nothing seems to epitomize extinction as much as dinosaurs. When we are dealing with large, relatively rare creatures such as dinosaurs, the problems of unraveling when, where, how much (magnitude), and how fast (rate) extinction occurred can be disheartening. Nevertheless, only after these questions have been addressed, even if not completely answered, can we turn to the question of possible causes. The ‘‘common knowledge’’ is that dinosaurs became extinct at the same time everywhere on earth. If one is to believe not only the popular press but also some scientists, this common knowledge comes from a global record of dinosaur extinction at the K–T boundary. The idea that the dinosaurs disappeared from earth at the same time in the wink of an eye is not new; but the proposition in 1980 that an asteroid impact caused this very rapid decline and extinction of dinosaurs has given it new life. Some proponents of this theory have explicitly stated that these extinctions were essentially instantaneous around the world. Such explicit assertions about global records of dinosaurs are patently false. Most people are surprised to learn that the geographic coverage of dinosaur extinction is appallingly bad. The only region where we currently have a reasonably large sample of dinosaurs and contemporary vertebrates extending to near or at the K–T boundary (and have a fossil record of similar quality above this boundary) is in the Western Interior of North America, especially well known in the eastern part of Montana and into southern Canada. This region formed the eastern coast of a great inland sea that split North America in half in the Late Cretaceous. This was certainly an extensive region stretching for thousands of miles; nevertheless, it is a very limited region to use to explore questions of extinction on a global scale.

Extinction One study showed that of 26 dinosaur localities from near the end of the Cretaceous, 20 are from the Western Interior of North America. This represents more than 75% (20/26) of all the information we have about dinosaurs leading up to their extinction. The record of the last dinosaurs is biased not only in where the localities are found but also in the taxa and numbers of specimens. Of the 20 genera of dinosaurs known from the latest Cretaceous, 14 (70%) are known only from North America. Turning to numbers of specimens (those represented by articulated individuals composed of several bones up to a complete skeleton found in very close association), 95% of the 100 specimens known from the latest Cretaceous are from North America. These figures once more emphasize that the record of dinosaurs near the K–T boundary is almost exclusively North American. In the next few years we may see a better global record of latest Cretaceous dinosaurs emerging. Especially promising are new finds in several sedimentary basins in China and localities in central South America. Until such time that we do have a more global record, arguments about the pace of dinosaur extinction on a global scale remain unsubstantiated speculation. For now, it must be emphasized that we simply have no record of dinosaurs that permits us to clearly show whether dinosaur extinction was catastrophically fast or glacially slow. Rather, the data we do have are more regional in scope and only permit us to examine questions of the magnitude and selectivity of these extinctions, but nothing of its pace. Although the vertebrate record of the K–T boundary is almost exclusively limited to the Western Interior, the uppermost Cretaceous HELL CREEK FORMATION in eastern Montana has yielded a taxonomically rich sample of 107 vertebrate species. The record includes 12 major vertebrate lineages—5 species of sharks and relatives, 15 of bony fishes, 8 of frogs and salamanders, 10 of multituberculate mammals, 6 of placental mammals, 11 of marsupial mammals, 17 of turtles, 10 of lizards, 1 of the unfamiliar champsosaurs, 5 of crocodilians, 10 of ornithischian dinosaurs, and 9 of saurischian dinosaurs (not including birds). Of these 107 vertebrate species or species, 49% (52 of 107) survived across the K–T boundary in the Western Interior. This is the minimum percentage survival of vertebrate species across the K–T boundary because some of the very rare species may have survived undetected. Twenty of the 107 are quite rare

223 species, represented by fewer than 50 identifiable specimens out of 150,000 specimens estimated to have been recovered from the Hell Creek Formation. Although an accurate estimate is not possible, certainly some of these very rare species must have survived. The extreme and very improbable case would be if all 20 survived. This would be 67% (72 of 107) survival. This provides the extreme maximum percentage survival in the region. An educated guess, though no more than a guess, would be that no more than 60% of vertebrate species survived the K–T boundary. When examined in greater detail, a very interesting pattern emerges within the record of survival and extinction for these Hell Creek Formation vertebrates. This is a pattern of disparity in species survival among the 12 major groups. Only 5 of the groups— sharks and relatives, marsupials, lizards, ornithischians, and saurischians—contributed to 75% of the extinction. What do sharks, lizards, marsupials, and the two lineages of dinosaurs have in common in these faunas, other than that each suffered at least 70% or more species extinction at the K–T boundary in western North America? If we are to understand causes of extinction at the K–T boundary we must

Late Cretaceous dinosaur bearing sites from around the world are reasonably common. These nonmarine lacustrine sites, however, rarely cross the Cretaceous/Tertiary boundary. Very few of the North American sites are known to span this important boundary (䊊 多). Localities that might span the boundary (䊊 ? ) are still being studied. Map by J. David Archibald (1996), from Dinosaur Extinction and the End of an Era. Columbia University Press. Reprinted by permission.

224 explain this disparate pattern of extinctions. Any theories about the cause(s) of extinction of dinosaurs and their contemporary vertebrates must be able to explain the previously discussed pattern. Any theories must explain why sharks, lizards, marsupials, ornithischians, and saurischians suffered very high levels of or even total extinction, whereas bony fishes, frogs and salamanders, multituberculate mammals, placental mammals, turtles, champsosaurs, and crocodilians suffered 50% or often much less extinction. There are as many as 80 theories or variants of theories that have been proposed for dinosaur extinction. Many of these are either frivolous (Martians hunted the dinosaurs to extinction) or untestable (a plague spread through dinosaur populations). Three particularly important and testable hypotheses are the impact, volcanism, and marine regression theories. To these recently has been added another called the Pele theory that suggests that, among other things, the decrease in atmospheric oxygen would have had a detrimental affect on dinosaurs. We must await further information to test this last theory adequately. Only the impact and marine regression theories have been relatively thoroughly tested with the vertebrate fossil record by their respective proponents, although proponents of the volcanism theory suggest that many of the biotic responses to an impact would also be found with massive volcanism. One method to assess the efficacy of these theories is to examine them in the context of the known K–T vertebrate record, starting with the impact theory. The original scientific paper in 1980 advocating the impact theory still offers the basic mechanism of how such an impact might cause extinction among both animals and plants, including vertebrates. The impact would create a dust cloud enveloping the globe for a few months to a year. Darkness would shroud the world as long as the dust remained in the atmosphere. Photosynthesis in the sea and on land ceased. As the plants died or became dormant, herbivores soon starved, followed by the carnivores. Some of the best physical evidence of such an impact is the discovery of anomalously high levels of the rare earth element iridium at the K–T boundary and the probable remains of an impact crater. Although iridium is found on the earth, especially deep in the earth, high levels of iridium are associated with extraterrestrial events. Generally, the larger the impacting object, such as an asteroid, the greater the increase in signature elements such as iridium. The

Extinction remains of a probable impact structure dubbed ‘‘Chicxulub,’’ approximately 110 miles across at the northern tip of the Yucatan peninsula from at or near the K–T boundary, are strong evidence for an impact. Neither elevated levels of iridium nor an impact crater, however, are direct evidence for specific causes of extinction at the K–T boundary. An incorrect assumption often made in testing the impact theory and its possible corollaries is that all major taxa show very high levels of extinction across the ecological spectrum on a global scale. As already discussed, for many organisms, but most notably dinosaurs and their contemporary vertebrates, there is no such global record at the species level. The impactgenerated scenario of extremely high levels of catastrophic extinction across most environments is so broad spectrum and tries to explain so much that it is difficult to test. The burden of proof for sweeping, catastrophic extinction scenarios rests with the proposers of the theory. The various corollaries of the impact theory, such as a sudden cold snap, highly acidic rain, or global wildfires, are more easily tested using the known K–T vertebrate record. A short, sharp decrease in temperature was not emphasized in the originally proposed hypothesis, but it soon became an important corollary of the impact theory. It is argued that if tremendous amounts of dust were injected into the atmosphere after a large impact, the darkness would not only suppress photosynthesis but also produce extremely cold temperatures. This hypothesized condition has become known as ‘‘impact winter.’’ It is argued that following a large impact, ocean temperatures would decrease only a few degrees because of the huge heat capacity of the oceans, but on the continents, however, temperatures would be subfreezing from 45 day up to 6 months. The temperature would remain subfreezing for about twice the time of darkness caused by the dust. If a suddenly induced, prolonged interval of subfreezing occurred in subtropical and tropical regions today, which vertebrates in this climate, which is similar to that of the latest Cretaceous, would be most affected? In general, ectothermic tetrapods would suffer most. Ectotherms, as the name suggests, heat or cool themselves using the environment. Endotherms such as mammals and birds generate their heat through metabolic activity. In endotherms approximately 80% of food consumption goes toward thermoregulation (regulation of body temperature).

Extinction Fishes, which are by and large ectothermic, would be generally less affected by a severe temperature drop. Today, the northern limit of turtles and crocodilians is controlled by temperature. These animals cannot tolerate freezing, becoming sluggish or immobile at 10–15⬚C. Various amphibians and reptiles do inhabit areas with low winter temperatures or severe drought, but they have evolved methods of torpor (estivation and hibernation) to survive. These are the exceptions, however, because species diversity for ectothermic tetrapods is far higher in warmer climates. More important, we should not assume that Late Cretaceous ectothermic tetrapods living in subtropical to tropical climates such as in eastern Montana were capable of extended torpor. Torpor is most often preceded by decreases in ambient temperature, changes in light regimes, and decreases in food supply. The ectotherms in eastern Montana could not have anticipated a short, sharp decrease in temperature. This is true even if the impact had occurred during a Northern Hemisphere winter when temperatures would been slightly lower. We must remember that this was a subtropical to tropical setting, and thus the extended, subfreezing temperatures advocated by proponents of this corollary would have been devastating to ectotherms even during a terminal Cretaceous winter in Montana. Except for a 70% decline in lizards, ectothermic tetrapods (frogs, salamanders, turtles, champsosaurs, and crocodilians) did very well across the K–T boundary. The corollary of a sudden temperature decrease simply does not fit with the vertebrate data at the K–T boundary. A latest (but not terminal) Cretaceous vertebrate fauna from northern Alaska strengthens the evidence that a hypothesized sudden temperature drop was not a likely cause of K–T boundary extinctions. Comparing the Late Cretaceous vertebrate faunas from Alaska and eastern Montana reveals a striking difference. Although the Alaskan fauna is decidedly smaller and with fewer species than that from eastern Montana, both have sharks, bony fishes, dinosaurs, and mammals. The Alaskan fauna, however, completely lacks amphibians, turtles, lizards, champsosaurs, and crocodilians. These taxa comprise 41 of 107 (38%) of the eastern Montana fauna. If even the fairly balmy temperature range of 2–8⬚C for Late Cretaceous Alaska was enough to exclude ectothermic tetrapods, a severe temperature drop to below subfreezing temperatures at the K–T boundary should have devastated the rich

225 ectothermic tetrapod faunas at midlatitudes. These species flourished. A second prominent corollary of the impact theory is highly acidic rain. The most commonly cited acids as products of an impact are nitric and sulfuric acid. It is argued that nitric acid would be produced by the combination of atmospheric nitrogen and oxygen as a result of the tremendous energy released by an impact. Sulfuric acid would be produced because large amounts of sulfur dioxide are vaporized from rock at the impact site. These acids would be precipitated in the form of rain. Estimates of the pH of these acid rains vary, but estimates reach as low as 0.0–1.5! It is suggested that global effects could have caused the pH of near-surface marine and fresh water to below 3. In today’s environment, rain below a pH of 5.0 is considered unnaturally acidic. Rain as low as 2.4 has been recorded, but annual averages in areas affected by acid rain range from 3.8 to 4.4. Acid fogs and clouds from 2.1 to 2.2 pH have been recorded in southern California and have been known to bathe spruce–fir forests in North Carolina. The biological consequences of such low pH values vary from one vertebrate group to another but are always detrimental. Aquatic species (fish, amphibians, and some reptiles) are the first and most drastically affected, with those reproducing in water being the first to suffer. If pH becomes lower than approximately 3.0, adults often die. The effects on aquatic vertebrates across the K–T boundary would have been very bad if a pH of 3.0 was reached and truly horrendous if it hit 0.0 as suggested by some authors. Some advocates of K–T acid rain argue that the surrounding soils or bedrock would have buffered the aquatic systems, and they even suggest that limestone caves could have been important refugia for birds, mammals, amphibians, and small reptiles. The only problem with this scenario is that there were none of these kinds of buffering soils or bedrock or limestone caves in eastern Montana in the Latest Cretaceous. Based on what we know of our modern biota’s reaction to acid rain, aquatic animals should have been devastated by acid rain at the K–T boundary. Of all the aquatic species, only sharks and their relatives show a drastic drop in eastern Montana. Thus, the likelihood of low pH rain is highly implausible. A third corollary of the impact theory that receives various levels of support is global wildfire resulting from the aftermath of the impact. Soot and charcoal have been reported from several sites at the K–T

226 boundary coincident with the enrichment of iridium noted earlier. It was argued that this pattern is unique and must come from the extremely rapid burning of vegetation equivalent to half of all the modern forests! Other scenarios argue that approximately 25% of the aboveground biomass burned at the end of the Cretaceous! Such a global conflagration is really beyond our comprehension. In order to grasp the magnitude of this scenario, imagine one-quarter to one-half of all structures on the globe engulfed in flames within a matter of days or weeks. This still would be only a fraction of what is argued to have been burned at the K–T boundary. In such an apocalyptic global wildfire, much of the aboveground biomass all over the world would have been reduced to ashes. In fresh water, those plants and animals not boiled outright would have faced a rain of organic and inorganic matter unparalleled in human experience. These organisms would have literally choked on the debris or suffocated as oxygen was suddenly depleted with the tremendous influx of organic matter. The global wildfire scenario is so broad in its killing effects that it could not have been selective, but, as discussed earlier, the vertebrate pattern of extinction and survival is highly selective. Thus, it is no surprise that this scenario of equal opportunity losers does not show any significant agreement with the pattern of vertebrate extinction and survival at the K–T boundary. Not only is there almost no fossil evidence supporting global wildfire but also the physical basis for such an event is suspect. It is argued that there is a global charcoal and soot layer that coincides with the K–T boundary, whose emplacement is measured in months. This also assumes that the sedimentary layer encasing the charcoal and soot was also deposited in only months. This is demonstrably not the case for at least one K–T section that continues to be cited in these studies—the Fish Clay of the Stevns Klint section on the coast of Denmark. The Fish Clay is a laterally discontinuous, complexly layered and burrowed clay reflecting the conditions at the time of its deposition. It is not the result of less than a year of deposition caused by an impact-induced global wildfire. Thus, carbon near the K–T boundary at Stevns Klint as well as in other sections is likely the result of much longer term accumulation during normal sedimentation. When all of the corollaries of an impact of an asteroid or comet are compared to the pattern of extinction

