Theagarten Lingham-Soliar Origin and Evolution

Home , Palaeonisciformes

Theagarten Lingham-Soliar The Vertebrate Integument Volume 1 Origin and Evolution The Vertebrate Integument Volume 1 The arid and barren landscape walled by plateaus, represents the hostile environment of the Cape Karoo in South Africa (a region pioneered by the author’s maternal family). The rocks of the Karoo System were deposited between the Carboniferous (360–286 million years ago) and Early Jurassic (208–187 million years ago) and are known for some of the most important ﬁnds of mammal-like reptiles in the world (Chap. 8). Photo, B Lingham (circa 1922, family archive) Theagarten Lingham-Soliar

The Vertebrate Integument Volume 1

Origin and Evolution

123 Theagarten Lingham-Soliar Life Sciences University of KwaZulu-Natal Durban South Africa

Present address Environmental Sciences Nelson Mandela Metropolitan University Port Elizabeth South Africa

ISBN 978-3-642-53747-9 ISBN 978-3-642-53748-6 (eBook) DOI 10.1007/978-3-642-53748-6 Springer Heidelberg New York Dordrecht London

Library of Congress Control Number: 2013957128

Ó Springer-Verlag Berlin Heidelberg 2014 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com) Dedicated to my mother LL-S, Who gave us dignity amidst the undigniﬁed world of pre-1994 South Africa Preface

The need for the present book became increasingly apparent as I worked on the integument and its structures over the past 14–15 years, first in fossil groups and later in extant vertebrates. Even among living vertebrates, in particular marine groups, it was only in the last quarter of the twentieth century that we were beginning to understand the structure of the dermis and its enormous biomechanical contribution to locomotion. Coming right up to 2013 my colleagues and I were able to demonstrate how little the biomechanical microstructure of the feather was known given that such knowledge would have important implications for bird flight. Yet we need to go back over 450 million years to the very beginning of vertebrate evolution to put our understanding of the vertebrate integument today in proper context. I was fortunate that my undergraduate lecturer and later Ph.D. supervisor, Beverly Halstead, at the University of Reading in the UK, was one of the key workers involved in research on vertebrate origins and that his seminal contributions on the extinct jawless vertebrates, on the origin of bone and on the origin of teeth formed a major part of his zoology lectures. This background characterises the direction and function of the book, which is aimed at both undergraduate and postgraduate students. The topic of the book at times may appear complex but the profusely illustrated text with diagrams, photos and some of my own artwork may help make the book easy to use and understand. While many of the illustrations are original I owe a debt to many authors cited whose illustrations have been used to a greater or lesser extent, unmodified or modified (if there are errors, they are mine). Fossilised integument and integumental structures although scarce have been preserved in vertebrate fossils, perhaps nowhere more so than in the famous Liaoning region in China. Although much of the integumental material on dinosaurs from China have, in recent years, been concerned with possible feather origins or feather-like structures, I hope the book will encourage investigations on a broader level because there is, I am sure, much material that may be neglected with respect to knowledge on the integument in general (i.e. not connected with feather origins). For instance, I have demonstrated this in the dermis of the dinosaur Psittacosaurus from Liaoning and with a colleague, Prof. Gerhard Plowdowski, on colour in the epidermis of another specimen, also from Liaoning. Feather origins are important, hence there is no suggestion that the search needs to be downplayed.

vii viii Preface

It seems safe to say that despite the importance of the integument in the lives of vertebrates and its changing morphological and functional role over time there is no other book that deals with the topic speciﬁcally, in particular its mechano- structural properties. Volume 2 will discuss the biomechanics of the integument in much greater detail. Acknowledgments

There are many individuals who have made my science possible, yet none more than my mother. My mother died unexpectedly before I received my Ph.D. She had an exciting early life in the Karoo—family racehorses, driving her parents Rolls Royce aged 14 and achieving recognition as a talented classical pianist while still in her teens. It all ended with the ‘legalized’ expropriation of our family’s assets and wealth during the apartheid years in South Africa, when I was just a boy. Despite the unimaginable hardships of those years, she single-handedly filled our lives with culture, kindness and hope. My brother Sagaren Soliar is 8 years older than I but thankfully he failed to recognise that when I was growing up. An avid reader, he would shove books at me to read. At age about 8, most memorably was Mika Waltari’s Sinuhe the Egyptian, regarded by authorities today as among the greatest historical novels written. It started my passion for Egyptology to the extent that when a few years later my brother raved about a book on hominid origins by the Leakey’s, I scathingly said, verbatim, that I would rather search for the tombs of pharaohs, even with the prospect of failure, than sift sand in a barren landscape only to find a few dried-up, old bones after 30 years—famous last words. His passion for knowledge and vibrant exchange of ideas had the most profound effect on my development and perceptions. My early career as an industrial research chemist in the UK, I owe to one man, Les Leedham, head of the research laboratory. He fostered my ability and arranged day-release to enable me to continue my part-time university studies (in chemistry). He was a giant of a man both in stature (over 6 feet 6 inches tall) and in human decency and confirmed why as an exile from South Africa I was proud to have chosen England as my new home. Several years later I met Beverly Halstead, who lectured the first year zoology undergraduate course at Reading University. I was not the keenest of students, preferring the fine pubs in Berkshire and interesting company to diligent course studies. However, Beverly Halstead’s passion for paleontology got through and in spite of myself I gained first class grades in it. Kindly, he recommended my study for a Ph.D. to the head of department with the words that he would rather chose me than someone with a double-first from Cambridge. Beverly Halstead died in a traffic accident before I could complete my Ph.D.

ix x Acknowledgments

During my Ph.D. I met Richard Estes (San Diego State University), one of the top herpetologists in the world who was on a year’s sabbatical at the Natural History Museum, London. We liked each other instantly and he had a major effect on my research. Before he flew back to the USA we agreed that he would for- malize himself as my second supervisor. I would never see him again. Shortly after returning home, Richard was diagnosed with inoperable intestinal cancer and he died within the year. Tom Kemp at Oxford University took over as bona fide supervisor. For his help I am grateful but also for his bravery, given that I had already got through two supervisors within 4 months of each other. Wolf-Ernst Reif was my host at Tubingen University on two successive Royal Society Post-doctoral Fellowships. He had a significant effect on my research on marine vertebrates, in particular on sharks, in which he was a world expert. I benefitted too from his extensive knowledge of philosophy. Alexi Yu Rozanov, a principled and generous man, was my host over 5 years at the Paleontological Institute of the Russian Academy of Sciences. He was extremely helpful to my research and instrumental in my appointment during this time as editor of Paleontological Journal. Alan Feduccia (University of North Carolina) has shown an interest in my work for 15 years now, first writing to me when I was in Russia. His distinguished career in the field of bird origins and evolution has been inspirational and is crowned by the honesty and scientific ethics with which he deals with opposing views in this controversial arena. My work on birds would have been the poorer were it not for discussions with Walter Bock (Columbia University), Storrs Olson (Smithsonian) and John Ruben (Oregon State University), to whom special mention must be made. There are many people over the years who provided valuable discussion and reviews on my research and others access to collections and material. They have all in various ways contributed to the making of this book. In no particular order they are Aggusto Azzaroli (University of Florence), Dirk Nolf, Annie Dhondt, Pierre Bultynck and Paur Sartenaer (all, Institute Royale, Belgium), Telles Antunes (New University of Lisbon), Erik Buffetaut (University of Paris), Sandra Chapman, Cyril Walker (both Natural History Museum, London), Neil Clark (Hunterian Musem, Glasgow), Arthur Cruickshank, (University of Leicester), Solweig Stuenes (Paleontological Museum, Uppsala), Gerhard Plodowski (Senckenberg Museum, Frankfurt), Joop van Veen (Teyler’s Museum, Harlem, The Netherlands), John Jagt (Natural History, Museum, Maastricht, The Netherlands), Sam Mukaratirwa (UKZN), Geremy Cliff (Natal Sharks Board), Tim and Trish Broderick (Jeremy Prince and Associates, Zimbabwe); Philipp Motta (University of Florida), Mike Everhart (Sternberg Museum), Chris McGowan (University of Toronto), Mary Hebrank (Duke University), Gerald Kooyman (Scripps Institute), Phillipe Janvier (Muséum d’Histoire Naturelle, URA 12 CNRS), Michael Wuttke (Referat Erdgeschichte, Generaldirektion Kulturelles Erbe Rheinland-Pfalz), Achim G. Reisdorf (Universität Basel Geologisch-Paläontologisches Institut). Contents

1 Introduction ...... 1 References ...... 8

2 The First Vertebrates, Jawless Fishes, the Agnathans ...... 11 2.1 Ostracoderms ...... 11 2.2 Marine or Freshwater Origins? ...... 13 2.3 Feeding ...... 16 2.4 Bone, a Chemical Store ...... 16 2.5 Heterostraci ...... 19 2.6 Osteostraci ...... 23 References ...... 30

3 The Earliest Jawed Vertebrates, the Gnathostomes ...... 33 3.1 The Evolution of Jaws ...... 33 3.2 The Evolution of Paired Fins ...... 37 3.3 Placoderms, Dominant Vertebrates of the Devonian...... 37 3.4 Evolutionary Significance of the Oldest Vertebrates: The Origin of Bone ...... 42 3.4.1 Aspidin ...... 43 3.4.2 Dentine ...... 46 3.5 The Origin of Teeth...... 49 3.5.1 Genes and the Origin of Teeth—Dermal or Oral...... 50 3.6 Evolutionary Comments ...... 54 References ...... 56

4 Evolution of Modern Fishes: Critical Biological Innovations ..... 59 4.1 Chondrichthyes, Advanced Cartilaginous Fishes ...... 61 4.1.1 Subclass: Elasmobranchii ...... 61 4.1.2 Subclass: Holocephali ...... 65 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes ...... 66 4.2.1 Class: Acanthodii ...... 66 4.2.2 Class: Actinopterygii ...... 68 4.2.3 Subclass: Cladista ...... 69 4.2.4 Subclass: Chondrostei ...... 69

xi xii Contents

4.2.5 Subclass: Neopterygii ...... 72 4.2.6 Class: Sarcopterygii...... 74 4.3 Scales: Organization of the Integumentary Skeleton in Gnathostomes ...... 77 4.4 Scales and Swimming Hydrodynamics ...... 81 4.5 Feeding in Fishes ...... 84 4.6 Reception of External Mechanical Stimuli in Fishes ...... 85 4.6.1 Lateral Line System ...... 87 4.6.2 Electroreception ...... 88 4.6.3 Function of Electroreception...... 93 4.6.4 Electrocommunication ...... 94 4.7 Hydrodynamics and Buoyancy in Fishes ...... 94 References ...... 97

5 Tetrapods and the Invasion of Land ...... 99 5.1 Global Environmental Changes ...... 99 5.2 Tetrapods and Cladistics...... 100 5.3 Paraphyletic Origins...... 105 5.4 Descendents of the Sarcopterygian Fishes and the Move onto Land ...... 108 5.5 The First Tetrapods ...... 114 5.6 The Evolution of Terrestrial Vertebrate Locomotion ...... 115 5.7 From Fins to Limbs ...... 117 5.8 The Problems of Breathing on Land ...... 123 References ...... 126

6 Crucial Vertebrate Innovations ...... 129 6.1 Best Foot Forward ...... 129 6.2 The Amniotic Egg ...... 131 6.3 The Reptile Integument ...... 137 6.4 Scale Types ...... 138 6.5 Temporal Openings and Classification of Reptiles...... 146 References ...... 149

7 Dinosaur Integument ...... 153 7.1 More than Skin Deep...... 153 7.2 Organic Analysis of the Dinosaur Integument ...... 167 7.3 Dinosaur Eggs and Embryos ...... 169 7.4 The Dermis in Dinosaurs ...... 176 7.5 Dinosaur Skin Modified as a Sail in Temperature Control? . . . . 180 7.6 Dinosaur Tracks and Traces ...... 181 7.7 Pterosaurs...... 189 References ...... 189 Contents xiii

8 Mammal-Like Reptiles ...... 193 8.1 In the Shadow of the Dinosaurs ...... 193 8.2 Pelycosaurs...... 194 8.3 Therapsids ...... 197 8.4 The Evolution of Hair ...... 201 8.5 Cynodonts and the Origin of Modern-Day Mammals ...... 205 8.6 The First Mammals ...... 207 8.7 Mammals Emerge into the Light ...... 210 References ...... 218

9 Reptiles Return to the Sea ...... 221 9.1 Problems of Developing Fresh Sea-Legs ...... 221 9.2 A Serpent Threaded Through a Turtle ...... 222 9.3 The Fastest of Them All ...... 233 9.4 Last of the Great Marine Reptiles ...... 236 9.4.1 Disease, Injuries, and Bone Repair in Mosasaurus Hoffmanni and Their Wider Implications...... 242 9.4.2 Mosasaur Integument...... 248 9.4.3 Extinction ...... 251 References ...... 252

Index ...... 255 Chapter 1 Introduction

In his book Darwinism, Alfred Russell Wallace (1899) minimized the part that chance played in evolution stating, ‘‘There may be something left to chance, but on the whole the fittest will survive.’’ Essentially he based this belief on variability in the characters of species that may give one individual an advantage over another in the struggle for survival. In certain respects this philosophy holds true but it is just one factor in the great process of evolution, which Dobzhansky (1962) commenting on the inheritance of advantageous characters referred to as the ‘‘dice of destiny.’’ The Darwinian concept of fitness in the idea of the ‘‘survival of the fittest’’ as the basis of organic evolution is a tautology—those organisms that survive by definition must be the fittest, hence the fittest survive. Implicit is a kind of design. We see during the course of the book that the more intensely the history of life on earth is investigated the more apparent it becomes that many vertebrates around today, including ourselves, are here by a good deal of chance than simply ‘‘good genes’’ as equally those that have become extinct may not necessarily have been victims of ‘‘bad genes.’’ The early evolution of the integument marked perhaps the most important time in vertebrate evolution as a whole. In most modern-day vertebrates the skin or integument is an organ that forms the outermost covering, which in many is the first line of protection or shield from mechanical force and environmental stresses for the underlying tissues. We also know that the skin by its variety of colour and patterns either via scales, hair, or feathers performs vital roles in camouflage and sexual behavior and in functions connected with the animal’s physiology such as retaining body heat. However, about 450 million years ago during the Ordovician (Fig. 1.1) the first vertebrates, small fish-like and jawless creatures, had a very different kind of integument. The integument was comprised of bony plates which gave these odd animals their collective popular name, ‘‘ostracoderms.’’ Because the armor formed the main barrier between the animal and its environment it can be regarded as the skin, but even more so when we realize that this outer bony layer is in fact dermal in origin (just as the scales in modern fishes). Despite their apparent simplicity, ostracoderms during the 100 million years of their existence contributed a crucial part to vertebrate history. In this time,

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 1 DOI: 10.1007/978-3-642-53748-6_1, Ó Springer-Verlag Berlin Heidelberg 2014 2 1 Introduction

Fig. 1.1 A geologic time scale (excluding the Precambrian). Modified from Enc. Britannica 1 Introduction 3 over 600 species of ostracoderms evolved and their relationships with fishes possessing jaws have been a critical question that paleontologists have been trying to answer to this day. The problems necessitated an understanding of the structure and function of the bony integument in these long extinct vertebrates, which are intrinsically tied to major steps in vertebrate evolution. For almost a century, these problems dominated the work of some of the greatest names in paleontology from the US, Canada, UK, Western Europe (notably Sweden and France), and the former Soviet Union. So important is the evolution of jaws that modern vertebrates are classified in two major groups, the agnathans (meaning ‘‘without jaws’’) and the gnathostomes (meaning ‘‘jaw-mouths’’). One of the first problems facing the early workers was in understanding the function of the bony armor in the agnathans. Some workers suggested defense, perhaps the most obvious explanation, others that it may have prevented water loss and yet others that it served as a store for calcium and phosphates. The latter argument was highly interesting for other reasons too. It is believed that the early vertebrate skeleton was largely unossified and the idea of a calcium and phosphate store in the dermis saw later workers shifting that explanation to the question of how ossification of the vertebral skeleton might have occurred in the early vertebrates. The careful appraisal of the anatomy of ostracoderms involved fine structural analysis, e.g., in the cephalic head shields of cephalaspids, which frequently were no larger than the nail on one’s little finger (Fig. 1.2). In the 1920s, the Swedish paleontologist Erik Stensio advanced many techniques including use of fine nee- dles and special refracting fluids, as well as adapting a technique of serial sections, invented by BS Sollas, to elucidate the fine structure of these animals. It laid the groundwork for investigations of these complex early vertebrates and contributed to fine detailed study in paleontology generally. Among the early problems facing workers was the question of feeding in these primitive vertebrates, which clearly would have been a problem given the absence of teeth and jaws. The general view is that agnathans were passive filter feeders lying on the mud of the sea bed. More recent findings show this view may fall short of the full story. Purnell (2001) proposed that the mouths of some of these jawless vertebrates were lined with dermal denticles that allowed a more active feeding behavior than previously thought namely microphagous filter feeding and that this was an advanced form of feeding compared to predation. The idea of being less fit and being outcompeted may in many instances be somewhat easy and trite explanations for complex evolutionary processes. Even what seemed like a foregone conclusion with respect to the extinction of the early jawless vertebrates, i.e., that they were outcompeted by jawed vertebrates, seems now questionable. For instance, Anderson et al. (2011) show that such explanations are anecdotal and lack scientific basis and that there may be other factors at work. It must be remembered too that the absence of jaws and teeth in vertebrates persisted quite remarkably for over a 100 million years, a reasonable assessment for success. 4 1 Introduction

Fig. 1.2 A photograph of a cephalic head shield of a cephalaspid (Kiearaspis). Lower Devonian, Spitsbergen. Ventral view. Endoskeletal component of shield prepared by Stensio to show cavities for brain nasohypophysial complex, eyes and auditory organ together with canals in its interior. From Stensio (1927)

Given the long history of jawless vertebrates, the inevitable important questions were, when and how did jaws evolve? This was closely tied to the debate over the origin of the vertebrate skeleton and centered around the issue of which group was the first to exhibit evidence of skeletonization and/or mineralization. It was a can of worms that raised further questions of when did the mineralized integument during the course of evolution become partitioned into the dermatocranium, into skeletal components of the pectoral apparatus, and into gill arches. The latter would be vital for the development of jaws, which would in turn pave the ground for the evolution of teeth and tooth-like elements of the oral and pharyngeal cavities. The complexity of the questions has meant that there is still no consensus of agreement on some of these critical issues. Among the most compelling evidence in recent years is that the jawless placoderms (literally ‘‘plate-skin’’) may have been the first vertebrates to show some of these characteristics with respect to the evolution of jaws (Donoghue and Sansom 2002). On the other hand, classic studies on the evolution of teeth have for some time been rather clear—that the skin denticles (scales) evolved first and merged inside the oral cavity to form teeth (the ‘‘outside-in’’ hypothesis). Somewhat surprisingly, we see that this hypothesis is not as safe as we thought and has recently been challenged by Fraser et al. (2010) with what they called the ‘‘inside-out’’ hypothesis, i.e., that teeth evolved from within the jaws. The evolution of tetrapods from sarcopterygian fishes is one of the major transformations in the history of life. The fish–tetrapod transition appears well 1 Introduction 5 documented, perhaps among the best in vertebrate evolution, with the first tetrapods apparently appearing during the Givetian, 391–385 million years ago in the Late Devonian. However, recently, a study by Niedzwiedzki et al. (2010) rocked the paleontological community back on its heels. The authors convincingly demonstrated, on the basis of fossilized quadruped tracks in Poland, that tetrapods had evolved much earlier than existing evidence showed, by some 18 million years in the Eifelian (Fig. 1.1), from a different line of ancestral fishes and in coastal marine rather than freshwater conditions. Here was an apparently advanced tetrapod whose ancestors were first off the starting block, when those from existing knowledge based on body fossils (skeletal etc.) were still at the early fish stage. Because there are no body fossils associated with the Polish tetrapod it is difficult to know what the ancestral fish group was or the stages leading to its evolution. Critically, the discovery throws wide open the bigger question of who exactly are the ancestors of tetrapods? In a purely genealogical system, each group must correspond to a single lineage (clade) composed of the common ancestor and all of its descendants which, until the discovery by Niedzwiedzki and colleagues, seemed to have been beautifully supported by the painstaking research of some of the most distinguished names in the field of tetrapod origins. The idea now is that tetrapods may be paraphyletic, i.e., a grade which does not contain all the descendants of a common ancestor—that several fish groups experimented with the move to land. From an evolutionary perspective, a paraphyletic origin of tetrapods muddies the waters but will no doubt shake up thinking and trigger off searches for new potential transitional forms. Once the fish to land transition by tetrapods was achieved, two major events had to happen for the early reptiles to survive. The first was the amniotic egg (that of birds and reptiles) which once and for all would remove reptiles from dependence on water and tadpole or larval stages of development. The amniotic egg provided the developing embryo with its own minipond, all the food and air it needed and a tough outer shell for protection. The other major event concerned the adult reptile itself—the reptilian integument. Everything changed. The epidermis now formed the outer region of the animal, giving rise to tough, protective, sometimes colorful, scales—taking over the role of the dermis in fishes. The momentous development was in the composition of the scales, of an entirely new material—b-keratin. b-keratin is extremely tough, the toughest natural elastomeric material known, and critically, extremely lightweight, unlike bone. It is also a highly stable material that would provide reptiles with maximal protection from the elements. With these two great innovations, the amniotic egg and the epidermis, the world was dramatically opened to the reptiles. It is probably safe to say that birds, the descendants of reptiles, would never have left the ground without the evolution in reptiles of b-keratin, the material of feathers (Fig. 1.3), not just light and tough but as latest research shows, with an inherent capacity for intracellular and extracellular hier- archical self-assembly (Lingham-Soliar and Murugan 2013). The popularly termed ‘‘Age of Reptiles’’ is most associated with the dominant fauna, dinosaurs. Yet, as Tom Kemp (1982) noted the first phase very largely belonged to the mammal-like reptiles and lasted throughout the latter part of the 6 1 Introduction

Fig. 1.3 Skeleton with feathers of Archaeopteryx, as it was found (Berlin specimen)

Carboniferous and the Permian. During the Triassic large dog-sized mammal-like reptiles, the cynognathids, gave rise to advanced forms such as Cynognathus and Probelesodon, well-known from the karoo beds of South Africa (Fig. 1.4). These forms Kemp considered could technically be regarded as mammals. However, during the Triassic they were gradually replaced by the dinosaurs. In fact the history of mammal-like reptiles shows, despite many advanced characteristics compared to the dinosaurs, that they had come close to extinction on more than one occasion (at the Permian–Triassic boundary and end of the Triassic), hence our own future can be said to have hung on a thread. By the end of the Triassic they were all extinct, leaving behind only their mammalian descendants. The future of mammals was reduced to tiny shrew-like animals such as Morganucodon. Their lifestyle involved being on the run, of skulking and secreting themselves to avoid danger, i.e., a species literally having to crawl under the proverbial rock to survive. This terror extended to being forced to emerge to feed only at night, when the dinosaurs slept. It was chance more than anything else that converted this pitiable lifestyle to advantage. George Gaylord Simpson (1950) one of the founding fathers of the neo-Darwinian synthesis emphasizes in The Meaning of Evolution the role chance plays in the evolution of species including in Homo sapiens, ‘‘Not all the chance favored his appearance, none might have, but enough did.’’ Ironically, the sorry state of the shrew-like mammals in the Triassic would result in the evolution of a complex suite of characters, development of the young 1 Introduction 7

Fig. 1.4 A life reconstruction of a cynodont, a mammal-like reptile about the size of a large dog, of early Triassic age. From Colbert (1955) within the mother, parental care including nourishment by the mother’s milk, multituberculate teeth aiding efficient feeding, a constant level of metabolism, a constant body temperature independent of external heat or cold, crowned by a unique body covering, hair—characters that would come to define the mammalian condition. Yet, none of these conditions could guarantee the survival of mammals—but they did and subsequently they would open up a wealth of possibilities for their descendants, culminating in the human species. What of the dermis and its new role since that of fishes? Less was known of the mechanical role of the dermis in reptiles, modern fishes, and mammals until research over the past 30 years in sharks, tuna, and dolphins showed that the dermis in many vertebrates, which had becomes much thicker than the epidermis, is comprised of multiple layers of oppositely oriented collagen fibers, helically wound—a construction that plays a major part in the locomotion of these groups of animals and in a number of other functions in terrestrial vertebrate. The cross-fiber architecture of the dermis is now firmly entrenched as a design principle of biomechanics (Fig. 1.5). The identical cross-fiber architecture of the dermis of extant vertebrates was subsequently found in fossil material of the fastest swimmers in the Jurassic seas, the ichthyosaurs. This group of reptiles, together with mosasaurs and plesiosaurs, is explored in the last chapter of the book because these reptiles had come full circle in their history, from the sea to the land, and finally back to the sea. Ichthyosaurs and plesiosaurs enjoyed a long history of about 150 million years and evolved to become some of the most spectacular marine vertebrates ever known. They became extinct in the later part of the Cretaceous at a time when the third group, the mosasaurs, with a much shorter history, was making a dramatic rise in the marine biota. Yet, at the height of mosasaur evolutionary development and diversification, they suddenly became extinct, in striking contrast to their contemporaries, detritivore crocodiles at the bottom of the food chain, which survived to give rise to present-day crocodiles. The lure of the sea did not end there. During the Eocene, another group of terrestrial animals would also return to the sea and initially fill the niche vacated by the mosasaurs. They would rise to become among the most dominant members of the marine biota today—dolphins and their relatives. 8 1 Introduction

Fig. 1.5 A hypothesized life reconstruction by the author of a group of Jurassic ichthyosaurs. Breach swimming may have been employed as a form of energy saving as in their thunniform counterpart, the cetaceans. Ichthyosaurs, dolphins, and thunniform sharks (mako and white shark), show classic convergent evolution in three phylogenetically distinct groups, reptiles, mammals, and ﬁsh, respectively

We will never stop asking questions How does it work? or What is its function? or How did it start? During the past 20–30 years we see that the vitality and vibrancy of the early workers on the vertebrate integument has returned in force and we can look forward to exciting times in the field in the future, but most importantly, as briefly demonstrated above in the history of research in the field— there will be no easy answers.