Extinction and survival for vertebrates at the K–T boundary in eastern Montana, there is relatively poor agreement. Without special pleading, these corollaries as currently proposed are unlikely causes of vertebrate extinction. This does not mean that all corollaries of an impact should be rejected, but it is imperative that those proposing the different corollaries separate those that are supported by the vertebrate fossil record from those that are not. The next hypothesis is the volcanism theory. Although some proponents of the impact theory do not agree, many advocates of both theories feel that a number of the same physical events would have occurred at the K–T boundary if either extensive volcanism or an impact took place. They also say that the biological results would be similar. Given the previous discussion of how most of the corollaries of the impact theory do not test well against the vertebrate fossil record, the volcanism and impact theories are equally weak in their biological predictions. The major difference in these two theories is in their timing. Whereas the impact theory measures most of the cataclysmic effects in months or years, with physical effects possibly lingering for a few hundred or a few thousand years, the volcanism theory measures effects into the millions of years. The effects of many volcanic eruptions, such as that of Mt. St. Helens, linger for a few months or a few years. Many other episodes of volcanism are very prolonged. These are flood basalt eruptions. The best known in the United States is the 16-million-year-old Columbia River flows in the northwest. In the past 250 million years, arguably one of the biggest flood basalt eruptions occurred on the Indian subcontinent. This was occurring during (and is probably related to) the collision of the subcontinent with the remainder of Asia. Its most obvious manifestation today is the tallest mountain range in the world, the Himalayas. These flood basalts, known as the Deccan Traps, cover an immense part of both India and Pakistan. Individual flows in the sequence cover almost 4000 square miles with a volume exceeding 2400 cubic miles. Individual flows average 30–160 ft thick, sometimes reaching 500 ft. In western India the accumulations of lava flows is 7800 ft, or 1.5 miles thick. The flows originally may have covered almost 800,000 square miles with a volume possibly exceeding 350,000 cubic miles (an area the size of Alaska and Texas combined, save about 30,000 square miles, to a depth of more than 2000 ft). Based on radiometric dating, paleomagnet-

Extinction ics, and vertebrate fossils, the bulk of the eruptions are centered around the K–T boundary, during a reversal in earth’s magnetic poles known as 29R or 29 reversed. The number 29 represents the 29th reversal of the earth’s magnetic field counting backwards from the present, which today has normal polarity by definition. The K–T boundary happens to fall in 29R. What would the effect have been on the global biota if something the magnitude of the Deccan Trap erupted for tens of thousands of years or longer? One of the greatest effects would have been to increase and maintain a much higher level of particulate matter in the atmosphere. Whether it would have caused warming through a greenhouse effect, cooling because of less light, or simply prettier sunsets is not certain. The amount of CO2 pumped into the atmosphere by the eruptions may have been a boon for green plants that require CO2 for photosynthesis, but a reduction in light reaching the surface because of particulate matter may have canceled the effects of increased CO2 . The effects of added particulate matter might have prevailed for no other reason than that they would linger longer after eruptions ceased, whereas the release of CO2 would have diminished much more rapidly after each eruption stopped. If the latter scenario is correct, the longer term effects over a million years or more would be to push a global cooling. Most estimates suggest that regional if not the global climate cooled through the K–T transition. Because the time frame is moderately long, many species, especially smaller ones, on land or in the sea could adapt to changes, whereas larger species such as dinosaurs may not have been as fortunate. Although it probably was not a cause of extinction for most species, the cooling across the K–T boundary would have been an added stress. A final long-term effect suggested for eruption of the Deccan Traps is reduced hatching success for eggs of herbivorous dinosaurs. Volcanic activity can release elements such as selenium that are highly toxic to developing embryos. Increased levels of selenium in the eggshells of dinosaurs are known from near the K–T boundary in southern France. Poisoning of eggs has also been reported from dinosaur eggs near the K–T boundary in Nanxiong Basin, southeastern China. The final hypothesis that has been tested with the vertebrate fossil record is the marine regression/habitat fragmentation theory, or simply marine regression theory. Many areas of the terrestrial realm were

227 repeatedly inundated by shallow epicontinental seas throughout geologic history. The term ‘‘epicontinental’’ refers to the occurrence of these very shallow seas upon the continental shelves and platforms rather than in deep ocean basins. Epicontinental seas reached depths of only 1500–2000 ft, very shallow compared to most large modern marine bodies. Epicontinental seas are almost nonexistent today, except for such bodies of water as Hudson Bay. It is known that during the Late Cretaceous, large areas of continents were submerged under warm, shallow epicontinental seas. It became clear only recently just how dramatic the loss of these seas was leading up to the K–T boundary. There is absolutely no mistaking that the K–T loss of shallow seas (or increase in nonmarine area) is greater than at any time in the past 250 million years. The nonmarine area increased from 42 million m2 to 53 million m2 —more than a 25% increase. This is the equivalent of adding the land area of all of Africa, the second largest continent today. The second largest increase in continental area in the past 250 million years occurred across the Triassic–Jurassic boundary. Like the K–T transition, this is also during one of the five universally recognized mass extinctions during the Phanerozoic or last 550 million years. Some of the most dramatic additions of nonmarine areas at or near the K–T boundary occurred in North America. Near the end of the Cretaceous, maximum transgression divided North America into two continents. As regression continued until at or near the K–T boundary, coastal plains decreased in size and became fragmented; stream systems multiplied and lengthened; and as sea level fell, land connections were established or reestablished. The driving force for these repeated inundations or transgressions of the lower-lying portions of continents is still not fully understood. The general consensus is that it is related to plate tectonics. It is thought that rises in sea level and inundations began as the motion of the plates increased. As this occurs, the margins along which the colliding plates converge are subducted or pushed downwards into the earth. This causes inundation by the seas upon shallow continental shelves and platforms. Whatever the geophysical factors driving the process, the physical manifestations of marine regression, like impacts and volcanism, are important ultimate causes of extinction. Although these processes of marine transgression and regression were global

228 in extent, a closer examination of North America is best because, as I have emphasized, this is where we have the vertebrate data at the K–T boundary. North America was split into two continents—a western continent (Laramidia) and an eastern continent (Appalachia)—by the Pierre Seaway for almost 40 million years during the Late Cretaceous. Most of our latest Cretaceous vertebrate fossils come from the east coast of Laramidia. The west coast of Appalachia as well as the eastern seaboard of Appalachia have also produced some specimens. In the last few million years of the Cretaceous the Pierre Seaway began to regress from both Laramidia and Appalachia. At or just shortly before the K–T boundary, the seaway reached its nadir. Placement of the receding coastlines both north and south have not been well established, but we know the southern coastline reached well into Texas. There is no question that there was a dramatic reduction in coastal plains. This is exactly the kind of environment from which we are sampling the last of the Late Cretaceous vertebrate community. A common refrain is that because the total amount of land increased with the regression, dinosaurs should have had more, not less, area and more environments in which to live. We know with considerable certainty that dinosaurs did live in other environments such as the higher, drier Gobi Desert in Mongolia during part of the Late Cretaceous. Currently, however, the only well-known vertebrate communities that preserve dinosaurs at the K–T boundary are coastal. Thus, arguments about what dinosaurs and other vertebrates may or may not have done in other environments are moot. It is simply incorrect to say that the dinosaurs and other vertebrates may have survived elsewhere when we have little or no information about other environments. The drastic reduction of coastal plains put tremendous pressure on some, especially large vertebrate species. Reduction of habitat, for example, in the Rift Valley system of East Africa, today first affects larger vertebrates, especially mammals. In the shrinking coastal plains of latest Cretaceous North America, the equivalent large vertebrates first affected were the dinosaurs. An additional problem, whether in East Africa today or the coastal plains of latest Cretaceous North America, is the fragmenting of the remaining habitat. This process, as a result of human activity, has become known as habitat fragmentation. In larger, undisturbed habitats, animals (and plants) can spread

Extinction more freely from one area to another. If the habitat is fragmented, although the amount of habitat may not have been greatly altered, it will reduce the flow of species from one fragment to another. For some species, even seemingly small barriers such as two-lane roads can be insurmountable. The results can be disastrous if viable populations cannot be maintained in the various fragments. Fragmentation can lead to extinctions. Barriers also arise in nature even among animals that would seem easily capable of dispersing. Although no doubt the result is very often extinction, we usually only see what survives in the form of differences between closely related species. Small arboreal primates in the rainforests of both South America and Africa form small fragmented groups that are isolated from each other often by rivers only tens of yards wide. Another example is the Kaibab squirrel on the North Rim of the Grand Canyon. Unlike its nearest relative, Abert’s squirrel, which is found on the south side of the Grand Canyon and in the western United States and Mexico, the Kaibab squirrel is restricted to an area of only 20 ⫻ 40 miles. Fragmentation, in this case the development of the Grand Canyon, helped produce the differences, but the margin between this and oblivion for the Kaibab squirrel is slim. The idea of habitat fragmentation not only extends to natural processes operating today but also to processes operating in the geological past. Although historical habitat fragmentation is not well understood by earth scientists, it is an all too real phenomenon among biologists studying the effects of human activity in modern rainforests and in urban settings. Declines of bird and mammal populations have been well documented in the city of San Diego as urban development divides and isolates habits in canyon areas. One would not expect that the natural equivalent of habitat fragmentation would be easily, if at all, preserved in the rock. The forcing factor for habitat fragmentation in the latest Cretaceous—marine regression—is a thoroughly documented fact during the waning years of the Late Cretaceous in North America. Globally, marine regression occurred within this same general time frame although how close in time it occurred in various regions is a matter of debate. Theory predicts that large species would be the most severely affected by habitat fragmentation for the reasons discussed previously. During the K–T transition in eastern Montana, only 8 of 30 large spe-

Extinction cies survived and these are partially or entirely aquatic (2 fishes, 1 turtle, 1 champsosaur, and 4 crocodilians), whereas all 22 large terrestrial species (and 1 aquatic species) became extinct (1 turtle, 1 lizard, 1 crocodilian, and 19 dinosaurs). Thus, predictions from habitat fragmentation fit the observed data very well. As noted previously, two other major physical events occur with marine regression in addition to habitat fragmentation—stream systems multiply and lengthen and, as sea level falls, land connections are established or reestablished. Only a few of the K–T boundary stream systems have been studied in detail in the Western Interior, and thus we do not know the exact drainage patterns for most latest Cretaceous and early Tertiary stream systems in the eastern part of Laramidia. Nevertheless, we are certain that as new land was added following marine regression in the early Tertiary, stream systems increased and lengthened. This process is another major corollary of marine regression. When freshwater habitats were bolstered following marine regression, most aquatic vertebrates did well, except those with close marine ties—sharks and some bony fishes. Such fishes may need to spend at least a portion of their life in a marine environment, in some instances to reproduce. The major group most likely to suffer would have been the sharks and their relatives. In fact, all five species of sharks disappeared. It is not clear, however, whether these disappearances from the Western Interior are actually extinctions at the K–T boundary or whether the shark species survived elsewhere in marine environments into the earliest Paleocene. The problem is that the definitively oldest marine sediments that postdate the K–T boundary in the Western Interior are no older than late early Paleocene in age. This means that there is a gap in marine sedimentation in the Western Interior of possibly 1 million years or more immediately after the K–T boundary. This pattern of disappearance and reappearance strongly suggests that as the Pierre Seaway regressed further and further away from eastern Montana, all sharks and relatives departed because connections to the sea became attenuated. New species of elasmobranchs did not occur in the area until a smaller transgression reached the Western Interior at or just before middle Paleocene times. This is known as the Cannonball Sea, which was a smaller seaway than the Pierre. The total disappearance of sharks and relatives is the only prediction that can

229 be made with any certainty as a result of the loss of marine connections and the lengthening of stream systems. The increase of stream systems was a positive factor helping to mitigate other stresses that may have been put on the freshwater system. New land areas were exposed as sea level lowered. In some cases this included the establishment or reestablishment of intercontinental connections. One such connection was the Bering land bridge joining western North America and eastern Asia. At various times during the Late Cretaceous this bridge appeared and then disappeared. This is suggested by similarities in parts of the Late Cretaceous vertebrate faunas in Asia and North America, especially the better studied turtles, dinosaurs, and mammals. Competition and extinction often result from biotic mixing, but predicting the fates of various taxonomic groups is usually not possible. An exception may have been the fate of marsupials in North America near the K–T boundary. The oldest marsupials are known from approximately 100-million-year-old sites in western North America. By some 85 Mya, we know of about 10 species of marsupial. This rose and stayed at about 15 species from approximately 75 Mya until the K–T boundary approximately 65 Mya, when it plummeted to one species. These were all quite small mammals, from the size of a mouse up to a very well-fed opossum or raccoon. Their teeth were very much like those of extant opossums, with slicing crests and welldeveloped but relatively low cusps (compared to contemporary placental mammals) for poking holes in insect carapaces, seeds, or whatever they found. Most did not appear to be specialists on any particular food. With the reestablishment of the Bering land bridge (or at least closer islands) near the K–T boundary, a new wave of placental mammals appeared in western North America. These mammals (traditionally known as condylarths) were the very early relatives of modern ungulates and whales. Their appearance in North America coincides with the very rapid decline of marsupials near the K–T boundary. Within a million years of the K–T boundary, 30 species of these archaic ungulates are known in North America, and their numbers kept on rising. Our best guess now is that the lineage that gave rise to these mammals first appeared in middle Asia between approximately 80 and 85 million years ago and reached North America near the K–T boundary. What is of interest is that the archaic ungulate invaders had dentitions very similar to contemporary marsupials and pre-


230 sumably ate similar things. Its seems more than coincidence that marsupials did well in North America for approximately 20 million years only to almost disappear with the appearance of the ungulate clade. It is ironic that both marsupials and ungulates were joint invaders of South America very soon after the K–T boundary. Their dentitions were already beginning to show differentiation, with the marsupials headed toward carnivory and ungulates headed toward herbivory. It shows what a little cooperation can do. These various physical events accompanying marine regression fit very well the pattern of extinction and survival described previously for the 107 vertebrate species from near the K–T boundary in eastern Montana. In fact, the patterns of extinction and survival for 11 of 12 of the major vertebrate groups agree very well with predictions from the marine regression theory discussed previously. Marine regression, extraterrestrial impact, and massive volcanism are all major environmental events that occurred near or at K–T boundary. None of these physical events appears sufficient by itself to be crowned the sole cause of extinctions at the K–T boundary. The evidence outlined here overwhelmingly supports this view. What is less certain is whether all three were necessary for the pattern of extinctions we see at the K–T boundary. No one knows for sure. Both marine regression and an impact were apparently necessary to give us the pattern of turnover at the K–T boundary for all species, not just the vertebrates discussed in this essay. The role of volcanism or the effects of the Pele hypothesis are less certain. The effects of massive volcanism are formidable, but the purported biological effects have not been as closely explored as those of the other two events. The Pele hypothesis simply has not yet been properly studied and tested. What is very clear is that at least three physical events—marine regression, extraterrestrial impact, and massive volcanism—did coincide near the K–T boundary, but this does not mean that all three events necessarily occurred simultaneously at this boundary. The greatest addition of nonmarine area in the past 250 million years brackets the Late Cretaceous mass extinctions. One of the largest known impact craters is thought to have been identified at very near the K–T boundary. Massive volcanism had been pouring out great quantities of lava sporadically for several million years during the K–T interval. This

was clearly one of the most geologically complicated and biologically challenging episodes in earth history. Given these challenges it is surprising that more species did not succumb. It suggests that life has been more resilient than we dreamed, or have a right to hope for, given the stresses we are placing it under today.

See also the following related entries: CRETACEOUS PERIOD ● EVOLUTION ● EXTINCTION, TRIASSIC

References Archibald, J. D. (1996). Dinosaur Extinction and the End of an Era: What the Fossils Say, pp. 237. Columbia Univ. Press, New York. Fastovsky, D. E., and Weishampel, D. B. (1996). The Evolution and Extinction of the Dinosaurs, pp. 460. Cambridge Univ. Press, Cambridge, UK. MacLeod, N., and Keller, G. (Eds.) (1996). The Cretaceous– Tertiary Mass Extinction: Biotic and Environmental Changes, pp. 575. Norton, New York. Raup, D. M. (1991). Extinction: Bad Genes or Bad Luck, pp. 210. Norton, New York.

II. Extinction, Triassic MICHAEL J. BENTON University of Bristol Bristol, United Kingdom

The oldest dinosaurs date from the Carnian Stage (see HERRERASAURIDAE; ORIGIN OF DINOSAURS), and the group split into recognizable theropods, sauropodomorphs, and ornithischians during that time interval. However, dinosaurs were minor elements of the Carnian faunas, and they were generally modest sized. Dinosaurs became abundant, and they diversified further, during the subsequent Norian Stage.

Competition or Mass Extinction? Until 1980, the origin of the dinosaurs was explained generally by a model of long-term competitive replacement. Inadequacies of stratigraphy, and of understanding of faunal compositions, suggested that dinosaurs had arisen some time during the Middle Triassic. The evidence for this consisted of a scattering of bony specimens from Europe and the Americas

Extinction and of footprints largely from Europe. Some of the footprints were from Early Triassic sediments, and the picture of a long slow rise of the Dinosauria throughout the first half of the Triassic was widely accepted. Competition Model The competitive model for Triassic faunal replacements (Bonaparte, 1982; Charig, 1984) proposed that the Early Triassic faunas, dominated by therapsid synapsids, gave way to Middle Triassic faunas dominated by rhynchosaurs and basal archosaurs, which were in turn replaced by Late Triassic dinosaur faunas. The evolutionary explanation for the successive replacements was that each group competed with its predecessors and rose to ascendancy because of some superior adaptations. The strongest argument for those superior adaptations was the change in archosaurian posture during the Triassic from the typical reptilian sprawling pattern to a fully erect posture in dinosaurs (Charig, 1972). Mass Extinction Model The alternative mass extinction model (Benton, 1983, 1986a,b, 1994b) states that none of this evolutionary relay took place, but that the dinosaurs radiated into empty ecologic space after a mass extinction event had wiped out preexisting groups. The particular proposal is that dinosaur faunas in fact showed relatively little change until the second half of the Carnian Stage, when the rhynchosaurs and dicynodonts, as well as several groups of amphibians, basal archosaurs, and smaller reptiles, disappeared at, geologically speaking, the same time. The dinosaurs radiated only after the slate had been wiped clean of other moderate to large-sized herbivores.

Triassic Faunal Evolution Pre-Carnian ‘‘Dinosaurs’’ Careful study of the supposed Early and Middle Triassic dinosaur remains showed that the bones were either definitely nondinosaurian or they were inadequate for identification (Benton, 1986b). Many so-called ‘‘dinosaur’’ remains from pre-Carnian rocks in Germany—named variously Teratosaurus or Zanclodon, for example—are teeth and vertebrae identifiable only as archosaurian. In addition, supposed dinosaur footprints from the Early and Middle Triassic of England turned out to be inorganic structures or clearly nondinosaurian (King and Benton, 1995), and the Middle Triassic material from France is equivocal.

231 Stratigraphy New studies of the stratigraphy of the Late Triassic and Early Jurassic dinosaur-bearing continental sediments (Olsen and Galton, 1977, 1984; Olsen and Sues, 1986; Hunt and Lucas, 1991; Benton, 1994a,b) have clarified the sequence of faunas (Fig. 1). The revisions have been based on new biostratigraphic evidence from palynomorphs and freshwater fishes, on rare ties with marine sequences dated by ammonoids, by sporadic magnetostratigraphic and radiometric tie points, and by comparisons of tetrapod faunas. The details of dating should improve considerably in the future as a result of further study of palynological evidence and of new data from magnetostratigraphy and scattered radiometric dates. Continental Tetrapod Evolution The faunal succession during the Triassic (Fig. 2) suggests, for Gondwanaland and Laurasia, that there was a substantial mass extinction event during the second half of the Carnian Stage, perhaps 225 Ma ago at or near the Carnian–Norian (Crn–Nor) boundary. The oldest Triassic faunas, from several parts of the world, were dominated by the dicynodont Lystrosaurus, which represented more than 90% of all animals found. This astonishing monoculture, with parallels today only in agricultural situations, was probably the result of the vast end-Permian extinction event, which had cleared the typical Late Permian faunas of the dominant groups, such as dinocephalians, gorgonopsians, and pareiasaurs (see REPTILES). During the Middle Triassic, faunas north and south became dominated by dicynodonts and/or rhynchosaurs as herbivores and cynodonts and basal archosaurs as carnivores. In no case did these archosaurs exceed 10% of the number of individuals in a fauna. This low representation was counter to the expectations of the classic competitive model for the rise of the basal archosaurs, and then for their progressive replacement by the dinosaurs. Most Carnian continental tetrapod faunas were comparable in composition. Substantial Middle to Late Carnian tetrapod assemblages are known from Argentina, Brazil, the United States (New Mexico, Arizona, and Texas), Britain, Morocco, and India (Fig. 1). Typically, these faunas were still dominated by rhynchosaurs and/or dicynodonts as bulky herbivores, with basal archosaurs and/or cynodonts as carnivores. Many of them contain one or two specimens of modest-sized dinosaurs. Tetrapod faunas of early Norian age from the



FIGURE 1 Stratigraphy of major Late Triassic and Early Jurassic continental tetrapod-bearing formations in Gondwanaland and Europe (A) and in North America (B). Abbreviations: CB, Cow Branch Formation; L, Lockatong Formation; M, McCoy Brook Formation; NH, New Haven Arkose; NO, New Oxford Formation; P, Passaic Formation; PR, Pekin Formation; PT, Portland Formation; T, Turkey Branch Formation; W, Wolfville Formation. (Based on various sources and summarized in Benton, 1994a.)


233 bers of Saurischia–Ornithischia) and the ornithischian Pisanosaurus mertii (Bonaparte, 1976; Sereno and Novas, 1992, 1994; Sereno et al., 1993; Novas, 1994; Sereno, 1994) (see HERRERASAURIDAE; PHYLOGENY OF DINOSAURS). The dinosaurs were rare, however, representing only 5.7% of the fauna (13 of 228 specimens, according to Rogers et al., 1993). The Ischigualasto fauna was dominated by the rhynchosaur Scaphonyx, a pig-sized grubbing herbivore, and Exaeretodon, a medium-sized herbivorous cynodont. Other rarer elements include the bulky herbivorous dicynodont Ischigualastia, the basal archosaurs Aetosauroides (a herbivore), Proterochampsa (a piscivore), and Saurosuchus (a carnivore), as well as the dinosaurs (Fig. 3). Rogers et al. (1993) established a radiometric date of 228 Mya for the lower part of the Ischigualasto Formation, which, they argue, places it in the middle of the Carnian Stage. Small dinosaur faunas are known elsewhere in

FIGURE 2 Relative abundances of different tetrapod groups through the Mesozoic. The patterns for carnivores (top) and herbivores (bottom) show evidence for a catastrophic loss of medium-sized to large animals in the late Carnian (225 Ma) and their replacement by dinosaurs. (Based on data in Benton, 1983.)

southwestern United States, Germany, and southern continents are quite different. Dinosaurs dominate some assemblages in abundance (consider the famous Ghost Ranch assemblage of Coelophysis in the upper Petrified Forest Member of the Chinle Formation and the rich assemblages of Sellosaurus and Plateosaurus from the German Stubensandstein) and to some extent in terms of diversity and size range. However, although some Norian assemblages are dominated by dinosaurs (25–60% of specimens), dinosaurs were not really diverse, with at most four or five genera at any time in the Stubensandstein. Sizes of Triassic prosauropods ranged up to 8 m for Plateosaurus and 10 m for Riojasaurus.

The Oldest Dinosaurs The best known early fauna of dinosaurs comes from the Ischigualasto Formation of San Juan Province, Argentina. It consists of Eoraptor lunensis and Herrerasaurus ischigualastensis (variously regarded as basal theropods, basal saurischians, basal dinosaurs, or the closest relatives of dinosaurs though not to be mem-

FIGURE 3 Range chart of reptiles through the thickness of the Ischigualasto Formation, Argentina, showing relative abundances, based on a collection of 228 specimens. Dinosaurs (Eoraptor and Herrerasaurus) occur rarely in the lower part of the formation. The most common animal is the rhynchosaur Scaphonyx, followed by the herbivorous cynodont synapsid Exaeretodon. The dicynodont Ischigualastia and the thecodontians Saurosuchus, Proterochampsa, and Aetosauroides are rare. The Herr Toba bentonite is dated 228 Ma. (Based on data in Rogers et al., 1993.)



FIGURE 4 Major events in the Late Triassic, showing phases of mass extinction (䊉) and turnover (i.e., high extinction and origination rates; 䊊). Climatic changes and the Manicouagan impact are indicated. (Based on data in Simms and Ruffell, 1990, and Benton, 1991.)

rocks of middle to late Carnian age: in the Santa Maria Formation of Brazil (Staurikosaurus, a herrerasaur), the Maleri Formation of India (Alwalkeria), the lower part of the Petrified Forest Member, Arizona (Coelophysis), the Argana Formation of Morocco (Azendohsaurus), and perhaps the Lossiemouth Sandstone Formation of Scotland (Saltopus). In these faunas, dinosaurs represent 1–6% of specimens found. Although specimens are rare, the three main lines of dinosaurian evolution—the theropods, sauropodomorphs, and ornithischians—were already laid down during Carnian times and they had apparently diverged from a single common ancestor within only a few million years. The Carnian dinosaurs in all lineages were moderately sized animals, all lightweight bipeds less than 6 m long.

Late Triassic Mass Extinctions During the Late Triassic, 20% or more of families of marine animals died out, scaling to some 50% of species, and this matches the severity of the K–T event (Sepkoski, 1990). Controversy revolves around

three issues: (i) whether there was a single extinction event in the Late Triassic or more than one, (ii) whether the event(s) was catastrophic and caused by a major extraterrestrial impact, and (iii) whether the extinction(s) had anything to do with the rise of the dinosaurs. The Triassic–Jurassic Boundary Event There is no question that there was a mass extinction at the end of the Triassic Period (Fig. 4), at the Triassic–Jurassic (Tr–J) boundary, 202 million years ago, when ammonoids and bivalves were decimated and when the conodonts finally disappeared (Sepkoski, 1990). On land too, several families of reptiles disappeared, particularly the last of the basal archosaurs and some nonmammalian synapsids (see REPTILES). Recent work on earliest Jurassic vertebrate faunas has led to the claim (Olsen et al., 1987, 1991; Hallam, 1990) that this Tr–J event was instrumental in triggering the radiation and huge success of the dinosaurs. In addition, a major impact crater site, the Manicouagan structure in Quebec, was identified as the

Extinction smoking gun for a catastrophic extraterrestrial impact at the Tr–J boundary (Olsen et al., 1987, 1991). Elevated levels of iridium were reported from a Tr–J boundary section in Austria (Badjukov et al., 1987), and shocked quartz has been found at a Tr–J boundary section in Italy (Bice et al., 1992). High levels of iridium in K–T boundary clays worldwide, and shocked quartz in many such sections, are, of course, taken as key evidence for a major impact (see EXTINCTION, CRETACEOUS). The case for an impact at the Tr–J boundary is far from certain. Hallam (1990) failed to find the iridium anomaly in the Austrian section, and the nature of the lamellae in the shocked quartz was not adequate to rule out other explanations, such as a volcanic source for the material (Bice et al., 1992). Furthermore, the Manicouagan impact structure was redated (Hodych and Dunning, 1992) away from the Tr–J boundary (202 Mya), well down in the Late Triassic, about 220 Mya. This redating means that the impact happened at a time when no mass extinctions were taking place. The Crn–Nor Event A commonly expressed view (Olsen et al., 1987, 1991; Hallam, 1990) has been that the postulated Crn–Nor extinction event (225 Mya) was restricted to nonmarine vertebrates and was a minor blip in the diversification of life compared to the Tr–J event. Recent studies of marine fossil records (Sepkoski, 1990; Simms and Ruffell, 1990) have indicated, however, that the foraminifera, ammonoids, bivalves, bryozoans, conodonts, reef corals, echinoids, and crinoids all showed global-scale extinctions either during the Carnian or at the Crn–Nor boundary (Fig. 4). Quantitative assessments of data on nonmarine tetrapods (Benton, 1986a,b, 1991, 1994a,b) have shown that the rate of loss of families was as great during the late Carnian interval as it was around the Tr–J boundary. Causes Causes of the two Late Triassic extinction events are hard to determine. Neither corresponds to the Manicouagan crater, and evidence for impact is nonexistent for the Crn–Nor event and limited to some shocked quartz at the Tr–J boundary in Italy. Earthbound causes have been posited for both events. Simms and Ruffell (1990) argued for a major climatic change across the tropical belt, from humid climates in the middle to late Carnian to arid climates in the Norian (Fig. 4). These might have caused major

235 stresses for land plants and herbivorous tetrapods and led to the widespread extinction of previously hugely abundant dicynodonts and rhynchosaurs. Extinctions in the sea may have been caused by increased rainwater runoff into the sea and spread of fresh waters in previously marine basins. The Tr–J event has been explained by a phase of oceanic anoxia in earliest Jurassic times (Hallam, 1990), an environmental crisis that would have affected mainly marine organisms. Effects on land animals would have been less severe, as seems to have been the case. The current state of knowledge about mass extinctions is astonishingly limited, and yet it is much advanced over the information available 20 years ago. In comparison to the K–T event that may or may not have finished the dinosaurs (see EXTINCTION, CRETACEOUS), very little effort has been devoted to gaining an understanding of Late Triassic extinction events that may have kick-started the radiation of the clade Dinosauria.


References Badjukov, D. D., Lobitzer, H., and Nazarov, M. A. (1987). Quartz grains with planar features in the Triassic– Jurassic boundary sediments from northern Limestone Alps, Austria. Lunar Planet. Sci. Lett. 18, 38. Benton, M. J. (1983). Dinosaur success in the Triassic: A noncompetitive ecological model. Q. Rev. Biol. 58, 29–55. Benton, M. J. (1986a). More than one event in the Late Triassic mass extinction. Nature (London) 321, 857–861. Benton, M. J. (1986b). The Late Triassic tetrapod extinction events. In The Beginning of the Age of Dinosaurs: Faunal Change across the Triassic–Jurassic Boundary (K. Padian, Ed.), pp. 303–320. Cambridge Univ. Press, Cambridge, UK. Benton, M. J. (1991). What really happened in the Late Triassic? Historical Biol. 5, 263–278. Benton, M. J. (1994a). Late Triassic terrestrial vertebrate extinctions: Stratigraphic aspects and the record of the Germanic Basin. Paleontol, Lombarda Nuova Ser. 2, 19–38. Benton, M. J. (1994b). Late Triassic to Middle Jurassic extinctions among continental tetrapods: Testing the pattern. In In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods (N. C. Fraser and H.-D. Sues, Eds.), pp. 366–397. Cambridge Univ. Press, Cambridge, UK.