References

Anderson PSL, Friedman M, Brazeau MD, Rayfield EJ (2011) Initial radiation of jaws demonstrated stability despite faunal and environmental change. Nature 2011. doi: 10.1038/ nature10207 Colbert EH (1955) Evolution of the vertebrates. Wiley, New York Dobzhansky T (1962) Mankind evolving. Yale University Press, New Haven Donoghue PCJ, Sansom IJ (2002) Origin and evolution of vertebrate skeletonization. Microsc Res Tech 59:352–372 Fraser GJ, Cerny R, Soukup V, Bronner-Fraser M, Streelman T (2010) The odontode explosion: the origin of tooth-like structures in vertebrates. BioEssays 32:808–817. doi: 10.1002/bies. 200900151 Kemp TS (1982) Mammal-like reptiles and the origin of mammals. Academic Press, London Lingham-Soliar T, Murugan N (2013) A new helical crossed-fibre structure of b-keratin in flight feathers and its biomechanical implications. Plos One (June, Issue 6):1–12 (e65849) References 9

Niedzwiedzki G, Szrek P, Narkiewicz, K, Narkiewicz M, Ahlberg PE (2010) Tetrapod trackways from the early middle Devonian period of Poland. Nature 463:43–48 Purnell MA (2001) Feeding in extinct jawless heterostracan ﬁshes and testing scenarios of early vertebrate evolution. Proc R Soc Lond B 269:83–88. doi: 10.1098/rspb2001.1826 Simpson GG (1950) The meaning of evolution. Yale University Press, New Haven Stensio EA (1927) The Devonian and Downtonian vertebrates of Spitsbergen, Part 1. Family cephalaspidae. Skr Svalbard Ishav 12(1927):1–391 Wallace AR (1899) Darwinism, 2nd edn. Macmillan and Co, London Chapter 2 The First Vertebrates, Jawless Fishes, the Agnathans

2.1 Ostracoderms

Vertebrates arose over 500 million years ago (MYA) but traces of their appearance only occur during the Ordovician period about 460 MYA. These primeval, small fish-like vertebrates are popularly known as the ostracoderms and during the 100 million years of their existence they were comprised of about 600 species. Ostracoderms are especially important in the history and evolution of vertebrates. As undoubted vertebrates they possessed a backbone, were bilaterally symmetrical and had a nervous system divided into brain and spinal cord (partly enclosed within the backbone). They are also characterized by the possession of no more than two pairs of limbs and muscular system consisting primarily of bilaterally paired masses and a well-developed coelom, which contained the organs. In appearance, ostracoderms were dorsoventrally flattened and, quite extraordinarily, they lacked jaws, a condition so important that the classification of modern vertebrates is recognized by two major groups, the agnathans (without jaws) and the gnathostomes (literally, jaw-mouths) (Figs. 2.1 and 2.2). Ostracoderms are now regarded as an artificial designation that includes perhaps four distinct superclasses of jawless craniate fishes, the Pteraspidomorphi, Anaspida, Thelodonti, and Osteostrachomorphi. Despite an apparently inauspicious beginning, these simple animals would be responsible for two events of major importance in the history of the vertebrates—the evolution of a tough outer protective layer, the integument, and the evolution of bone. Remarkably, in the beginning the histories of these seemingly disparate structures were intrinsically linked. Ostracoderms possessed an external bony head shield or armor but its internal skeleton was probably not ossified to any great extent. Debate over the origin of the vertebrate skeleton has revolved around the question of which group was the first to exhibit evidence of skeletonization and mineralization. The vertebrate skeleton is comprised of at least two distinct skeletal systems: the dermoskeleton (arising from the dermis; widely perceived to encompass teeth, scales, fin spines, etc.) and the endoskeleton (the braincase, branchial skeleton, axial, and appendicular skeletons). It has been argued that although elements of the skeleton can be

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 11 DOI: 10.1007/978-3-642-53748-6_2, Ó Springer-Verlag Berlin Heidelberg 2014 12 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.1 Ostracoderms evolved into a wide variety of shapes and sizes during their 100 MYR reign on earth. The nearly 600 species of ostracoderms can be classiﬁed into one of nine or ten groups. From Forey and Janvier (1994)

Fig. 2.2 The fossilized impression left by the head shield and body scales of a 400-million-year- old ostracoderm from Great Britain. The specimen is about 15 cm long and shows the location of the eye sockets and special sensory ﬁelds on the head. From Forey and Janvier (1994) interchangeably derived from either system, the two systems have remained distinct throughout vertebrate phylogeny (Donoghue and Sansom 2002), of which more will be said later (Fig. 2.2). 2.2 Marine or Freshwater Origins? 13

2.2 Marine or Freshwater Origins?

The environment of early vertebrates in the Ordovician was of enormous importance in the question of how these vertebrate characters came about and how their roles would later take very separate paths. There had been much early discussion by geologists and zoologists on whether the first vertebrates evolved in a marine or freshwater habitat (Halstead 1969a). The prevailing view among zoologists was that vertebrates had originated in the sea. This view was challenged when fragments of bony armor were discovered in the freshwater Middle Ordovician Harding Sandstone of Colorado. This new evidence presented a number of questions on how the early vertebrates would cope with living in a freshwater habitat (Halstead 1985), which literally involved a sea change in scientific thinking. Halstead (1985) showed how despite an originally purely marine habitat these early vertebrates, e.g., the thelodonts showed an ability to deal with variations in salinity and were able to colonize both brackish and fresh waters and by the lower Devonian were able to spread globally. Among other groups that made freshwater colonizations of non-marine habitats during the Wenlock or Ludlovian times were the galeaspids, a group of cephalaspidomorphs, known only from South China (Figs. 2.3 and 2.4). In the sea the concentration of salts in the body fluids of animals and that of the water is approximately the same, hence there is no appreciable osmotic gradient between the two. In freshwater on the other hand the concentration of salts is negligible. The idea that a bony integument could serve as waterproofing was proposed by Berrill (1955) and Homer Smith (1963), i.e., to prevent waterlogging of the body. The drawback to this hypothesis is that the bony integument was present when the ostracoderms colonized marine environments where there was no need for waterproofing. In support of the idea of a freshwater habitat, Homer Smith and others suggested that the glomerular kidney was developed principally to control osmoregulation. Prior to the freshwater problems of osmoregulation the development of a bony armor, as the name implies, led to the hypothesis that it may have evolved in vertebrates for protection. Alfred Sherwood Romer (Romer and Grove 1935; Romer 1971) was one of the strongest supporters of this hypothesis. Romer also believed in a freshwater origin of vertebrates and contended that the bony integument would have protected the ostracoderms from the formidable pincers of the giant freshwater scorpions or eurypterids that inhabited the lakes and rivers of the time, although other workers such as Halstead (1969a) suggested that there was no evidence to show the eurypterids and ostracoderms ever shared the same locations. Romer, like Homer Smith, believed that the glomerular kidney originated to help control osmoregulation. The stumbling block to this hypothesis is that hagfishes, which possess a glomerular kidney have always had a marine environment. The British physiologist Robertson (1957) further showed that a similar filtration system was also present in a number of marine invertebrates that are stenohaline (low tolerance to salinity changes) and consequently have little need to modify intracellular osmolality. Maintaining internal salt conditions like that of the sea water was widely 14 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.3 Thelodont faunal provinces and palaeogeography. a Silurian. b Lower Devonian. 1 South America; 2 Africa: 3 Angaraland; 4 North America; 5 Baltica; 6 North China; 7 Australia; 8 South China; 9 Antarctica; 10 India. From Halstead (1985) regarded as a strong argument against the freshwater origin of vertebrates. However, the physiological arguments and counter arguments seemed redundant when it was subsequently shown that the Harding Sandstone, rather than a freshwater deposit, extended over thousands of miles and was in fact an offshore deposit. Indeed, the matter should have been settled by the even older deposits from the Lower Ordovician rocks of Russia that showed unequivocally shallow marine conditions. Halstead (1985) noted that it was unfortunate that the Harding Sandstone’s mistaken association with a freshwater system had led effectively to such a waste of time and energy and a number of erroneous zoological explanations and he unequivocally 2.2 Marine or Freshwater Origins? 15

Fig. 2.4 Radiation of galeaspids from South China. From Halstead (1985) stated that the origin and early evolution of the vertebrates took place in exclusively marine conditions. He cited little evidence of non-marine vertebrate faunas extending as far back as the Middle-Cambrian chordate Pikaia (a possible cepha- lochordate from the Middle-Cambrian Burgess Shale, about 550 million years ago), through to the Upper Cambrian, Ordovician, and early Silurian records. Despite this apparent resolution, over time it is not unusual to see that ideas once unfashionable become acceptable, or at least in part, with renewed research. For instance, Griffiths (1985) has suggested that there is more to this than just the simple salt content of the habitat and that the glomerular kidney may have originally been a feature of a freshwater and ion-regulatory function. Griffiths also suggested that the first vertebrates may have been anadromous (marine fish that migrate to freshwater to breed) like the lamprey and that the rivers and estuaries may have provided a safe haven for reproduction and early development of the young with a move to the coastal marine waters to feed after they had grown big enough to compete with marine species (see Foreman et al. 1985). This is an intriguing hypothesis and whether the marine stage came first or the freshwater what seems undeniable is that the invasion of fresh waters, as Beverly Halstead (1985) emphasized, marked perhaps one of the most important advances in the evolution of the physiology of vertebrates and that these jawless oddly looking animals, microphagous detrital feeders, would herald the invasion of predators, the benthonic placoderms, and nektonic aconthodian fishes. 16 2 The First Vertebrates, Jawless Fishes, the Agnathans

2.3 Feeding

Many zoologists believe that the bony ostracoderms were in the main line of vertebrate evolution (Moy-Thomas and Miles 1971). While defense or even predatory behavior are not excluded because they almost certainly would have helped to protect the animal, the consensus of opinion seems to be that as a primary function, a heavy bony armor would be incompatible with an active predatory lifestyle. At a major stage in their evolution, vertebrates had arrived at a crossroads. Glover et al. (2011) proposed that they had to make an evolutionary choice between nutrient absorption via the skin as in hagfish and many invertebrates or forms of filter feeding (Jollie 1982), to a form of feeding that limited exchanges across permeable surfaces and was associated with more active feeding methods and specialized digestive systems. The view that there was a long-term ecological trend toward increasingly active and predatory habits in the heterostracans was contradicted by Purnell (2001) who contended that heterostracans were microphagous suspension feeders and that this was a relatively advanced development in feeding. He based it on his observations on heterostracan feeding structures, which exhibit recurrent patterns of in vivo wear and covered internally by microscopic oral denticles. The functional significance of the denticles derives from the fact that their tips are consistently directed outwards, i.e., the entrance to the heterostracan mouth would have been lined with imbricate rows of anteriorly directed barbs, which would have prevented grasping, biting, or any other form of macrophagy (Fig. 2.5).

2.4 Bone, a Chemical Store

As we saw above, the seemingly obvious answer that bone on the outside of an animal provides protection has nevertheless been a source of much debate. Many of the notable paleontologists and biologists of the time were not distracted by what might have seemed the most obvious answers. Halstead (1969a) put his ﬁnger on the pulse when he observed that during the course of evolution a structure necessary for survival under new conditions may invariably already be present or had begun its development in its original environment although for a different reason. Answers with respect to the function of the bony armor in the jawless vertebrates came from the most unlikely source. On the basis of work on proto- zoans, it was shown that a conveniently accessible phosphate store is likely to have been needed by an animal with much muscular activity (Pautard 1961). We know that bone in humans and many terrestrial vertebrates is a reservoir for calcium and phosphate and it is reasonable to postulate that bone evolved originally as such a store. Calcium phosphate, e.g., is a convenient way of storing phosphate. The need for a phosphate store in the early vertebrates may be connected with the seasonal changes in the sea and different levels of availability. Free phosphates 2.4 Bone, a Chemical Store 17

Fig. 2.5 Oral plates of heterostracans. a Oblique anterior view of a pteraspid heterostracan showing the configuration of the oral plates forming the lower margin of the mouth. The illustration is based on a reconstruction of Errivaspis waynensis Blieck. b Oblique anterior view of the mouth of Protopteraspis vogti preserving oral plates in situ. The image is a montage of scanning electron micrographs of an epoxy replica of specimen A28720/2 (Paleontologisk Museum, Oslo; Devonian, Ben Nevis Formation, Spitzbergen). c Oblique view of isolated oral plate of Loricopteraspis dairydinglensis (White), specimen NHM P43713. d Oblique view of isolated oral plate of L. dairydinglensis, specimen NHM P43711. Both (c) and (d) show typical patterns of wear developed on the ventral surface of oral plates (anterior to left). The enlarged views of the areas outlined by the boxes show worn dentine ridges and parallel scratches (NHM P43711 and NHM P43713, Lower Devonian, Ditton Group, Dairy Dingle, near Neenton, Shropshire, UK). Scale bars, 1 mm. From Purnell (2001) are most plentiful in the sea during the winter months. Phytoplankton (Phylum Bacillariophyta) have the ability to utilize the free phosphates. Because they are free floating, these microscopic algae of the ocean are capable of developing in the surface layers since light of photosynthetically effective wavelengths is largely filtered out at a depth of about 50 feet (Stanier et al. 1989) (Fig. 2.6). Where the environment is favorable their growth is largely limited by the relative scarcity of two elements phosphorus and nitrogen. These elements are made available as phosphates and nitrates by the runoff of rainwater from the continents and subsequent distribution by ocean currents, when profuse development of 18 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.6 Diatoms. Microscopic one-celled alga (may be colonial) found in water. They are frequently called jewels of the seas because of their shapes and are among the most important and proliﬁc sea organisms. They are photosynthetic and serve directly or indirectly as food for many animals

phytoplankton occurs. In the past, as now, a heavy intake of phosphates occurs during spring and summer months by the phytoplankton. Free phosphates on the other hand, even when abundant, cannot be directly utilized by animals. Animals obtain their phosphates by feeding on phytoplankton and others by feeding on the latter the whole process serving to lock the phosphates into the marine plant and animal fauna. Thus at the end of the summer months when there was a dearth of phosphates in the sea, their presence in the animal cycle could be utilized by the bottom-dwelling ostracoderms feeding either on the decaying phytoplankton that sinks to the bottom or other animals. Thus, a means for storing phosphates at times when it is abundant in their diet would be of considerable advantage in the relatively long-living ostracoderms. Workers such as Pautard (1961, 1962) and Halstead (1969a, b) concluded that bone originated as a simple store of phosphates laid down in the skin of the earliest vertebrates. Griffiths (1985) agreed that a calcium and phosphate store has advantages during times of food shortage. He demonstrated in a number of animal including in fishes in which phosphates stored in bone are reabsorbed into the blood that they play a vital role during reproduction (in humans mineralization and reabsorption are under the control of hormones from the parathyroid and thyroid glands). He made a compelling case for his hypothesis concerning an anadromous freshwater origin of vertebrates. Most living anadromous fishes feed little during their migrations. He compared this with periods in the early Paleozoic when the rivers had even less food and the hypothetical anadromous ancestral vertebrates would have been unable to acquire phosphate or other minerals through the diet during their upstream spawning migrations. The capacity of the female proto vertebrates to synthesize vitellin, which was vital to reproductive success, would have depended on body phosphate reservoirs. This would have provided a strong selective pressure for the evolution of dermal bone as a reservoir. Either way, whether the first vertebrates had a freshwater or marine origin, the above arguments indicate the most likely reasons for the origin of bone was that it served as a store of vital minerals. It is not difficult to see that such a store of calcium phosphate in the outermost layer of these animals would rapidly have 2.4 Bone, a Chemical Store 19 assumed the secondary role of a protective armor. Thus this would conform to the classic of an original function being co-opted to a new one, i.e., protection of the body. Yet, even more profound was how that store of calcium phosphate shifted from the outside of the animal to the inside and eventually to a fundamentally different primary role, that of an internal bony skeletal support system, which would revolutionize life on earth. This transition took a while and it was only considerably later in vertebrate evolution that bone came to replace the cartilaginous endoskeleton (not in an evolutionary sense)—and change the course of vertebrate evolution. It is for this very important reason that the earliest jawless vertebrates are pivotal to the story of vertebrate evolution and why so many eminent paleontologists, zoologists, and geologists devoted so many years of research to trying to find answers to vertebrate origins. Also important, in resolving these mechanostructural and physiological problems we are coming closer to understanding the phylogenetic origins of modern fishes. Simpson (1950) in his book The Meaning of Evolution showed how the acci- dents that occur during evolution may have resulted in the development and rise of novel and successful types of organization that allow unimagined possibilities and enable certain groups of animals to overshadow all other life forms put together. The evolution of bone may be regarded as one of those accidental events in nature that defines evolutionary theory. We may be surprised that bone started its journey not on the inside of our vertebrate ancestors but on the outside and more so for the seemingly mundane purpose of a storehouse for calcium and phosphate. This is what the evidence points to, as we see amply demonstrated in the fossil record, but it brought with it a whole bundle of new problems, which is why we need to look a little closer at these curiously odd ancestors of ours. Fragments of bony armor are found historically in horizons of Ordovician age in the US and Russia, and more recently Australia. They belong to the primitive jawless vertebrates the ostracoderms whose nearest living relatives are the present- day cyclostomes, which, however, are naked (more will be said later). The ostracoderms possessed a bony carapace and as we have already indicated, their vertebrate status is beyond question.

2.5 Heterostraci

The heterostracans, among the most primitive agnathans, include the earliest known vertebrates, the arandaspids from the late Cambrian of the US and the earliest Middle Ordovician of Spitsbergen and Australia and the astraspids and eriptychids (Figs. 2.7 and 2.8) from the Middle Ordovician of America (Jarvik 1980). Their long held importance as a possible ancestor of the extant myxinoids has recently been called in question and a number of workers have placed them 20 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.7 Heterostracan ostracoderms. a Astraspis (Ordovician). b Cyathaspid (Silurian- Devonian). c Eglonaspis, showing tubular mouth (Devonian). d Drepanaspis (Lower Devonian). e Pycnosteus (Middle Devonian). f Pteraspis, showing oral plates splayed out to form a scoop (Lower Devonian). g Doryaspis (Lower Devonian). From Halstead (1969a) outside the main line of potential ancestors (Janvier 1996a). Unlike in the cyclostomes, in heterostracans the head and anterior part of the trunk are encased in a bony armor, the shield or carapace (Fig. 2.8). To anyone who has ﬂipped through a 2.5 Heterostraci 21

Fig. 2.8 Imperfect dorsal shield of the astraspid Astraspis desiderata. From Jarvik (1980)

book with pictures of these animals they could easily be perceived as formidable monsters of the oceans. However, most were small, no more than a few centi- meters long. At the anterior end there is an opening for the mouth, generally bounded ventrally by oral plates, and small laterally placed orbital fenestrae but unlike the osteostracans (Fig. 2.8) there is no nasohypophysial opening on the dorsal side of the carapace and only one branchial opening on each side. Behind the carapace, the trunk and tail are covered in scales. There are no paired fins but spines or ridge-scales in the position of the median and ventrolateral fin fold. There is nothing to indicate that true moveable folds with radial muscles were developed (Jarvik 1980). However, these agnathans had developed a number of features that affirmed their vertebrate status even more such as an efficient nasal apparatus, eyes, a pineal organ, an advanced acoustic-lateralis system and a lateral line system (Halstead 1969a; Jarvik 1980; see Chap. 4). The various genera and species can be distinguished mainly by differences in proportion and in the development of the superficial layer of the dorsal and ventral shields. From studies in the development of the armor of the carapace in the 22 2 The First Vertebrates, Jawless Fishes, the Agnathans various groups the overall shape can be reconstructed. It shows that despite the apparently limited potential of the animals, they underwent a remarkable radiation, and a surprisingly large number of unusual modifications of the basic form of the carapace evolved. Halstead (1962) demonstrated that it was possible to show from the carapace the way in which these groups are related to one another, and hence to be able to trace the main evolutionary lineages within the Heterostraci (Fig. 2.6). Halstead (=Tarlo 1960) also showed that what had once been considered to be three distinct types of growth of dermal plates in the Heterostraci, are in fact all related. This will be dealt with in detail further on. From the perspective of mobility it would appear that the heterostacans were rather limited. The absence of fins indicates that they had little active maneuverability in the three-dimensional environment. Compared to the flattened and broad anterior regions of the body the posterior trunk was much narrower and deeper, suggesting that the animal was propelled forward by lateral movements of the tail. Halstead (1969a) suggested that the markedly convex ventral surface of the carapace (Fig. 2.6) meant that forward movement would have automatically lifted the body free of the bottom as the animal bounded from one mud patch to another, a form of movement he described as resembling the slow-motion bounding motion of a finch. Consistent with this type of movement, it seems likely that the convex shaped ventral surface was more connected with the body’s hydrodynamics and to remaining above the mud for a longer period. Jarvik (1980) summed up the importance of the heterostracans with respect to six major developments, four of which are mentioned here within the context of the integument and the origin of bone. (1) In the microstructure of the dermal skeleton of the earliest heterostracans they show practically all the types of hard tissue (calcified cartilage, bone, dentine, and enamel-like structures), characteristics of the later appearing vertebrates. In the histologic structure of the skeleton they are thus typical vertebrates. (2) In pre-Silurian and later heterostracans, the bone is of the acellular type known as aspidin which has been the source of considerable debate and which some authors considered primitive (Halstead 1973) while others considered to be a secondary derivative of bone (Orvig 1967) (see below). (3) Distinct grooves for sensory lines present in Astraspis and Arandaspis prove a lateral line system was developed and that modifications of the brain had already taken place. (4) Given that the dorsal shield in both astraspids and eriptichids is composed of polygonal scales (‘‘tessarae’’), it has been considered by several workers to be primitive. Despite numerous attempts based on this hypothesis, often based on the time of the first appearance of the various groups in the geological record, there is uncertainty. For instance, certain authors (Denison 1951, 1964, 1967; Miles, and others; see Jarvik 1980) have indicated that statements as to primitiveness based on geological age is not necessarily the case, a subject that has been the source of much debate (below). 2.6 Osteostraci 23

2.6 Osteostraci

The cephalaspids (Osteostracomorphi) are the best known group of the ostracoderms (Fig. 2.9). For a long time, they were the only early members in which the internal anatomy of the head was known (Jarvik 1980). Cephalaspids possessed a solid cephalic shield and a scale covered trunk and tail. A major advance in some forms was the presence of paired pectoral fins, replaced by a ventrolateral crest, one or two dorsal fins and a heterocercal caudal fin with occasionally a paired horizontal flap (Heintz 1939). These features indicate a more efficient benthic mode of life than in the heterostracans. The cephalic shield varies considerably in shape and is comprised of an outer exoskeletal and inner endoskeletal part. The endoskeletal component furnishes equivalent comparison data not only with the neurocranium of gnathostomes but also with the dorsal parts of the visceral endoskeleton, the shoulder girdle and possible anterior parts of the pectoral fins (Jarvik 1980), which will be discussed further below. Near the center of the shield is a pair of orbital fenestrae (located as far away from the mud as possible in a benthic mud-grubber) and between them lies the pineal foramen. In front of the latter is the nasohypophysial opening. The fact that distinct grooves are present in Astraspis (Jarvik 1980) (Fig. 2.8) as well as in Arandispis (Fig. 2.1) proves not only that the lateral line system (we will see it more fully developed in advanced fishes) developed in these early vertebrates but also implies that the modifications of the brain and cranial nerves connected with the development of this system had already taken place. It seems almost certain that these forms had undergone the same specializations as later heterostracans in most important respects. Ventrally, there is a large, hollow oralo-branchial chamber, which apart from the mouth and paired series of external branchial apertures, is covered by small dermal bones (Fig. 2.10). A third major group of ostracoderms were the anaspids (Anaspida), a group that includes the early Silurian Jamoytius kerwoodi (Fig. 2.11). The body shape is much more fish-like than the other groups, indicating a more active mode of life. Possession of both a hypocercal tail and paired fins (Ritchie 1964) may indicate that the group was surface feeders. Although the anaspids were agnathan, there are signs that the lower margins of the mouth were bounded by strong plates of dermal armor analogous to jaws and teeth (Ritchie 1964). Despite these novelties Halstead (1969a) thought this group was closest to the living cyclostomes and several authors believed that the origin of modern agnathans may be found among them. However, Strahan’s (1958) view, based on plotting developmental stages, which suggests that it is not possible to derive the modern lampreys, let alone the hagfishes from the anapsids, is becoming increasingly apparent in recent studies. The taxonomic position of Jamoytius kerwoodi has been a source of considerable discussion (Ritchie 1968, 1984). Most recently, Sansom et al. (2010) re- examined the well-known specimen (Figs. 2.12 and 2.13). Taking into account 24 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.9 Cephalaspid ostracoderms. a Tremataspis (Silurian). b Thyestes (Silurian). c, d Aceraspis (Silurian-Devonian). e Hemicyclaspis (Silurian-Devonian). f Cephalaspis (Devonian). From Halstead (1969a) taphonomic studies, topological analysis, model reconstruction, and elemental analysis, they arrived at a more rigourously tested anatomical interpretation. On the basis of their analysis they provided a new cladistic analysis which showed that Jamoytius, a jawless vertebrate, and Euphanerops (=Endeiolepis) form a clade, 2.6 Osteostraci 25

Fig. 2.10 Cephalaspis signata (Lower Devonian). a Photograph of natural cast of oralo-branchial chamber in dorsal view. Cephalaspid Nectaspis areolata (Lower Devonian). b ventral aspect. From Jarvik (1980)

which they named Jamoytiiformes. They proposed that they are stem-gnathostomes rather than proto-lampreys or Anaspida. Their cladogram also revealed the Anaspida to be similarly along the line to gnathostomes, and not related to Petr- omyzoniformes (lampreys) (Fig. 2.14). Philippe Janvier (1996a) has also shown that the anatomy and physiology of lampreys and hagfishes are so different that it is difficult to reconstruct an ancestral morphotype of the cyclostomes, and there is no clear evidence of any fossil taxon that is neither a fossil hagfish or a fossil lamprey, but would be more closely related to the cyclostomes than to the 26 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.11 Anaspids. a, b Jamoytius (Silurian). a Restoration (complete). b Head region to show branchial basket, eyes and mouth. c, d Pharyngolepis (Silurian), (c) ventral view showing lateral ﬁns, (d) lateral view. From Halstead (1969a) gnathostomes. For instance, he points out that the dorsal, median, nasohypophysial complex of osteostracans, which has been regarded as identical and homologous to that of lampreys, instead that recent investigations (notably on the galeaspid braincase) now suggest that this resemblance is in fact a convergence. 2.6 Osteostraci 27

Fig. 2.12 Jamoytius kerwoodi White holotype (NHM P11284a) immersed in 90 % ethanol with incident polarized light and ﬁlter, illustrating the conﬂicting interpretations of White (1946), in bold, and Ritchie (1960, 1963, 1968, 1984) in plain text. Scale bar represents 10 mm. From Sansom et al. (2010) and references therein. Copyright Palaeontology 28 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.13 Body parts and topological interpretation of the holotype (NHM P11284a) of Jamoytius. Scale bar represents 10 mm. From Sansom et al. (2010). Copyright Palaeontology

Recent ﬁndings also show that although there are variations as to the position of certain taxa, the Galeaspida and Osteostraci constantly group together with the Gnathostomes (Forey and Janvier 1993, 1994, Janvier 1996b). Until now an intermediate stage between the separation of the olfactory ducts of agnathans and gnathostomes was not known but evidence from the jawless galeaspids, a 435–370 million-year-old group of ostracoderms from China to Vietnam (Gai et al. 2011), provide the earliest evidence for the clear separation of the paired olfactory organs from the hypophyseal duct as in jawed vertebrates but unlike in the cyclostomes and osteostracans (Fig. 2.15). In vertebrate phylogenetic terms it is regarded as a prerequisite condition for evolutionary developmental biology models for the origin of complete diplorhiny (clear separation of the olfactory organs from the 2.6 Osteostraci 29

Fig. 2.14 Jamoytius. a single most parsimonious tree from the unconstrained phylogenetic analysis with decay support indices. b Strict consensus of trees resulting from analysis constrained for cyclostome monophyly with decay indices. From Sansom et al. (2010). Copyright Palaeontology hypophyseal duct) and jaws (Gai et al. 2011). These authors propose this as an intermediate condition in the establishment of diplorhiny and jaws in which this barrier to the forward growth of (neural-crest-derived) craniofacial (ectomesen- chyme) development was removed. The way to the possession of jaws was paved by these forms and in the next chapter we will see the emergence of true jaws. 30 2 The First Vertebrates, Jawless Fishes, the Agnathans

Fig. 2.15 Shuyu zhejiangensis, Silurian of Zhejiang, China. a Restoration of external morphology. b Synthetic restoration of nasal and hypophyseal region. na nasal sacs; no nostril; olf.b olfactory bulb; et.r ethmoid rod; orb orbital opening; pi pineal organ; vcl lateral head vein or dorsal jugular vein. Reprinted with permission of Gai et al. (2011). Copyright Macmillan Publishers, Ltd

References

Berrill NJ (1955) The origin of vertebrates. Oxford University Press, New York Denison RH (1951) The exoskeleton of early Osteostraci. Fieldiana Geol 11:199–218 Denison RH (1964) The cyathaspididae: a family of Silurian and Devonian jawless vertebrates. Fieldiana Geol 13:309–473 Denison RH (1967) Ordovician vertebrates from western United States. Fieldiana Geol 16:131–192 Donoghue PCJ, Sansom IJ (2002) Origin and early evolution of vertebrate skeletonization. Microsc Res Tech 59:352–372 Foreman E, Gorbman A, Dodd JM, Olsson R (eds) (1985) Evolutionary biology of primitive fishes. Plenum, New York Forey PL, Janvier P (1993) Agnathans and the origin of jawed vertebrates. Nature 361:129–134 Forey PL, Janvier P (1994) Evolution of the early vertebrates. Am Sci 82:554–560 Gai Z, Donoghue PCJ, Zhu M, Janvier P, Stampanoni M (2011) Fossil jawless fish from China foreshadows early jawed vertebrate anatomy. Nature 476:324–327 Glover CN, Bucking C, Wood CM (2011) Adaptations to in situ feeding: novel nutrient acquisition pathways in an ancient vertebrate. Proc R Soc B 278(1721):3096–3101 Griffith RW (1985) Habitat, phylogeny and the evolution of osmoregulatory strategies in primitive fishes. In: Foreman RE, Gorbman A, Dodd JM, Olsson R (eds) Evolutionary biology of primitive fishes. Plenum, New York, pp 69–80 Halstead LB (1962) The classification and evolution of the Heterostraci. Acta Palaeontol Pol 7:249–290 Halstead (1969a) Pattern of vertebrate evolution. Oliver and Boyd, Edinburgh Halstead LB (1969b) Calcified tissues in the earliest vertebrates. Calc Tiss Res 3:107–134 Halstead LB (1973) The heterostracan fishes. Biol Rev 48:279–332 Halstead LB (1985) The vertebrate invasion of fresh water. Phil Trans R Soc B 309:243–258 Heintz A (1939) Cephalaspida from downtonian of Norway. skr norske vidensk-acad, 1 mat- naturv kl 5:1–119 References 31