236 Bice, D. M., Newton, C. R., McCauley, S., Reiners, P. W., and McRoberts, C. A. (1992). Shocked quartz at the Triassic–Jurassic boundary in Italy. Science 255, 443–446. Bonaparte, J. F. (1976). Pisanosaurus mertii Casamiquela and the origin of the Ornithischia. J. Paleontol. 50, 808–820. Bonaparte, J. F. (1982). Faunal replacement in the Triassic of South America. J. Vertebr. Paleontol. 21, 362–371. Charig, A. J. (1972). The evolution of the archosaur pelvis and hind-limb: An explanation in functional terms. In Studies in Vertebrate Evolution (K. A. Joysey and T. S. Kemp, Ed.), pp. 121–155. Oliver & Boyd, Edinburgh, UK. Charig, A. J. (1984). Competition between therapsids and archosaurs during the Triassic period: A review and synthesis of current theories. Symp. Zool. Soc. London 52, 597–628. Hallam, A. (1990). The end-Triassic mass extinction event. Geol. Soc. Am. Spec. Paper 247, 577–583. Hodych, J. P., and Dunning, G. R. (1992). Did the Manicouagan impact trigger end-of-Triassic mass extinction? Geology 20, 51–54. Hunt, A. P., and Lucas, S. G. (1991). The Paleorhinus Biochron and the correlation of the nonmarine Upper Triassic of Pangaea. Palaeontology 34, 487–501. King, M. J., and Benton, M. J. (1995). Dinosaurs in the Early and Mid Triassic, fact or fiction? The footprint evidence from Britain. Submitted for publication. Novas, F. E. (1994). New information on the systematics and postcranial skeleton of Herrerasaurus ischigualastensis (Theropoda: Herrerasauridae) from the Ischigualasto Formation (Upper Triassic) of Argentina. J. Vertebr. Paleontol. 13, 400–423.

patterns of the Triassic–Jurassic tetrapod transition. In The Beginning of the Age of Dinosaurs (K. Padian, Ed.), pp. 321–351. Cambridge Univ. Press, Cambridge, UK. Olsen, P. E., Shubin, N. H., and Anders, M. H. (1987). New Early Jurassic tetrapod assemblages constrain Triassic–Jurassic tetrapod extinction event. Science (New York) 237, 1025–1029. Olsen, P. E., Fowell, S. J., and Cornet, B. (1991). The Triassic/Jurassic boundary in continental rocks of eastern North America; A progress report. Geol. Soc. Am. Spec. Paper 247, 585–593. Rogers, R. R., Swisher, C. C., III, Sereno, P. C., Forster, C. A., and Monetta, A. M. (1993). The Ischigualasto tetrapod assemblage (Late Triassic) and 40Ar/ 39Ar calibration of dinosaur origins. Science 260, 794–797. Sepkoski, J. J., Jr. (1990). The taxonomic structure of periodic extinction. Geol. Soc. Am. Spec. Paper 247, 33–44. Sereno, P. C. (1994). The pectoral girdle and forelimb of the basal theropod Herrerasaurus ischigualastensis. J. Vertebr. Paleontol. 13, 425–450. Sereno, P. C., and Novas, F. E. (1992). The complete skull and skeleton of an early dinosaur. Science 258, 1137– 1140. Sereno, P. C., and Novas, F. E. (1994). The skull and neck of the basal theropod Herrerasaurus ischigualastensis. J. Vertebr. Paleontol. 13, 451–476. Sereno, P. C., Forster, C. A., Rogers, R. R., and Monetta, A. M. (1993). Primitive dinosaur skeleton from Argentina and the early evolution of Dinosauria. Nature 361, 64–66. Simms, M. J., and Ruffell, A. H. (1990). Climatic and biotic change in the late Triassic. J. Geol. Soc. London 147, 321–327.

Olsen, P. E., and Galton, P. M. (1977). Triassic–Jurassic extinctions: Are they real? Science 197, 983–986. Olsen, P. E., and Galton, P. M. (1984). A review of the reptile and amphibian assemblages from the Stormberg Group of Southern Africa, with special emphasis on the footprints and the age of the Stormberg. Palaeontol. Africana 25, 87–110. Olsen, P. E., and Sues, H.-D. (1986). Correlation of continental Late Triassic and Early Jurassic sediments, and

Extraterrestrial Impact Theory see EXTINCTION, CRETACEOUS

F Fabrosauridae PENG GUANGZHAO Zigong Dinosaur Museum Sichuan, People’s Republic of China

The Fabrosauridae was originally proposed by Galton (1972) as a group of basal ornithischians. It comprises some of the small, unarmored, and most primitive-looking ornithischians yet discovered—an early lineage in the ornithischian radiation. Fabrosaurids are known from southern Africa in the Upper Triassic and Lower Jurassic and from China in the Middle and Upper Jurassic. Because they bear the generalized states for many anatomical characters of the Ornithischia, they have attracted considerable attention in phylogenetic studies. However, most genera previously referred to the Fabrosauridae are represented by fragmentary remains, sometimes no more than isolated teeth, and their affinities are debatable. In 1990, specimens of Agilisaurus from the Middle Jurassic of Zigong, Sichuan, China, provided new evidence for resolving the systematic problems of the Fabrosauridae. The genus Fabrosaurus, type and only species F. australis, was established by Ginsburg (1964) on a partial right dentary with several teeth from the Upper Red Beds of Basutoland (Lesotho). Thulborn (1970, 1972) successively described the cranial and postcranial material of two more specimens from the same horizon as the type specimens in Lesotho, which he inferred were congeneric and conspecific with Ginsburg’s specimens. He transferred F. australis from the Scelidosauridae to the Hypsilophodontidae. Galton (1972) removed both Ginsburg’s and Thulborn’s specimens of F. australis from the Hypsilophodontidae to the newly created family Fabrosauridae. Subsequently, Galton (1978) removed Thulborn’s

specimens from F. australis and created a new genus as well as a new species for them, Lesothosaurus diagnosticus. However, Gow (1981) published a sharp criticism of Galton’s procedure, considering the genus Lesothosaurus a ‘‘myth.’’ Even so, the genus Lesothosaurus was adopted without comment by most authors. Sereno (1991) considered F. australis a nomen dubium and the Fabrosauridae invalid. He used Lesothosaurus rather than the Fabrosauridae in discussions of ornithischian phylogeny. However, by ICZN rules a family name does not become invalid with the loss of its eponymous genus, although these rules may not apply within the PHYLOGENETIC SYSTEM. Thulborn (1992) opined that F. australis Ginsburg 1964 is a determinate species founded on diagnostic material, and that Lesothosaurus diagnosticus Galton 1978 is a subjective junior synonym. In his view, therefore, all fabrosaurid specimens so far described from the Elliot Formation and Clarens Formation of southern Africa are referable to the species F. australis, or at least to the genus Fabrosaurus. Initially, Galton (1972) created the Fabrosauridae on the primitive character of marginally positioned cheek teeth. Later, Galton (1978) revised its diagnosis on the basis of approximately 40 characters, but the majority are plesiomorphic for Ornithischia rather than apomorphic for Fabrosauridae. Consequently, the inclusions of Alcodon kuehnei and Trimucrodon cuneatus from Portugal, Tawasaurus minor and Xiaosaurus dashanpensis from China, Nanosaurus agilis, Revueltosaurus callenderi, Technosaurus smalli, and Scutellosaurus lawleri from North America, and Echinodon becklesii from England have made this group a paraphyletic assemblage. Plesiomorphy, as commonly acknowledged, is phylogenetically uninformative; therefore, Sereno (1991) concluded that A. kuehnei, T. cuneatus, X. dashanpensis, N. agilis, and R. callenderi are ornithischians of unknown relation or nomina dubia, E. becklesii and S. lawleri are more closely related to other ornithischian subgroups, and T. minor




FIGURE 1 Reconstructions of skulls in lateral view. (A) Fabrosaurus australis (after Sereno, 1991); (B) Agilisaurus louderbacki.

and T. smalli are indeterminate Prosauropoda. This left F. australis (⫽ Lesothosaurus diagnostics) as the only genus and species diagnosable on apomorphic features. The recent discovery of a Middle Jurassic primitive ornithischian, Agilisaurus louderbacki Peng 1990, represented by a beautifully preserved skeleton from the Xiashaximiao Formation of Zigong, Sichuan Basin, China, has shed light on the systematic relationships of fabrosaurids. It is evident that Fabrosauridae is a monophyletic taxon diagnosed by the following unequivocal features: 1. lacrimal inserts into a narrow slot in the apex of the maxilla (Figs. 1A and 1B); 2. mandible with peculiarly salient finger-like retroarticular process (Figs. 1A and 1B); 3. especially short forelimb that is approximately 40% of the hindlimb in length; 4. ilium with a supra-acetabular flange over the anterior half of the acetabulum (Fig. 2); 5. posterior process of the ilium with a distinct brevis shelf that first turns medially and then downwards (Fig. 2); 6. ischium with a dorsal groove on the proximal shaft (Fig. 2); 7. pedal digit I reduced, with the splint-like shaft of metatarsal I and the ungual extending just beyond the end of the second metatarsal. In addition to the characters listed previously, fabrosaurids may be unique in tooth morphology and structure. They differ from prosauropods in their less symmetrical tooth crown, with fewer erect denticles of unvarying size on the anterior and posterior edges.

Unlike other primitive ornithischians, such as heterodontosaurids and hypsilophodontids, the teeth of fabrosaurids are thinly and uniformly enameled on either the buccal or lingual sides, each with a single round central vertical ridge. Compared to Agilisaurus, Fabrosaurus is more primitive in having six premaxillary teeth, a flat maxilla, a lower coronoid eminence, a large external mandibular fenestra, and a very short prepubis. Agilisauris is clearly more derived. It differs from Fabrosaurus in being larger and more complex in structure, as well as in having the specializations of a well-developed rod-like palpebral, a longitudinal depression along the sutural line between the nasals, and a nutritive foramen on the femoral shaft. Apart from the two genera mentioned previously, the Upper Jurassic ornithischian Gongbusaurus from

FIGURE 2 Reconstruction of the pelvic girdle of Agilisaurus louderbacki in lateral view.

Fabrosauridae China may be more closely related to fabrosaurids than to hypsilophodontids. The type species, G. shiyii Dong et al. 1983, from the Shangshaximiao Formation of Rongxian County in Sichuan Basin, is represented only by isolated teeth, but the included species G. wucaiwanensis Dong 1989, from the Shishugou Formation of Junggar Basin, Xinjiang, is represented by some fragmentary cranial and postcranial elements that show two apomorphic resemblances to fabrosaurids: the reduced pedal digit I with the splint-like metatarsal I and the ilium with a distinct brevis shelf. The most unequivocal feature in Gongbusaurus is that there are four distal tarsals. Although the other diagnostic characters of fabrosaurids are uncertain in Gongbusaurus, the long, deep posterior process of the ilium suggests that it is more closely related to Agilisaurus than to Fabrosaurus. The phylogenetic relationships of the three genera that constitute Fabrosauridae can be briefly indicated in Fig. 3. The fabrosaurids are small (1- or 2-m body length). The skull is triangular, with large circular orbits on the sides that suggest that the eyes were huge and directed laterally. A long and rod-like palpebral (supraorbital) extends across the dorsal margin of the orbit. It might have been used as a support for the eyeball. Like all other ornithischians, fabrosaurids have a small predentary bone joined to the tips of the two mandibular rami. The first premaxillary tooth is set a little back from the tip of the premaxilla, and the general structure of this area suggests that it may have had a small horny beak. The lower beak fitted inside the upper beak and premaxillary teeth, forming a more effective cropping device. In fabrosaurids, the leaf-like tooth crowns are mediolaterally compressed, with several coarse marginal denticles on either side of the apex and with

239 thinly and uniformly enameled buccal and lingual sides. Single or paired, highly inclined wear facets are usually present. A single wear facet indicates that the tooth was worn directly against a single counterpart tooth in the opposing jaw; paired wear facets were produced when the tooth interlocked with two teeth in the opposing jaw. Therefore, the jaw action of fabrosaurids was strictly vertical. Such a jaw mechanism is the simplest among known ornithischians. The forelimb of fabrosaurids was very much smaller than the hindlimb and terminated in a diminutive hand, whereas the hindlimb was distinctively elongated. The length of the forelimb is only approximately 40% of that of the hindlimb. Such a small forelimb could not have been used for locomotion, and it is obvious that fabrosaurids were bipedal. Adaptations for bipedalism are also evident in other skeletal features. The entire skeleton is very lightly built, with a largely fenestrated skull, a very short neck and trunk, and slender, hollow, and thin-walled limb bones, whereas the tail is much longer, occupying nearly half the total body length. Lightening of the skeleton implies weight reduction, which is most marked in front of the hips. The long tail presumably acted as a counterbalance for the weight of the body in front of the hips and as a compensating mechanism for shifts in the center of gravity. In ornithischians, as in other dinosaurs and ornithodirans, the lengthening of the distal parts of the hindlimb is usually associated with rapid bipedal progression. The hindlimb of fabrosaurids is also somewhat unusual because the tibia is considerably longer than the femur, and metatarsal III exceeds half of the femur length. These higher hindlimb ratios suggest that fabrosaurids were adapted for bipedal, fast running.

See also the following related entries: HYPSILOPHODONTIDAE ● ORNITHISCHIA

References Colbert, E. H. (1981). A primitive ornithischian dinosaur from the Kayenta Formation of Arizona. Museum Northern Arizona Bull. 53, 1–61.

FIGURE 3 Phylogenetic relationships within the Fabrosauridae.

Dong, Z. (1989). On a small ornithopod (Gongbusaurus wucaiwanensis sp. nov.) from Kelamaili, Junggar Basin, Xinjiang, China. Vertebr. Palasiatica 27, 140–146. [In Chinese] Dong, Z., and Tang, Z. (1983). Note on the new mid-


240 Jurassic ornithopod from Sichuan Basin, China. Vertebr. Palasiatica 21, 168–172. [In Chinese] Dong, Z., Zhou, S., and Zhang. Y. (1983). The dinosaurian remains from Sichuan Basin, China. Palaeontol. Sinica New Ser. C 23, 1–145. [In Chinese] Galton, P. M. (1972). Classification and evolution of ornithopod dinosaurs. Nature 239, 464–466. Galton, P. M. (1974). The ornithischian dinosaur Hypsilophodon from the Wealden of the Isle of Wight. Bull. Br. Museum (Nat. History) Geol. 25, 1–152. Galton, P. M. (1978). Fabrosauridae, the basal family of ornithischian dinosaurs (Reptilia: Ornithopoda). Palaontol. Zeitschrift 52, 138–159. Ginsburg, L. (1964). Decouverte d’un Scelidosaurien (Dinosaure ornithischien) dans le Trias superieur du Basutoland. Comptes Rendus Acad. Sci. Paris 258, 2366–2368. Gow, C. E. (1981). Taxonomy of the Fabrosauridae (Reptilia, Ornithischia) and the Lesothosaurus myth. South Africa J. Sci. 77, 43. Hunt, A. P. (1989). A new ornithischian dinosaur from the Bull Canyon Formation (Upper Triassic) of East Central New Mexico. In The Dawn of the Age of Dinosaurs in the American Southwest (S. G. Lucas and A. P. Hunt, Eds.), pp. 355–358. New Mexico Museum of Natural History, Albuquerque. Marsh, O. C. (1877). Notice of some new vertebrate fossils. Am. J. Sci. 14, 249–256. Owen, R. (1861). On the fossil Reptilia of the Wealden and Purbeck Formations. Part V. Palaeontogr. Soc. 7, 31– 39. [Monograph] Padian, K. (1990). The ornithischian form genus Revueltosaurus from the Petrified Forest of Arizona (Late Triassic: Norian; Chinle Formation). J. Vertebr. Paleontol. 10, 268–269. Peng, G. (1990). A new small ornithopod (Agilisaurus louderbacki gen. et sp. nov.) from Zigong, Sichuan, China. Newaleters Zigong Dinosaur Museum 2, 19–27. [In Chinese] Peng, G. (1992). Jurassic ornithopod Agilisaurus louderbacki (Ornithopoda: Fabrosauridae) from Zigong, Sichuan, China. Vertebr. Palasiatica 30, 39–53. [In Chinese] Santa Luca, A. P. (1984). Postcranial remains of Fabrosauridae (Reptilia: Ornithischia) from the Stormberg of southern Africa. Palaeontol. Africana 25, 151–180. Sereno, P. C. (1986). Phylogeny of the bird-hipped dinosaurs (Order Ornithischia). Natl. Geogr. Res. 2, 234–256. Sereno, P. C. (1991). Lesothosaurus, ‘‘fabrosaurids,’’ and the early evolution of Ornithischia. J. Vertebr. Paleontol. 11, 168–197. Thulborn, R. A. (1970). The skull of Fabrosaurus australis, a Triassic ornithischian dinosaur. Palaeontology 13, 414–432. Thulborn, R. A. (1971). Origins and evolution of ornithischian dinosaurs. Nature 234, 75–78.