Janvier P (1996a) The dawn of the vertebrates: characters versus common ascent in the rise of current vertebrate phylogenies. Palaeontology 39:259–287 Janvier P (1996b) Early vertebrates. Oxford monographs in geology and geophysics 33. Oxford University Press, Oxford Jarvik E (1980) Basic structure and evolution of the vertebrates (volume 1). Academic Press, London Jollie M (1982) What are the ‘‘calcichordata’’? and the larger question of the origin of the chordates. Zool J Linn Soc 75:167–188 Moy-Thomas JA, Miles RS (1971) Paleozoic fishes. WB Saunders, Philadelphia Ørvig T (1967) Phylogeny of tooth tissues: evolution of some calcified tissues in early vertebrates. In: Miles AEW (ed) Structural and chemical organization of teeth, vol 1. Academic Press, London, pp 45–110 Pautard FGE (1961) Calcium phosphate and the origin of backbones. New Sci 12:364–366 Pautard FGE (1962) The molecular-biologic background to the evolution of bone. Clin Orthopaed 24:230–244 Purnell MA (2001) Feeding in extinct jawless heterostracan fishes and testing scenarios of early vertebrate evolution. Proc R Soc Lond B 269:83–88. doi:10.1098/rspb.2001.1826 Ritchie A (1964) New light on the Norwegian anaspida.Skr. norske, Vidensk-Acad., 1. Mat.- naturv. Kl., 14:1–35 Ritchie A (1968) New evidence on Jamoytius kerwoodi white, an important ostracoderm from the Silurian of Lanarkshire, Scotland. Palaeontology 11:21–39 Ritchie A (1984) Conflicting interpretations of the Silurian agnathan, Jamoytius. Scot J Geol 20:249–256 Robertson JD (1957) The habitat of the earliest vertebrates. Biol Rev Cambr Philos Soc 32:156–187 Romer AS, Grove BH (1935) Environment of the early vertebrates. Am Midl Nat 16:805–856 Romer AS (1971) Vertebrate paleontology. University of Chicago Press, Chicago Sansom RS, Freedman K, Gabbott SE, Aldridge RJ, Purnell MA (2010) Taphonomy and affinity of an enigmatic Silurian vertebrate, Jamoytius kerwoodi White. Palaeontology 53:1393–1409 Simpsom GG (1950) The meaning of evolution. Yale University Press, New Haven Smith HW (1963) From fish to philosopher. Doubleday, New York Stanier RY, Ingraham JL, Wheelis ML, Painter PL (1989) General microbiology. Macmillan Education Ltd, Basingstoke Strahan R (1958) Speculations on the evolution of the agnathan head. Proc Cent Bicent Congr Biol Singapore 83–94 Tarlo [Halstead] LB (1960) The downtonian ostracoderm corvaspis woodward, with notes on the development of dermal plates in the heterostraci. Palaeontology 3:217–226 Chapter 3 The Earliest Jawed Vertebrates, the Gnathostomes

3.1 The Evolution of Jaws

Edwin Colbert (1955) remarked that the history of life, like human history, has been marked by certain great developments that rise above the general level of events. One such momentous event or revolution in the history of vertebrates was the appearance of jaws. Nelson (2006) showed that more than 99 % of the roughly 58,000 living vertebrate species have jaws. This revolution would take vertebrates from purely bottom feeders or parasites to forms in which new anatomical elements evolved enabling the exploiting a wide variety of habitats that would radically transform life on earth. The first jawed vertebrates appeared in the upper Silurian and lower Devonian. They comprised two principle extinct groups of primitive jawed vertebrates, the placoderms (Class Placodermi) and the acanthodians, (Class Acanthodii, formerly placed within the placoderms) (Figs. 3.1 and 3.2). Their basal origins are the Acanthothoraciformes from the lower Devonian, and are therefore the oldest known jawed vertebrates. They possessed an armored head shield and trunk shield of overlapping bony plates, very much like the jawless ostracoderms. They are, however, readily distinguished from the ostracoderms by their paired fins and presence of jaws (Donoghue and Sansom 2002). The classical view is that jaws evolved via modifications of the anterior gill arch cartilages (viscerocranial elements) that lost their function as gill support and assumed the role as an articulated lower jaw. In modern fishes sharks demonstrate the surviving gills and gill arches best (Fig. 3.3). For jaws to function effectively they needed to be stiff, i.e., ossified. Ossifications that form in the visceral cartilages are like those of the chondrocranium and nearly all of the postcranial skeleton replacement bones (Fig. 3.4). The skeleton of each visceral arch in the gnathostomes comprised usually four segments per arch. The middle two are the epibranchial and ceratobranchial (Fig. 3.3). The first visceral arc enlarges to become the jaws, it forms the mandibular arch. The epibranchial forms the key element of the upper jaw or palatoquadrate and the ceratobranchial forms the lower jaw or mandibular cartilage. The jaws for the first time become firmly

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 33 DOI: 10.1007/978-3-642-53748-6_3, Ó Springer-Verlag Berlin Heidelberg 2014 34 3 The Earliest Jawed Vertebrates, the Gnathostomes

Fig. 3.1 Acanthodians. Restorations of representative acanthodii illustrating some characteristics. After Hildebrand (1995) anchored to the chondrocranium. This form of joint is autostylic or self-supported as it is attached to the chondrocranium by ligaments and not by the hyomandibular. The second visceral arch or hyoid arch is at some stage later in the evolution of jaws sequestered for further support. In placoderms, the head and trunk shields were articulated by bony joints, allowing the forward part of the skull to rotate upward, thereby increasing the gape and allowing an intake of bigger prey items. These characters set them apart from their earlier jawless relatives for it enabled them to seize and take larger and more effective bites of prey rather than simply suck up organic particles from the mud. However, that was just one part of predatory behavior—first it was necessary to catch the prey (below and chap. 4). Halstead (1969a) discussed at some length jaw origins and traced a basic condition from which the higher vertebrates evolved namely in the heterostracans; they were the ‘‘pre-gnathostomes.’’ In these forms the mandibular arch had no associated gill slit, but acted to support the mouth. As Halstead (1969a) observed any increase in this role and consequent enlargement of the elements concerned would produce de facto a jaw. It was, however, in the acanthodian fishes of the Silurian that the role of the gill arches in the evolution of jaws really began to take shape. In a classic paper DMS Watson (1937) was among the first to investigate the problem. In some aspects, particularly with respect to whether or not the second gill arch, the hyoid arch, participated in jaw support, he strongly differed with Stensio (1927). Miles (1964, 1965), however, subsequently demonstrated 3.1 The Evolution of Jaws 35

Fig. 3.2 Placoderms. Reconstructions of some representative illustrating some characteristics of the class (boldface labels) and of two principle orders (standard labels). After Hildebrand (1995) fairly conclusively that Stensio was right and that the dorsal part of the hyoid arch, the hyomandibular was concerned with jaw suspension. This view was also supported by Mallatt (1996) who argued that the ﬁrst gill arch moved forward in the mouth. These gill arches that occurred in the more derived agnathans, and still occur in the gnathostomes, are jointed. Mallat proposed that the origins of a jointed rod supported by a musculature serving as a structure for more effective ventilation (and ﬁltering food through the gills), could then be co-opted as jaws in a proto- 36 3 The Earliest Jawed Vertebrates, the Gnathostomes

Fig. 3.3 Primitive visceral skeleton showing evolution of jaws, represented by a stylized elasmobranch. After Hildebrand (1995)

Fig. 3.4 Replacement bones of the skull and mandible at the early stage of tetrapod evolution. Ossifications of the chondrocranium are shaded, those of the mandibular arch are hatched. All bones are paired except those under the brain and supraoccipital gnathostome. The first stage in the development of jaws was thus to develop jointed branchial arches rather than the one-piece branchial basket seen in lampreys. This more efficient means of oxygenation in a fish would enable a more active pelagic predator. Over generations these joints would become more pronounced, until eventually their development was such that they allowed the next step in the journey toward jaws. Thorough investigations by a number of authors, including a series of publications by Miles (1964, 1973a, b, 1997), indicated that acanthodians are closely 3.1 The Evolution of Jaws 37 related to teleostomes or osteichthyian fishes (those having an internal bony skeleton) and referred them to the subgroup Teleostomi, a view that was widely accepted. However, Jarvik (1977) suggested that no significant characters indicating relationship with the Teleostomi have been found and that features such as the absence of the outer dental arcade, the subterminal mouth, and the histological structure and mode of growth of the teeth debar them from relationship with the Teleostomi. He proposed that the acanthodians agree more closely in these features with elasmobranchs or possibly selachians. Opinions on the question of shark and ray classification, however, differ considerably and classification of acanthodians within elasmobranchii has largely been rejected.

3.2 The Evolution of Paired Fins

Active feeding in early vertebrates demanded precise control of swimming motions. As we saw in Chap. 2, a major advance in some forms of cephalaspids was the presence of paired pectoral fins, one or two dorsal fins, and a heterocercal caudal fin, which preempted the development of jaws and more active feeding. Nevertheless in the beginning the fins initially acted as stabilizers in which the main function was to prevent unnecessary roll. They were triangular, with broad bases and without much capability of independent movement. In the acanthodians there were pectoral and pelvic fins and in some a series of small accessory fins between the main paired fins. In the arthrodires (placoderms), a large and varied group of armored, jawed fishes, the fins were no more than lateral spines, which they may have used to anchor themselves or to move about on the sea bottom (Fig. 3.5). We will see in Chap. 4 how the development of fins led to the success of modern-day fishes in both marine and fresh water biotas.

3.3 Placoderms, Dominant Vertebrates of the Devonian

Placoderms, were the most diverse group of Devonian fishes and were globally distributed in all habitable freshwater and marine environments, like teleost fishes in the modern fauna. The relationship of placoderms to the two major extant groups of jawed fishes—osteichthyans (bony fishes) and chondrichthyans (cartilaginous sharks, rays, and chimeras)—remains uncertain. The group’s evolutionary history is complicated, and recent claims that they form a paraphyletic group, apparently based on highly derived Late Devonian forms was challenged by Gavin Young (2010) whose analysis supports placoderm monophyly. Besides the origin of jaws, placoderms are considered to show evidence of a number of firsts in vertebrate skeletal development (Donaghue and Sansom 2002). They are the earliest vertebrates to exhibit evidence of ossification of the splanchnocranium-proper (oral skeleton, i.e., bones forming part of the cranium 38 3 The Earliest Jawed Vertebrates, the Gnathostomes

Fig. 3.5 Arthrodires. a Radotina, primitive tessellated arthrodire, dorsal view of skull (Lower Devonian). b Gemuendina, dorsal view (Lower Devonian). c Coccosteus, lateral view (middle Devonian). d, e Ctenurella, restoration and skeleton in lateral view (Middle Devonian). After Halstead (1969a) 3.3 Placoderms, Dominant Vertebrates of the Devonian 39 and mandible) in the form of an ossified mandibular arch skeleton and hyoid; there is little evidence of ossification of the branchial arches; the first evidence of dermal bone association with the splanchnocranium in the form of dermal jaw-bones that support small dental element in some, but not all groups; the earliest evidence of an ossified axial skeleton, although it is not preserved in many groups, presumably because it was not mineralized. The prevailing view is that along with the acanthodes, the placodermi also became extinct. However, a recent study by Brazeau (2009) suggests that neither group became extinct but that they might be paraphyletic groups (grade, which does not include all the descendants of the common ancestor) that evolved into higher fish. While the acanthodii or ‘‘spiny sharks,’’ were long thought to be ancestral to modern groups of fish, the idea of the placoderms, which are more specialized and distinctive, giving rise to modern fishes is more radical. Brazeau describes a genus of acanthodes, Ptomocanth anglicus a 100 years older than Acanthodes. They found that the braincase of P. anglicus has a radically different morphology from Acanthodes, which has several important implications for the relationships of acanthodians and later vertebrates. For this reason the braincase of Acanthodes was thought to most closely resemble that of early bony vertebrates that would ultimately give rise to the land vertebrates, i.e., the acanthodians were thought to share a closer ancestor with bony vertebrates than with sharks. How- ever, the braincase of Ptomacanthus was found to more closely resemble that of early shark-like fishes, and shares very few features in common with Acanthodes and the bony vertebrates. Thus, Brazeau and colleagues’ results suggest that Ptomacanthus was either a very early relative of sharks, or close to the common ancestry of all modern jawed vertebrates. A phylogenetic analysis of early gnathostome braincases also placed them as stem chondrichthyans and stem gnathostomes, closer to chondrichthyans than to osteichthyans (Fig. 3.6). Concerning evolutionary trends, there is an underlying belief that less efficient forms are replaced by more efficient forms as mentioned in Chap. 1. Such a belief may sometimes be based on unfounded assumptions. For instance, findings by Anderson et al. (2011) indicate that long-held assumptions concerning the replacement of jawless fishes by newly evolved jawed forms are not justified and they provide a new perspective on debates concerning the Devonian shift from agnathan- to gnathostome-dominated fossil assemblages. A range of schemes, largely derived from anecdotal evidence, have sprung up in response to this pattern of turnover. Faunal data clearly show that gnathostomes shared habitat space evenly with ‘‘ostracoderm’’-grade agnathans well into the Early Devonian. It is only during and after the Emsian that gnathostomes became taxonomic dominants in most fossil assemblages, several million years after the evolutionary development of mandibles was firmly entrenched (Fig. 3.7). The authors state that the long history of coexistence between gnathostomes and ‘‘ostracoderms’’ argues against the direct ecological replacement of jawless fishes by jawed forms. It seems that the variety of feeding mechanisms in early jawed animals had little to no affect on the diversity of jawless fishes, which shared ecological space with the jawed fishes for at least 30 million years before beginning to notably decline. When the jawless 40 3 The Earliest Jawed Vertebrates, the Gnathostomes 3.3 Placoderms, Dominant Vertebrates of the Devonian 41 b Fig. 3.6 Ptomacanthus anglicus NHM P 24919a. a, b Interpretive sketch of specimen a with accompanying photograph b. c, d Close-up photograph of neurocranium, tooth row, and anterior part of palatoquadrates c and interpretive sketch of neurocranium d. B.art basal articulation; Br.A branchial arches; Ch notochordal notch; Glen? possible occipital glenoid; Hyp hypophyseal opening; I.car.a foramen for the internal carotid artery; L.Pq left palatoquadrate; Mk mineralized Meckelian cartilage; NVIIpal? possible foramen for the palatine ramus of the facial nerve; Nc neurocranial mineralizations; Pch parachordal mineralizations; R.Pq right palatoquadrate. Scale bar, 1 cm. Reprinted with permission of Brazeau (2009). Copyright Macmillan Press, Ltd

Fig. 3.7 Functional mandibular disparity among Silurian/Devonian gnathostomes. All horizontal axes show time as indicated at bottom. a Disparity (sum of variances) across eight time bins. The dark-grey region spans the 95 % conﬁdence intervals based on 1,000 bootstrap pseudoreplicates. b Relative contributions (partial disparity) of major gnathostome groups to overall functional disparity. Orange ‘‘Acanthodii’’; green Sarcopterygii; blue ‘‘Placodermi’’; yellow Chondrichthyes; red Actinopterygii. c Faunal composition data for the late Silurian and Devonian. Disks represent individual vertebrate assemblages plotted as a function of time and proportion of gnathostomes that comprise those faunas (disks jittered within time bins for clarity). The area of each disk is proportional to the total number of vertebrate genera represented, ngenera. Eifel. Eifelian stage; Givet. Givetian stage; Loc. Lockhovian stage; Lu. Ludlow series; Pr. Pridoli series; Pra. Pragian stage. Reprinted with permission of Anderson et al. (2011). Copyright Macmillan Publishers, Ltd 42 3 The Earliest Jawed Vertebrates, the Gnathostomes

ﬁshes did decline, there was no apparent indication that their jawed cousins took up new functional roles, which according to Anderson and colleagues calls into question old ideas of ecological replacement.

3.4 Evolutionary Signiﬁcance of the Oldest Vertebrates: The Origin of Bone

The success of vertebrates, which includes the considerable size that was attained in both marine and terrestrial members, is in large measure due to the mechanical properties that were imparted by an internal skeleton made of bone, although its original function was quite unconnected as we saw earlier. The most primitive type of growth of bone is seen in the dermal plates in which independent cyclomorial plates increase in size until they form a complete carapace. The next major stage is the condition in which the complete median plate is a single synchronomorial unit. The final stage of the unit can be acquired at an early point in life as a primordium around which cyclomorial growth can occur until the animal is fully developed. Halstead (1960) showed that the Downtonian heterostracan Corvaspis exhibited all three forms of growth patterns. Synchronomorial growth of the large median plates occurred in addition to which rows of small rounded tubercles developed by cyclomorial growth, the latter forming dermal plates by simple increase in size or becoming incorporated into the median plates. Thus, he was able to establish that what had been formerly considered to be three distinctive types of growth of the dermal plates in the Heterostraci were in fact all related (Fig. 3.8). Since the bony armor forming the outer covering of the animal was the main barrier between it and its environment, it can justifiably be regarded as skin. The preservation of the bony armor of the earliest vertebrates has made it possible to examine their microscopic structure in considerable detail. Halstead (1969b) reviewed the origins of calcified hard tissue among the earliest vertebrates, from which many details here are taken. The bone-like nature of the armor of the first vertebrates was established early on in its study. Agassiz (1845, for reference see Halstead 1969b) was among the first to note that the main tissue of the armor of Psammosteus was a homogeneous substance without bone-cell spaces. A short while later in a classic paper on the microstructure of early fossil vertebrates, which included the jawless vertebrate Pteraspis, Thomas Huxley (1858) noted that the armor of this animal possessed no bone-cell spaces. Lankester (1870) used this character to distinguish the group of jawless fishes, which he named the Heterostraci. Then in 1899 Traquair proposed that Psammosteus and Pteraspis should by virtue of their histology both be included in the Heterostraci, a view confirmed by subsequent work. Later Gross (1930, 1935), who gave a detailed and thorough account of psammosteid 3.4 Evolutionary Significance of the Oldest Vertebrates: The Origin of Bone 43

Fig. 3.8 Ornamentation of dermal plates. a Astraspis sp., showing independent polygonal tesserae and zones of tubercles produced by cyclomorial growth. b Poraspis sp., showing longitudinal dentine ridges produced synchronomorially. c Tolepelepis sp., showing individual units produced by cyclomorial growth, with adjacent synchronomorial area. From Tarlo Halstead (1960) microstructure, especially the specializations of its dentine, introduced the noncommittal term aspidin for the tissue making up the bony armor. The importance of aspidin as a form of vertebrate hard tissue is implicit in the considerable debate that was devoted to its structure and function over the years.

3.4.1 Aspidin

Aspidin forms the main part of the dermal armor in the heterostracans. It lies between the external dentine tubercles and the basal lamellar layers and consists of a spongy textured tissue with the macroscopic appearance of cancellous bone (Fig. 3.9). Although the general architecture of the heterostracans armor and bone seems to be formed in a similar way, there is controversy as to whether the hard tissue in the middle layer in the heterostracans should be designated bone or whether the noncommittal term aspidin should be used. It is worth spending a little time on two layers of heterostracan armor, starting with aspidin. Aspidin was shown to be made up of a three-dimensional scaffolding of cal- ciﬁed trabeculae of varying size with the apposition of successive lamellae. In some heterostracans the trabeculae form vertical walls that enclose polygonal spaces. In later heterostracans such as Psammosteus, in addition to the initial trabeculae there are concentric layers of lamellar aspidin surrounding the so-called vascular spaces, named aspidones by Gross (1961) (Fig. 3.9). There has been some controversy on the difference between aspidin and bone, mainly on the nature of the ﬁne structures preserved in aspidin. As Halstead noted (1969b), aspidin has generally been held to have been an acellular bone and as such has been considered primitive by both himself (Halstead-Tarlo 1963) and Denison (1963). Stensio (1927) and Orvig (1951), on the other hand, believed it to have been secondarily derived from cellular bone in the same way as the acellular bone of the 44 3 The Earliest Jawed Vertebrates, the Gnathostomes

Fig. 3.9 Block diagram of heterostracan armor, showing superﬁcial dentine tubercles, spongy aspidin, and basal lamellar aspidin (From Halstead 1969b)

modern teleosts. Later Orvig (1967) took up a neutral position on this question. Moss (1961), however, proposed that the spaces in aspidin were simply oblique sections of uncalcified fiber bundles, which had been mistaken for spindle-shaped spaces. Halstead (1969b) pointed out that the spindle-shaped spaces in aspidin had been interpreted by many workers as the former sites of bundles of collagen. Other workers claimed that they housed the scleroblastic cells responsible for the production of the organic matrix on which calcification took place; that they represented the lacunae of aspidinocytes (Halstead-Tarlo 1963, 1964; Tarlo 1965; Halstead 1969a, b), a view accepted by Obruchev (1964) and Novitskaya (1966). The idea that the spindle-shaped spaces in aspidin represented collagenous fiber bundles or Sharpey’s fibers held much support for many years although later research on the subject would show that the explanation was unlikely. Sharpey’s fibers are not components of internal systems but are usually found anchoring one type of tissue to another. This was vigorously opposed by Orvig who suggested that a new term, i.e., intrinsic fibers for those that are an integral part of bone tissue matrix. However, true Sharpey’s fibers of varying caliber are commonly found obliquely aligned in the basal layers of the heterostracan plates (Figs. 3.10 and 3.11). They are also found at the edges of plates and presumably held the latter firmly in the dermis. Hence, the idea of spaces in aspidin representing collagen fibers is not entirely lacking in merit Further evidence suggests that the spaces represented extensions of the cell processes of the aspidinocytes in an endeavor to maintain contact with the source of nutriment in the vascular spaces (Halstead 1969b). Fine tubules found in a 3.4 Evolutionary Significance of the Oldest Vertebrates: The Origin of Bone 45

Fig. 3.10 Weigeltaspis godmani Halstead Tarlo, sharpey’s ﬁbers in basal aspidin (x128). From Halstead (1969b)

Fig. 3.11 Heterostracan dentine. a Psammosteus megalopteryx (Trautschold), ornamentation of superficial tesserae, with fracture healed by line of new dentine tubercles, b Ganosteus stellatus Rohon, second generation dentine tubercles, developing aspidin and resorption of summits of first generation tubercles, c Tartuosteus maximus Mark-Kurik, pleromic dentine, d Ganosteus stellatus Rohon, ornamentation of dentine tubercles with second generation tubercles situated in resorption cavities. From Halstead (1969a) 46 3 The Earliest Jawed Vertebrates, the Gnathostomes number of heterostracans, e.g., the Ordovician Astraspis, were found to be similar to those in dentine and suggested that they were canaliculae, which had housed the cell processes of retreating osteoblasts. It seemed that these spaces marked the site of the aspidin forming cells which on producing the matrix were trapped within it to become the aspidinocytes (Halstead 1963). This led to Denison’s (1963) view that if it were to be accepted, it would mean that aspidin could readily change into trabecular dentine although Halstead (1969b) suggested a more parsimonious interpretation, i.e., that the difficulties in interpretation could be readily overcome by recognizing that at the very beginning of the evolutionary history of the vertebrates the tissues aspidin and dentine were hardly to be distinguished. In some ways the various views with respect to the spaces enclosed in the tissue aspidin reached a stalemate among some of the main players—they can be interpreted as having once housed bundles of collagenous fibers (e.g., Orvig 1951) or concordant in some points with cell processes (Denison 1963) or cell bodies (Halstead 1969b). What the various studies have shown, however, is that aspidin has all the attributes of bone and it is reasonable to consider it as a primitive type of bone or its precursor.

3.4.2 Dentine

For an animal to be aware of its surroundings and viable, its covering tissue must be sensitive. Dentine as we know is the material that gives teeth their sensitivity, to heat, cold, and pain and even pleasure. As in teeth the surviving system of radiating tubules which extend from the pulp cavity to the outer surface of the dentine and are the pathway for sensation, may well be connected with the original skin-like role of this tissue in the earliest vertebrates (Halstead 1969a). Dentine formed the superficial part of the earliest vertebrates. In the Ordovician heterostracans there appears to be two different types of dentine. In Astraspis there is a glassy, sculptured cap to the tubercles that has been identified as enamel (Bryant 1936), but which Orvig (1958) showed contained minute tubules and hence could be identified as dentine. While there has been debate on whether the outer surface of the earlier heterostracans was in fact dentine or a form of aspidin, in the majority of the later heterostracans the dentine tubules are notable for their modernity. The psammosteid Tartuosteus has tubules with fine lateral and terminal branches that are very reminiscent of human dentine (Fig. 3.12). The presence of tubules in dentine has been related by Tarlo (1965) to the probable original role of dentine as a tissue primarily concerned with sensitivity. For dentine to perform the role of sensitivity in the outermost skin layer in the heterostracans it would also be the most vulnerable and susceptible to damage. It follows that dentine must have possessed the property of repair. In early studies in the late 1900s it was considered that a notable feature of dentine tubercles was the presence of successive generations and that by analogy with tooth succession that 3.4 Evolutionary Significance of the Oldest Vertebrates: The Origin of Bone 47

Fig. 3.12 Asterolepis ornata EICHWALD. 1-6 juveniles individuals. 1a, 2a details of the dermal plate LMD 260/80 ; 1b trabecula and vascular canal ; 2b vascular canal ; 3, 4 squamation in visceral view ; 3 fulcral scales, LMD 260/236 ; 4 flank scale, LMD 260/236 ; 5 isolated flank scale in external view LMD 260/243 ; 6 flank scale in external view, LMD 260/244 from specimen LMD 260/10 ; 7 transversal cross-section of adult Asterolepis scale from zone D (caudal fin), LP 15/1. Trionyx sinensis Weigmann. 8 longitudinal section of a costal plate, subadult, VZ 2306A ; 9 transversal section of xiphiplastron, subadult, VZ 2305. From Ivanov et al. (1995) 48 3 The Earliest Jawed Vertebrates, the Gnathostomes the underlying tubercle was an un-erupted secondary tubercle. However, Gross (1930, 1935) clearly demonstrated that the new generations were positioned on top of the old and that the intervening space was filled by a mass of dentine through which meandered numerous long dentine tubules. Apparently, a network of epidermal tissue around the bases of the tubercles was stimulated so that it prolif- erated over the site of injury or damage. Beneath it new dentine tubercles developed—but, instead of simply replacing the worn or damaged units, they were positioned on top of the old surface, i.e., the second generation tubercles were the more superficial and situated over the resorbed cavities of the older tubercles as observed, e.g., in Ganosteus (Fig. 3.11). The placing of new generations on top of the old seems quite anomalous although other authors such as Jarvik (1959) have shown that this was unquestionably the case. This apparent reversal of the successional sequence compared with teeth will be explained further on. Recently, Alexandre Ivanov and colleagues (1995) used comparative anatomy in living soft-shelled turtles (Trionyx) to propose that the histological structure of the dermal ossifications of juvenile and adult Asterolepis (Placodermi, Antiarchi). Asterolepis material is represented by different stages of dermal bone development, i.e., stages of the dermal ossifications. They presented a hypothesized reconstruction of similarity of the dermal plates and scales development in Asterolepis (Fig. 3.12) compared with the complete cycle of the dermal ossification development in a complete development series (embryos and postembryos) of the living soft-shelled turtle (Trionyx sinenesis Weigmann). It is clear that even after a century of research on the jawless vertebrates, the controversy with respect to the origin of the dermoskeleton is far from over. Donoghue and Sansom (2002) reject all the main hypotheses for the origin of the skeleton. They state that what is overlooked is the point that the vertebrate skeleton has an invertebrate origin and that it not linked to developing an osmotic barrier, ion storage, or the storage of biolimiting elements, i.e., the earliest phylogenetic appearance of the skeleton is not linked to buffering acid by-products, disposing of waste products, sensory reception, or protection. Rather, the earliest representation of the vertebrate ‘‘skeletal system’’ they say relates entirely to feeding and respiration. While this they acknowledge brings together skeletonization with calcification, nevertheless they suggest that the distinction between the two is hard to draw. Rather they present a systematic framework in which the data they propose imply a scenario for the origin of the skeleton that is diametrically opposed to the hitherto prevalent ‘‘armor’’ hypothesis. Rather than arguing that the skeleton evolved to protect filter- feeding early vertebrates from predation, the phylogenetic hypothesis presented they argue is in favor of a hypothesis in which the skeleton evolved first to perform a feeding function in a predator/scavenger, and was only secondarily co-opted to perform a protective role among the ostracoderms. This leads directly to another problem connected with the early origin of vertebrates, the origin of teeth. 3.5 The Origin of Teeth 49

3.5 The Origin of Teeth

New generations of teeth, unlike as we have seen with the dermal tubercles, almost always erupt from below, resorbing the roots and never the summits of the first generation. However, as Halstead (1969a) pointed out, this seemingly fundamental difference is more apparent than real. In the early stages the germ of the second generation tooth is positioned above and to the side of the first, much the same spatial arrangement as found in the heterostracans tubercles. It is only later in development that the relative positions of the two teeth change and the new tooth comes to lie beneath the one it will eventually replace. The canonical view for the origin of teeth is that they derive from separate skin denticles during the evolution of jaws in vertebrates, i.e., from the bony dermal armor or external skin tubercles that originated in the jawless vertebrates during the Ordovician. This view, i.e., that the oropharyngeal denticles were derived secondarily from external skin denticles and moving into the mouth has been challenged by Zerina Johanson and Moya Smith (2003), who suggest that they evolved independently. Their theory is that sets of denticles on the pharyngeal (gill) arches were precursors of the organized tooth families developing within a specialized dental epithelium (dental lamina) along the jaw margin and not a consequence of denticles migrating from the outside. They base their conclusions on an intensive study of the early jawed vertebrates, the placoderms, and a number of later fishes in which they show that these denticles in the pharyngeal region clearly differ from the external tubercular dermal ornament covering the head and trunk shield plates. They propose that neither pharyngeal denticles nor teeth derive from the latter during the evolution of jawed fishes. Rather, they speculate that teeth might have originated three or more times among jawed vertebrates. While genes governing it may have arisen just once they suggest that the coordination that creates ordered patterns may be unique to each group (Fig. 3.13). Despite the rich record of fossil tooth-like structures, the actual sequence of events accounting for the evolution of oral versus dermal tooth-like units continues to evoke controversy. Recently, Fraser et al. (2010) proposed a somewhat new hypothesis on the origin of teeth from these tooth-like units. These they classified after Orvig as simply all structures that comprise a mineralized hard tissue unit consisting of attachment bone, dentine, and sometimes with a superficial layer of enamel/enameloid, formed from a single papilla—as odontodes, whether dermal or oral in origin. They suggest that the coordination that creates ordered patterns may be unique to each group of vertebrates. This is interesting with respect to our understanding the development of teeth across a wider range of jawed vertebrates. They considered the two hypotheses (i) that skin denticles evolved first and odontode-inductive surface ectoderm merged inside the oral cavity to form teeth (the ‘‘outside-in’’ hypothesis) and (ii) patterned odontodes evolved first from endoderm deep inside the pharyngeal cavity (the ‘‘inside-out’’ hypothesis) as proposed by Zerina Johnson and Moya Smith (2003) and others. However, the point in common between the two theories they suggested was that odontodes as structures 50 3 The Earliest Jawed Vertebrates, the Gnathostomes

Fig. 3.13 Tooth sets along the lower jaw of the chondrichthyan Carcharhinus melanopterus. a staggered or offset positions of tooth sets, particularly with regard to the alternation of the functional teeth at the jaw margin (functional teeth). Scale bar 1.0 cm. b Denticle whorl of the agnathan (jawless ﬁsh) Loganellia (Thelodonti). Scale bar _ 1.0 mm. From Johanson and Smith (2003)

shared a deep molecular homology, i.e., co-expressed genes that determine whether odontodes develop ‘‘inside and out,’’ wherever and whenever these co-expressed gene sets signal to one another, and importantly do not rely on a primacy of location, i.e., that teeth might have originated three or more times among jawed vertebrates although, as Johanson and Smith also proposed, genes governing it may have arisen just once. Both Johanson and Smith (2003) and Fraser et al. (2010) present interesting new insights to our understanding the development of teeth across a wider range of jawed vertebrates (Figs. 3.14 and 3.15).