Thulborn, R. A. (1972). The postcranial skeleton of the Triassic ornithischian dinosaur Fabrosaurus australis. Palaeontology 15, 29–60. Thulborn, R. A. (1973). Teeth of ornithischian dinosaurs from the Upper Jurassic of Portugal, with a description of a hypsilophodontid (Phyllodon henkeli gen. et sp. nov.) from the Guimarota Lignite. Mem. Servicos Geol. Portugal 22, 89–134. Thulborn, R. A. (1992). Taxonomic characters of Fabrosaurus australis, an ornithischian dinosaur from the Lower Jurassic of southern Africa. Geobios 25, 283–292. Weishampel, D. B., and Witmer, L. M. (1990). Lesothosaurus, Pisanosaurus, and Technosaurus, In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 416–425. Univ. of California Press, Berkeley.

Farmington Museum, New Mexico, USA see MUSEUMS



Fatima MARTIN LOCKLEY University of Colorado at Denver Denver, Colorado, USA


Middle Jurassic dinosaur tracksite in the vicinity of Fatima, Portugal, reveals the two longest sauropod trackways in the world, measuring 142 and 147 m. The tracks are also distinctive because of the wellpreserved impressions of the manus claws (digit I). Such claw impressions are not normally seen in sauropod trackways, and their position in life has been the subject of much debate (Lockley et al., 1994).

(Video) DK Knowledge Encyclopedia Dinosaur

See also the following related entries: EUROPEAN DINOSAURS ● FOOTPRINTS AND TRACKWAYS

Reference Lockley, M. G., Santos, V. F., Meyer, C. A., and Hunt, A. P. (Eds.) (1994). Aspects of Sauropod Biology. Gaia: Geoscience Magazine of the National Natural History Museum, Lisbon, Portugal.

‘‘Feathered’’ Dinosaurs



Fernbank Museum of Natural History, Georgia, USA see MUSEUMS



‘‘Feathered’’ Dinosaurs PHILIP J. CURRIE Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada In 1996, several skeletons of a 1-m-long animal were found in Liaoning (People’s Republic of China) that show feather-like structures covering the head, trunk, tail, arms, and legs. Named Sinosauropteryx prima, this animal is closely related to Compsognathus from the Upper Jurassic of Europe. The integumentary structures were simpler than true feathers, and each seems to be composed of a central rachis and branching barbs but lacks the aerodynamic quality of avian feathers. The longest ‘‘feathers’’ were about 3 cm in length, and it has been suggested that they were more suitable for insulation than they were as display structures. The discovery of these specimens has given additional support to the hypotheses that theropod dinosaurs were the direct ancestors of birds, and that some theropods were endothermic.

See also the following related entries: BIRD ORIGINS ● PHYSIOLOGY ● SKIN

Fernbank Science Center, Georgia, USA see MUSEUMS



Field Museum of Natural History, Illinois, USA see MUSEUMS



Flaming Cliffs A commonly used reference for a famous Mongolian locality known as Bayn Dzak.


Floras, Mesozoic

Feeding see DIET; TEETH




Footprints and Trackways MARTIN LOCKLEY University of Colorado at Denver Denver, Colorado, USA


he study of dinosaur footprints can be considered a branch of vertebrate paleontology, but it is also a branch of ichnology, the study of trace fossils. As such, dinosaur ichnology conforms to different procedures and conventions from those applied to the naming and description of dinosaur body fossils. Ichnotaxonomy, or the naming of trace fossils, allows ichnologists to give distinctive names (ichnogenera and ichnospecies) to footprints and other traces (see COPROLITES; TRACE FOSSILS). For example, the ichnospecies Tyrannosauripus pillmorei (referring to a tyrannosaur track discovered by the geologist Pillmore) was probably made by the species Tyrannosaurus rex, the only known animal of that type capable of making the track. Endings such as -ipus, -opus, -podus, or -ichnium indicate a trace fossil rather than the animal itself. In most cases, the species that made a particular track is not known, even though available evidence may narrow down the probable trackmaker to a more or less specific taxon with varying degrees of confidence. The naming of tracks conforms to the same general principles as the naming of body fossils. That is, they are named on the basis of distinctive morphologies, with the designation of type specimens (holotypes), and must include descriptions and appropriate documentation. In practice there are clear morphological differences between tracks made by sauropods, theropods, and other major groups, such as ornithopods or ceratopsians. However, the difference between various theropod morphotypes may be more subtle. In addition, a significant percentage of tracks in any sample may be poorly preserved. Consequently, guidelines have evolved that recommend that tracks be named only when sufficient supplies of ‘‘wellpreserved’’ material are available. Such guidelines recommend having sequences of tracks (trackways) available for study rather than isolated footprints (Sarjeant, 1989).

Unlike the traces of invertebrate animals that often represent behavior such as burrowing, traces of vertebrates usually reveal foot morphology and so reflect the anatomy of the trackmaker to some degree. For this reason, vertebrate tracks are often named after the inferred trackmaker as in the example of Tyrannosauripus or Brontopodus (meaning brontosaur track). There is no rule about the choice of such names, and many tracks are named after places, rock formations, people, and so on. Edward Hitchcock (1793–1864) is considered the father of vertebrate ichnology. He described many dinosaur tracks from the Lower Jurassic of New England, but attributed most to birds or nondinosaurian reptiles. His most complete work (Hitchcock, 1858) contains many ichnogenus names (notably, Grallator and Eubrontes) still in use today. His work was revised by Richard Swan Lull in a series of papers (notably, Lull, 1953). Baron von Nopsca has been credited by some as the first European to produce a seminal work on vertebrate ichnology (Nopsca, 1923). Other notable encyclopedic and glossary contributions include Haubold (1971, 1984) and Leonardi (1987). The past decade has witnessed a renaissance in dinosaur ichnology, beginning with the publication of the proceedings of the First International Symposium on Dinosaur Tracks and Traces (Gillette and Lockley, 1989), two general books on dinosaur tracks (Thulborn, 1990; Lockley 1991), and regional studies (Leonardi, 1994; Lockley and Hunt, 1995). Tracks or footprints (same meaning) are described on the basis of footprint morphology as one, two, three, four, or five toed (i.e., mono-, di-, tri-, tetra-, or pentadactyl, respectively). Hind footprints are referred to as pes or pedal impressions and front foot impressions as manus or manual impressions. Most quadrupedal dinosaurs had larger hind footprints than front footprints (⫽ heteropody). Some dinosaurs, notably theropods and ornithopods, were bi-


Footprints and Trackways pedal. Trackmaker morphology and movement can be understood in part by noting whether the long axis of the foot rotates outward, points forward, or rotates inwards. Similarly, some dinosaurs have longer inner digits (I or II), a larger central digit (III), or longer outer digits (IV or V) referred to as entaxonic, mesaxonic, or ectaxonic, respectively. Trackmakers may have put their entire foot flat on the ground (plantigrade) or walked on their toes (digitigrade). All such features must be studied in relation to the overall pattern of the trackway (consecutive steps made by an individual trackmaker), in which one can measure step or pace (left to right or right to left), stride (left to left or right to right), and pace angulation (angle between lines of two consecutive steps). Different trackway patterns are characteristic of particular groups and reveal the stance of the trackmakers. For example, wide trackways indicate sprawling and narrow trackways indicate erect or upright stance. Because tracks were made by living animals, they are useful for interpreting behavior. A basic feature of behavior is locomotion, and various formulae have been proposed to estimate dinosaur speed (애) from trackways. The most widely used is 애 ⫽ 0.25 g0.5 ⭈ ␭1.67h⫺1.17 proposed by Alexander (1976), where ␭ is stride length, h is hip height (4 ⫻ footprint length), and g is acceleration due to gravity. Other formulae are available (Thulborn, 1990), but none can calculate absolute speed owing to the fact that duration of step and stride (cadence) is not known. Thus, all calculated velocities are relative speed estimates. Trackways with increasing or decreasing step length may indicate acceleration or deceleration, whereas those with alternating long and short steps suggest limping. Other reports of unusual individual behavior, such as hopping or attacking, are not supported by the trackway evidence or are, at best, very controversial. Parallel trackways of the same type on a single surface may indicate gregarious or herd behavior (Lockley, 1991, 1995). Such examples are particularly common among sauropod and ornithopod tracks in the Late Jurassic through Cretaceous and are characterized by regular spacing of trackways. Trackway orientations may also reflect ancient geography and are often recorded running parallel to ancient shorelines. Trackways often occur in similar proportions in numerous track assemblages (ichnocoenoses) from par-

243 ticular sedimentary facies or rock formations. Such recurrent associations can be used to define particular ichnofacies that reflect the composition of animal communities in particular environments (Lockley et al., 1994a). Current evidence suggests that tracks record distinctive animal communities in different settings, ranging from desert dunes to desert playas, humid swamps, or carbonate environments. Thus, tracks are useful for paleoecological census studies. There is danger of misinterpreting tracks if students do not take adequate care to understand how they are preserved. For example, tracks can be transmitted downwards from the surface on which the foot came to rest, into one or more underlayers. Such transmitted tracks are known as ghostprints or underprints and are generally less distinct and somewhat larger than the true tracks in younger layers above. Although undertracks lack details of skin or pad impressions, they sometimes show well-preserved claw marks that are not seen in association with the true tracks. Some underlayers preserve only hind or front footprints from trackways of quadrupeds. Sauropod trackways that show only front footprints were for a long time interpreted as the result of swimming behavior but are now known to be underprints caused by the front feet having sunk deeper than the hind feet. It is important to understand not only how tracks are preserved but also how they are made. The form of a track is more often than not a representation not only of the animal’s foot anatomy but also of the kinematic pattern of its step cycle and the condition of the substrate when the track was made. All these features have to be considered in interpreting and taxonomizing trackways (Padian and Olsen, 1984). Other aspects of understanding track preservation pertain to bias in the fossil record in general. In most cases, small tracks are rarer and less well preserved than large tracks. Tracks are also preferentially preserved along shorelines where substrates were wet. Thus, animals that habitually frequented dry areas are less well represented. However, because most major dinosaur groups are well represented in the Mesozoic track record, it can be inferred that the record is not excessively biased. The tracks of carnivorous dinosaurs (predators) are often more common than those of herbivores (prey) and relatively more common than the predator:prey ratios inferred from the bone record. Such inconsistencies have been at-

244 tributed to higher activity levels among theropods, but other explanations are possible. For example, different predator:prey ratios may reflect the inclination of predators to patrol shorelines and hunt by sight. Predator:prey ratios based on trackway numbers should be corrected for estimated body size to give biomass estimates for ecological interpretations. Comparison of footprint and bone records in various formations is also revealing for at least two reasons. First, tracks mainly provide evidence of terrestrial faunas, whereas some skeletal assemblages are dominated by the remains of aquatic vertebrates (fish, turtles, crocodiles, etc.). Such evidence establishes that tracks must be included in paleontological censuses in order to obtain as complete a picture as the fossil record will allow. Second, recent studies show that tracks are many orders of magnitude more abundant than bones in many formations. For example, in some deposits, hundreds or thousands of trackways have been recorded, whereas the skeletal record consists of only one or two or, in some cases, no individual remains. Such evidence gives us a new perspective on the importance of tracks by filling in many gaps in the fossil record (Lockley, 1996). In recent decades ichnologists have recognized that the same track assemblages occur in rocks of the same or similar age over wide areas. Thus, tracks are useful for correlation or track stratigraphy referred to as palichnostratigraphy. It has become evident that dinosaur track correlations can be made from North America to Europe and Asia throughout much of the Mesozoic, but especially in the Late Triassic and Jurassic. In several cases the correlations can be made using two or three ichnogenera in formations or time intervals where no skeletal remains exist. Because tracks are abundant and trackmakers potentially mobile over wide areas, they fulfill the important criteria for use in stratigraphic correlation. Such correlations also reveal a prolonged record of faunal interchange between continents that are often shown as having no connections on paleogeographic maps. Recent studies have often shown that tracks occur in exceptional abundance over wide areas in particular thin sedimentary units or even on single surfaces. Such extensive sites are referred to as megatracksites or even dinosaur freeways and can be measured on the order of thousands or even tens of thousands of square kilometers. Examples have been identified in both the Jurassic and Cretaceous of North America

Footprints and Trackways and Europe. Such megatracksites are associated with sediment aggradation during periods of rising sea level. Indeed, some studies indicate that both the track and bone records are more prolific and complete at times of elevated sea level. Such an abundant track record provides the raw data necessary for making track correlations. It should be emphasized, however, that megatracksites are essentially continuous track layers in a particular facies in a given region, whereas track correlations identify the same tracks (ichnogenera) in different regions that are often on different continents and in different facies. Despite the fact that Edward Hitchcock first described dinosaur tracks (as those of birds) in 1836 before dinosaurs were known, dinosaur ichnology is still a relatively young and immature science. During the past decade, however, rapid progress has been made in the field, resulting in a significant new literature. Possibly the most important results pertain to the realization that tracks are incredibly abundant at literally thousands of sites on most continents. Such abundance provides a large database but also reveals the extent to which numerous formations have a better record of tracks than skeletal remains. For historical reasons track ichnotaxonomy is somewhat confused, but recently progress has been made toward revising nomenclature. This important first step allows valid names to be applied in paleoecological and biostratigraphic studies, and so demonstrates important patterns of track distribution in time and space. It is also important to realize that ichnotaxonomy is a separate system from body fossil taxonomy, and that matching tracks with trackmakers can be done in some circumstances (in which both bones and tracks are known from a given formation or in which track morphologies are distinctive) but not in others (in which there are no bones to match the track record). Even so, it is desirable, though not essential, to have track names that reflect the morphology or affinity of the trackmaker. Such names must be applied judiciously, for example, only when trackmaker affinity can be deduced with high levels of confidence. Many old track names are invalid because they are duplicative ( junior synonyms) or are based on poorly preserved material (nomina dubia). As a result, compilations from old track literature (Weishampel et al., 1990) can be misleading, as demonstrated by Lockley et al. (1994b). Such problems, however, are well within

Forelimbs and Hands the abilities of ichnologists to solve. Through careful study of distinctive track assemblages, containing well-preserved material, we can hope that ichnotaxonomy will be progressively simplified and revised. Through comparison of the bone and track records, the affinity of trackmakers will become much clearer in most cases. Without doubt, the recent renaissance in dinosaur ichnology has brought the field into the paleontological mainstream. This integration of the track and skeletal records is taking place at a fundamental level, owing to the compelling evidence that the body fossil record is impoverished and less complete than it would be without track evidence and vice versa.