3.5.1 Genes and the Origin of Teeth—Dermal or Oral

We have seen that the teeth are comprised of two important mineralized tissues: dentine and enamel. Cementum is the third component. There is growing evidence that genes encoding extracellular matrix proteins involved in the biomineralization of bone, dentin, and enamel diverged from a common ancestor gene. The initial event was the generation of SPARCL1 (Sparc-like 1) from SPARC (secreted protein, acidic, and rich in cysteine; BM-40/osteonectin) in a chromosome-wide large segmental duplication that spawned the chromosomal region ancestral to the long arm of human chromosome 4. The new gene (SPARCL1) gave rise to the secretory calcium-binding phosphoprotein (SCPP) gene family. The role of genes in distinctive roles, i.e., used in surface tissue or in body tissue was proposed by Kawasaki et al. (2007) who hypothesized that the duplication histories of SCPP genes and their common ancestors, SPARC, and SPARCL1 at around the same time that Paleozoic jawless vertebrates ﬁrst evolved mineralized skeleton. 3.5 The Origin of Teeth 51

Fig. 3.14 Theories of odontode evolution. Schematic diagrams represent a generalized (hypothetical) early vertebrate/ﬁsh in lateral/sagittal view: a1 Outside-in theory; ectodermal tissue is hypothesized to have integrated (green arrow) into the oropharyngeal cavity (opc), leading to the evolution of oral odontodes and subsequently oral and pharyngeal teeth. a2 Modiﬁed outside-in theory; ectodermal tissue integrated (green arrow) into the endodermal oral cavity via the mouth opening (the anterior boundary of the endoderm and ectoderm) and the gill slits (gs) in early vertebrates to initiate/transfer dental competence (arrow) to the endoderm of the oropharyngeal cavity. The point is made that ectoderm must be in regional contact with endoderm for teeth to form. b Inside-out theory; skin denticles and teeth are structures forming independently from ectoderm and endoderm, respectively. This theory states that teeth originated in the posterior pharyngeal endoderm of jawless vertebrates; a dental competence that was co-opted anteriorly (red arrow) in concert with the evolution of oral jaws. This theory states that skin denticles did not grade into teeth. e eye; n nasal placode; opc oropharyngeal cavity. From Fraser et al. (2010)

They suggested SPARCL1 arose from SPARC by whole genome duplication. Then both before and after the split of ray-finned fish and lobe-finned fish, tandem gene duplication created two types of SCPP genes, each residing on the opposite side of SPARCL1 giving them specific functions, i.e., the formation of surface or body tissue. Kawasaki et al. (2007) argue that the origin of SPARCL1 coincided with the innovation of mineralized skeleton and occurred in a genome-wide duplication in the stem jawed vertebrates (after the divergence of jawless fish). They suggest that the tooth initially co-opted the gene-regulatory circuits that had been already used for the development of dermal skeleton, and gene duplication has played a significant role in skeletal mineralization, both as it first arose and in sustaining it in current lineages. Gene duplication provides new genetic materials, which might be recruited by an adaptive trait. In contrast, Sire et al. (2007) argue for a much more 52 3 The Earliest Jawed Vertebrates, the Gnathostomes

Fig. 3.15 The inside and out gene regulatory hypothesis for odontode evolution a Schematic diagram represents a generalized early vertebrate/fish in lateral/sagittal view: Regardless of tissue origin (endoderm or ectoderm), the ingredients for odontode evolution, instigated by the appearance of the putative odontode gene regulatory network (oGRN), involved the collaboration of two pre-existing gene co-expression groups: (i) the neural crest-derived ectomesenchymal co- expression group (mesCEG) and (ii) the epithelial co-expression group (epCEG), which operates within both the endoderm and ectoderm b, c The evolution of both skin denticles and teeth were separate operations of the combination of epCEG and mesCEG in alternative locations, the epidermis and the oropharngeal cavity (opc). Within the opc, co-option of the oGRN potential was transferred to the oral jaws during the transition from jawless (agnathans) to jawed vertebrates (gnathostomes). From Fraser et al. (2010) ancient origin for SPARCL1. Living vertebrates include a great diversity of mineralized elements, comprising not only endochondral and dermal bone (including osteoderms and scutes), mineralized cartilage, and teeth (dentin and enamel), but also scales, fin rays, otoliths, and egg shells They propose that because enamel and enameloid were identified in early jawless vertebrates, about 500 million years ago (MYA) that it suggests that enamel matrix proteins (EMPs) have at least the same age. They propose that the evolutionary analysis of five genes (amelogenin (AMEL), ameloblastin (AMBN) and enamel in (ENAM), amelotin (AMTN), and odontogenic ameloblast associated (ODAM)) indicates that they are related: AMEL is derived from AMBN, AMTN and ODAM are sister genes, and all are derived from ENAM. Using molecular dating they showed that AMBN/AMEL duplication occurred more than 600 MYA (Fig. 3.16). 3.5 The Origin of Teeth 53 54 3 The Earliest Jawed Vertebrates, the Gnathostomes b Fig. 3.16 Vertebrate phylogeny and the evolution of skeletal mineralization. There are many extinct vertebrates (dashed lines) split from our ancestral lineage after the modern jawless vertebrates (lampreys and hagfish) but before cartilaginous fish. The most ancient mineralized tissue has been found in the feeding apparatus of conodonts (Donoghue and Sansom 2002, 2006). Subsequent jawless vertebrates evolved dermal skeleton. Cartilaginous fish have teeth but some placoderms, the first jawed vertebrates, may have independently innovated teeth (Johanson and Smith 2005). Two WGD are thought to have taken place, first in the stem jawless vertebrates (WGD1) and second in the stem jawed vertebrates (WGD2). SPARC descended from the common ancestor of protostomes and deuterostomes, whereas SPARCL1 arose through the WGD2. SCPP genes originated from SPARCL1 but have been found only in teleosts and tetrapods to date. Genes for acidic SCPPs and Pro/Gln-rich (P/Q) SCPPs arose before the divergence of ray-finned fish and lobe-finned fish. The vertebrate phylogeny is based on Donoghue et al. (2006). Tunicates are shown as the outgroup. From Kawasaki et al. (2007)

Fig. 3.17 Gnathostome genus level diversity curves for the Givetian to Serpukhovian (n = 1,018) (SI Appendix). Tetrapoda is deﬁned here as all taxa closer to crown Tetrapoda than Rhizodontida based on the tree in the article by Coates et al. (2008). This includes elpistostegalians. All other sarcopterygians are referred to Sarcopterygii. Reprinted from Sallana and Coates (2010). Courtesy PNAS, Open Access

3.6 Evolutionary Comments

From an evolutionary perspective there seems no clear answer as to why the placoderms died out and in turn why their roles were filled by chondricthians and actinopterygians. Was it simply that they were outcompeted by more advanced fishes or were there other factors at work? The results of a recent study by Lauren Cole Sallana and Michael Coates (2010) raised questions about this. It includes delving into the peri-Hangenberg extinction selectivity among gnathostomes and the failure of previously diverse clades to reradiate and the ultimate causes of this global event. The subsequent recovery of previously diverse groups (including placoderms, sarcopterygian fish, and acanthodians) was minimal. In contrast, among fishes, actinopterygians, and chondrichthyans, all scarce within the Devonian, undergo 3.6 Evolutionary Comments 55 56 3 The Earliest Jawed Vertebrates, the Gnathostomes b Fig. 3.18 Histogram of species-level faunal composition for 66 well-sampled macrofossil localities (n = 1,267) (Dataset S1). Localities are arranged temporally, although some are concurrent. Restricted intervals used in some analyses are noted above the time scale. End- Devonian sites are from formations conformable to Hangenberg sediments and are treated as contemporaneous. The Hangenberg and Kellwasser events are represented by black lines at the relevant stage boundaries. Reprinted from Sallana and Coates (2010). Courtesy PNAS, Open Access

large diversification events in the aftermath of the extinction, dominating all subsequent faunas. The authors suggest that significantly, the post-Hangenberg configuration of vertebrate biodiversity persists to the present day: chondrichthyans, actinopterygians, and tetrapods thrived. They conclude that narrative explanations of early vertebrate evolution with only passing reference to a long-term biotic crisis toward the end of the Devonian, the Hangenberg event, which they emphasize represents a previously unrecognized bottleneck in the evolutionary history of vertebrates as a whole that shaped the roots of modern biodiversity, are no longer sufficient (Sallana and Coates 2010) (Figs. 3.17 and 3.18).

References

Anderson PSL, Friedman M, Brazeau MD, Rayfield EJ (2011) Initial radiation of jaws demonstrated stability despite faunal and environmental change. Nature 476:206–209 doi:10.1038/nature10207 Brazeau MD (2009) The braincase and jaws of a Devonian ‘acanthodian’ and modern gnathostome origins. Nature 457:305–308 Bryant WL (1936) A study of the oldest known vertebrates, Astraspis and Eriptychius. Proc Amer Phil Soc 76:409–427 Coates MI, Ruta M, Friedman M (2008) Ever since owen: changing perspectives on the early evolution of tetrapods. Annu Rev Ecol Evol Syst 39:571–592 Colbert EH (1955) Evolution of the vertebrates. Wiley, New York Denison RH (1963) The early history of the vertebrate calcified skeleton. Clin Orthop 31:141–152 Donoghue PCJ, Sansom IJ (2002) Origin and evolution of vertebrate skeletonization. Microsc Res Tech 59:352–372 Donoghue PC, Sansom PCJ, Downs JP (2006) Early evolution of vertebrate skeletal tissues and cellular interactions, and the canalization of skeletal development. J Exp Zool B (Mol Dev Evol) 306:278–294 Fraser GJ, Cerny R, Soukup V, Bronner-Fraser M, Streelman T (2010) The odontode explosion: the origin of tooth-like structures in vertebrates. BioEssays 32:808–817. doi:10.1002/ bies.200900151 Gross W (1930) Die Fische des mittleren Old Red Sub-Livlands. Geol Paläont Abh, N F 18:121–156 Gross W (1935) Histologische Studien am Aultenskelett fossiler Agnathen und Fische. Palaeontographica 83A:1–60 Gross W (1961) Aufbau des Panzers obersilurischer Heterostraci und Osteostraci Norddeutsch- lands (Geschiebe) und Oesets. Acta zool (Stoekh) 42:73–150 Halstead LB (1969a) Pattern of Vertebrate Evolution. Oliver and Boyd, Edinburgh Halstead LB (1969b) Calcified tissues in the earliest vertebrates. Calc Tiss Res 3:107–134 References 57

Halstead-Tarlo LB (1963) Aspidin: the precursor of bone. Nature (Lond) 199:46–48 Halstead-Tarlo LB (1964) The origin of bone. In: Blackwood HJJ (ed) Bone and tooth. Pergamon Press Oxford, New York, p 3–15 Hildebrand M (1995) Analysis of Vertebrate structure. Wiley-Blackwell, Wiley, New York Huxley TH (1858) On Cephalaspis and Pteraspis. Quart J geol Soc Lond 14:267–280 Ivanov A, Cherepanov G, Luskevics E (1995) Ontogenetic development of antiarch dermal ossifications. Geobios 19:97–102 Jarvik E (1959) Dermal fin-rays and Holmgren’s principle of delamination. Kgl svenska Vet- Akad Handl 4:1–51 Jarvik E (1977) The systematic position of acanthodian fishes. In: Andrews SM et al. (eds) Problems in vertebrate evolution. Academic Press, London pp 199–255 Johanson Z, Smith MM (2003) Placoderm fishes, pharyngeal denticles, and the vertebrate dentition. J Morphol 257:289–307 Johanson Z, Smith MM (2005) Origin and evolution of gnathostome dentitions: a question of teeth and pharyngeal denticles in placoderms. Biol Rev Camb Philos Soc 80:303–345 Kawasaki K, Buchanan A, Kenneth V, Weiss M (2007) Gene duplication and the evolution of vertebrate skeletal mineralization. Cells Tissues Organs 186:7–24. doi:10.1159/000102678 Lankester ER (1870) A monograph of the fishes of the Old Red Sandstone of Britain. Part 1. The Cephalaspidae Palaeontogr Soc 1868:1–62 Mallatt J (1996) Ventilation and the origin of jawed vertebrates: a new mouth. Zool J Linn Soc 117:329–404 Miles RS (1964) A reinterpretation of the visceral skeleton of Acanthodes. Nature 204:457–459 Miles RS (1965) Some features in the cranial anatomy of Acanthodians and the relationships of the Acanthodii. Acta Zool 46:233–255 Miles RS (1973a) Articulated acanthodian fishes from the Old Red Sandstone of England, with a review of the structure and evolution of the acanthodian shouldergirdle. Bull Br Mus Nat Hist (Geol) 24:111–213 Miles RS (1973b) Interrelationships of fishes. In: Greenwood PH, Miles RS, Patterson C (eds), p 63–103 Miles RS (1977) Dipnoan (lungfish) skulls and the relationships of the group: a study based on new species from the Devonian of Australia. Zool J Linn Soc 61:1–328 Moss ML (1961) The inital phylogenetic appearance of bone : an experimental hypothesis. Trans NY Acad Sci 23:495–500 Nelson JS (2006) Fishes of the world. Wiley, New York Obruchev D (1964) Subclass Heterostraci (Pteraspides). Osn Paleont 20:45–82 Ørvig T (1951) Histologic studies of Placoderms and fossil Elasmobranchs. I: The endoskeleton, with remarks on the hard tissues of lower vertebrates in general. Ark Zool 2:321-454 Ørvig (1958) Pycnaspis splendens, new genus, new species, a new ostracoderm from the Upper Ordovician of North America. Proc US Nat Mus 108:1–23 Ørvig T (1967) Phylogeny of tooth tissues: evolution of some calcified tissues in early vertebrates. In: Miles AEW (ed) Structural and Chemical Organisation of Teeth. Academic Press, New York/London Novitskaya LI (1966) The microstructure of certain Psammosteida. In: Obruchev D, Mark-Kurik E (eds) Psammosteids (Agnatha, Psammosteidae) of the Devonian of the USSR. Tallinn: Geological Institute, Academy of Sciences of Estonian SSR, p 257–28 Sallana LC, Coates M (2010) End-Devonian extinction and a bottleneck in the early evolution of modern jawed vertebrates. PNAS 107:10131–10135 Sire J-Y, Donoghue PCJ, Vickaryous MK (2007) Origin and evolution of the integumentary skeleton in non-tetrapod vertebrates. J Anat 214:409–440 Stensio EA (1927) The Devonian and Dowtonian vertebrates of Spitsbergen. I. Family Cephalaspidae. Skr Svalbard Ishav 12(1927):1–391 Tarlo Halstead LB (1960) The Downtonian ostracoderm Corvaspis kingi Woodward, with notes on the development of the dermal plates in the Heterostraci. Palaeontology 3:217–226 58 3 The Earliest Jawed Vertebrates, the Gnathostomes

Tarlo LB (1965) Psammosteiformes (Agnatha), a review with descriptions of new material from the Lower Devonian of Poland. I. Gen Part Palaeont Pol 13:1–13 Traquair R (1899) Report on fossil ﬁshes collected by the geological survey of Scotland in the Silurian Rocks of the South of Scotland. Trans Roy Soc Edinb 39:827–864 Young GC (2010) Placoderms (Armored Fish): dominant vertebrates of the devonian period. Annu Rev Earth Planet Sci 38:523–550 Watson DMS (1937) The Acanthodian ﬁshes. Phil Trans R Soc B 228:49–146 Chapter 4 Evolution of Modern Fishes: Critical Biological Innovations

Modern fishes comprise the largest and most successful group of all vertebrates on the planet. The higher fishes are divided into two major groupings, Chondrichthyes or cartilaginous fishes and the Osteichthyes or bony fishes. Their extraordinary rise is marked by among other things important anatomical and mechanical developments that went beyond those of the placoderms with which they originally shared the oceans. Both groups originated in the Devonian. Chondrichthyans, as with the sarcopterygian and actinopterygian divergence of the bony fishes (below), also belong to two major subclasses, the Elasmobranchii (thought to arise from a group of placoderms near the Stensioelliformes) to which sharks belong, and the Holo- cephali (believed to arise from certain arthrodires), to which chimaerids belong. Goujet (2001) proposed a number of advanced conditions in support of the relationship of placoderms as sister group of chondrichthyes (Fig. 4.1; note, cyclostome monophyly now seems better corroborated than previously (Philippe Janvier personal commun. 2013)). At present fossil evidence showing intermediate stages in development between Osteichthyes and Chondrichthyes is tenuous and the relationships are uncertain. Colbert (1955) suggested that although sharks are generally considered to be primitive fishes it is doubtful that they are more truly primitive than the bony fishes, Actinopterygii, and that the view may be based on the idea that the cartilaginous skeleton of sharks is considered more primitive than the bony skeleton of other fishes. Given that cartilage and its occurrence in sharks appears later in the fossil record than bone and bony fishes it seems reasonable to think that the cartilaginous skeleton of sharks is a secondary development and that bone with its older history is truly primitive. This question is not as simple as it first appears (Donoghue et al. 2006) and is discussed at greater length further on. It is clearly a topic that generates much debate still. In this context, the order of the following discussion is not meant to have any evolutionary implications. We begin with sharks for no other than practical reasons—the bony fishes ideally follow because they lead better into the succeeding section namely the tetrapod invasion of land.

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 59 DOI: 10.1007/978-3-642-53748-6_4, Ó Springer-Verlag Berlin Heidelberg 2014 60 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.1 Interrelationships of the major living and fossil craniates. Arandaspids, astraspids, and eriptychiids are not shown here as they are currently included with heterostracans in the clade Pteraspidomorphi. After Janvier (2001) 4.1 Chondrichthyes, Advanced Cartilaginous Fishes 61

4.1 Chondrichthyes, Advanced Cartilaginous Fishes

Chondrichthyes (cartilaginous fishes) arose during the early Paleozoic. They radiated during the Mesozoic and continue to be well represented to the present. Of the chondrichthyans namely the sharks, rays, and chimaerids, the sharks by the Carboniferous made up about 60 % of the species of fishes in shallow tropical habitats (Lund 1990). Their evolution followed two major lines of development that have led to their great success. They involve body structure and behavior—a streamlined body form that includes strategic location of the dorsal, pectoral, and pelvic fins and a large high aspect ratio caudal fin and an aggressive predaceous lifestyle. The body form was honed in sharks to produce what is frequently referred to as the ultimate predator in the world, the white shark, Carcharodon carcharias (Lingham-Soliar 2005a, b, c) (Fig. 4.2). Shark swimming biomechanics will be dealt with in Volume 2.

4.1.1 Subclass: Elasmobranchii

The first true shark-like fossils of elasmobranchs occur in the early Devonian. One of the first sharks from one of the two extinct infraclasses (Cladoselachimorpha) was Cladoselache, the other was the freshwater Xenacanthus (Xenacanthimorpha), which was found in the upper Devonian Cleveland shales. Cladoselache had a typical shark-like streamlined, elongated body shape, and a large heterocercal tail (Figs. 4.3 and 4.4). The head was pointed, with a ventrally located mouth fur- nished with large sharp teeth, each of which consisted of a high central cusp and low lateral cusps on each side. Opposed to this was the single element of the lower jaw or mandible. To compensate for the protection that dermal bones gave to the brain in ostracoderms and placoderms, the chondrocranium became a highly solid structure, despite never ossifying, comprising lateral walls and a roof. The jaw articulation was amphistylic (involving the hyoid and quadrate bones and relatively primitive with just two articulations to the braincase, which consisted of a single element the palatoquadrate or, it may be hyostylic (Fig. 4.5). There are two fairly low dorsal fins. The pectoral and pelvic fins were well developed although the broad bases suggested that their movements were rather restricted. On the other hand the large size of the pectoral fins suggests that they must have been of considerable importance in maintaining stability and in steering. The structural pattern of Cladoselache, while primitive for most sharks, is important because it shows the bauplan from which modern sharks would develop. This would take place with the advent of the Mesozoic era, particularly during the Jurassic period when many of the features that define the evolutionary developments of modern sharks, and which would lay the foundations for their extraordinary success, occurred. Such features would include significant changes that would greatly increase swimming speed and maneuverability namely a more 62 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.2 The white shark, Carcharodon carcharias. After Lingham-Soliar (2005b)

Fig. 4.3 Diversity in the body form of Paleozoic sharks from two extinct infraclasses. a Cladoselache a cladoselachid (Cladoselachimorpha). b Xenocanthus, a xenocanthid (Xeno- canthimorpha). After Helfman et al. (2009) hydrodynamic body shape and greater mobility and control of the pectoral and pelvic fins, aided by basal fin supports and flexible rays (ceratotrichia) supporting the fin web (rather than a more-or-less single mass of cartilage). The attachment to 4.1 Chondrichthyes, Advanced Cartilaginous Fishes 63

Fig. 4.4 Sharks allied with the infraclass Euselachii. a Ctenacanthus, an Upper Devonian ctenacanthid (Ctenacanthiformes); b Hybodus, a hybodontid, representative of the order Hybodontiformes, the most diverse elasmobranch group in the Triassic and Jurassic; and c Squalus, a modern squaliform shark in the division Neoselachii. After Helfman et al. (2009) the body also became much narrower enabling greater freedom and control of movements, steps leading to the highly maneuverable ﬁns of modern sharks. With respect to feeding, these sharks also had a selective advantage over the early gnathostomes. In contrast to the placoderms and most acanthodians, sharks evolved a tooth replacement system. Teeth grew in whorls in which the foremost exposed functional teeth were backed up by several replacement teeth that developed from basic germinal forms as they moved forward to replace the functional teeth at the front that were either lost or damaged, in effect behaving as in a chain-belt (see Chap. 3). 64 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.5 Principle types of jaw suspension. Mandibular arch and derivatives are hatched, hyomandibular and derivatives are black, chondrocranium is shaded, teeth and extent of membrane bones are indicated in outline. The types of suspension may intergrade. After Hildebrand (1995)

Elasmobranch infraclasses and developmental patterns are to a large part deﬁned by skull structure and tooth forms (Helfman et al. 2009). From the three infraclasses the only surviving one, the Euselachii, contained one extinct Order, Ctenacanthiformes, which contained extinct Division Hybodonta, and one extant Order Hybodontiformes, which contained the extant Division NeoSelachii (modern sharks and rays) (Fig. 4.6). 4.1 Chondrichthyes, Advanced Cartilaginous Fishes 65

Fig. 4.6 Phylogenetic relationships among living chondrichthyans. Relationships among the batoid rays remain a matter of debate, including discussion of whether the rhinobatiform guitarﬁshes are in fact monophyletic. After Helfman et al. (2009)

4.1.2 Subclass: Holocephali

Holocephalans and elasmobranchs are considered to be a monophyletic group. Nevertheless, holocephalans differ in significant respects from elasmobranchs. Notably, is a single opercular opening for the four gill openings, non-protrusible jaws because jaw suspension is strengthened by being firmly fused with the 66 4 Evolution of Modern Fishes: Critical Biological Innovations braincase (autostylic suspension) as adaptation to feeding on shellfish whereas in elasmobranchs the mobility is gained by a hyostylic jaw suspension (Fig. 4.5). Modern chimeras belong to a previously diverse clade, most diversification occurring in the Mesozoic but with only one order surviving beyond the close of the Cretaceous. Along with ratfishes (and most fossil forms) the teeth consist of six pairs of grinding plates on the jaw margins. The tail form is often diphycercal and the dorsal fin and spine are erectile. A common ancestral lineage of the modern holocephalans originated in the late Silurian about 420 MYA, having survived from the end-Permian (Lund 1990; Inoue et al. 2010). It has been suggested that holocephalans are closely related to the Selachii because both selachians and holocephalians have many characters in common, such as placoid scales, pelvic claspers, and absence of true bone. It has also been suggested recently that both holocephalians and selachians are related to the acanthodians on the basis of the gill arch structures (Nelson 2006) (Fig. 4.7). In the 1970s contributions to acanthodian classification such as by Denison, Miles, and Jarvic (see Chaps. 2 and 3) have variously proposed that they are most closely related to the elasmobranchs or that they are the sister group to chondrichthyans, placoderms, and osteichthyes or the modern view of Nelson (2006) that they are the sister group to the Eutel- eostomi (Sarcopterygii and Actinopterygii) (Fig. 4.7).

4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes

The Osteichthyes or bony fishes represent the greatest class of fishes. They are divided into three distinct classes, the acanthodians, sarcopterygians, and actinopterygians, the latter two historically (albeit without official rank) referred to as Osteichthyes, literally ‘‘bony fishes.’’ Actinopterygii (ray-fins) are comprised of bichirs, sturgeons, gars, bowfins, and teleosts, and the sarcopterygians (fleshy-fins) of coelacanths, lungfishes, and tetrapods. The success of fishes is reflected by the fact that their total numbers are more than half of the total number of approximately 55,000 recognized living vertebrate species (Nelson 2006). In contrast to other classes, the bony fish have a well-ossified internal skeleton (except in the sturgeons and spoonbills), lower jaws connected to the cranium through the hyoid arch, and swim bladders or lung structures.

4.2.1 Class: Acanthodii

Acanthodians or spiny sharks are an extinct group from the late Ordovician with no extant members. They contain the earliest known jawed vertebrates. They survived until the Early Permian about 100 million years after the major 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes 67

Fig. 4.7 Acanthodian fishes. This group is characterized by conspicuous fin spines anterior to the pelvic, pectoral, anal, and dorsal fins. Primitive genera have several pairs of spines without accompanying fin structures between the pectoral and pelvic fins. a Climatius, family Climatiidae, in lateral view. b Euthacanthus, family Climatiidae, in ventral view. c Diplacanthus, family Diplacanthidae. d Ischnacanthus, Ischnacanthidae. e Homalacanthus, family Acanthod- idae. f Acanthodes showing internal skeleton. After Carroll (1988) ostracoderms groups. They are identified by the stout median and paired spines. Early acanthodians had multiple gill covers, broad unembedded spines anterior to all fins except the caudal, as well as additional spines between the pectoral and pelvic fins (Fig. 4.8). Acanthodians apparently possessed a less efficient tooth replacement and tooth structure than the sharks and the bony fishes, similar to the Placodermi, although a close relationship with this group is unlikely (Helfman et al. 2009). The relationships of the acanthodians to other jawed vertebrates are also obscure, given that they possess features found in both sharks and bony fishes. Like early bony fishes they possessed ganoid-like scales and a partially ossified internal skeleton and certain aspects of the jaw appear to be more like those of bony fishes than sharks. On the other hand, the bony fin spines and certain aspects of the gill apparatus would seem to favor relationships with early sharks. Such similarity to modern sharks is, however, considered by most authors to be superficial and few believe they are related to modern chondrichthyans (see Jarvik 1980). For a general descriptions of this diverse group see Carroll (1988). 68 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.8 Cladogram showing the relationships of the extant actinopterygians as presented here. The Clupeomorpha and Ostariophysi compose the subdivision Ostarioclupeomorpha (=Otocephala), sister to the Euteleostei (many fossil clades omitted). After Nelson (2006)

4.2.2 Class: Actinopterygii

The Actinopterygii are the most successful group of fishes today. They are among the most ancient of fishes and are readily identified with the primitive ray-finned fishes in size and shape, which is helpful in establishing their taxonomy and functional biology (Fig. 4.7). 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes 69

4.2.3 Subclass: Cladista

Formerly two subclasses were recognized, the Chondrostei and the Neopterygii. However, more recently, and not without controversy (Nelson 2006; Helfman et al. 2009), a third subclass, the Cladista, was established to contain modern polyp- teriforms (birchirs and reedﬁsh).

4.2.4 Subclass: Chondrostei

The Chondrostei have a plus-300 million year history. Their origins are obscure, possibly late Silurian albeit based only on fragmentary scales. This would make them older than the sarcopterygians and as old as the placoderms and elasmobranchs with only the acanthodians among jawed, bony fish older. Only with complete fossil finds from the Mid to Late Devonian was it possible to distinguish chondrosteans from sarcopterygians by their possession of a single triangular dorsal fin, asymmetrical tail, thick rhomboidal, ganoid scales that extended onto the fins, and ridge scales along the back (Helfman et al. 2009) (Fig. 4.9). They gave rise to a great variety of types, with elongate bodies and jaws, bottom-living types that fed on microorganisms, deep-bodied marine reef fishes, and coral-eating reef fishes. The group (collectively known as palaeoniscoids) increased in number and complexity and flourished throughout the latter part of the Paleozoic when the ostracoderms, acanthodians, and placoderms had disappeared and sarcopterygians diminished in abundance. Thereafter they declined, becoming almost extinct, by the middle of the Cretaceous. The most diverse order of chondrichthyans, the Palaeonisciformes, are considered the most basal actinopterygian stock from which all other chondrosteans and the holosteans evolved. They were the most common fishes of their time. The evidence suggests that changes in the jaw, including reorientation of the hyomandibular from the oblique to the vertical, and fin structure (Helfman et al. 2009 and references therein) (Fig. 4.10) led to diversified feeding habits and increased mobility and to actinopterygian success and dominance. Other changes were in scale structure from heavy, thick interlocking diamond-shaped units to thinner, lighter, circular cycloid forms. The reduction was achieved by elimination of the dentine, vascular, and ganoine layers (see Chap. 3). The condition of jointed scales and reduction in scale thickness meant that the fins (further aided by the dermal fin rays) became flexible, highly mobile structures in both vertical and lateral directions. This was followed by increased ossification of the vertebral column, with development of distinct centra and neural and haemal arches. Finally, modifications in the caudal region followed the trend toward a symmetrical, homocercal tail. All these modifications would lead to a body plan in fishes with maximum 70 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.9 Actinopterygian fishes at different grades of development. a Moythomasia and b Mimia, two primitive palaeoniscoid fishes from the Upper Devonian, with thick rhomboidal scales extending onto the fins, broadly triangular dorsal and anal fins, fulcral (ridge) scales along the back, a long mouth, and an asymmetrical heterocercal tail. c Parasemionotus, a pre-teleostean neopterygian from the Triassic, showing more flexible fins, shorter mouth, and abbreviate heterocercal tail. d Eolates, an advanced euteleost from the Lower Eocene, with characteristic teleostean diversified dorsal and anal fins, shortened vertebral column, premaxillary dominated upper jaw and homocercal tail. After Helfman et al. (2009) swimming efficiency and maneuverability, which together with more efficient feeding mechanism, would produce more advanced prey and predatory forms. Despite the important trends in paleoniscoid evolution, only one order has survived to the present with 10 becoming extinct. Surviving Chondrostei are the bottom-feeding marine and freshwater sturgeons, the strange plankton-feeding paddlefishes of the Mississippi of North America and the Yangtze River of China. As we saw earlier the birchirs and reedfishes formerly in this subclass were placed in the new subclass, Cladista. Nevertheless, the trends they initiated, would continue with even greater improvements later in the more advanced actinopterygians, in the subclass Neopterygii, a group that may be considered the most successful of all vertebrates. 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes 71

Fig. 4.10 Morphological (and ecological) convergence in fish evolution. Palaeoniscoids were ancestral to early neopterygians, which were ancestral to modern teleosts. Certain body designs or plans have apparently been repeatedly favored in actinopterygians leading to convergent designs among unrelated lineages. These striking convergences in body shape and presumably function are depicted for representative palaeoniscoids, early neopterygians, and teleosts. 1 Elongate piscivores with long tooth-studded jaws and dorsal and anal fins placed posteriorly for rapid starts; 2 compressed-bodied, predatory, shallow water fishes with deeply forked tails and trailing fins; 3 broad-finned bottom feeders, with subterminal mouths; 4 eel-like benthic forms; and 5 compressed circular forms and large fins for maneuverability in shallow water habitats with abundant structure. Gliding fishes such as the Triassic chondrostean Thoracopterus can also be equated with modern teleostean flyingfishes. After Helfman et al. (2009) 72 4 Evolution of Modern Fishes: Critical Biological Innovations

4.2.5 Subclass: Neopterygii

The Neopterygii represents the great majority of living fishes and the culmination of the long evolution of a bauplan that maximized swimming and feeding efficiency. It is safe to say that they continued the evolutionary developments from where the primitive actinopterygians left off and consequently the latter’s replacement by more advanced forms. The descendants, the neopterygians (literally ‘‘new fins’’) indicate the dramatic change in tail shape from what was still structurally heterocercal with a fairly stiff axial lobe that resulted in an asymmetrical or unequal driving because the driving force is unevenly distributed relative to the body axis (Wilga and Lauder 2002). An initial radiation of neopterygians during the Triassic and Jurassic and even greater expansion in the Cretaceous would lead to numerous modifications toward the teleost condition among which would be changes that brought about equal flexibility of the upper and lower lobes and greatly improved tail mechanics. The steady improvement in tail shape over 400,000,000 years is one of the prominent features of fish evolution. In primitive fishes the tail (vertebral) axis turned upward (heterocercal) or downward (hypocercal) and a lobe of flesh pro- jected from it (Fig. 4.11). Figure 4.11 shows the evolutionary developments from pre-teleosteon neopterygians to modern teleosts and shows the major evolutionary refinements in structure and function in the different groups such as diversification of the dorsal and paired fins (Figs. 4.12 and 4.13) (see Volume 2 on thunniform swimming). Modern teleosts show a reduction of bony elements compared to pre-teleostean groups (Nelson 2006) through fusion or loss of bones. Higher teleosts have the following features (Helfman et al. 2009). (1) There are fewer but more ossified vertebrae with the percomorphs showing the greatest reduction of 20–30. A shorter more ossified axial skeleton allows for the attachment of a stronger trunk musculature, which enhances locomotion (see Volume 2). (2) There are fewer vertebral accessories, e.g., ribs and intermuscular bones and replacement of the latter with fewer, thicker zygopophyses. (3) There are fewer bones in the skull. (4) There is a reorganization and reduction in the number of bones in the tail, including fusion of the supporting bones (epurals, hypurals, centra) and a reduction in the number of fin rays in the tail. (5) There is a reduction in the number of biting bones in the upper jaw from two to one. (6) There is a reduction in the number of fin rays in paired fins. (7) There is a reduction in the amount of bones in scales (see Sect. 4.3). One probable reason was a shift from mechanical protection to increased mobility that favored lighter quicker fishes with improvements in both predation and predator avoidance. 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes 73 74 4 Evolution of Modern Fishes: Critical Biological Innovations b Fig. 4.11 A phylogeny of some actinopterygian fishes showing tail morphologies of living and fossil forms within the neopterygian clade (rooted at N). The fossil suggest that the morphologies ancestral to the bowfin and gar lineages are much more symmetrical than their living representatives reveal. Thus it would appear that imperfect caudal symmetry (lepidotrichial symmetry between the top and bottom corners of the tail but with fulcral scales above the body axis) originated at the base of the Neopterygii, and that the extant gar and bowfin are not so intermediate in morphology between teleosts and primitive actinopterygians as they might seem to be. This example also demonstrates the importance of incorporating fossil data into comparative analysis whenever possible. No consensus exists as to the relationships between the major neopterygian clades. After Metscher and Ahlberg (2001)

In the evolution of neopterygians the dorsal fin shows a number of changes in structure position and use (Helfman et al. 2009). The developments are most apparent during teleostean phylogeny (Fig. 4.14). The pectoral fins move further up from an abdominal position onto the sides of the body. These relocations of the pectoral fins have several functions such as slow swimming and hovering, and backing in the water. Placement of the pelvic fins forward helps in braking and reduces pitching. A major problem in the evolution of high-speed swimming fishes and other vertebrates involves oscillations of the tail, which induce recoil or lateral oscillations at the head. The location of the dorsal fin and increase in its surface area to counteract these forces anteriorly, and the part played by the dermis, are further characteristics of neopterygian evolution. This, together with the progressive development of a symmetrical tail fin leading to the homocercal condition and ultimately to the crescent-shaped fin of the fastest swimmers in the ocean (see Volume 2).

4.2.6 Class: Sarcopterygii

The Sarcopterygii are extremely ancient in origin with members appearing in the early Devonian. The oldest articulated primitive fish from the Ludlow of Yunnan, China, Meemannia eos, that represents the oldest near-complete gnathostome (jawed vertebrate) provides insights into the origin and early divergence of osteichthyans, and indicates that the minimum date for the actinopterygian–sarcopterygian split was no later than 419 million years ago (Zhu et al. 2010). The ancestral sarcopterygians are one of the most actively studied fossil groups of fishes, not least because of their place in tetrapod evolution. While there is still much debate the following summary of the major hypotheses is taken from Forey (1998), presenting 13 different phylogenies by different authors in the past 30 years and Forey’s concluding analysis is presented in his cladogram (Fig. 4.15) and revolves largely around the relative positions of lungfish (subclass Dipnoi), coelocanths (subclass Coelacanthomorpha) and the osteolepiform-porolepiform- panderichthyid (subclass Tetrapodomorpha) lineages relative to tetrapods (see Chap. 6). 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes 75

Fig. 4.12 Diversification of the dorsal fin in modern teleosts. a Primitively the dorsal fin is a single, spineless, subtriangular structure that serves as an anti-roll device and pivot poin during swimming as in herrings (Clupeidae). However, this simple fin has been greatly modified in more advanced fishes and can serve in locomotion, predator protection, and many other functions. b In cods (Gadidae) there are three dorsal fins. c More commonly a spiny anterior and soft rayed posterior separation exists as in the the squirrelfishes (Holocentridae). d In frogfishes (Antenar- iidae) modified dorsal spine serve as lures and camouflage. e The sucking disk of the sharksucker (Echeneidae) is derived embryologically from the spiny dorsal fin. After Helfman et al. (2009) 76 4 Evolution of Modern Fishes: Critical Biological Innovations 4.2 Osteichthyes, Advanced Bony Fishes: Teleostomes 77 b Fig. 4.13 The phylogeny of paired fin locations in teleosts. The locations and functions of the pectoral and pelvic girdles have changed during evolution of the Teleostei. Pectoral fins move from a ventral to a lateral position and the pectoral fin base changes its orientation from horizontal to vertical. Pelvic fins move from abdominal to thoracic and even jugular locations. Extant representatives of phases in this observed trend are represented as follows. a An elopomorph (bonefish, Albulidae). b A primitive paracanthopterygian (troutperch, Percopsidae). c A generalized acanthopterygian (cichlid, Cichlidae). This trend is by no means absolute. Many specialized, relatively primitive teleosts have laterally placed pectorals (e.g., catfishes) and advanced teleosts may have pelvics in abdominal positions (e.g., atherinomorphs), but overall the trends describe a progressive change during teleostean phylogeny. After Helfman et al. (2009)

The tetrapodamorpha comprised the Rhizodontiformes, the Osteolepidiformes, and the infraclass the Elpistostegalia, which is particularly important to us as tetrapods. Tetrapodomorphs were large predatory fishes that are characterized by sarcopterygian traits, such as two dorsal fins, cosmine covered bones and scales, kinetic skulls, fleshy lobed fins, and replacement teeth on the jaw margins. Most is known of the osteolepidiforms, especially Eusthenopteron foordi (Jarvik 1980) because of excellently preserved material (Figs. 4.16 and 4.17). However, although the osteolepiforms possessed many homologues with later tetrapods they are considered unlikely to have been transitional forms to living on land. Rather, it is the elpistostegalians that are generally considered the most likely sister group of modern tetrapods (Helfman et al. 2009). The transition from lobe- finned aquatic sarcopterygians to limb and digit-bearing stem tetrapods (tetrapodomorphs; Ahlberg 1991) is thought to have been narrowed further in recent years with the discovery and description of the elpistostegalian Tiktaalik roseae (Daeschler et al. 2006; see Chap. 6) and several intermediate or more tetrapod-like skeletal features. Coinciding with the acquisition of skeletal features permitting an increasingly terrestrial existence, the integumentary skeleton of tetrapodomorphs underwent a number of important changes. These pivotal changes leading to tetrapods and life on land is the subject of much controversy, which will be discussed in the next chapter. Before that we will look at below some crucial characteristics that are associated with the evolution of fishes.

4.3 Scales: Organization of the Integumentary Skeleton in Gnathostomes

Fish scales are formed of bone from the deeper, dermal, skin layer, unlike those of reptiles, which are epidermal (see Chap. 6). The skull and integumentary elements of placoderms (Silurian to late Devonian, *435–360 MYA) are characterized by cellular bone covered by a cellular dentine with polarized cell processes (semidentine). The scalation of chondrichthyans that includes the elasmobranchs (e.g., sharks) are placoid scales or dermal denticles; these are bony, spiny projections with a superﬁcial layer of enameloid (=odontodes (placoid scales)), 78 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.14 Phylogenetic relationships among the actinopterygian ﬁshes. After Helfman et al. (2009) 4.3 Scales: Organization of the Integumentary Skeleton in Gnathostomes 79

Fig. 4.15 An interpretation of relationships among euteleostome fishes (‘Osteichthyes’) showing actinopterygians as the sister group to the various extant and extinct sarcopterygian taxa. After Forey (1998) capping an orthodentine crown, attached to the dermis by bone (Vickaryous and Sire 2009). A pulp cavity lies within the scales. Ganoid scales, which are found on such fishes as gars and the bowfin, are similar to placoid scales but are covered with a peculiar enamel-like substance called ganoin. The advanced fish have either cycloid scales (e.g., carp) or ctenoid scales (e.g., perch; sunfish). These are typical overlapping fish scales; cycloid scales are large, thin, and round or oval in shape, and exhibit growth rings along their free edges. Ctenoid scales resemble cycloid scales but have comb-like teeth on their overlapping edge. For osteichthyans, sister group to Chondrichthyans, the plesiomorphic integumentary skeleton consists of large numbers of robust overlapping rhombic scales organized into obliquely 80 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.16 Eusthenopteron foordi. Top, photographs of well preserved specimens from lowermost Upper Devonian, Escuminac Bay, Canada. After Jarvik (1980)

Fig. 4.17 Eusthenopteron foordi. a Restoration in lateral view. b Neurocranium, postcranial endoskeleton, and lepidotrichia in lateral aspect. After Jarvik (1980) oriented rows. These rhombic scales have two structural forms corresponding to the ray-finned–lobe-finned dichotomy. Basal actinopterygians have ganoid scales, whereas basal sarcopterygians have cosmoid scales. Ganoid scales are identified by the presence of ganoine, a shiny, acellular hypermineralized tissue structurally identical to enamel, typically overlying layers of orthodentine and basal plate of 4.3 Scales: Organization of the Integumentary Skeleton in Gnathostomes 81 lamellar bone. Unlike enamel, ganoine is multilayered and, as evidenced by modern taxa, always localized deep to an epithelium. Based on the study of scale development, ganoid scales are hypothesized to have given rise to elasmoid scales. The second major structural form of rhombic scale is the cosmoid scale of basal sarcopterygians. At first glance, cosmoid scale histology compares well with that of the ganoid scale: a shiny superficial tissue comparable with enamel or enameloid, overlying a stacked sequence of dentine and lamellar bone. However, cosmoid scales (and the tissue complex known as cosmine) are uniquely characterized by an intrinsic, interconnected canal system, with numerous flask-shaped cavities and superficial pores. The oldest known examples of cosmine come from various Devonian sarcopterygians, dated *415 to 410 MYA, including Psarolepis romeri, Achoania jarvikii, Styloichthys changae and Meemannia eos (Zhu et al. 2010). Cosmoid scales were present among basal-most members of both Actinistia (coelacanths) and Dipnomorpha (lungfish) but were independently lost in each lineage. The cosmoid scale (and cosmine tissue) is extinct, and is no longer found in living species (Vickaryous and Sire 2009). As we have seen there has been much discussion that true teeth developed from placoid scales (see Chap. 2) (Figs. 4.18 and 4.19).

4.4 Scales and Swimming Hydrodynamics

We know that fish scales afford protection and help in hydrodynamic streamlining. However, placoid scales may play a very specific role in the locomotion of sharks. The morphology of the scale crown varies considerably not only across the body but also among species. The majority of scales on faster swimming sharks such as the shortfin mako Isurus oxyrinchus, blacktip shark Carcharhinus limbatus, and silky shark Carcharhinus falciformis have a series of parallel riblets (also termed micro-ridges, ridges, or keels; Fig. 4.20) that run in an anterior–posterior direction, often terminating in cusps on the trailing edge of the scale (Motta et al. 2012). Reif and Dinkelacker (1982; also Reif 1985) first proposed because of the highly specific orientation of these riblets and consistent height and spacing especially in fast swimming sharks that they have a hydrodynamic role, a hypothesis that has been well supported. Longitudinal riblets reduce drag because they impede the fluctuating turbulent crossflow near the wall, and in this way reduce momentum transfer and shear stress. Moin and Bewley (1994) showed by simulations of trajectories of marker particles in a turbulent water flow between parallel plates exactly how they work. The riblets appear to inhibit the motion of eddies by preventing them from coming very close to the surface (within about 50 microns). By keeping the eddies this tiny distance away, the riblets prevent the eddies from transporting high-speed fluid close to the surface, where it decelerates and saps the animals momentum. In a similar way a golf ball’s dimples augment the turbulence 82 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.18 Top left and right, Hiemenia sp. Photographs of two scales. Upper Lower or Lower Middle Devonian, Northwestern Canada. Top left, rhombic scale; Top right, cycloid scale, orientated with anterior part upwards. Bottom, Hiemenia ensis Orvig. After Jarvik (1980) very close to the surface, bringing the high-speed airstream closer and increasing the pressure behind the ball (Moin and Kim 1997). The coefficient of drag is much lower for a dimpled ball (Shark swimming biomechanics and hydrodynamics will be discussed in detail in Volume 2). Motta et al. (2012) have shown specialization of scale morphology and patterns in the flanks of the high-speed swimming short fin mako shark, I. oxyrinchus that may further act to reduce drag. The majority of the shortfin mako shark scales have three longitudinal riblets with narrow spacing and shallow grooves. They found that the shortfin mako shark has a region of highly flexible scales on the lateral flank that can be erected to at least 50° compared to their other test animal the 4.4 Scales and Swimming Hydrodynamics 83

Fig. 4.19 Diagram showing transformation of scales. a Rhomboid scale of osteolepids into, c cycloid scales of rhizodontids by growth of individual scales toward overlapped areas and successive modification of articular ridge into central boss. b Hypothetical intermediate stage. d Part of external portion of scale of Middle Devonian osteolepid to show cosmine layer with external openings (pores) of the mucus canals. e Eusthenopteron foordi, part of vertical section of dermal bone showing outer tuberculated and inner laminated areas. After Jarvik (1980). f Transverse sections of Meemannia skull roof. Meemannia possesses a pore–canal network combined with three or four layers of odontodes and enamel in dermal bone surface. This condition preceded the appearance of typical cosmine found in crown-group sarcopterygians. od1–od3, layers of odontodes; e1–e3 enamel layers; p, pore openings; puc, pulp cavity. After Zhu et al. (2010) blacktip shark, C. limbatus. The pivoting and erection of flank scales and resulting drag reduction they hypothesized to be passively driven by localized flow patterns over the skin. 84 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.20 Scanning electron micrograph of scales in the mako shark, Isurus oxyrinchus. Angled to show the dissected internal section of part of the dermis that gives rise to the scales. The surface of the dermis shows scales with parallel riblets or keels that run in an anteroposterior direction. (Authors unpublished data)

4.5 Feeding in Fishes

The subject of feeding in fishes is a complex and challenging topic. One of the problems of feeding in water as opposed to air is that the motion of a predator as it approaches its prey may deflect the prey out of reach. Another problem is the considerable energy needed to move the often large amounts of water taken in with foods. Feeding in fishes has evolved to turn such apparent disadvantages to their advantage, which has resulted in more versatility and higher level of opportunistic feeding than in terrestrial animals. Two major changes characterize teleost feeding (Helfman et al. 2009). 1. Suction feeding becomes highly advanced with the development of a protrusible pipette mouth capable of generating powerful suction force. The head length may be increased by one-third at a rapid speed, thereby greatly increasing the fish’s speed of attack. This is enabled by major modifications and remodeling of the head bones which ultimately enables rapid expansion of the orobranchial chamber and jaw protrusion (Hildebrand 1995). 2. Development of a protrusible mouth however came at a sacrifice to up-and- down chewing motions. Soft-bodied, relatively small prey were not the problem because they could be manipulated to enable head-first swallowing. However, harder exoskeletons of, e.g., mollusks and crustaceans require mechanical breakdown before digestive enzymes in the gut can have an effect on the internal softer parts of the prey’s body. The soft protrusible mouth is ineffective for this. In order to retain the benefits of the protrusible mouth a new form of dentition evolved to compensate for its negative effects—the pharyngeal jaw comprising upper and lower crushing dermal tooth plates aided by a strong musculature (Figs. 4.21 and 4.22). This enabled mechanical breakdown of prey posterior to the marginal jaws and just anterior to the esophagus, a remarkable development in fishes. We have seen in the previous chapter that this topic is of considerable interest and is not without controversy. 4.6 Reception of External Mechanical Stimuli in Fishes 85

Fig. 4.21 Stages in the evolution of the jaws of ray-ﬁnned ﬁshes. a Primitive chondrostean, Pteroniscus. b Less specialized teleost, Salmo. c More specialized teleost, Sebastes. After Hildebrand (1995)

4.6 Reception of External Mechanical Stimuli in Fishes

For adaptations in the aquatic medium there are deviations from the terrestrial paradigm. For underwater communication, sound is often more appropriate than a visual signal because of significant absorption of light even in short distances. In water, as in any medium, sound is a pressure variation (biological range 10 mPa to 1 kPa; 1 atm, the atmospheric pressure at sea level, is 101.3 kPa) accompanied by a longitudinal oscillation of the medium particles (0.01 nm to 1 lm) (Schellart and Wubbels 1998). Many fishes (also invertebrates and marine mammals) produce sounds that, within a limited range, exceed the ambient noise level and communication is hardly impeded by underwater barriers (rocks, coral reefs) because the wavelength of biological sounds is long (Schellart and Wubbels 1998). The sounds originate from or reflect of prey, predators, other fishes in a school, and environmental obstacles (Helfman et al. 2009). A detailed discussion of underwater acoustics can be found in Kalmijn (1988). Vibrations from sound and other movements are capable of traveling long distances in water. Because of an aquatic environment many fish species have adapted auxillary structures to enhance their pressure sensitivity for the reception of such external stimuli. Mechanoreception in fishes comprises two major sensory systems, the lateral line (LL) system and the 86 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.22 Diagram showing jaw expansion and protrusion in a teleost ﬁsh that enables food intake. The buccal cavity is modeled as a truncated cone, which can be expanded to create progressively more negative pressure as one moves back toward the base of the cone. Principle movements are indicated with arrows. DO Dilator operculi; EP Epaxial musculature; GH Geniohyodeus; LAP Levator arcus palatine; LO Levator operculi; SH Sternohyoideus. After Liem (1990)

inner ear. We look at the role the skin plays in fishes in detecting two components that make up the lateral line system in most fishes, other than teleosts, mechanical and electrical signals, which incidentally have a similar developmental origin. As we saw in Chap. 2, the LL system is an exceedingly old feature of fishes that evolved in the ostracoderms during the Silurian. 4.6 Reception of External Mechanical Stimuli in Fishes 87

4.6.1 Lateral Line System

The mechanoreception component of the LL system is present in all fishes. Detection of mechanical signals in the water is made by a complex arrangement of end organs or neuromasts located in the skin, usually distributed over the entire body. Each individual neuromast comprises a cluster of innervated hairs inter- spersed with supporting cells and covered by a fragile jelly-like projection, the cupula. The hair cells are typically oriented in opposite directions, which result in two populations of hair cells and afferent nerve fibers responding 180° out of phase (Schellart and Wubbels 1998) (Fig. 4.23). The neuromasts of most bony fishes are set in a series of interconnected depressions, forming a canal, with openings at intervals to the environment. The large accessory structure, the cupula, which is mechanically coupled to the underlying hair cells, improves the signal to noise ratio. Neuromasts are of two types (1) they can be found on the entire body including the tail (superficial neuromasts, SNs) with the cupulae extending into the water and (2) enclosed in canals (canal neuromasts, CNs) with the cupulae surrounded by canal fluid. The canals of the head consisting of bony grooves (‘‘cephalic lateral line canals’’) generally are covered by skin whereas the single (usually) trunk canal (from just behind the operculum to the tail) is formed by scales (the lateral line). Several hundred fibers may enervate a single CN while SNs are enervated by fewer. The LL system is enervated by three nerves each with its own ganglion which serve to enervate the head, jaws, cheek, around the eye, and the trunk and tail (Fig. 4.24). Superficial neuromasts being more exposed make them more sensitive to water movements across the skin and particularly receptive for detecting water currents for orientation (rheotaxis) or, movement of the fish itself in areas with little water velocity or unidirectional or low frequency vibrations (below 20 Hz), but not very useful for detecting small stimuli in areas of swift or turbulent water (Helfman et al. 2009). The spatial distribution suggests that gradients play an important role in LL functioning e.g. in prey capture when vision is poor. Moreover, stimulus direction for the neuromasts in one part of the fish body is reversed with respect to the neuromasts in another part of the body. The gradients and points of reversal depend on the size of the source. Together, phase information and amplitude gradients may provide detailed information about the source type and its location. Information may be obtained from stationary objects if the fish is swimming around or moving past or toward objects which causes strong changes of the gradients of particle motions along the body of the fish as shown by mathematical analysis (Schellart and Wubbels 1998 and references therein). Superficial neuromasts are more abundant in fishes that are sedentary or slow swimmers that inhabit quiet areas e.g. goldfish, Carassius auratus. Canal neuromasts on the other hand, are shielded from constant stimulation by water moving across the skin and are better at detecting stimuli if the fish or the water around it is moving quickly. They are therefore more effective in detecting transient currents, or currents of higher frequency (20–100 Hz; Helfman et al. 2009 and references therein). Canal 88 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.23 Lateral line (LL) organ. a Schematic neuromast structure. b Example of the distribution of the neuromasts of the blind cave fish. c Distribution and size of LL receptor fields of medullar neurons of the rainbow trout. The solid lines indicate the trunk canal. Circles: small receptor fields; broken lines: elongated receptor fields along the trunk canal. After Schellart and Wubbels (1998)

neuromasts therefore tend to be better developed in fast swimming fish or fish living in turbulent waters such as rainbow trout (Oncorhynchus mykiss). Collectively, free neuromasts and those of canals and pits form a system of distance touch that is so well coordinated that it gives both spatial and temporal information to fishes whatever their habitat or lifestyle and is of major importance for their survival. The LL system not only enables detection of ‘active sources, but may also provide clues about the static (mechanical) environment. Spatial pro- cessing of information is obtained from the input from the afferent fibers that converge in the medulla from receptive fields that can be very large.