References Alexander, R. M. (1976). Estimates of speeds of dinosaurs. Nature 261, 129–130. Gillette, D. D., and Lockley, M. G. (Eds.) (1989). Dinosaur Tracks and Traces, pp. 454. Cambridge Univ. Press, Cambridge, UK. Haubold, H. (1971). Ichnia Amphibiorium et Reptiliorum fossilium. In Handbuch der Palaeoherpetologie, Teil 18 (O. Kuhn, Ed.), pp. 1–124. Fischer–Verlag, Stuttgard. Haubold, H. (1984). Saurierfa¨hrten. Wittenberg Lutherstadt, Die Neue Brehm-Bucherei, pp. 232. Hitchcock, E. (1858). A Report on the Sandstone of the Connecticut Valley Especially Its Fossil Footmarks, pp. 220. White, Boston. [Reprinted by Arno Press; in the Natural Sciences in America Series] Leonardi, G. (Ed.) (1987). Glossary and Manual of Tetrapod Footprint Palaeoichnology, pp. 117. Departamento Nacional da Produca˜o Mineral, Brasilia. Leonardi, G. (1994). Annotated Atlas of South American Tetrapod Footprints (Devonian to Holocene), pp. 248. Companhia de Pesquisa de Recursos Minerais, Brasilia. Lockley, M. G. (1991). Tracking Dinosaurs: A New Look at an Ancient World, pp. 238. Cambridge Univ. Press, Cambridge, UK. Lockley, M. G. (1995). Track records. Nat. History 104, 46–51. Lockley, M. G. (1997). The paleoecological and paleoenvironmental utility of dinosaur tracks. In Dinosaurs: A Sesquicentennial Celebration (J. O. Farlow and M. BrettSurman, Eds.), in press.

245 Lockley, M. G., and Hunt, A. P. (1995). Dinosaur Tracks and Other Fossil Footprints of the Western United States, pp. 338. Columbia Univ. Press, New York. Lockley, M. G., Hunt, A. P., and Meyer, C. (1994a). Vertebrate tracks and the ichnofacies concept: Implications for paleoecology and palichnostratigraphy. In The Paleobiology of Trace Fossils (S. Donovan, Ed.), pp. 241–268. Wiley, New York. Lockley, M. G., Santos, V. F., Meyer, C. A., and Hunt, A. P. (1994b). Aspects of Sauropod Biology, pp. 266. Gaia: Geosciencies Journal of the Natural History Museum of the University of Lisbon, Lisbon, Portugal. Lull, R. S. (1953). Triassic life of the Connecticut Valley. State Geological and Natural History Survey of Connecticut Bull. 81, pp. 336. Nopsca, F. (1923). Die familien der Reptilien. Fortschr. Geol. Palaont. Heft 2, 1–210. Padian, K., and Olsen, P. E. (1984). The track of Pteraiehuus: not pterosaurian, but crocodilian. J. Paleontol. 58, 178–184. Sarjeant, W. A. S. (1989). Ten paleoichnological commandments: A standardized procedure for the description of fossil vertebrate footprints. In Dinosaur Tracks and Traces (D. D. Gillette and M. G. Lockley, Eds.), pp. 369–370. Cambridge Univ. Press, Cambridge, UK. Thulborn, R. A. (1990). Dinosaur Tracks, pp. 410. Chapman & Hall, London. Weishampel, D. B., Dodson, P., and Osmo´lska, M. (1990). The Dinosauria, pp. 733. Univ. of California Press, Berkeley.

Forelimbs and Hands PER CHRISTIANSEN Københavns Universitet Copenhagen, Denmark

Plesiomorphically dinosaurs were bipedal animals (see BIPEDALITY), and although most members of several lines later became obligatory quadrupedal, the QUADRUPEDALITY probably evolved convergently. The earliest dinosaurs had long, slender forelimbs, about half as long as the hindlimbs, with a long, low deltopectoral crest, and rather small hands suited for manipulation, not progression. The relative forelimb length became reduced in most lines not adopting a quadrupedal posture, except MANIRAPTORANS (see BIRD ORIGINS). The earliest dinosaurs were all small animals and as some lines grew to immense sizes,

246 the quadrupedal posture probably became a necessity. The mobility of the forelimbs of all dinosaurs was always somewhat parasagittal as in most larger extant mammals. SAUROPODS in particular appear to have had limbs that hardly allowed any lateromedial mobility at all, as in extant artiodactyls and perissodactyls and, of course, elephants (Fig. 1; Table I). The antebrachium of dinosaurs was usually morphologically quite similar to those of extant mammals or

Forelimbs and Hands reptiles in overall view, but there was never a somewhat posteriorly directed, rounded humeral caput with a distinct neck as in most living mammals, and the proximal part of the humerus was lateromedially expanded, taking on a spoon-like shape. Other differences between dinosaurs and mammals were also apparent, but overall the functional anatomy of dinosaurian forelimbs is closer to mammals than reptiles. In quadrupedal dinosaurs the forelimbs appear never

FIGURE 1 Comparative morphology of dinosaurian forelimbs. (1–7) Humeri (1–6, cranial view; 7, lateral view): 1, theropod (Syntarsus); 2, prosauropod (Plateosaurus); 3, sauropod (Dicraeosaurus); 4, stegosaur (Stegosaurus); 5, ankylosaur (Euplocephalus); 6, ornithopod (Camptosaurus); 7, ceratopsian (Centrosaurus). (8–14) Radii (8, 10–12, cranial view; 9 and 14, lateral view; 13, medial view): 8, theropod (Gallimimus); 9, prosauropod (Plateosaurus); 10, sauropod (Brachiosaurus); 11, stegosaur (Stegosaurus); 12, ankylosaur (Euplocephalus); 13, ornithopod (Dryosaurus); 14, ceratopsian (Chasmosaurus). (15–21) Ulnae (15, cranial view; 21, caudal view; 16, 18, and 19, lateral view; 17 and 20, medial view): 15, theropod (Gallimimus); 16, prosauropod (Plateosaurus); 17, sauropod (Brachiosaurus); 18, stegosaur (Stegosaurus); 19, ankylosaur (Euplocephalus); 20, ornithopod (Dryosaurus); 21, ceratopsian (Triceratops). (22–27) Manus (cranial view): 22, theropod (Syntarsus); 23, prosauropod (Plateosaurus); 24, sauropod (Brachiosaurus); 25, stegosaur (Stegosaurus); 26, ornithopod (Iguanodon); 27, ceratopsian (Centrosaurus).

Forelimbs and Hands


TABLE I Forelimb Proportions in Some Dinosaurs Compared to Extant Mammals Size range of Taxon Mammalia Carnivora Ursidae Canidae Artiodactyla Cervidae Bovidae Perissodactyla Rhinocerotidae Proboscidea Elephantidae Dinosauria Prosauropoda Plateosauridae Sauropoda Diplodocidae Camarasauridae Ceratopsia Ceratopsidae Ornithopoda Hadrosauridae a


Propodia (mm)

Epipodia (mm)


7 (12) 15 (24)

142–422 70–255

108–385 65–275

0.76–0.94 0.82–1.08

11 (21) 32 (49)

116–385 90–409

115–416 83–361

0.88–1.08 0.87–1.34

4 (8)




2 (10)




2 (3)




3 (6) 2 (4)

750–1150 450–1004

525–840 298–702

0.69–0.78 0.64–0.71

5 (8)




6 (7)




n is number of species and in parentheses are number of specimens.

to have sprawled as in modern reptiles but rather to have been mammalian in function, and the gait was largely parasagittal.

Prosauropoda Archaic PROSAUROPODS, such as most Plateosauridae, Massospondylidae, Anchisauridae, and Yunnanosauridae, were at least facultatively bipedal, as indicated by their considerably longer and stronger hindlimbs. The Melanorosauridae and probably Blikanasauridae were quadrupedal because their appendicular anatomy approached that of sauropods. The humerus is the longest bone in the forelimb and the distal epiphysis is somewhat anteriorly and slightly medially angled compared to the long axis of the diaphysis, indicating elbow flexure and slight medial orientation of the epipodium. The deltopectoral crest is pronounced, indicating a rather strong upper arm and shoulder musculature, in accordance

with the fairly large proximal part of the scapula. The epicondyles are moderately well developed, indicating quite powerful carpal flexors. The propodium : epipodium ratio is usually 0.5–0.7. The ulna has an anterior proximal depression for the radius and the two bones probably did not cross over much. The distal epiphysis is usually triangular and slightly angled, and the olecranon process is quite well developed. The proximal epiphysis of the radius is concave or saddle shaped as in many extant mammals, indicating fairly well-developed anteroposterior and less well-developed lateromedial mobility. The carpus may have included as many as six elements (Young, 1941) but usually the proximal carpals are rarely preserved and largely unknown. Galton (1990) suggested that they were cartilaginous, which seems reasonable. Frequently found is the small, semicircular intermedium and three distal carpals, of which distal carpal I usually is the largest.

Forelimbs and Hands

248 The metacarpus is distinctly asymmetrical with metacarpal I being the strongest and the rest decreasing in size numerically. Metacarpal IV and especially V are reduced. The phalangeal formula is usually 2-34-2-0. The articulating facets are usually of the condylar–cotylar type but are often somewhat ginglymoid. Digit I is massive and bears a large, somewhat recurved ungual, and the morphology of the hand indicates that slight supination was possible. Unguals are present only on digits I–III, with II and III much smaller and straighter than I. In the quadrupedal Melanorosauridae, the limb bones are more massive and longer compared to the hindlimb in other prosauropods, the epiphyses are less angled in relation to the diaphyses, and the asymmetry of the hand is less pronounced. However, no prosauropod appears to even have approached the highly apomorphic metacarpal anatomy of sauropods.

Sauropoda The sauropod dinosaurs included the largest terrestrial animals of all time and thus faced problems of support of mass not experienced by most other dinosaurs. Despite the great diversity and longevity of the clade, their appendicular anatomy was rather conservative. All sauropods were obligatory quadrupeds with forelimbs primarily adapted for support of mass, and the limb bones appear never to have been hollow as in theropods. Sauropod locomotion has often been compared to that of elephants. A very elephantine

trait is that all epiphyses of the three long bones of the forelimb are perpendicular to the long axes of their respective diaphyses, which lack curvature. This suggests that the pillar-like limb posture and graviportal mode of locomotion of recent elephants was also present in all sauropods. The forelimb bones were massive, more so than in modern elephants, and radius and ulna were both well developed. The propodial:epipodial proportions (0.55–0.79, highest in certain diplodocids) also point to a graviportal mode of locomotion, compared to 0.76–0.81 in recent elephants (Fig. 2). Archaic sauropods, such as the Euhelopodidae (sensu Upchurch, 1995), Camarasauridae, and paraphyletic Cetiosauridae, have forelimbs of moderate length compared to the hindlimbs. The Dicraeosauridae and Diplodocidae have shorter forelimbs. In the Brachiosauridae, the forelimbs are usually as long as or longer than the hindlimbs, mimicking the condition in proboscideans and giraffids. The deltopectoral crest is usually pronounced, indicating rather powerful propodial and scapular musculature. This is confirmed by the rather large size of the shoulder girdle, which is similar to those of elephants in this respect. The humeral epicondyles are modest, indicating modest manus flexors. The ulna is usually the longer of the two epipodial bones and has a more reduced olecranon process than in elephants, indicating less epipodial extensor power in accordance with a more limited capacity for forward propulsion. The epipodial bones probably crossed each other only

FIGURE 2 Propodial robusticity in sauropod dinosaurs (n ⫽ 29). 䊉, sauropods; 䊊, elephants; 䉭, rhinoceroses; 䊐, bovids.

Forelimbs and Hands modestly. The carpus was not elephantine; elephants have a double row of squarish or more rounded carpals where intercarpal movement occurs, whereas in sauropods there is only one row of block-like carpals (a large scapholunare and a smaller radial). This must have made the manus less flexible. Furthermore, the metacarpals are fully erect and form a semicircle, and the phalanges are reduced in number and size. Only digit I bears an ungual. Unlike those of prosauropods, all five metacarpals are stout and unreduced in size. The phalangeal formula is usually 2-1/2-1/2-1/2-1. Unlike elephants, sauropods did not have a posterior heel pad, and they were unique among very large terrestrial vertebrates in having an unguligrade forelimb posture. This stance probably made a pushoff against the ground impossible, unlike all extant mammals including elephants. The propulsive force must have come from propodial retractors exclusively. The single claw was probably not used in locomotion but possibly for grasping tree trunks during high browsing in a tripodal posture (Upchurch, 1994); hence its modesty in the giraffe-like Brachiosauridae.

Theropoda Theropod forelimbs were always considerably shorter than the hindlimbs (except in Maniraptora) and unsuited for terrestrial progression. A general trend during theropod evolution (except Maniraptoriformes) was relative shortening of the forelimb and shortening of the hand compared to the arm. This trend reached its peak in tyrannosaurids, whereas the ORNITHOMIMOSAURIA, DROMAEOSAURIDAE, and OVIRAPTOROSAURIA reversed this trend. Proximally, the humerus was lateromedially expanded by the large medial tuberosity and prominent deltopectoral crest. This crest reached quite far down the diaphysis in Herrerasaurus and ornithisichians (Sereno, 1993) but was more proximal in theropods. In ornithomimids and tyrannosaurids, except Tyrannosaurus rex, the deltopectoral crest was markedly reduced. The humeral diaphysis was usually somewhat sigmoidal. The distal epiphysis is lateromedially expanded, often with a shallow olecranon fossa posteriorly, and cranially and slightly medially directed causing permanent elbow flexure and supination of the arm during flexion. The epicondyles were rather prominent, indicating quite powerful manual flexors. In ornithomimids, the diaphysis was straighter and the distal expansion and epicondyles were more modest. The antebrachium

249 was always shorter than the propodium, the ulna being the longer of the two epipodial bones. The ulnohumeral ratio varied from 0.53–0.61 in tyrannosaurs, 0.7–0.9 in ornithomimids, to 0.95 in Herrerasaurus and theropod outgroups. The olecranon process was quite prominent, except in gracile ceratosaurs (Rowe and Gauthier, 1990). The ulnar diaphysis was cranially slightly convex and tapered distally. The radius was straighter, with a concave or saddle-shaped proximal epiphysis, and the radiohumeral ratio varied from 0.42–0.5 in tyrannosaurs, 0.76 in Deinonychus, to about 0.9 in Herrerasaurus and certain ornithomimids. The carpus is not well known in many taxa. In Herrerasaurus there are two proximal carpals, probably radial and ulnar, and five smaller distal carpals (Sereno, 1993). Ornithomimids have five carpals in all, whereas Allosaurus appears to have four—an intermedium, radial, and two unidentified carpals (Madsen, 1976). Deinonychus appears to have had only two ossified carpals (Ostrom, 1969), originally identified as radial and ulnar but recently interpreted as distal carpals I and II (fused) and III (see BIRD ORIGINS). The semilunate carpal also occurs in oviraptorosaurs, troodontids, and some less well-known taxa. In Herrerasaurus and theropod outgroups there are five metacarpals, the fifth being vestigial, and the phalangeal formula is 2-3-4-1-0. In all theropods metacarpal V is absent. Theropod unguals are always lateromedially compressed and recurved, but are straighter in ornithomimids. Ceratosaurs still have four metacarpals and a phalangeal formula of 2-3-41-X. However, tetanurans are tridactyl, and tyrannosaurs are didactyl with only a vestigial third metacarpal. Metacarpal I is short and stout, and metacarpal II is the longest. Phalanges are usually ginglymoid, especially in dromaeosaurs, and the unguals recurve with prominent flexor tubercles. The manus reaches its relative peak of development in dromaeosaurs, in which the hand is large and mobile and the phalanges, especially the large unguals, have prominent flexor tubercles. The manus is least developed in tyrannosaurs and certain abelisaurids.