4.6.2 Electroreception

Perhaps the most interesting specialization of the lateral line system is the formation in several groups of fish of deeply buried, single electrically sensitive organs or ampullae. Such structures are found among almost all non-teleost taxa in the Petromyzontiformes, Elasmobranchii, Holocephali, Dipneusti, Crossopterygii, Polypteriformes and Chondrostei (lampreys, sharks and rays, lungfishes, reedfishes, coelocanths, sturgeons, paddlefishes, bony fishes). It is absent in Myxini- formes and Holostei and more rarely found in teleosts with exceptions such as . eeto fEtra ehnclSiuii ihs89 Fishes in Stimuli Mechanical External of Reception 4.6

Fig. 4.24 Drawing of a section through the skin containing and ampullary (left) and tuberous (right) electroreceptor organ of the weakly electric gymnotiform ﬁsh, Eigenmannia. After (von der Emde 1998) 90 4 Evolution of Modern Fishes: Critical Biological Innovations

Plotosus. The electric organ is represented by structures called mormyromasts in freshwater African fish (mormyrids) and in electric eels (gymnotids); small pit organs of catfishes (silurids), and ampullary or lateral-line organs in, e.g., sharks and rays. In the latter group the elasmobranchs, they are most abundant and were first identified as the ampullae of Lorenzini. On the other hand the common ancestor of Teleostei and Holostei might have lost the ability to detect weak electric currents which may have been reinvented several times, i.e., independently, in the past 400+ million years of its evolution and at least twice in the four teleost groups in which it is found (Mormyriformes, Gymnotiformes, Siluriformes, and the Xenomystinae) (von der Emde 1998). Electric fishes may be connected to their environment by both abiotic and biotic signals. Abiotic fields are low frequency of only a few Hertz and arise from geological and seismological processes and the flow of seawater through the earth’s magnetic field. They may provide electroreceptive fish with information on important landmarks for spatial orientation. Electric fields of biotic origin may be of low or high frequencies. Biotic fields originate from contracting muscles, epithelia around the gills associated, e.g., with breathing and biochemical processes. There are two main types of electric organ discharge (EOD) of weakly electric fishes (1) a brief pulse-like discharge, and (2) continuous wave-type discharge. Thus an actively moving fish is surrounded by a constantly changing electric field that not only gives it information about its own activities but it may also provide electric signals to other electroreceptive animals. Weakly electric fishes use their EODs for active electrolocation and electrocommunication. An advantage in the use of electric signals for such activities in contrast to, e.g. acoustic signals, is that the waveform is only slightly distorted by the environment. Acoustic signals are more susceptible to distortions by the medium and objects in the environment (by reflection, refraction, scattering, attenuation), whereas electric signals and their waveform pass almost unaffected through the medium, conditions used by weakly electric fish during both electrocommunication and electrolocation (von der Emde 1998 and references therein). The remarkable sensitivity of fishes to electrical stimuli has been shown by experimentation—i.e., to minute, local potential differences in the surrounding water at their body surface. In behavioral experiments with sharks and rays, sensitivity to changes of 0.01 microvolt per centimeter (one microvolt = 1/1,000,000 of a volt) along the body surface has been found for the ampullae of Lorenzini (Kalmijn 1974). Similar, though somewhat higher, values have been recorded from the ampullary nerve fibers. A decrease in voltage at the opening of the ampulla causes an increase of the spontaneous nerve-impulse frequency; an increase in voltage at the opening produces the opposite response. Through their electrical sensitivity, such fish can detect and locate other organisms in darkness, in turbid water, or even when these organisms are hidden in the sand or in the mud of the sea bottom (Kalmijn 1971, 1974). The sensory cells or electroreceptor organs are located in the skin and are all part of the acousticolateralis system. It is suggested that the electroreceptive organ arise from the same embryonic precursors as the neuromasts of the lateral line system (see Helfman et al. 2009 and references therein). There are two classes of 4.6 Reception of External Mechanical Stimuli in Fishes 91 electroreceptor organs, ampullary, and tuberous receptors. In ampullary receptors the salt water makes it a good conductor of electricity. In elasmobranchs these receptors are the ampullae of Lorenzini. Ampullary organs may be found in all electroreceptive fish and respond to signals of low frequencies up to about 50 Hz. Tuberous organs contain high frequency receptors and are restricted to weakly electric fishes that produce their own high frequency signals (von der Emde 1998) (Fig. 4.24). Ampullary receptor organs are now thought to function as both thermo and electroreceptors (Braun et al. 1994). The ampullae of Lorenzini of marine species are withdrawn from the body and lie in chambers buried in the epidermis. They communicate with the surface pores by means of long canals that may be about 1 mm in diameter and up to 20 cm in length. These pores, which may be widely separated, are connected to the ampullae in clusters and surrounded by a capsule. Hundreds of electroreceptor cells are grouped together at the base of a single epidermal pit with only the apical portions protruding into the lumen. The basal portion of the receptor cells forms a chemical synapse with the afferent neurons and with the brainstem via the lateral line nerves. The lumen and also the inside of the canal are filled with a jelly-like substance with a resistance of the same magnitude as the outside sea water, i.e., with excellent electrical conductivity. There are only a few differences between the ampulary receptor organs of other non-teleost groups and the ampullae of Lorenzini of elasmobranchs (Fig. 4.25). However teleost ampullary receptors are found in only a few species, mostly in fresh water. They are less sensitive to electrical fields and respond to voltage gradients of about 100 lV/cm. One of the main uses for ampullary receptors is for prey detection. The ampullae are often concentrated around the head and in some fishes such as sharks they are concentrated in large numbers around the snout and mouth while in skates and rays they are also found on the pectoral fins and in the eel-tailed catfish they are found all over the body. Kalmijn (1971) demonstrated that sharks would attack prey emitting even mild electric signals but not those in which the signals were covered by a barrier. The unique head morphology of hammerhead sharks (sphyrnids) prompted Kajiura and Holland (2002) to hypothesize that it might have evolved to enhance electrosensory capabilities. They tested their hypothesis by comparing the behavioral responses of a similarly sized sandbar shark, Carcha- rhinus plumbeus and hammerhead shark, Sphyrna lewini, to prey-simulating electric stimuli. Although the response threshold and orientation pathways were similar in the two species, the behaviors were markedly different. Scalloped hammerheads typically demonstrated a pivot orientation in which the edge of the cephalofoil closest to the dipole remained stationary while the shark bent its trunk to orient to the center of the dipole. By contrast, sandbars swam in a broader arc toward the center of the dipole. They attributed this to the hydrodynamic properties of the cephalofoil, which enables the hammerheads to execute sharp turns at high speed and a greater probability of prey encounter and prey capture. Tuberous electroreceptor organs are found in fresh water fishes, South Amer- ican Gymnotiformes and African Mormyriformes that produce their own electric 92 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.25 Ampullae of Lorenzini in the snout of the white shark, Carcharodon carcharias (anterior view). a Surface cut away to show how each vesicle (or ampulla) opens to the surface through a tubelike duct which is ﬁlled with an amber colored gelatinous substance that has excellent electrical conductivity (scale = 10 cm). b Enlarged dorsolateral view. (Author’s unpublished data)

fields for active electrolocation. The frequency range is between 50 and 2,000 Hz and occasionally with very short pulses up 18 kHz. Like ampullary organs, tuberous organs have epidermal chambers with a sensory epithelium at the base comprising electroreceptor cells, which may be entirely inside the sensory 4.6 Reception of External Mechanical Stimuli in Fishes 93 chamber or only small portions of the apical membrane are exposed to the chamber lumen. However, the canal leading to the surface of the skin instead of containing electro-jelly contains loosely assembled cells—the air spaces keeping the electrical impedance relatively low. A number of layers of flattened cells composing the chamber and canal provide efficient electric isolation (von der Emde 1998). There are two main classes of tuberous organs, one specialized to encode the precise timing of an electric organ discharge and the other to encode stimulus amplitude. All fishes possess one type of electroreceptor of each class, e.g., pulse markers units and burst and burst duration coders. With two populations of receptor cells, one receiving time and the other amplitude information, weakly electric fish can analyze both their self-produced EODs and those of neighboring fish, and hence can be employed for two main functions, for electrolocation and for electrocommunication.

4.6.3 Function of Electroreception

Electricity in fishes is most frequently used for communication and the signals are species specific and aspects of the EOD such as amplitude, frequency, and pulse length are capable of imparting information that serves an important function in the lives of these more ‘‘primitive’’ fishes compared to the majority of teleosts. Kalmijn (1974, 1988) demonstrated in laboratory experiments that marine sharks and rays use their electroreception to detect weak electric fields of prey (e.g., prey fish buried in the sand produce electric currents through their gill muscles) and for orientation and migration in their environment (e.g., by measuring the magnetic field). The weak electric fields of conspecifics were also found to be important social signals in the breeding behavior of stingrays. Some electroreceptive fishes able to detect natural occurring weak electric currents are among the Sarcopterygii (see Sect. 4.6.4) namely the coelacanth, Latimeria. Experiments showed that that they were responsive to weak electric fields and that they could be conditioned to respond to local electric dipole fields of \0.2 lV/cm (von der Emde 1998 and references therein). On the other hand most teleosts are non-electrogenic but a few electroreceptive forms are found among siluriforms. They can use electrodetection for prey detection and orientation. Species such as Ictalurus were found to be able to locate goldfish hidden behind a layer of electrically transparent agar (Kalmijn 1974). Perhaps most interesting are several catfish species of the genus Clarius that are able to employ a form of pack-hunting behavior by electrolocation of weakly electric mormyrid fishes in the Okavango Delta of Botswana. However, the problem is that the catfish are sensitive to low frequencies of about 50 Hz whereas the prey emits short pulse- type signals of several kilohertz. It is thought that the catfish get around this problem by ‘‘filtering’’ out the ‘‘noise’’ and detecting only the low frequency signals of the prey. 94 4 Evolution of Modern Fishes: Critical Biological Innovations

In active electrolocation the fishes produce the electric signals that they are then able to detect and analyze. One example is of elasmobranchs that swim through magnetic fields. However, one of the more general problems involves the delin- eation of information that the fish receives from the electric signals, i.e., not only being able to tell an objects distance but to independently determine its size during electrolocation, particularly difficult in the absence of a focusing mechanism. It is now believed that fishes may enhance electrolocation by using behavioral methods to resolve questions of size and amplitude of electrical signals. For example the motions of the fish can enhance the electrical image and its contrast. Fishes may also use time measurements during electrolocation. This is particularly useful in detecting the capacitive features of an object because they are primarily found in living animals. Thus fishes, e.g., mormyrids, are capable of distinguishing between living and nonliving objects and in addition two objects that caused identical amplitude changes at the skin surface of the fish (von der Emde 1998). Living animals and plants emit complex waveform distortions which mormyrids have a means of recognizing, which is especially helpful in prey identification.

4.6.4 Electrocommunication

In addition to mormyriforms and gymnotiforms (and electric eels) at least two other electroreceptive fishes have developed weak electrogenic organs solely for the purpose of electrocommunication, the synodontid catfish and some rays. Signals are specific to individual species and aspects of the EOD e.g., amplitude, frequency, and pulse length can be utilized to exchange information such as species, sex, size, courtship, dominance interactions, territoriality, location, distance, and possibly individual identity (von der Emde 1998; Helfman et al. 2009). All this information requires considerable coordination by the fish’s central nervous system. In the South American gymnotid, Apteronotus, and the African mormyrids the electric organs are controlled by pacemaker cells in the medulla, which are regulated by two clusters of neurons elsewhere in the brain (Helfman et al. 2009)—a remarkable case of convergence given that the two groups are believed to have evolved their EOD capabilities independently.

4.7 Hydrodynamics and Buoyancy in Fishes

Animals are denser than either fresh or sea water and therefore tend to sink unless they have mechanisms that keep them buoyant. Among vertebrates, the problem of avoiding sinking is achieved principally by the use of hydrofoils or buoyancy aids (Lingham-Soliar 2005c, Fig. 1). 4.7 Hydrodynamics and Buoyancy in Fishes 95

Sharks are negatively buoyant and must continually swim to stop sinking. Yet, many sharks are capable of low swimming speeds of about 0.1–0.7 Ls-1 (Fish and Shannahan 2000) with rates of movement obtained from tracking of C. carcharias ranging from 0.21 to 0.25 Ls-1 (Strong et al. 1992). Slow swimming enables opportunistic predation and taking advantage of unpredictable food sources (Webb 1997). Besides hydrofoils, Wilga and Lauder (2002) have shown, within pre- scribed circumstances, that flattening of the anterior body in certain shark species accounts for hydrodynamic lift anteriorly as opposed to lift from the pectoral fins. This is unlikely to play any significant part in the body trim of C. carcharias because of the rounded cross-section of the head region and most of the body in such lamnid sharks. The use of large hydrofoils or body flattening in white sharks may be ruled out in any dominant role in the lifestyle and predatory behavior of larger members (below). It was shown that even the dynamic caudal fin in the thunniform shark Car- charodon carcharias may be reduced in size in larger members of the species. Allometric scaling analysis which was employed to investigate the consequences of size evolution on hydrodynamic performance and ecology in the white shark C. carcharias (Lingham-Soliar 2005c) found, contrary to expectation, that the caudal fin was negatively allometric. Discriminant analysis using the power equation y = axb was negative for caudal fin span (S) versus fork length (FL). In other words as the animal got larger the caudal fin became relatively smaller. In contrast in two delphinid species, Delphinus capensis, and Tursiops aduncus, the span of the flukes versus fork length rises in positive allometric fashion, and strong positive allometry of S versus HA (area) was also recorded. In C. carcharias negative caudal fin allometry contrasts with positively allometric liver mass (weight and length), suggesting a size-dependent reversal in the role of lift from, for example, the caudal fin versus a buoyancy agent in resisting the tendency to sink. Liver weight may be as high as 28 % of body weight in C. carcharias [400 kg weight ([3 m FL), exceeding that of smaller animals \150 kg weight (\2 m FL) by as much as 2.5-fold (Lingham-Soliar 2005c). By implication, since buoyancy is low in small white sharks, and given a conical head (body flattening is minimal), very large pectoral fins (Thomson and Simanek 1977) are reasonably considered to be the main source of lift anteriorly (Fig. 4.26). Differences in hydrodynamic performance between small–moderate and large C. carcharias may result in feeding niche partitioning that could reduce intra- specific conflict and predation. For example, agonistic aggression in white sharks was recorded by Strong et al. (1992) whereby feeding by smaller C. carcharias may be thwarted by larger conspecifics. Although niche partitioning is largely theoretical at this point, support may be inferred from well-documented size- related shifts in diet in larger C. carcharias, attributed to such factors as decreasing agility with size and changes in tooth morphology (Tricas and McCosker 1984)— smaller sharks \2.5 m predating predominantly on fish, with marine mammals becoming more prominent in the diet of large sharks (Cliff et al. 1989; Bruce 1992 96 4 Evolution of Modern Fishes: Critical Biological Innovations

Fig. 4.26 White shark, Carcharodon carcarias, specimen (over 5 m long) caught in the safety nets along the coastal waters of KwaZulu-Natal. The dissected specimen (lower image) shows the massive size of the liver. (Author’s unpublished data)

and references therein). Ecomorphological interpretations here are consistent with the view that locomotor repertoire and habitat characteristics would be expected to be matched: (1) the properties of propulsive systems predict the type of habitat in which these properties would be beneﬁcial and (2) characteristics of the habitat predict the locomotor and hence morphological characteristics that are expected to facilitate its occupation (Webb 1997). References 97

References

Ahlberg PE (1991) Tetrapod or near tetrapod fossils from the upper devonian of Scotland. Nature 354:298–301 Braun HA, Wissing H, Schaefer K, Hirsch MC (1994) Oscillation and noise determine signal transduction in shark multimodal sensory cells. Nature 367:270–273 Bruce BD (1992) Preliminary observations in the biology of the white shark, Carcharodon carcharias, in South Australian waters. Aust J Mar Freshw Res 43:1–11 Carroll RL (1988) Vertebrate paleontology and evolution. Freeman, New York Colbert EH (1955) Evolution of the vertebrates. Wiley, New York Cliff G, Dudley SFJ, Davis B (1989) Sharks caught in the protectivegill nets off Natal, South Africa 2. The great white shark Carcharodon carcharias (Linnaeus). S Afr J Mar Sci 8:131–144 Daeschler EB, Shubin NH, Jenkins FA Jr (2006) A Devonian tetrapod-like fish and theevolution of the tetrapod body plan. Nature 440:757–763. doi:10.1038/nature04639 Donoghue PC, Sansom IJ, Downs JP (2006) Early evolution of vertebrate skeletal tissues and cellular interactions, and the canalization of skeletal development. J Exp Zool B (Mol Dev Evol) 306:278–294 Fish FE, Shannahan LD (2000) The role of the pectoral fins in body trim of sharks. J Fish Biol 56:1062–1073 Goujet D (2001) Placoderms and basal ganthostome apomorphies. In: Ahlberg PE (ed) Major events in early vertebrate evolution. Taylor and Francis, London and New York, pp 209–222 Helfman GS, Collette BB, Facey DE, Bowen BW (2009) The diversity of fishes, 2nd edn. Wiley- Blackwell, Chichester Hildebrand M (1995) Analysis of vertebrate structure. Wiley-Blackwell, John Wiley and Sons, Inc. New York Inoue JG, Miya M, Lam K, Tay B-H, Danks JA et al (2010) Evolutionary origin and phylogeny of the modern holocephalans (chondrichthyes: chimaeriformes): a mitogenomic perspective. Mol Biol Evol 27:2576–2586. doi:10.1093/molbev/msq147 Janvier P (2001) Ostracoderms and the shaping of gnathostome characters. In: Ahlberg PE (ed) Major events in early vertebrate evolution. Taylor and Francis, London, pp 172–186 Jarvik E (1980) Basic Structure and Evolution of the Vertebrates (Vol 1). Academic Press, London Kajiura SM, Holland KN (2002) Electroreception in juvenile scalloped hammerhead and sandbar sharks. J Exp Biol 205:3609–3621 Kalmijn AJ (1971) The electric sense of sharks and rays. J Exp Biol 55:371–383 Kalmijn AJ (1974) The detection of electric fields from inanimate and animate sources other than electric organs. In: Fessard A (ed) Handbook of sensory physiology, vol 3, Electroreceptors and other specialized receptors in lower vertebrates. Springer, Berlin, pp 147–200 Kalmijn AJ (1988) Hydrodynamic and acoustic field detection. In: Atema J, Fay RR, Popper AN, Tavolga WN (eds) Sensory biology of aquatic animals. Springer, New York, pp 83–130 Liem KF (1990) Aquatic versus terrestrial feeding modes: possible impacts on the trophic ecology of vertebrates. Amer Zool 30:209–221 Lingham-Soliar T (2005a) Dorsal fin in the white shark, Carcharodon carcharias: a dynamic stabilizer for fast swimming. J Morphol 263:1–11 Lingham-Soliar T (2005b) Caudal fin in the white shark, carcharodon carcharias (lamnidae): a dynamic propeller for fast, efficient swimming. J Morphol 264:233–252 Lingham-Soliar T (2005c) Caudal fin allometry in the white shark carcharodon carcharias: implications for locomotory performance and ecology. Naturwissenschaften 92:231–236 Lund R (1990) Shadows in time—a capsule history of sharks. In: Gruber SH (ed) Discovering sharks American Littoral Society special publication 14. American Littoral Society, Highlands 98 4 Evolution of Modern Fishes: Critical Biological Innovations

Metscher BD, Ahlberg PE (2001) Origin of the teleost tail:phylogenetic frameworks for developmental studies. In: Ahlberg PE (ed) Major events in early vertebrate evolution.Taylor Francis, London, p 333–349 Moin P, Bewley T (1994) Feedback control of turbulence. Appl Mech Rev (part 2) 47:3–13 Moin P, Kim J (1997) Tackling turbulence with supercomputers. Am Sci 276(1):46–52 Motta P, Habegger ML, Lang A, Hueter R, Davis J (2012) Scale morphology and flexibility in the shortfin mako Isurus oxyrinchus and the blacktip shark Carcharhinus limbatus. J Morphol. doi:10.1002/jmor.20047 Nelson JS (2006) Fishes of the World, 4th edn. Wiley, Hoboken Reif WE (1985) Squamation and ecology of sharks. Cour Forsch-Inst Senckenberg 78:1–255 Reif WE, Dinkelacker A (1982) Hydrodynamics of the squamation in fast swimming sharks. N Jb Geol Paläont Abh 164:184–187 Schellart NAM, Wubbels RJ (1998) The auditory and mechanosensory lateral line system. In: Evans DH (ed) The physiology of fishes. CRC Press, New York, pp 283–312 Strong R, Murphy RC, Bruce BD, Nelson DR (1992) Movements and associated observations of bait-attracted white sharks, Carcharodon carcharias: a preliminary report. Aust J Mar Freshw Res 43:13–20 Thomson KS, Simanek DE (1977) Body form and locomotion in sharks. Am Zool 17:343–354 Tricas TC, McCosker JE (1984) Predatory behavior of the white shark (Carcharodon carcharias), with notes on its biology. Proc Calif Acad Sci 43:221–238 Vickaryous MK, Sire J-Y (2009) The integumentary skeleton of tetrapods: origin, evolution, and development. J Anat 214:441–464. doi:10.1111/j.1469-7580.2008.01043.x von der Emde G (1998) Electroreception. In: Evans DH (ed) The physiology of fishes. CRC Press, New York, pp 313–343 Webb PW (1997) Swimming. In: Evans DH (ed) The physiology of fishes. CRC Press, New York, pp 3–24 Wilga CD, Lauder GV (2002) Three-dimensional kinematics and wake structure of the pectoral fins during locomotion in leopard sharks Triakis semifasciata. J Exp Biol 203:2261–2278 Zhu M, Wang W, Yu X (2010) Meemannia eos, a basal sarcopterygian fish from the lower devonian of China—expanded description and significance. In: Elliott DK, Maisey X, Yu JG, Miao M (eds) Morphology, phylogeny and paleobiogeography of fossil fishes. Verlag Dr. Friedrich Pfeil, Munich, pp 199–214 Chapter 5 Tetrapods and the Invasion of Land

5.1 Global Environmental Changes

The Devonian gives the first widespread evidence for nonmarine conditions and is remarkable for evidence of the colonization of land as well as freshwater rivers and lakes by both plants and fishes. It was a period that saw the extensive evolutionary radiation and proliferation of fishes in both fresh and marine waters, hence why it is popularly known as the ‘‘Age of Fishes.’’ These conditions were well represented on all the continents and are responsible for iconic images of murky swamp lands and a luxuriant flora of mosses and giant ferns among which lurked strange fish-like creatures, benignly unaware of their oddities. Such habitats included coastal lagoons and large brackish river deltas, presenting increasingly complex food webs that afforded new feeding opportunities. These conditions combined, afforded an environment that may have been highly conducive to the evolution of tetrapods. It is in this type of environment that early lobe–-finned fishes such as Eus- thenopteron (see Chap. 4) were thought to have initiated the transformation from water to land—as the pools they inhabited dried up or became less hospitable they moved across tracts of land to other nearby, more hospitable pools of water. This is in contrast to the earlier view that fish had deliberately invaded land from the sea in search of a ‘‘brave new world’’, i.e., an unfamiliar but new niche in which to find food, mate, produce offspring, and enable the survival of their species. The chunky limbs of Eusthenopteron seemed to support this view of a kind of pond- hopping during times of need (Fig. 5.1). However, taking the morphology of Eusthenopteron as a whole, in particular a streamlined body-form with dorsal, anal, and pelvic fins placed near the back of the body, Clack (2012) proposed that its lifestyle was much more like that of the modern pike (Esox), a fully aquatic ambush predator. She and her colleagues extended this view even further when they showed that the earliest alleged tetrapods such as Acanthostega were wholly aquatic and quite unsuited to life on land. The question at the core of such recent developments that challenges some time held views is when is a tetrapod a tetrapod, i.e., when does it cease to be a fish?

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 99 DOI: 10.1007/978-3-642-53748-6_5, Ó Springer-Verlag Berlin Heidelberg 2014 100 5 Tetrapods and the Invasion of Land

Fig. 5.1 A life reconstruction of Eusthenopteron

It is a question that Laurin et al. (2000) try to answer from a phylogenetic perspective. They regard it as part of a larger problem between practitioners of Linnean taxonomy and those of phylogenetic systematics (cladistics). The former define ‘‘Tetrapoda’’ as the taxon that includes all vertebrates that bear digits whereas the phylogenetic definition of Tetrapoda is ‘‘the most recent common ancestor of extant lissamphibians and amniotes and all of its descendants’’ but differing radically in that it actually excludes some digit-bearing vertebrates. In this way the authors attempt to truly discriminate between limbs that are still tied to the aquatic medium despite some apparent innovations and those that are firmly planted on land. Daeschler and Shubin (1998) were among the first to question the classically portrayed view of the limb-fin transition as a major adaptive shift that requires a strict dichotomy between an aquatic and a terrestrial existence stating that basal tetrapods and some sarcopterygians have structures that can be interpreted to be adaptive in both arenas, i.e., that one structure is not necessarily a transitional stage to another (Fig. 5.2). The transformation from fin to limb is one of the crucial developments that allowed vertebrates to become truly terrestrial and change their lives thereafter completely. To make their point Laurin et al. (2000) also show that the initial function of digits, which have usually been interpreted as an adaptation to the terrestrial environment is belied by morphological evidence that suggests that Acanthostega might have retained internal gills and was primitively aquatic and that digits appearing in an aquatic environment would only be an exaptation to the terrestrial environment (Fig. 5.3).

5.2 Tetrapods and Cladistics

Laurin et al. (2000) and others such as Daeschler et al. (1994; 2006) enable us to reassess our ideas on the evolution of the pentadactyl (five-fingered) limb in the context of a number of potential tetrapod candidates, which is discussed (below) by a number of key workers in the field. Another very important development we will see is that of changes to oxygenating the body necessitated by the move from an aquatic to a land-based animal and the transformation from gills to lungs. 5.2 Tetrapods and Cladistics 101

Fig. 5.2 Strict consensus tree from a phylogenetic analysis of 114 characters and nine taxa. Tiktaalik is the sister group of AcanthostegaþIchthyostega in one of the two most parsimonious trees, and clades with Elpistostege as sister to the tetrapods in the other. Reprinted with permission from Daeschler et al. (2006). Copyright Macmillan Publishers, Ltd

However, Laurin and colleagues advocacy for a phylogenetic definition of tetrapods to ‘‘the most recent common ancestor of extant lissamphibians and amniotes and all of its descendants’’ is a question that raised its head before and indeed it seems an appropriate time to look at a major, controversy that broke out in the late 1970s/early 1980s that also involved phylogenetic systematics and the origin of tetrapods—before going on to ground-breaking research that has led to our present knowledge of the first tetrapods (Fig. 5.4). In response to a highly critical article in Nature by Halstead (1978) ‘‘The cladistics revolution—can it make the grade?’’ Gardiner et al. (1979) presented an interesting response entitled, ‘‘The salmon, the lungfish and the cow: a reply.’’ They presented a diagram in which the lungfish and cow shared a common node with the salmon branching off a separate node (Fig. 5.4). However, we have to go back several decades to understand what the fuss was all about. In the 1940s when the German entomologist Willi Hennig was faced with the almost insurmountable problem of classifying thousands of dipteran (fly) species, he came up with a solution, i.e., that genealogical relationships should exclusively reflect organisms that are grouped strictly on the basis of the historical sequences by which they descended from a common ancestor, which he described fully in his book Grundzüge einer Theorie der phylogenetischen Systematik (1950). He called the system Phylogenetic Systematics, which has popularly come to be known as Cladistics. It diverged significantly from evolutionary systematics, the traditional school of thought which holds that taxonomic classifications ought to be based on genetic as well as genealogical affinities. One of the defining principles of 102 5 Tetrapods and the Invasion of Land

Fig. 5.3 Sarcopterygian limbs: a the forelimb of Eusthenopteron; b the forelimb of a recently discovered rhizodontid, probably showing convergent similarities with the tetrapod limb; and c the forelimb of Acanthostega. In all three limbs, the shading indicates the maximal potentially homologous regions using only the topological argument; the elements identiﬁed as homologous to metacarpals and phalanges in (a) and (b) might be homologous with distal carpals or have no homologues in stegocephalians. Anterior is to the left. Abbreviations: h humerus; in intermedium; r radius; u ulna; un ulnare. After Laurin et al. (2000)

Hennigs’ cladistics is that only advanced characters (apomorphies) could be used to define new forms from an ancestral species and not primitive characters (plesiomorphies) or a mixture of primitive and advanced characters. Another distinction is that rather than changes occurring within a lineage, in cladistics the lineage splits into two (producing a dichotomy within a cladogram) or more separate lines and a choice has to be made concerning which of two groups (sister groups) are more closely related to the third (the third group or nearest common ancestor remains on the cladogram trunk). The seemingly obvious answer to the question above (Gardiner et al. 1979), i.e., which two of the salmon, the lungfish and the cow would group together more closely would according to tradition be that the salmon and lungfish as fishes would group together and the cow (a tetrapod representative) would form an outgroup. Tied to this apparently simple set of relationships was the controversial question involving the origin of tetrapods. But there was more to it. Behind the question was a movement by a small group at the time led by some of the strongest advocates of cladistics (following the appearance of the English translation of Hennig’s book in 1966) at the Natural History Museum, London and the AMNH to 5.2 Tetrapods and Cladistics 103

Fig. 5.4 Phylogeny of Devonian and Lower Carboniferous stegocephalians. The type of limb present in many poorly known Devonian stegocephalians is uncertain, as shown by the ambiguous optimization of the character ‘type of limb’ (the absence of data for a given taxon is indicated by the absence of a square data box below its name). Phylogeny is mostly based on the work by Ahlberg, but the position of Tulerpeton, and uncertainties about the position of Ichthyostega and Acanthostega reflect findings by Laurin (1998); ref. in Laurin et al. (2000). After Laurin et al. (2000) revolutionize plant and animal systematics forever. It would not be plain sailing as the ensuing barrage of opposition would prove. As we see, Gardiner et al.’s (1979) cladogram shows the cow grouped together with the lungfish and the salmon as the outgroup, in striking contrast to perceived wisdom. Their argument was that the primitive characters that define fishes are arbitrary because they not only can be found in all fishes but also in other groups of non-fish vertebrates. Thus, they said the term fishes would not accurately define their phylogeny, i.e., their evolutionary origin because there are no shared derived or advanced characters (synapomorphies) that unite all fishes that cannot be applied to non-fish vertebrates. This was 104 5 Tetrapods and the Invasion of Land archetypal Hennigian philosophy applied to vertebrates for the first time, i.e., that a new group could only be defined by its possession of new characters that other groups of organisms lacked and the pattern of evolution would be revealed over time by animals possessing successively novel characters. Thus, phylogenetic systematics would involve the search for pairs of organisms (sister taxa) most closely related by their shared possession of advanced homologous (genealogically related) characters or synapomorphies. Given this, the cladistics interpretation of the above problem is that the lungfish is more closely related to the cow than it is to the salmon because ‘‘lungfish and cows share derived characters such as internal nostrils, an epiglottis, a two-chambered auricle, and so on… not found in salmon’’—i.e., the lungfish forms a sister group with the cow (clade 1) while the ray- finned fishes form the more distant relationship or outgroup. Rosen et al. (1981), which included some members of the original salmon, lungfish, and cow debate, performed the first thorough cladistics investigation of sarcopterygian interrelationships and arrived at a rather controversial conclusion— they dismissed osteolepiforms as an ill-defined assemblage of primitive lobe- finned fishes—remote from tetrapods. They proposed that the anatomical characters used for relating the Devonian fish Eusthenopteron foordi (and all osteolepiforms) to tetrapods were merely primitive characters of bony fishes. They stated instead that lungfishes were the closest relatives of tetrapods based on derived characters absent in ancestral groups and added scathingly, ‘‘had the study of lungfish relationships developed directly from comparisons among living gnathostomes without interruption by futile paleontological searches for ancestors, there would have been no ‘rhipidistian barrier’ to break through,’’ and that their solution to the problem could have happened decades earlier. It was clear that the butt of their attack was directed at Erik Jarvik who had just a year before published a major study on Eusthenopteron foordi, including the apparent position of osteolepiforms in tetrapod evolution, in his Basic Structure and Evolution of the Vertebrates (1980, in 2-volumes). Jarvik (1981) responded to Rosen et al. (1981) in no less robust terms. On anatomical grounds based on some of the most authoritative studies at the time he dismissed as unfounded the presence of nostrils (see Zhu and Ahlberg 2004), an epiglottis and a two-chambered auricle in ancestral lungfish as proposed by Rosen’s British and French colleagues (Gardiner et al. 1979), stating that they had ‘‘arbitrarily picked out three characters from Kesteven’s unreliable list (see Jarvik 1981) and, without checking their reliability…’’ had assumed ‘‘that, by application of cladistic terms, they can prove relationships’’. Clearly incensed by what he regarded as a lack of scholarship Jarvik added: ‘‘This irresponsible manner of dealing with facts is found in many other cladistic papers and characterizes also the paper by Rosen et al.’’ Jarvik’s (1981) strong feelings need to be viewed in the context that in much of the twentieth century research into the origin of tetrapods, including those at the top end of the stem of the subclass Tetrapodomorpha, focused on the osteolepiforms as potential ancestors with particular emphasis on the extensively described Eusthenopteron. The latter was frequently used as a starting point for explaining the origin of tetrapod characters. Nevertheless, while views on the osteolepiform 5.2 Tetrapods and Cladistics 105

Fig. 5.5 Cladogram showing relationships of salmon, lungﬁsh, and cow. The lungﬁsh and the cow are more closely related to each other than either is to the salmon. After Fastovsky and Weishhampel (1996)

position in the tetrapod stem group ﬂuctuated over the years, osteolepiforms in more recent studies have generally been avoided with respect to the issue of the group’s monophyly (Figs. 5.5, 5.6).

5.3 Paraphyletic Origins

In some ways things return full circle to Jarvik’s hypothesis (1981). Ahlberg and Johanson’s (1998) re-examination of old material and incorporation of perfectly preserved three-dimensional osteolepiforms Medoevia and Gogonasus, and the articulated rhizodont Gooloogongia, shows the Osteolepiformes to be paraphyletic. This as it happens they view as a boon rather than a shortcoming in trying to understand the complex story of the transformation of ﬁsh to tetrapod (we will see more evidence in support of this later in the chapter). Of the two traditionally accepted osteolepiform subgroups, the Tristichopteridae (=Eustenopteridae) are closer to tetrapods than are the Osteolepididae, and the latter are themselves paraphyletic. In other words, osteolepiforms are an ensemble of primitive tetrapodomorphs, a large group of large predatory ﬁshes that also includes the rhizodontids, elpistostegids, and tetrapods (Fig. 5.7). As Janvier (1998) says, if osteolepiforms were monophyletic, the discovery of increasingly primitive species of this group would merely have told us about the history of osteolepiforms, but because Ahlberg and Johanson (1998) show that osteolepiforms seem to be an ensemble of basal tetrapodomorphs, we can now expect the discovery of early and primitive forms (such as Kenichthys from China) to tell us about the early history of the entire lineage that leads, ultimately, to the tetrapods (see Chap. 8). 106 5 Tetrapods and the Invasion of Land

Fig. 5.6 New phylogeny of the tetrapodomorph stem group. Thin black lines represent outgroups; gray lines indicate ‘‘osteolepidids’’ (including the problematic genus Kenichthys30); empty outline indicates rhizodonts; positive slope hatching indicates tristichopterids; thick black lines indicate elpistostegids ? tetrapods. ‘‘Tetrapoda’’ indicates the clade ‘vertebrates with limbs’ which contains, but is more inclusive than, the tetrapodomorph crown group. Reprinted with permission of Ahlberg and Johanson (1998). Courtesy, Macmillan Publishers, Ltd

Ahlberg and Johanson (1998) were able to discriminate between the stem lineage, and thus ‘‘ancestral characters’’ for the Tetrapoda ? Elpistostegalia and the crownward position of Eusthenopteron. The stem lineage they show would include inter alia a hinged braincase where the hinge ran through the profundus nerve foramen, a unique pattern of dermal bones, a small tripodal scapulocoracoid (endoskeletal shoulder girdle), anal, and posterior dorsal fin supports comprising a basal plate and three unjointed radials,and a pectoral fin skeleton comprising four axial elements (humerus, ulna, ulnare, IV), preaxial unjointed radials, and a postaxial flange on the ulnare. On the other hand, the comparatively crownward 5.3 Paraphyletic Origins 107

Fig. 5.7 Ancestral characters for clade Elpistostegalia ? Tetrapoda. These precise morphologies, illustrated by Eusthenopteron, characterize the tetrapod stem lineage between the Osteolepis ? Gogonasus node and the Tristichopteridae node. a Braincase (known in many genera, including Osteolepis, Gogonasus, Ectosteorachis and Medoevia). b Dermal skull bones (known in most genera). La lacrimal; Ju jugal; Po postorbital; Sq squamosal; Qj quadratojugal; Pop preopercurlar; De dentary; Mx maxilla. c Pectoral girdle and fin, mesial view (pectoral fin skeleton known in Megalichthys, Sterropterygion and tristichopterids; girdle known in many genera). Cla, clavicle; clei, cleithrum; Sca, scapulocoracoid. d Posterior dorsal fin support (known in Megalichthys, Rhizodopsis and tristichopterids). Reprinted with permission of Ahlberg and Johanson (1998). Courtesy of Macmillan Publishers, Ltd position of Eusthenopteron the authors say agrees surprisingly well with precla- distic perceptions (e.g., Jarvik 1980) of this group and shows the derived characters which originate within the ‘‘osteolepiform’’ as part of the Tetrapodomorpha. Most of the characters appear at the (Tetrapoda ? Elpistostegalia) and (Tristi- chopteridae) node (Fig. 5.5). They include loss of cosmine, loss of extratemporal bones, narrowing of the otic part of the skull, and lengthening of the snout, orbitotemporal region, and corresponding parts of the lower jaw. The last character complex may be functionally associated with the reduction in intracranial joint mobility and the adoption of a snapping mode of prey capture. It is interesting that these findings largely vindicate Jarvik’s (1981) sharp response to what he referred to as the ‘‘pompous oration’’ of ‘‘the four cladists 108 5 Tetrapods and the Invasion of Land

(Rosen et al. 1981), not doubting their own infallibility.’’ As many notable prec- ladistic studies had shown years earlier, Ahlberg and Johanson’s (1998) cladogram on the origin of tetrapods relegates the Dipnoi (which includes lungﬁshes, central to Rosen and colleagues’ origin of tetrapods hypothesis) to an outgroup (Fig. 5.5). They also show parallel evolution within the tetrapod stem group.

5.4 Descendents of the Sarcopterygian Fishes and the Move onto Land

The move onto land by the descendents of the sarcopterygian fishes was a momentous step that opened up unprecedented changes and great developments in vertebrate structure, mechanics, and physiology. Two major developments had to take place before the transition could be truly said to have bridged the divide between water and land. The sarcopterygian fins had to be transformed into legs sturdy enough to support the weight of the Devonian tetrapod on land (Hall 2007) and the animal would have to be able to breathe predominantly by means of lungs. In the beginning the transition to land may have been more gradual than at first thought (Ahlberg and Johanson 1998). Until recently, it was assumed that nearly all early Devonian tetrapods and other stegocephalians lived only in freshwater bodies and on dry land (in a similar manner to extant amphibians, which generally cannot tolerate the marine environment).This assumption was supported partly by the freshwater paleoenvironmental interpretation of many localities in which early Devonian tetrapods, other stegocephalians, and their sarcopterygian relatives were found. However, Laurin et al. (2000) have drawn attention to the fact that many of these localities have recently been reinterpreted as estuarine, deltaic, or even as coastal marine environments. Hence, these recent interpretations raise the possibility that the intolerance of lissamphibians to the marine environment is a relatively recent specialization of this clade. The morphological gap between osteolepiform lobe-finned fishes and the tetrapods has been frustratingly wide (Clack 2006). The gap was bounded at the top by primitive Devonian ‘‘tetrapods’’ such as Ichthyostega and Acanthostega from Greenland, and at the bottom by Panderichthys, a tetrapod-like predatory fish from the latest Middle Devonian of Latvia (Ahlberg and Clack 2006) (Fig. 5.8). Ahlberg and Johanson’s (1998) prediction that closer study of the parallel radiations in the ‘‘osteolepiforms’’ should cast much new light on the ecological background to the origin of tetrapods is particularly pertinent with respect to an understanding of the crownward position leading to tetrapods in the light of more recent findings. As we saw above, the osteolepiforms were a critical part of the paraphyletic stem group ancestors of tetrapods (large aquatic predators) with many derived characters occurring in the crownward Eusthenopteron that was shared with later groups (Tetrapoda ? Elpistostegalia) and (Tristichopteridae) (Ahlberg and Johanson 1998; Fig. 5.5). However, between Panderichthys and the early stem tetrapods 5.4 Descendents of the Sarcopterygian Fishes and the Move onto Land 109

Fig. 5.8 Interrelationships of bony ﬁshes and the four- legged vertebrates (tetrapods). a The status of osteolepiforms (red) and rhizodontids (yellow) in most of the previous phylogenies. b The revised status of osteolepiforms and rhizodontids according to the new analysis by Ahlberg and Johanson (1998). Reprinted with permission of Janvier (1998). Copyright Macmillan Publishers, Ltd

Acanthostega and Ichthyostega there is a gap, both anatomically and in time (by a period of about 20 million years). A major discovery of the fossil elpistostegalian sarcopterygian fish Tiktaalik roseae, announced to the world with great fanfare, apparently changes that with features that are both morphologically and functionally transitional (Daeschler et al. 2006, Shubin et al. 2006). Tiktaalik roseae apparently provides unique insights into how and in what order important tetrapod characters arose—importantly a mosaic of character that show how they resemble and differ from their nearest evolutionary neighbors. What stands out about Panderichthys is, unlike the rather conventional osteolepiform fishes farther down the tree, it is vaguely crocodile-shaped and looks like a fish-tetrapod transitional form, but not quite. It is here that Tiktaalik comes into its own—it is more tetrapod-like than Panderichthys. Tiktaalik was a heavy-bodied fish about 3 m long without a dorsal fin, a broad head with a pointed snout, a neck, robust, limb-like pectoral fins with digit-like elements connected by transverse joints including a mobile wrist (clearly enabling them to haul themselves from the water) (Fig. 5.9), thick imbricated ribs. It also has a shortened skull roof and elongated snout, a modified ear region, a mobile neck, a functional wrist joint, and other features that presage tetrapod conditions. The pectoral skeleton is particularly significant in that it shows the most advanced transitional features leading to the tetrapod limb (Shubin et al. 2006) (cf. other 110 5 Tetrapods and the Invasion of Land

Fig. 5.9 Tiktaalik in context. The lineage leading to modern tetrapods includes several fossil animals that form a morphological bridge between fishes and tetrapods. Five of the most completely known are the osteolepiform Eusthenopteron; the transitional forms Panderichthys and Tiktaalik; and the primitive tetrapods Acanthostega and Ichthyostega. The vertebral column of Panderichthys is poorly known and not shown. The skull roofs (left) show the loss of the gill cover (blue), reduction in size of the postparietal bones (green), and gradual reshaping of the skull. The transitional zone (red) bounded by Panderichthys and Tiktaalik can now be characterized in detail. These drawings are not to scale, but all animals are between 75 cm and 1.5 m in length. They are all Middle–Late Devonian in age, ranging from 385 million years (Panderichthys) to 365 million years (Acanthostega, Ichthyostega). The Devonian–Carbonifer- ous boundary is dated to 359 million years ago. Reprinted with permission from Ahlberg and Clack (2006). Copyright Macmillan Publishers, Ltd sarcopterygian fishes including Gogonasus (Long et al. 2006), with a limb form intermediate between Eusthenopteron and Panderichthys). As Ahlberg and Clack (2006) note, in some respects, Tiktaalik and Panderichthys are straightforward fishes: they have small pelvic fins, retain fin rays in their paired appendages and have well-developed gill arches, suggesting that both animals remained mostly aquatic. But what is striking is the apparently intermediate characters exhibited by Tiktaalik roseae and tetrapods, in particular limb development. It is perhaps not surprising therefore that they propose Tiktaalik as a link between fishes and land vertebrates, which might in time become as much of an evolutionary icon as the proto-bird Archaeopteryx. Yet, just as the vertebrate transition from water to land seems to be as crystal clear as we could hope, new findings (Niedz´wiedzki et al. 2010) literally shows it may be as clear as mud. 5.4 Descendents of the Sarcopterygian Fishes and the Move onto Land 111

Fig. 5.10 a Reconstruction of the right pectoral fin of Tiktaalik., Dorsal view; b ventral view. Elements with stipple shading were preserved in articulation in NUFV 109 and prepared in the round. Elements with a dashed outline are reconstructed based on their presence in the articulated distal fin of NUFV 110. It is not known how many radials lie distal to the first, second, and fourth in the proximal series. Note the dorsal expansion of the distal articular facets on the ulnare and third distal radial/mesomere. The dorsal expansion of these facets would have facilitated extension of the distal fin. Reprinted with permission Shubin et al. (2006) Copyright Macmillan Publishers, Ltd

Janvier and Clemént (2010) commenting on the paper by Niedz´wiedzki and colleagues state that the fish-tetrapod transition was seemingly quite well documented. There was a consensus that the divergence between some elpistostegalians (such as Tiktaalik or Panderichthys) and tetrapods might have occurred during the Givetian, 391–385 million year ago. Coeval with the earliest fossil tetrapods, trackways dating to the Late Devonian was evidence for their ability to walk or crawl on land (Clack 2002a, b). But, Niedz´wiedzki et al.’s (2010) paper may have changed this view so dramatically that the question of the fish-tetrapod transition may have to be radically reevaluated. The authors describe fossil tracks made by a four-limbed animal (tetrapod) in marine tidal flat sediments (not freshwater as previously contended) from the Eifelian (early Middle Devonian) of southeastern Poland–predating the oldest tetrapod skeletal remains, quite unambiguously according to the authors, by 18 million years and, yet more surprisingly, the earliest elpistostegalian fishes by about 10 million years (Fig. 5.10). The tracks show distinct manus (‘‘hand’’) and pes (‘‘foot’’) prints (somewhat different size) in 112 5 Tetrapods and the Invasion of Land

Fig. 5.11 Tetrapod trackways. a Muz PGI 1728.II.16. (Geological Museum of the Polish Geological Institute). Trackway showing manus and pes prints in diagonal stride pattern, presumed direction of travel from bottom to top. A larger print (vertical hatching) may represent a swimming animal moving from top to bottom. b At the bottom is a generic Devonian tetrapod based on Ichthyostega and Acanthostega (from Daeschler et al. (2006) ﬁtted to the trackway. Above, Tiktaalik (from Daeschler et al. (2006) with tail reconstructed from Panderichthys) is drawn to the same shoulder–hip length. Positions of pectoral ﬁns show approximate maximum ‘stride length’. c Muz. PGI 1728.II.15. Trackway showing alternating diagonal and parallel stride patterns. In (a) and (c), photographs are below, interpretative drawings are above. Thin lines linking prints indicate stride pattern. Dotted outlines indicate indistinct margins and wavy lines show the edge of the displacement rim. Scale bars, 10 cm. Reprinted with permission of Niedz´wiedzki et al. (2010). Copyright Macmillan Publishers, Ltd 5.4 Descendents of the Sarcopterygian Fishes and the Move onto Land 113

Fig. 5.12 Foot morphologies. a Laser surface scan of Muz. PGI 1728.II.1, left pes. b Complete- articulated left hind limb skeleton of Ichthyostega, MGUH f.n. 1349, with reconstructed soft tissue outline. c Left hind limb of Acanthostega, reconstructed soft tissue outline based on skeletal reconstruction in Coates (1996). We note the large size of the print compared to the limbs of Ichthyostega and Acanthostega, and that the print appears to represent not just the foot but the whole limb as far as the knee. d digit; fe femur; ti tibia; fi fibula; fib fibulare. Scale bars, 10 mm. Reprinted with permission Niedz´wiedzki et al. (2010). Copyright Macmillan Publishers, Ltd individual prints, which clearly possessed digits (Fig. 5.11). Janvier and Clemént (2010) state that the match between these tracks and the limb anatomy of Ichthyostega and Acanthostega is so impressively close that were they found in Famennian/Frasnian rocks they would be readily attributed to an Ichthyostega-like animal, as were the previously reported Late Devonian trackways (Fig. 5.12). In simple terms it means that there was a fully fledged tetrapod walking on terra firma at a time when according to perceived wisdom (a well-documented series of body fossils) fleshy lobed-finned fishes were barely able to haul themselves onto land. One of the problems of coming to terms with any reappraisal of tetrapod origins is that the long preceding Emsian period has yielded very few lobe-finned fishes, none of which could be regarded as potential tetrapod or elpistostegid ancestors. However, it seems that perhaps the most conceivable explanation is that workers have almost certainly overestimated the reliability of the fossil record and vice versa underestimated the vagaries of preservation. One answer for the apparent mismatch in the timing of elpistostegalian-tetrapod divergence may be consistent with the hypothesis proposed by Ahlberg and Johanson (1998) namely that much of the lower part of the tetrapodomorph stem lineage was paraphyletic and consisted of ‘osteolepiform’ fishes in which parallel evolution led to several groups of large predatory fishes that could potentially have been ancestors of the tetrapods. This is not too dissimilar to the polyphyletic grade system (groups that do not contain the common ancestor, and therefore had two separate origins) proposed by 114 5 Tetrapods and the Invasion of Land

Erik Jarvik although his character assessment with respect to homology may have been a trifle unconventional. It may be that paraphyletic forms may not simply have disappeared or become extinct but have left a much poorer evolutionary record. This may explain unexpected blips in the timing of tetrapod evolution. It opens up the question, should Tiktaalik be demoted from its transitional fish- tetrapod status? Yet, Tiktaalik may not directly represent a one-time sequential transformation of fish to tetrapod but it certainly shows how the characters that enabled the step from water to land occurred and that the steps may have occurred in other forms perhaps earlier and with more erratic fossil records. This kind of thing occurs in convergent evolution all the time, in the most distantly related forms, so there is every reason to think it occurred in more closely related forms. Chaos theory may apply to evolutionary events as it does to most other complex natural systems in which any small initial changes in the evolutionary process may have unexpected and profound final effects. To put in context what Janvier and Clemént (2010) refer to as a grenade lobbed into the picture, a brief glimpse of the massive progress made in tetrapod origins over many years of intensive studies is presented below so as not lose sight of their importance.

5.5 The First Tetrapods

The term tetrapod has been considerably reviewed in recent years but for the moment we will continue with the old usage. The earliest tetrapods, the labyrinthodonts, are from the Upper Devonian to the Triassic and include Ichthyostega, discovered by Gunar Save-Soederbergg in 1932 and considered as the first tetrapod. It was archetypal in possessing many common characters that connect the higher sarcopterygian fishes with the primitive tetrapods that enabled the first great step from the sea to the land. For about 50 years, Ichthyostega was the iconic sole representative of a tetrapod that had become adapted to life on land—and where things stood with little change in all that time. However, during the past 25 years or so the study of the origin of tetrapods has undergone a revolution that has transformed their study from essentially a single Devonian taxon, Ichthyostega, known from a single geographical area, East Greenland, to a rapidly expanding field represented by an increasing number of specimens and taxa from around the world (Clack 2006) (Fig. 5.13). The work of Clack and her colleagues/students during the mid-1980s saw a renaissance in work connected with early tetrapods. Her descriptions of the Late Devonian tetrapod Acanthostega marked the beginning of this vital period of work. Acanthostega had been collected by geologists at the same Greenland site as Ichthyostega, which had been long considered to be the key to the fish-tetrapod intermediate transition (in this context the term amphibian has been discontinued). They discovered several anatomical features that turned out to be diagnostic for early tetrapods. Acanthostega, despite being based on two fragmentary skull roofs, fulfilled almost exactly expectations for a transitional form between a fish and a 5.5 The First Tetrapods 115 tetrapod (Clack 2002b), a form closely related to Ichthyostega. But importantly as she states it also showed how incomplete was the knowledge on Ichthyostega itself. However, in contrast to Acanthostega, Ichthyostega was represented by a wealth of material, its skull had only been described in a preliminary study and only the tail and hind limb had been described in monographic detail (Jarvik 1952). This all changed during the past 25 years. The anatomy of Acanthostega has radically changed ideas about how this transition from fish to tetrapod took place. More recently the anatomy of Ichthyostega was reassessed and knowledge expanded almost exponentially with detailed descriptions of old material as well as of exciting new finds, which included important developments on its locomotory abilities. Studies by Jennifer Clack, Per Arlberg, Philippe Janvier, and others cited here on the origin of tetrapods have in the past 15 years or so resulted in some interesting conclusions, e.g., on the unique construction of the otic (ear) regions. This evolutionary event is now represented by at least nine named genera of Devonian tetrapods, several new near-tetrapods, and a number of new tetrapods from the Early Carboniferous (Clack 2006).

5.6 The Evolution of Terrestrial Vertebrate Locomotion

Jarvik (1952) figured Ichthyostega with relatively small paddle-like hind limbs. Later there was an interesting transmutation of Ichthyostega into a more stoutly terrestrial animal of popular image in which a different reconstruction by Jarvik, showed the hind limb as backwardly directed and paddle-like (1980), while in a later reconstruction by Jarvik (1996) it is stouter and more anteriorly directed (Fig. 5.13a, b). Part of the problem lies in a dearth of tetrapod fossil finds in an apparent gap in the fossil record known as Romer’s gap which extends from approximately 360 to 245 million years ago. However, with more material becoming available there has been a departure from this view (Coates and Clack (1995). It appears that in Ichthyostega at least the forelimbs appear to have been weight bearing and it was more likely to have been at least occasionally capable of terrestrial excursions. Unlike other labyrinthodonts, Ichthyostega had dermal fin rays and a dorsomedial tail fin supported by endoskeletal elements, indicating that it was still tied to the water. Speculations are that Ichthyostega was probably an obligate bellywalker, given that its tail was too short and its center of mass too far forward to employ the tail-supported walking trot which, on the other hand may have been employed by Acanthostega (Clack 1997). It is important to distinguish that the fish-tetrapod transition and the conquest of land were quite separate evolutionary steps (Clack 2002a). Despite bridging the water-land divide the Devonian tetrapods, including Ichthyostega, were essentially aquatic. They retained fish-like gills, fish-like tail, and despite a number of advanced osteological structures in the limbs that included digits, they clearly allowed no more than an ability to make brief, awkward incursions onto land 116 5 Tetrapods and the Invasion of Land

Fig. 5.13 Genoa River tracks. a Two adjacent tracks after Warren and Wakeﬁeld (1972). Arrow shows supposed direction of travel. Scale bar 50 mm. b Individual pes footprints drawn from the specimens, not to scale. (Scale may be judged from (a))

rather than adaptation to a terrestrial lifestyle, which probably occurred later in the carboniferous. This view preceded a new reconstruction of Ichthyostega by Ahl- berg et al. (2005) based on extensive re-examination of original material and augmented by newly collected specimens (Fig. 5.13c). Their reconstruction differs substantially from those previously published in particular with respect to the vertebral column and a variable neural spine height, suggesting that it was adapted for dorsoventral rather than lateral flexion. Despite such significant findings the authors concluded that Ichthyostega appeared to be an early and ultimately unsuccessful attempt at adapting the tetrapod body plan for terrestrial locomotion, divergent but not very remote from the lineage that successfully solved these adaptive problems (below) and ultimately gave rise to all living tetrapods. In a recent three-dimensional reconstruction of limb joint morphology in Ichthyostega, Pierce et al. (2012) show that Ichthyostega could not have employed typical tetrapod locomotory behaviors, such as lateral sequence walking and importantly that it lacked the necessary rotary motions in its limbs to push the body off the ground and move the limbs in an alternating sequence (Fig. 5.14). Interestingly, they concluded that early tetrapods such as Ichthyostega (long considered an iconic tetrapod) were unlikely to have made some of the recently described Middle Devonian trackways (above). The shock really was that the transition from fins to limbs despite earlier views had not quite been achieved in Ichthyostega, a transition that will be looked at a little closer below. 5.7 From Fins to Limbs 117

Fig. 5.14 Ichthyostega reconstructions. a, b Earlier reconstructions, skeletal, and life respectively. After Jarvik (1980). c Recent skeletal reconstruction. Reprinted with permission Ahlberg et al. (2005). Courtesy of Macmillan Publishers, Ltd

Fig. 5.15 Three-dimensional reconstruction of Ichthyostega from lCT scan data. Anterolateral view. Scale bar, 10 cm. Reprinted with permission Pierce et al. (2012). Copyright Macmillan Publishers, Ltd

5.7 From Fins to Limbs

As Clack (2012) has shown on the basis of recent analysis, Ichthyostega has some highly specialized features, particularly with respect to the limbs, that make it unsuitable as a representative of a Devonian tetrapod and consequently has to be 118 5 Tetrapods and the Invasion of Land moved down the evolutionary ladder in this important event (Fig. 5.15). Here, we shall look at the limb characters that gradually evolved to make the distinction between a fish and a tetrapod through the work of Coates et al. (2002 and a number of studies cited therein). Rhizodontids are regarded as the most basal group of stem tetrapods showing paired fin structures with particular resemblance to primitive tetrapod limbs. Their pectoral fins were large, broad distally, with well-developed lobes, endoskeletons, extensive scale cover, and elongate fin rays. The classic example is of Sauripterus from the Upper Devonian. The major axis of the fin consists of three mesomeres, presumed homologs of the humerus, ulna, and ulnare. Four radials articulate distal to the ulnare. The humeral shaft is subcylindrical with a strongly convex head and prominent ventral/postaxial and dorsal processes. The dorsal processes resemble the ectepicondyle and supinator process of limb humeri. In comparison with rhizodontids, the osteolepiform Eusthenopteron is by far the best known fish-like stem tetrapod. The major axis of the pectoral fin consists of four mesomeres. Only preaxial radials are present as well as a pair of small terminal radials beyond the fourth mesomere. First, third, and occasionally fourth mesomeres bear prominent post axial processes—thus the humerus analog has a large, well-developed entepicondyle, which narrows and curves to a slight hook. The pelvic fin is smaller than the pectoral, with an axis of only three mesomeres and with a postaxial process extending from only the second of these. Fin rays in osteolepiforms consist of conventional lepidotrichia. In panderichthyids the pectoral and pelvic fins are located in unusually ventral positions relative to species described so far (Boisvert 2005) (Fig. 5.16). The humerus and ulna homologs can be described as first and second mesomeres, but there is no obviously axial characteristic to the plate distal to the ulna and the relationship to that of tetrapod limbs is uncertain. The humerus is uniquely, for a fin, like those of early limbs. Most significantly it is dorsoventrally compressed and has separate flexor and extensor surfaces instead of the cylindrical shafts present in rhizodonts and osteolepiforms. The only prominent postaxial process in the entire fin endoskeleton is the entepicondyle of the humerus. The ulna is also dorsoventrally flattened, but like osteolepiforms and other finned examples, subequal in length to the rod-like radius. Fin rays consist of apparently conventional lepidotrichia that segment and branch only distally. Acanthostega, from the Upper Devonian of East Greenland, is the only basal tetrapod with digits known in any detail and whose girdles and limbs differ significantly from that of osteolepiforms (Fig. 5.17). The humerus has the characteristic l-shaped form of many early tetrapod limbs. Several canals open on its surface including the entepicondyle and ectepicondyle foramina. A number of processes are more prominent than in previous species. In anterior aspect, a slanting, short deltopectoral crest lies considerably below the level of the level of the radial facet. A distinct trough separates radial and ulna facets. The radius is long and spatulate and dorsoventrally flattened distally. Eight digits are present and form an anteroposteriorly arranged set or radiating series. The hindlimb is also profoundly different from all pelvic appendages of other species. The hind limb is 5.7 From Fins to Limbs 119

Fig. 5.16 Illustrated cladogram of the transformation of fins to feet with Eusthenopteron, Panderichthys, Tiktaalik, Acanthostega, Ichthyostega, and Dendreppeton. From Clack (2012). Image appears Courtesy of the publisher, Indiana University Press. All rights reserved larger than the forelimb and the bones of the hindlimb are zeugopod, the tibia and fibula are significantly shorter than the stylopod, the femur. Like the humerus, the femur bears a number of larger processes indicating much greater elaboration of the appendicular muscles. Unlike the humerus, the femur has a distinct shaft region, broadened proximillay to produce a femoral head. Distally the femur expands to produce articular surfaces for the tibia and fibula. The tibia and slightly smaller fibula are broad, flat, and subrectangular, with no semblance of a shaft region in either and contribute most to the paddle shape. Next in development were the limbs of Ichthyostega, which were described earlier in the chapter, but Tul- erpeton from the Upper Devonian of Central Russia, showing yet more advanced features, deserves mention. Tulerpeton differs in limb morphology in several important ways from Acanthostega and Ichthyostega. The humerus has slightly more elegant proportions than those of the above species, with moderate torsion between proximal and distal extremities, and the beginnings of a shaft region are apparent. The ectepicondyle is aligned proximally with the latissimus dorsi process and projects distally slightly anterior to the ulnar process. The deltopectoral crest lies 120 5 Tetrapods and the Invasion of Land 5.7 From Fins to Limbs 121 b Fig. 5.17 Pictures and drawings of Panderichthys rhombolepis, specimen GIT434-1. a Outline of the body of Panderichthys. Gray shading indicates preserved portions of Panderichthys rhombolepis specimen GIT 434-1. b Panderichthys rhombolepis specimen GIT 434-1 with head (h) and body (b) outlined. The pelvic girdle and fin are shaded in orange. c Pelvic girdle and fin. The matrix is distinguished from the fossil by an overlay of gray shading. d Specimen drawing. F femur; Fi, fibula; Fre fibulare; Int intermedium (proximal end of the); Pel pelvic girdle; T tibia. Vertical hatching indicates broken bone; grey shading indicates matrix; circles indicate thin dermal bone covering. e Reconstruction of the pelvic fin. Thick outline indicates preserved margin, thin outline indicates inferred margin, dotted lines indicate uncertain margin. Solid black scale bars, 10 mm. Reprinted with permission Boisvert (2005). Copyright Macmillan Publishers, Ltd

subcentrally along the anterior margin, and a distinct notch separates the supinator process from the bulbous radial condyle. Radius and ulna are slender and elongate, slightly shorter than the humerus and subcylindrical in cross-section. Six digits are present and the individual phalanges are elongate and very dissimilar to the stout phalanges of Acanthostega. In the hind limb the femur is particularly well-formed. It has a broad intertrochanter, a robust internal trochanter, a robust adductor blade and broad, elongate intercondylar fossa, and small but distinct ﬁbular fossa. The ankle bones are well developed and the phalanges and metatarsals are elongate and more robust than the corresponding elements of the manus and again there are apparently six digits (Fig. 5.18).