Stegosauria Stegosaurids were all obligatory quadrupeds with moderately long limbs and hindlimbs considerably longer than the forelimbs. The limb proportions were somewhat elephantine, indicating limited locomotory potential, and the radiohumeral ratio was ap-

Forelimbs and Hands

250 proximately 0.67–0.81, compared to 0.76–0.81 in recent elephants. The humerus was stout and greatly expanded proximally and distally, and the diaphysis was much more robust than in elephants. Also, unlike elephants, the humeral muscle scars were large and the distal epiphysis was somewhat anteriorly directed compared to the long axis of the diaphysis, causing permanent elbow flexure. The deltopectoral crest is usually very pronounced, reaching one quarter or more down the diaphysis, and the medial tuberosity and epicondyles are pronounced, especially the medial epicondyle. Distally the posterior olecranon fossa was usually distinct but often rather shallow. The radius had moderately to greatly expanded proximal and distal epiphyses almost perpendicular to the long axis of the diaphysis. The proximal epiphysis was usually concave or more planar, whereas the distal epiphysis was inclined. The ulna is more robust and longer than the radius, and the olecranon process is large or very large, which indicates massive triceps musculature, causing the ulnohumeral ratio sometimes to exceed 1 (0.82–1.03). The ulnar diaphysis is somewhat triangular and tapers distally. There is a proximal anterior radial fossa. The manus is reminiscent of that of sauropods: The carpus consists of the block-like radial and ulnarulnar, both often subequal in size or the radial can be the larger. In juveniles of some species there appear to have been four carpals (Gilmore, 1914). There are five short and stout metacarpals that appear to be almost erect and form a semicircle as in sauropods. The phalanges are reduced in size and number, and only digit I appears to have borne a rather straight, lateromedially wide and hoof-like ungual. The phalangeal formula is usually 2-2-2-2-1. Ankylosauria Ankylosaur forelimbs were shorter than the hindlimbs and stout. The humerus is short and massive with a short, thick diaphysis and great proximal and somewhat less distal expansion. The deltopectoral crest is large and is usually relatively larger in ankylosaurids than in nodosaurids. The epicondyles were well developed and the distal epiphysis was angled to the long axis of the diaphysis, causing permanent elbow flexure. The radius is the shorter of the two epipodial bones and has an ellipsoid or circular diaphysis. The proximal epiphysis is rounded and concave, and there are usually prominent proximal rugosities, perhaps for insertion of the interos-

seus. In ankylosaurids, the radiohumeral ratio is approximately 0.60. The ulna resembles that of stegosaurs in being proximally massive, in tapering distally, in being anteroposteriorly concave with a distinct proximal fossa for the radius, and in having a very prominent olecranon process. The ulnohumeral ratio is considerably higher than the radiohumeral ratio, generally approximately 0.85 or even more, due to the size of the olecranon process. The manus is not well known and most genera appear to have had a pentadactyl manus with short and stout phalanges, superficially resembling the hand of stegosaurs. The phalangeal formula appears to have been 2-3-3/42. There were dorsally flattened, hoof-like claws on digits II–IV.

Ornithopoda The Ornithopoda was a large, long-lived, and successful dinosaurian clade (Fig. 3). Nonetheless, their appendicular anatomy did not vary much. All had considerably shorter forelimbs than hindlimbs, but most probably progressed quadrupedally at least some of the time. Ironically, the most archaic ornithopods, the Heterodontosauridae, had proportionally the longest forelimbs but do not appear to have used these in locomotion. The humerus is generally longer than the antebrachium in more archaic ornithopods, but in hadrosaurs the radiohumeral and ulnohumeral ratio often exceeds 1. The humeral diaphysis is slightly bowed or sigmoid, and the proximal and distal lateromedial expansions are moderate. The caput is anteroposteriorly thickened and distinct and is usually located more or less centrally. The anteriorly directed deltopectoral crest is strong, especially in heterodontosaurids and lambeosaurine hadrosaurs, in which it extends almost halfway down the diaphysis. The medial tuberosity is pronounced in heterodontosaurids but less well developed in iguanodontids and hadrosaurs. The distal epiphysis is cranially directed, more so in Heterodontosaurus than in iguanodontids or hadrosaurs. In well-preserved specimens the condyles are round and well developed. The olecranon fossa is weakly developed in Heterodontosaurus but is usually more pronounced in most iguanodontids and hadrosaurs. The radiohumeral ratio is 0.61 in Camptosaurus, approximately 0.7 in Heterodontosaurus (Weishampel and Witmer, 1990), and much higher in hadrosaurs (0.92–1.16). The proximal radial epiphysis is rounded and concave, the proximal and distal

Forelimbs and Hands


FIGURE 3 Propodial robusticity in ornithopod dinosaurs (n ⫽ 30). 䊉, ornithopods; 䊊, elephants; 䉭, rhinoceroses; 䊐, bovids.

expansions are modest to moderate, and the distal epiphysis is usually more squarish and planar. The diaphysis is straight and often rounded in iguanodontids and hadrosaurs but is more triangular in hypsilophodontids. The ulna is always longer than the radius, and the ulnohumeral ratio in Camptosaurus is 0.67–0.70. In Heterodontosaurus it is 0.80 (Weishampel and Witmer, 1990), and in hadrosaurs it is 1.02–1.21. The olecranon process is large in heterodontosaurids and most iguanodontids but more modest in hypsilophodontids and hadrosaurs. The diaphysis is relatively straight and tapers somewhat distally, although there is an expansion at the distal epiphysis. The distal epiphysis is usually slightly convex. The carpus of Heterodontosaurus includes no fewer than nine bones, including the ulnar, radial, and pisiform in the proximal row of carpals, one medial carpal (tentatively identified as the centrale), and five distal carpals (Santa Luca, 1980). The carpals do not fuse to the metacarpals or to each other. The radial and ulnar were probably immovably attached to the antebrachium, and the large number of carpals must have provided the manus with substantial flexibility. In more derived conditions, the carpal number is reduced to three proximal carpals (of which the ulnar is the largest) and two smaller distal carpals in Tenontosaurus (Forster, 1990), three proximal carpals in Dryosaurus (the radial, intermedium, and ulnar, all approximately subequal in size), and only two maximal carpals in hadrosaurs (the radial and ulnar). The

metacarpus includes five metacarpals in heterodontosaurids, hypsilophodontids, dryosaurids (Galton, 1981), and iguanodontids, but metacarpal I is absent in hadrosaurs. Unusual among dinosaurs, metacarpal I of Heterodontosaurus was directed medially, with the proximal lateral metacarpal condyle at right angles to the long axis of the bone and the lateral side longer than the medial side. In hypsilophodontids and iguanodontids, metacarpal V was directed laterally, and in iguanodontids metacarpal I was also directed medially. Anatomically, it appears different from the condition in Heterodontosaurus, suggesting convergence. The phalanges in Heterodontosaurus were quite long with ill-defined condylar–cotylar joints, and the phalangeal formula is 1-3-4-3-2. Digits I–III bore large, somewhat recurved unguals with large flexor tubercles (Santa Luca, 1980). In hypsilophodontids, iguanodontids, and hadrosaurs, the phalanges were shorter and stouter and the interphalangeal joints condylar–cotylar or ginglymoid, especially in the distal phalanges. The claws were hoof-like, except in Iguanodon and Ouranosaurus, in which digit I was transformed into a large, spike-like protuberance, probably a defensive weapon. Digits II–IV in hadrosaurs and iguanodontids (except Camptosaurus) were the longest and had their large metacarpals closely appressed to each other. These were probably enclosed in connective tissue to form a common foot. Digits I and V projected medially and laterally from this unit, respectively, and these animals frequently

Forelimbs and Hands

252 walked quadrupedally, as shown by trackways (e.g., Paul, 1987; Currie, 1989).

Ceratopsia Among the CERATOPSIA, the small PSITTACOSAURIDAE were bipedal and the Protoceratopsidae were at least facultatively quadrupedal. The forelimb is considerably shorter than the hindlimbs in all ceratopsians. The humerus is gracile in small forms and becomes increasingly robust as linear dimensions increase. The humeral caput is located approximately centrally on the proximal part and extends somewhat caudally, indicating that the long axis of the bone was inclined posteriorly. The humerus is always longer than the antebrachium. The diaphysis is rather straight and approximately circular in psittacosaurids and protoceratopsids but has greater lateromedial than anteroposterior diameter in ceratopsids, unlike large modern herbivores such as large bovids or rhinoceroces (in which either the reverse is the case or the diaphysis is circular) (Fig. 4). The deltopectoral crest is quite large in psittacosaurids, larger still in protoceratopsids, very large in ceratopsids, and is truly gigantic in large chasmosaurines such as Triceratops and Torosaurus. The medial tuberosity is usually quite pronounced, especially among larger forms. The distal expansion is moderate in psittacosaurids and protoceratopsids but large in ceratopsids, in which the epicondyles are usually prominent. The distal con-

dyles are rounded and well defined in uncrushed specimens and set at an angle compared to the long axis of the diaphysis, indicating permanent elbow flexure. The ulna is quite stout in smaller forms and massive in large ceratopsids. The diaphysis is somewhat triangular and tapers distally. The olecranon process is large in smaller forms and gigantic among large chasmosaurines. The radius is shorter than the ulna, rather straight, has moderate proximal and distal expansions in smaller taxa, and has large expansions in large taxa. The proximal epiphysis is usually somewhat medially inclined, whereas the distal epiphysis is more planar. There has been dispute about the forelimb posture of ceratopsids for a long time. The morphology indicates that the elbow was directed posteriorly as in large modern herbivores, and trackways show an erect posture similar to large mammals, with manus prints only slightly wider than the hindlimb prints (Lockley and Hunt, 1995). The manus prints probably had to be slightly more lateral because the anterior ribcage was so wide. The carpus includes three unfused proximal carpals, identified as ulnar, intermedium, and radial, and a smaller distal carpal in psittacosaurids. Ceratopsids also have four carpals, whereas Protoceratops had five. Metacarpals I–III in psittacosaurs are well developed, but IV is reduced and V is vestigial. The phalangeal formula is 2-3-41-0. This is a problem because psittacosaurids are regarded as the sister group of the Neoceratopsia,

FIGURE 4 Propodial robusticity in ceratopsian dinosaurs (n ⫽ 25). 䊉, ceratopsians; 䊊, elephants; 䉭, rhinoceroses; 䊐, bovids.

Forelimbs and Hands which do not have these manus reductions, suggesting reversals will have to be accepted. In neoceratopsids, the hand is stout and wide, and metacarpals IV and V are somewhat reduced, but not to the same extent as in psittacosaurids. The metacarpals become progressively larger as linear dimensions increase. The phalanges are short and stout with cotylar–condylar articulating facets, and the terminal claws are broad, flat, and hoof-like in ceratopsids but are less pronounced in protoceratopsids. The phalangeal formula is 2-3-4-3-1/2.

References Currie, P. J. (1989). Dinosaur footprints of Western Canada. In Dinosaur Tracks and Traces (D. Gillette and M. G. Lockley, Eds.), pp. 293–300. Cambridge Univ. Press, New York.

253 Santa Luca, A. P. (1980). The postcranial skeleton of Heterodontosaurus tucki from the Stormberg of South Africa. Ann. South Africa Museum 79, 159–211. Sereno, P. C. (1986). Phylogeny of the bird-hipped dinosaurs (order Ornitischia). Natl. Geogr. Soc. Res. 2, 234–256. Sereno, P. C. (1993). The pectoral girdle and forelimb of the basal theropod Herrerasaurus ischigualastensis. J. Vertebr. Palaeontol. 13(4), 425–450. Upchurch, P. (1994). Manus claw function in sauropod dinosaurs. Gaia 10, 161–171. Upchurch, P. (1995). The evolutionary history of sauropod dinosaurs. Philos. Trans. R. Soc. London Ser. B 349(1330), 365–390. Weishampel, D. B., and Witmer, L. M. (1990). Heterodontosauridae. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 486–497. Univ. of California Press, Berkeley. Young, C. C. (1941). A complete osteology of Lufengosaurus huenei (gen. et sp. nov.) from Lufeng, Yunnan, China. Palaeontol. Sinica (N.S.) Ser. C 7, 1–53.

Forster, C. A. (1990). The postcranial skeleton of the ornithopod dinosaur Tenontosaurus tilletti. J. Vertebr. Palaeontol. 10(3), 273–294. Galton, P. M. (1981). Dryosaurus, a hypsilophodontid dinosaur from the Upper Jurassic of North America and Africa. Postcranial skeleton. Palaeontol. Z. 55(3/4), 271–312. Galton, P. M. (1990). Basal Sauropodomorpha– Prosauropoda. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 320–344. Univ. of California Press, Berkeley. Gilmore, C. W. (1914). Osteology of the armoured Dinosauria in the United States National Museum, with special reference to the genus Stegosaurus. Bull. U.S. Natl. Museum 89, 1–136. Lockley, M. G., and Hunt, A. P. (1995). Ceratopsid tracks and associated ichnofauna from the Laramie Formation (Upper Cretaceous: Maastrichtian) of Colorado. J. Vertebr. Palaeontol. 15(3), 592–614. Madsen, J. H., Jr. (1976). Allosaurus fragilis: A revised osteology. Bull. Utah Geol. Mineral. Survey 109, 1–163. Ostrom, J. H. (1969). Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Peabody Museum Nat. History Bull. 30, 1–165. Paul, G. S. (1987). The science and art of restoring the life appearance of dinosaurs and their relatives: A rigorous how-to guide. In Dinosaurs Past and Present (S. J. Czerkas and E. C. Olson, Eds.),Vol. II, pp. 5–49. Natural History Museum of Los Angeles County, Los Angeles. Rowe, T., and Gauthier, J. (1990). Ceratosauria. In The Dinosauria (D. B. Weishampel, P. Dodson, and H. Osmo´lska, Eds.), pp. 151–168. Univ. of California Press, Berkeley.

Fort Peck Power Project, Montana, USA see MUSEUMS



Fort Worth Museum of Science and History, Texas, USA see MUSEUMS




Freilichtmuseum, Germany see MUSEUMS





Fruita Paleontological Area JAMES I. KIRKLAND Dinamation International Society Fruita, Colorado, USA

Created in 1976, the Fruita Paleontological Area (FPA) was the first area set aside in the United States by the Bureau of Land Management as a specially protected management area based solely on its fossil resources. The FPA has been known locally for 100 years for its fossil dinosaur remains. In 1975, George Callison, Jim Clark, and Mark Norell, then of California State University at Long Beach, discovered significant occurrences of small vertebrate remains (Callison, 1987). Their discovery was followed by the discovery of the second known Ceratosaurus skeleton by Lance Ericksen of the Museum of Western Colorado. Over the ensuing years a diverse, well-preserved vertebrate fauna has been identified from the basal Brushy Basin Member of the Morrison Formation (Upper Jurassic, Kimmeridgian) at the FPA (Fig. 1). Taking advantage of the remarkable three-dimensional outcrop at the FPA, the principles of highresolution event stratigraphic methodologies have been used in the analysis of the periodic flood deposits (crevasse splays) to facilitate a detailed analysis of the interrelationships of individual fossil sites to each other and to specific local environments. These local environments include bank-controlled, low-sinuosity, gravelly river channel sandstones (ribbon sandstones), clay-dominated levee deposits, wellsorted sandstone proximal crevasse splay complexes, increasingly distal graded crevasse splays superimposed by carbonate-rich paleosoils, permanent abandoned channel ponds and/or springs, and temporary alkaline ponds characterized by barite nodules. Depositional base level was controlled by the distribution of active and abandoned channel–levee complexes (Kirkland et al., 1990). Within this framework, a taphonomic basis for the

Fruita Paleontological Area distribution of fossil remains can be discerned. Virtually all ribbon sandstones extend across the FPA, with flow to the north-northeast to the east-northeast, and preserve isolated dinosaur bones. One ribbon sandstone (ceratosaur channel) enters the area from the southeast curves due west before turning north and finally exiting the area to the northeast. The ceratosaur channel is very rich in disarticulated to at least semiarticulated dinosaur remains: Ceratosaurus, Allosaurus, Camarasaurus, Apatosaurus, Stegosaurus, and Dryosaurus have been identified to date. In addition, goniopholid crocodilians, turtles, and unionid clams occur in these sandstones, indicating permanent water flow. Sediments indicating abandoned channel ponds and/or springs often overlie the ribbon sandstones. These environments are most readily identified by the abundance of carbonaceous plant materials preserved within them. They are also characterized by variable amounts of dinosaur bone, theropod teeth, and aquatic snails (mainly Viviparous reesidei). Isolated splays within these ponds sometimes preserve abundant unionid clams, which suggests that the abandoned channels were reoccupied during flooding events. At the top of the abandoned channel pond sequences, conchostracans become common with the loss of carbonaceous plant material. At one pond an algal limestone preserves several intact examples of the fish Hullettia hawesi with abundant silicified Viviparous reesidei and serves as the type locality for both taxa (Yen, 1952; Kirkland, 1997). Proximal crevasse splay complexes are best developed at bends in the ceratosaur channel and are identified as being the source of several fossiliferous distal crevasse splays. In one case, theropod tracks were found in a proximal crevasse splay complex. Rooted gray calcareous splay slurries are associated with disarticulated and associated skeletons of microvertebrates, whereas more distal splays associated with oxidized, more mature soils lack vertebrate remains and are more strongly rooted. Better articulated microvertebrate remains have been found in association with laminated smectitic claystones identified as temporary alkaline pond deposits. These microvertebrates represent a mixed transported assemblage; the terrestrial elements tend to be better articulated, often with complete skulls and articulated jaws. The terrestrial taxa include one species of very small theropod dinosaur, one species of very small ornithischian di-

Fruita Paleontological Area


FIGURE 1 The Fruita Paleontological Area in Colorado is famous for Jurassic dinosaurs preserved in the Morrison Formation. James Kirkland, pictured above, has explored these sediments in search of dinosaur fossils. (Photo by Franc¸ois Gohier.)

nosaur (cf. Echinodon) (Callison and Quimby, 1984), common small terrestrial mesosuchian crocodilians, a small pterosaur, three species of rhynchocephalians, four species of lizards, the earliest snake?, and at least eight species of mammals (Callison, 1987). Small eggshell fragments have also been recovered (Hirsch, 1994). The aquatic elements include turtle, frog, lungfish, actinopterygians, crawfish, and the remains of caddisfly cases and were most likely transported from the river during flooding episodes. The alkaline ponds often preserve common conchostracans. Most of these taxa are new and are being described. An accumulation of juvenile and hatchling-sized Dryosaurus bones and small egg fragments and associated small terrestrial crocodile bones is associated with a horizon of small calcareous nodules in a soil horizon. It has been interpreted as a site close to

a Dryosaurus nesting site, where small mesosuchian crocodilians were preying on eggs and/or hatchlings (Kirkland, 1994). This well-documented interplay of local environments and significant fossil localities makes the FPA an excellent natural laboratory for the study of Upper Jurassic faunas, floras, sedimentology, taphonomy, ecology, and climatology.

See also the following related entry: MORRISON FORMATION

References Callison, G. (1987). Fruita: A place for wee fossils: In Paleontology and Geology of the Dinosaur Triangle (W. R. Averett, Ed.), pp. 91–96. Museum of Western Colorado, Grand Junction, CO.


Fruitland Formation

Callison, G., and Quimby, H. M. (1984). Tiny dinosaurs: Are they fully grown. Journal of Vertebrate Paleontology 3(4), 200–209. Hirsch, K. F. (1994). Upper Jurassic eggshells from the Western Interior of North America: In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 137–150. Cambridge Univ. Press, Cambridge, UK. Kirkland, J. I. (1994) Predation of dinosaur nests by terrestrial crocodilians. In Dinosaur Eggs and Babies (K. Carpenter, K. F. Hirsch, and J. R. Horner, Eds.), pp. 124– 133. Cambridge Univ. Press, Cambridge, UK. Kirkland, J. I. (1997). Morrison fishes. In The Morrison Formation: An interdisciplinary study (K. Carpenter, D. Chure, and J. I. Kirkland, Eds.), in press. Geological Society of America Special Paper. Kirkland, J. I., Mantzios, C., Rasmussen, T. E., and Callison, G. (1990) Taphonomy and environments; Fruita Paleont. Resource Area, Upper Jurassic, Morrison Formation, W. Colorado: J. Vertebr. Paleontol. 10(Suppl. to No. 3), 31A. Yen, T.-C. (1952). Molluscan fauna of the Morrison Formation. United States Geological Survey Professional Paper No. 233-B, pp. 21–51.

Fruitland Formation MICHAEL J. RYAN Royal Tyrrell Museum of Palaeontology Drumheller, Alberta, Canada

The San Juan Basin of northeastern New Mexico encompasses strata of Late Cretaceous, Paleocene, and early Eocene age and contains rich vertebrate fossil assemblages. Excellent reviews of this region include Hunt and Lucas (1993) and Lucas and Williamson (1993). Throughout the Cretaceous, New Mexico was located at the western margin of an epicontinental seaway that stretched from the Gulf of Mexico to the Arctic Ocean. The late Campanian (Judithian) Fruitland Formation (Fig. 1) was named by Bauer (1916) and is at least 170 m thick. It is composed of a series of complexly interbedded channel sandstones, carbonaceous gray shales and siltstones, and coal beds that can reach thicknesses of 9 m and extend laterally over several kilometers. To the northwest, the formation was deposited as the landward facies of a marshy, delta complex that was poorly drained and subjected to frequent flooding. To the southeast,

FIGURE 1 Chronology, lithostratigraphy, vertebrate fauna, and general depositional environment of the late Campanian and early Maastrichtian of the San Juan Basin, New Mexico (modified from Lucas and Williamson, 1993).

the formation displays a barrier shoreline signature. Paleochannels are lenticular, up to 6 m thick and represent high-sinuosity, mostly unbraided river channels with numerous vertebrate remains. Paleoflow was to the northeast at right angles to the shoreline that trended NW–SE. Unionid bivalves and nonmarine gastropods are common in the Fruitland indicative of the freshwater channel system of the local environment. Plant material is common throughout the Fossil Forest region of the formation with at least eight successive fossil forest horizons. These horizons preserve more than 400 in situ tree stumps and logs. In the north, the easily recognizable, persistent brown-colored ribbon sandstones mark the top of the formation, whereas in the south the stratigraphically highest, thick (⬎1 m ) coal can be used as the formational contact (Hunt, 1992). The climate of the Fruitland was warm, humid, and seasonal based on the associated fossil leaf flora and fossil treering evidence. The upper Fruitland and lowermost Bisti Member of the overlying Kirtland Formation constitute one of two stratigraphic intervals containing vertebrate fossils in the Fruitland–Kirtland stratigraphic sequence, the other being the Naashoibito Member of the Kirtland. In the Fruitland Formation, fossils are

Fruitland Formation equally common in both channel and interchannel environments. The Fruitland contains a diverse vertebrate assemblage including 10 chondrichthyans, 13 osteichthyans, two anurans, three urodeles, 10 chelonians, two teiid and two anguid lizards, one aniilid snake, five crocodilians, at least 15 dinosaurs, and more than three dozen mammals. This fauna comes primarily from the upper two-thirds of the formation in an outcrop belt between the San Juan River and the Kimbeto Wash. The dinosaurs are as follows: Saurischia: Tyrannosauridae (?Albertosaurus libratus and ?Albertosaurus sp., gen. et. sp. nov.), Ornithomimidae (cf. Ornithomimus sp. and Ornithomimidae indet.), Dromaeosauridae indet., Troodontidae indet., Titanosauridae (undescribed n. gen et sp.); Ornithischia: Hypsilophodontidae (?Thescelosaurus sp.), Lambeosauridae (Parasaurolophus crystocristatus and ?Corythosaurus sp.), Hadrosauridae indet., Hadrosauroidea footprints, Nodosauridae indet., Ankylosauridae indet., Pachycephalosauridae indet., Ceratopsidae [Pentaceratops sternbergii, Ceratopsidae indet. (⫽ cf. Chasmosaurus Gilmore 1935)]. Microvertebrate localities are more common in the Fruitland than in the overlying Kirkland and account for almost all fish, amphibian, and lower reptiles recovered to date. Dinosaur specimens are largely made up of disarticulated and isolated material (limbs, ribs, and vertebrae) with partial skeletons and skulls being rare. Hydrodynamic sorting appears to be common. Theropods are poorly known from isolated material. The ankylosaurs are known from scutes. The hadrosaurs are known from partial skeletons with the Hadrosauridae known from a juvenile dentary (Wolberg, 1993). Eggshell fragments and nests (Wolberg, 1992) suggest that dinosaurian nesting may have occurred in the area.

See also the following related entries: CRETACEOUS PERIOD ● KIRTLAND FORMATION


References Bauer, C. M. (1916). Contributions to the geology and paleontology of San Juan County, New Mexico. 1. Stratigraphy of a part of the Chaco River Valley. United States Geological Survey, Professional Paper No. 98P, pp. 271–278. Gilmore, C. W. (1935). On the Reptilia of the Kirtland Formation of New Mexico, with descriptions of new species of fossil turtles. Proc. United States Natl. Museum 83, 159–188. Hunt, A. P., and Lucas, S. G. (1993). Cretaceous vertebrates of New Mexico. In Vertebrate Paleontology in New Mexico (S. G. Lucas and J. Zidek, Eds.), Bull. 2, pp. 77– 91. New Mexico Museum of Natural History. Lucas, S. G., and Williamson, T. E. (1993). Late Cretaceous to Early Eocene vertebrate biostratigraphy and biochronology of the San Juan Basin. In Vertebrate Paleontology in New Mexico (S. G. Lucas and J. Zidek, Eds.), Bull. 2, pp. 92–104. New Mexico Museum of Natural History. Wolberg, D. L. (1992). Dinosaur nesting near the Cretaceous (Campanian–Maastrictian) marine shoreline; Evidence from the first infant dinosaur discovered in the San Juan Basin, of New Mexico. Geol. Soc. Am. 24(7), 270. [Abstract] Wolberg, D. L. (1993). A Juvenile hadrosaurid from New Mexico. J. Vertebr. Palaeontol. 13, 367–369.

Fryxell Geology Museum, Illinois, USA see MUSEUMS



Fukui Prefectural Museum, Japan see MUSEUMS



Functional Morphology XIAO-CHUN WU ANTHONY P. RUSSELL University of Calgary Calgary, Alberta, Canada


unction in a biological context may generally be defined as the use or action of structures (such as bones, dermal armor, muscles, etc.) of organisms. The study of functional morphology attempts to explain the diversity of structure and function exhibited by organisms by proposing theories of how structures work and how they have come to be. The limb skeletons of terrestrial tetrapods (such as dinosaurs) function to provide a rigid frame on which muscles can act. The actions of particular subsets of these muscles (which are controlled by the nervous system) on the elements of the legs enable these animals to effect locomotion. The wings of birds can be studied from the viewpoint of how they are involved in flight, but other attributes of their morphology and the basis of their design can be understood by realizing that such wings have evolved from the forelimbs of advanced theropod dinosaurs. It has long been accepted that function and form (structure) are intimately related, despite little consensus on just what this relationship is. Much (current and previous) work on functional morphology is based on the assumption that form and function are intimately correlated and, thus, biological function can be deduced from a study of form or structure. Recently, however, some functional morphologists (such as Lauder, 1995) have postulated that form and function are related only at a very general level, and that they are often not tightly linked. In contrast to neontology (the study of extant taxa), studies of the functional morphology of extinct organisms, and hence the DINOSAURIA, are often limited to deductions based on fossilized hard tissues (bones, teeth, dermal armor, eggs, etc.) or, less often, on traces of animal activities (impressions of external structures, trackways, etc.). However, functional reconstructions of dinosaur morphology can help us enrich our understanding of their paleobiology and, sometimes, reveal some of the subtleties of their evolutionary history (Weishampel, 1995). Methods currently

employed in studies of dinosaur functional morphology may be referred to three categories: the nonhistorical approach, the historical (phylogenetic) approach, and the synthetic approach.

Nonhistorical Approach The nonhistorical approach has a long history and thus numerically dominates studies of functional morphology in VERTEBRATE PALEONTOLOGY. This method takes as its base aspects of biomechanics that are theoretically well substantiated and emphasizes analogies (with beams, machines, or extant animals as models or modern analogies, etc.) as universals from which to assess their functional relevance to extinct animals. Studies using mechanical analogies are inherently nonhistorical, whereas studies using extant animals as models or modern analogies usually do not consider the phylogenetic relationships among the taxa of interest. For any particular model to be applied, it has to be established on the basis of explicit, well-understood parameters that come from engineering, biology, or other related sources. There are many variants of the nonhistorical approach. Of them, the modeling method based on machine analogies is the most often used in studies of the functional morphology of extinct vertebrates. These are directed toward the understanding of the operation of a particular anatomical system. Other variants include experiments, graphic representation and mathematical computation, and computer-based simulations (see Weishampel, 1995). The following are examples of the most commonly used nonhistorical approaches in the study of the functional morphology of the Dinosauria. Modeling Lever Devices The dinosaurian jaw, as it does in all vertebrates, operates as a