Fig. 5.18 Comparison of pectoral and pelvic fins. Pectoral (a, c, e) and pelvic fins (b, d, f)of Eusthenopteron (a, b), Panderichthys (c, d) and Acanthostega (e, f) all in ventral view. F femur; Fi fibula; Fre fibulare; H humerus; Int intermedium; R radius; T tibia; U ulna; Ure ulnare. Thick outline indicates preserved margin; thin outline indicates inferred margin; dotted lines indicate uncertain margin. Scale bars, 10 mm. Reprinted with permission Boisvert (2005). Copyright Macmillan Publishers, Ltd 122 5 Tetrapods and the Invasion of Land

Fig. 5.19 Cladogram of the pectoral fins of taxa on the tetrapod stem. Unlike other tetrapodomorph fishes (1), Tiktaalik has reduced the unjointed lepidotrichia, expanded the radials to a proximal, intermediate, and distal series, and established multiple transverse joints in the distal fin. The fin also retains a mosaic of features seen in basal taxa. The central axis of enlarged endochondral bones is a pattern found in basal sarcopterygians and accords with hypotheses that a primitive fin axis is homologous to autopodial bones of the tetrapod limb. In some features, Tiktaalik is similar to rhizodontids such as Sauripterus. These similarities, which are probably homoplastic, include the shape and number of radial articulations on the ulnare, the presence of extensive and branched endochondral radials, and the retention of unjointed lepidotrichia. Reprinted with permission of Shubin et al. (2006), Copyright Macmillan Publishers, Ltd

The compelling evidence is that the vast majority of changes involved in the ﬁn-limb transition were initiated and largely completed in animals that were mostly, and primitively, aquatic (Fig. 5.19). The pectoral and pelvic girdles underwent a concomitant development with the anterior and posterior limbs (Clack 2012). Together with limb development many other characters were developing as ‘‘preparation’’ for living on land. More 5.7 From Fins to Limbs 123 intermediate characters and information are being discovered on the sequence of character acquisition of early tetrapod lower jaws as vertebrates moved from aquatic to terrestrial habits. Daeschler et al. (2006) show that the mandibular canal of Densignathus is partially enclosed, an intermediate condition between the fully enclosed primitive condition seen in Ventastega and Ichthyostega and the open mandibular canal of most post-Devonian forms. In another study Daeschler and Shubin (1998) show through a number of new fossils that the origin of tetrapods was not a simple progression from aquatic ﬁsh to terrestrial amphibian—it involved a series of evolutionary experiments with structure, function, and ecology. Ahlberg (1991, 1995) has also shown from postcranial stem tetrapod remains from the Late Frasnian of Scat Craig, Morayshire, Scotland a combination of advanced and primitive characters in indeterminate material that may belong to Elginerpeton, Ichthyostega and Acanthostega. They suggest that on the basis of a postorbital bone that may belong to Elginerpeton and resembling the postorbitals of Ichthyostega and Acanthostega that the typical stem tetrapod facial morphology had evolved before the end of the Frasnian. Among such developments were those connected with the vitally important ability to breathe on land.

5.8 The Problems of Breathing on Land

Lungs may have evolved for a quite different purpose, to assist in an aquatic animal’s buoyancy. The heavy scale armour of the early bony fishes would certainly weigh the animals down. Cartilaginous fishes lack a swim bladder but to compensate against sinking, as we saw in Chap. 4, many have enormous livers with oils less dense than the surrounding medium, e.g., in the white shark, Car- charodon carcarias, the liver may occupy a volume about a third that of the body. Also, the pectoral fins may act as hydrodynamic organs to prevent sinking. Open sea sharks need to swim constantly to avoid sinking into the depths. The swim bladder in bony fishes may perform another important function, i.e., as a store for oxygen (as in some modern fishes with swim bladders), which was rather low in the Devonian, perhaps half that of the present day (Clack 2012). This would be important for animals with high energy demands such as active vertebrate predators and prey. The swim bladder may therefore be considered as a primitive de facto lung. For instance, in lungfishes such as Protopterus, bowfins, and bichirs the swim bladder is supplied by paired pulmonary arteries that branch from the posterior most aortic arc—a similar system is also found in terrestrial salamanders. We will look at this in a little more detail below. The transition from water to land would have necessitated aerial respiration and greater dependence on the lungs. How and when this occurred is not fully clear. However, Coates and Clack (1991) reported the discovery of a fish-like branchial skeleton in Acanthostega gunnari, from the Upper Devonian of East Greenland, one of the earliest ‘‘tetrapods’’ known. A number of features such as a proximally expanded ceratohyal and large, ventrally grooved ceratobranchials (as in modern 124 5 Tetrapods and the Invasion of Land

fishes) and shoulder girdle bearing a postbranchial lamina as in fishes, which supports the posterior wall of the opercular chamber, all suggest that Acanthostega may have retained fish-like internal gills and an open opercular chamber for use in aquatic respiration (Coates and Clack 1991). This implies that these early tetrapod forerunners were basically aquatic. Edwin Colbert (1955) suggested that the problem of respiration out of the water had been preempted for the tetrapods by their fish ancestors, the crossopterygians, in which lungs were well developed and probably frequently used. If this is so the early would-be tetrapods may have had a ‘‘head start’’ on the problem of breathing in air, so it was actually not so much of a problem for them and that they merely had to go on using the lungs when they became predominantly terrestrial. Consistent with this is the view that the rhipidistian fishes probably had two respiratory apparatuses, a branchial (gill) system for aquatic respiration and a pulmonary (lung) system for air breathing (as in the Australian lungfish). Recent findings suggest that a similar dual breathing system may also apply to fishes such as Acanthostega (Clack 2012). One problem that requires consideration in these early tetrapod forerunners such as Acanthostega is the dynamics of ventilation. Lung ventilation in vertebrates occurs by a process of decreasing and increasing body volume. The pleural cavity in mammals and pleuroperitoneal cavity in other vertebrates decrease during inspiration and increase during inspiration. From a mechanical point of view, decreasing the volume of a body is a relatively simple process—surround it with muscles that can be contracted so as to squeeze. Increasing the body cavity poses a more complex problem because muscles generate force only in the direction of shortening. The solution is a system of levers—reptiles and mammals use pleural ribs to convert forces in the required direction—the diaphragm in mammals being perhaps the most sophisticated system of levers. Air-breathing fishes and amphibians on the other hand do not possess an abdominal or thoracic lever system to increase their body volume for inspiration. Instead they use a lever system in the head, the hyoid apparatus, to expand the mouth cavity and suck in air, after which the mouth cavity is closed and air is pumped into the lungs. However, Brainerd (1994) has shown one family of air-breathing fish, polypterids, that do not use the hyoid apparatus for lung ventilation but rather a system referred to as recoil aspiration (Fig. 5.20). The body ventral scale jacket is deformed as the body volume decreases and as the fish opens its mouth to inhale the integument recoils thereby sucking air into the lungs. This occurs by a combination of the movement of the scales and collagen fibers during exhalation as a consequence of which energy is stored in the stretched collagen fibers which enables the body to return to its original shape, thereby solving the problem of absence of levers as well as conserving energy (Fig. 5.21). While the development of lungs in polypterids is different from that of sarcopterygian vertebrates including lungfish, amphibians, and amniotes, and in the constraints with respect to the aerating mechanism, the significance of an elastic mechanism in amphibians in reinflating the lungs would be extremely useful (Fig. 5.20). The role of collagen fibers, particularly as a crossed fiber architecture in the dermis, acting as springs has been documented in many vertebrates and 5.8 The Problems of Breathing on Land 125

Fig. 5.20 A diagrammatic view of Polypterus sp. The dorsal and ventral aspects of the scales overlap to form scale rows, which wind in a helical pattern around the long axis of the ﬁsh. From Brainerd (1994)

Fig. 5.21 Diagrammatic transverse section through a polypterid combined with a pressure trace from the pleuroperitoneal cavity of a 22 cm long P. senegalus. Zero pressure is the ambient hydrosostatic pressure at the level of the transducer and zero time is the point of minimum pressure which also corresponds to the beginning of mouth opening for inhalation. In the diagram the fish’s head is pointing toward the observer. The sections correspond to a position approximately halfway along the fish’s body, and thus only the right lung, which is much longer than the left is represented. In life, the body cavity space outside of the lungs is filled with water- based tissues which cannot change volume. During exhalation, active emptying of the lungs deforms the stiff integument causing the body cavity pressure to decrease. Upon inspiration, the deformed integument recoils to a round shape, thus sucking air into the lungs and resolving the reduced body cavity pressure. From Brainerd (1994) 126 5 Tetrapods and the Invasion of Land invertebrates and will be considered in greater detail in Volume 2. With respect to modern amphibians there have been few similar studies although Schwinger (1995) and Zanger et al. (1995) have demonstrated the presence of a crossed fiber pattern of collagen fibers in the dermis of Xenopus laevis and Rana esculenta. While the latter authors consider the fiber system similar in structure compared with that of fishes they suggest a more passive function than in fish locomotion. However as a number of authors have shown the function of a crossed fiber architecture of collagen in the dermis extends beyond axial locomotion as, e.g., seen in polypterids (Brainerd 1994) and in a number of animals where it acts as a spring (Pabst 1996) including in the linea alba in humans (Axer et al. 2001). Evidence of such fiber patterns acting as springs in others animals suggests a similar potential in amphibians including in fossil forms such as Acanthostega. Future integrated studies coupling mechanical data with kinematics may broaden our understanding of the problems of air breathing in primitive vertebrates.

References

Ahlberg PE (1991) Tetrapod or near tetrapod fossils from the Upper Devonian of Scotland. Nature 354:298–301 Ahlberg PE (1995) Elginerpeton pancheni and the earliest tetrapod clade. Nature 373:420–425 Ahlberg PE, Johanson Z (1998) Osteolepiforms and the ancestry of tetrapods. Nature 395:792–794 Ahlberg PE, Clack JA (2006) A firm step from water to land. Nature 440:747–748 Ahlberg PE, Clack JA, Blom H (2005) The axial skeleton of the Devonian tetrapod Ichthyostega. Nature 437:137–140 Axer H, Keyserlingk DG, Prescher A (2001) Collagen fibers in linea alba and rectus sheaths (II). J Surg Res 96:127–134 Boisvert CA (2005) The pelvic fin and girdle of Panderichthys and the origin of tetrapod locomotion. Nature 438:1145–1147 Brainerd E (1994) Mechanical design of polypterid fish integument for energy storage during recoil aspiration. J Zool Lond 232:7–19 Clack JA (1997) Devonian tetrapod trackways and trackmakers; a review of thefossils and footprints. Palaeogeogr Palaeoclimatol Palaeoecol 130(227):250 Clack JA (2006) The emergence of early tetrapods. Palaeogeogr Palaeoclimatol Palaeoecol 232:167–189 Clack JA (2002a) The dermal skull roof of Acanthostega, an early tetrapod from the Late Devonian. Trans R Soc Edin (Earth Sciences) 93:17–33 Clack JA (2002b) An early tetrapod from Romer’s Gap. Nature 418:72–76 Clack JA (2012) Gaining ground: the origin and early evolution of tetrapods, 2nd edn. Indiana University Press, Indiana Coates MI, Clack JA (1991) Fish-like gills and breathing in the earliest known tetrapod. Nature 352:234–236 Coates MI, Jeffery JE, Ruta M (2002) Fins to limbs: what the fossils say. Evol Dev 4:390–401 Coates MI, Clack JA (1995) Romer’s Gap—tetrapod origins and terrestriality. In: Arsenault M, Lelievre, H, Janvier P (eds) Bulletin du Museum National d’Histoire Naturelle, Paris, 373–388 Colbert EH (1955) Evolution of the vertebrates. John Wiley & Sons, Inc., New York Daeschler EB, Shubin N (1998) Fish with fingers? Nature 391:133 References 127

Daeschler EB, Shubin NH, Jenkins FA (2006) A Devonian tetrapod-like fish and the evolution of the tetrapod body plan. Nature 440:757–763 Daeschler EB, Shubin NH, Thomson KS, Amaral WW (1994) A Devonian tetrapod from North America. Science 265:639–664 Fastovsky DE, Weishhampel DB (1996) The evolution and extinction of the dinosaurs. Cambridge University Press, Cambridge Gardiner B, Janvier P, Patterson C, Forey P, Greenwood PH et al (1979) The salmon, the lungfish and the cow: a reply. Nature 277:175–176. doi:10.1038/277175b0 Hall BK (ed) (2007) Fins Into limbs: evolution, development, and transformation. Chicago University Press, Chicago Halstead LB (1978) The cladistics revolution—can it make the grade? Nature 276:758–760 Hennig W (1950) Grundzüge einer Theorie der phylogenetischen Systematik. Deutscher Zentralverlag Hennig W (1966) Phylogenetic Systematics (trans: Davis D, Zangerl R, reprint. 1979). University of Illinois Press, Urbana Janvier P (1998) Forerunners of four legs. Nature 395:748–749 Janvier P, Clemént G (2010) Muddy tetrapod origins Nature 463:40–41 Jarvik E (1952) On the fish-like tail in the ichthyostegid stegocephalians. Meddelelser om Grønland 114:1–10 Jarvik E (1980) Basic structure and evolution of vertebrates. Academic Press, London Jarvik E (1981) Lungfishes, tetrapods, palaeontology and plesiomorphy: a reply. Syst Zool 30:378–380 Jarvik E (1996) The Devonian tetrapod Ichthyostega. Fossils Strata 40:1–206 Laurin M, Girondot M, de Ricqlès A (2000) Early tetrapod evolution. Tree 15:118–123 Long JA, Young GC, Holland T, Senden TJ, Fitzgerald EMG (2006) An exceptional Devonian fish from Australia sheds light on tetrapod origins. Nature 444:199–202. doi:10.1038/ nature05243 Niedz´wiedzki G, Szrek P, Narkiewicz, K, Narkiewicz M, Ahlberg PE (2010)Tetrapod trackways from the early Middle Devonian period of Poland. Nature 463:43–48 Pabst DA (1996) Springs in swimming animals. Amer Zoo 36:723–735 Pierce SE, Clack JA, Hutchinson JR (2012) Three-dimensional limb joint mobility in the early tetrapod Ichthyostega. Nature 486: 523–526 Rosen DE, Forey PL, Gardiner BG, Patterson C (1981) Lungfishes, tetrapods, palaeontology and plesiomorphy. Bull Am Mus Natur Hist 167:163–275 Schwinger G (1995) mechanical properties of the skin of Xenopus laevis (Anura, Amphibia). J Morphol 224:15–22 Shubin NH, Daeschler EB, Jenkins FA Jr (2006) The pectoral fin of Tiktaalik roseae and the origin of the tetrapod limb. Nature 440:764–771 Warren JW, Wakefield NA (1972) Trackways of tetrapod vertebrates from the Upper Devonian of Victoria, Australia. Nature 238: 469–470 Zanger K, Schwinger G, Greven H (1995) Mechanical properties of the skin of Rana esculenta (Anura, Amphibia) with some notes on structure related to them. Ann Anat 177:509–514 Zhu M, Ahlberg PE (2004) The origin of the internal nostril of tetrapods. Nature 432:94–97 Chapter 6 Crucial Vertebrate Innovations

6.1 Best Foot Forward

The plight of the tetrapod forebears at the closing stages of the Carboniferous was quite dire. They had become the unfortunate victims of a change in the earth’s climate that irrevocably reversed their earlier good fortune. The lush swamps were drying up completely or being reduced in size, heralding more arid conditions and areas of desert and near-desert (DiMichele et al. 2006). The rich life afforded to the large numbers of early tetrapod forebears living in the aquatic and semi-aquatic environment was under extreme stress (Fig. 6.1). Those vertebrates, the ‘‘amphibians,’’ that survived were restricted to the small areas of lakes and rivers that remained. They were ill-equipped for dry conditions and tightly bound to the water. Their skin, the epidermis in particular, was extremely thin to allow gaseous and nutrient exchange whereas out of water for any measure of time highly prone to rapid dehydration. The mucous glands would only help to stall drying up over short periods of time (Fig. 6.2). Hence, the generally arid conditions in the Carboniferous proved catastrophic for most of the ‘‘amphibians’’ with only a few groups surviving. However, their misfortune would force one of the most successful radiations on land, that of the reptiles, which includes the dinosaurs and the mammal-like reptiles, the latter would ultimately give rise to our own forebears (Chap. 8). Later, we will look at how a major change in the structure of the skin would allow the conquest of land. But first we will look at another remarkable development. If one were compelled to choose a single definitive breakthrough that would save the day for the vertebrate conquest of the land it might well be something so small that one could hold it in the palm of one’s hand. It is the amniotic egg (sometimes referred to as the cleidoic egg). Early tetrapods made the first great step in clambering onto land but they were still tied to the water, most importantly because of one vital factor involving their mode of reproduction—their eggs. Amphibian eggs are shed directly into the water, i.e., they are fertilized externally which means they must be accessible to the male’s sperm. It also means that there is a larval stage, as e.g., the frog tadpole, which must fend for itself at a rather

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 129 DOI: 10.1007/978-3-642-53748-6_6, Ó Springer-Verlag Berlin Heidelberg 2014 130 6 Crucial Vertebrate Innovations

Fig. 6.1 Reconstruction of the closing stages of the Carboniferous

Fig. 6.2 Skin of a typical modern-day amphibian

immature and small state of development. The amniotic egg of reptiles in contrast is fertilized internally and then laid on land. Apart from gas exchange, the amniotic egg constitutes a closed system, which gives nothing to the outside and requires nothing from it (besides gas exchange). It is a marvel in many ways (Fig. 6.3). 6.2 The Amniotic Egg 131

Fig. 6.3 Cross-section of a typical amniotic or cleidoic egg. a Structural components. b Various membranes and internal structures shown at a late embryonic stage

6.2 The Amniotic Egg

As we saw in Chap. 2, the ﬁrst jawless vertebrates were protected against the harsh outer world by a tough outer integument. The amniotic egg simply took that protection a step further—protection of the unborn young, the embryo, by a tough outer integument. The outer shell of the egg in effect plays the same role as the integument in the adult animal. In reptiles the shell is frequently a soft leathery structure whereas in birds it is a calcareous shell. The shell may be so tough that in embryonic birds the beak has a structure known as the eggtooth or eggbreaker (disappears in adulthood), which aids in breaking through the shell at the appropriate time. This eggtooth is thought to have been identiﬁed in the embryo of sauropod dinosaurs (Garcia 2007) (Fig. 6.4). Within the shell the developing embryo is further protected by a series of membranes. The most important of these is the amnion, which envelops the embryo and encloses it in a watery environment, mimicking the external pond that was necessary for the survival of the larval or 132 6 Crucial Vertebrate Innovations

Fig. 6.4 Premaxillary embryonic morphology of titanosaurs from Auca Mahuevo (photograph and interpretative drawing). Both premaxillae in articulation (M CF-PVPH-659). a ventral; b rostroventral views showing the egg-tooth-like structure. a alveoli, dt displaced tooth, et eggtooth, is interpremaxillary symphysis, t tooth. Scale bar equals 4 mm. c skull reconstruction of titanosaurian embryos. d interpretive drawing of a hatching titanosaur (note the egg-tooth-like structure (arrowed) in both figures. After Garcia (2007) 6.2 The Amniotic Egg 133 tadpole stage of amphibians. Connected to the amnion is a large yolk sac, which is surrounded by albumen, or egg white. The albumen in turn is surrounded by two shell membranes (inner and outer). A further large sac is connected to the amnion and responsible for the removal of embryonic waste products due to the active metabolism of the growing embryo. This membrane also forms an embryonic ‘‘lung’’ enabling the absorption of oxygen from the atmosphere via the porous shell. The amnion thus frees reptiles from dependence upon the pond forever by eliminating the need for a living and vulnerable larval stage (Fig. 6.3). However, the amniotic egg is also more energetically cost-effective enabling reptiles and birds to produce fewer eggs but which are more protected, known as the K-strategy of reproduction, i.e., more energy is spent on care of fewer offspring (most fishes and amphibians compensate by producing large numbers of eggs with low quality of care, known as the r-strategy in which the vast majority never reach fruition). This development of the amniotic egg is so important that it unites many of the higher vertebrates that include the magnificent dinosaurs and present-day birds in a group, the Amniota. As far as physical protection is concerned the eggshell performs more or less the same role as the integument of many adult animals. Bird eggshells may be considered as typically adapted to dry terrestrial conditions. The eggshells consist mainly of calcite, unlike the majority of vertebrate hard tissue, which is phos- phatic. The structure of the domestic chicken (Gallus gallus) egg is typical. Under a microscope, beneath an outer protein cuticle can be seen a ‘‘palisade’’ layer, a series of prisms arranged normal to the surface while on the inner side is a cone layer that inserts into an outer shell membrane. When decalcified the organic matrix seems to form an unstructured mass which is about 70 % protein, containing little cysteine and no hydroxyproline. Avian eggshells are pierced by innumerable narrow pores connecting the inside to the outside through which gaseous exchange takes place (Wainwright et al. 1976, p. 218; Deeming 2006) (Fig. 6.5). There is an apparent gap in our knowledge on how the amniotic egg evolved— at one end there is the unprotected egg of amphibians, which included early near- tetrapods, e.g., Ichthyostega, which must be fertilized externally in water and at the other end we have a shelled egg with internal fertilization. However, work by Charles Deeming and others is beginning to show how the transformation came about. Because of investigations into the chemistry of the shell we may be getting a glimpse into the chemical changes that occurred during the evolution of the amniotic egg. In contrast to typical bird eggs, most snakes and lizards produce eggs with flexible shells that allow interactions with the environment to maintain water balance, i.e., it is still partially dependent on an external water source (Deeming 1991, 2006). Geckos, in contrast produce rigid eggshells that are independent of an external source of water and can be oviposited in more open, dryer locations (Pike et al. 2011) (Fig. 6.6). A study by Sexton et al. (2007) showed that the extent of permeability via the eggshell in squamate reptiles depends on its composition, particularly with respect to the amino acid distribution. Rigid gecko eggshells they found had significantly lower levels of 7 of the 17 amino acids evaluated 134 6 Crucial Vertebrate Innovations

Fig. 6.5 Diagram of a section through a typical amniotic/cleidoic eggshell (chicken, Gallus gallus). a Detail of components prior to development of embryo. b Diagram of section showing embryo, yolk sac, and major membranes

Fig. 6.6 Geckos, uniquely among lizards lay hard- shelled eggs with calcareous shells. From Alexander and Marais (2007)

and that proline was the most important amino acid in distinguishing between these two groups of eggshells. Proline occurred at significantly higher levels in flexible eggshells and high levels of proline had also been observed in the eggshells of other species. The findings suggested that proline and other amino acids may, as in plants, be associated with the alleviation of water and salt stress. The flexible shell, and more specifically its chemistry, may provide us with clues to the evolutionary stages that led to the transformation from the amphibian eggs to the advanced amniotic egg of advanced reptiles and birds. In contrast to the eggs in Lepido- sauromorpha (lizards), crocodile, non-avian dinosaur, and bird eggs are all 6.2 The Amniotic Egg 135

Fig. 6.7 Hyphalosaurus baitaigouensis eggs and hatchling. a Unhatched egg with embryo (LPM-R169). b Hatched egg with associated neonate (LPM-R168). Both to same scale. Semitransparent outline in (a) indicates approximate position of embryo. Close-ups in (b) show details of skeleton including open notochordal canals (nc) in cervical vertebrae, shape of interclavicle (ic), and poor ossification of tibia (t) and fibula (f). Scale bar, 10 mm. c Relative size of eggs, neonate, and adult H. baitaigouensis. Scale bar = 5 mm. From Hou et al. (2010) characterized by a rigid shell formed from a thick outer layer of inverted calcareous wedges, with organized pore openings positioned between the calcareous wedges and patent on the surface. In the fossil record flexible, or soft-shelled, eggs are rare, leaving large gaps in our knowledge of the reproductive biology of many tetrapod groups. However, a recent study by Hou and colleagues (2010) reports on two flexible-shelled eggs of the hyphalosaurid choristodere (enigmatic aquatic reptiles that died out in the Miocene), Hyphalosaurus baitaigouensis, from the Early Cretaceous of China, one containing an embryo and the second associated with a neonate (newborn) (Fig. 6.7). Results from SEM of the samples of the eggs showed details of eggshell ultrastructure. In cross-section the external mineralized layer of 136 6 Crucial Vertebrate Innovations

Fig. 6.8 A female individual of Darwinopterus associated with an egg from the Tiaojishan Formation of Liaoning Province, China (ZMNH M8802). a Skeleton with fractured forearm (arrow) and associated egg (double-headed arrow). Scale bar, 5 cm. b Sacrum, pelvis, and the associated egg. Scale bar, 2 cm. as articular end of sacral rib, ca caudal vertebrae, cv cervical vertebrae, dv dorsal vertebrae, f femur, h humerus, il ilium, ip ischiopubis, impression of egg, i.e., md mandible, ps pes, pp prepubis, ra radius, ri rib, sc scapulocoracoid, sk skull, sv sacral vertebrae, u ulna, ta tail, t tibia, wph2 wing-phalanx 2. Reprinted from Lu et al. (2011). Courtesy AAAS the egg is very thin (\10 lm). While there may have been connecting irregularly spaced connection pores with the inner layers of the egg they cannot be determined with certainty. Regularly spaced, structured pores appear to have been absent. Below the mineralized layer is a thick layer of irregular structure that probably represents the degraded fibrous portion of the eggshell. The very thin eggshell seems reasonable if ideal conductivity of water vapor is important. Hou and colleagues note that incubation strategy is dictated mainly by water vapor conductance properties, and calcification is negatively correlated with conductance as indicated earlier by Deeming and colleagues. Their findings also agree with research that has shown that flexible-shelled eggs have high mass-specific water vapor conductance values and increase greatly in mass over incubation. The thin external layer of the fossil eggshell implies high conductance values and a requirement of significant moisture for successful incubation, i.e., the egg is still tied to the environment in some respects. H. baitaigouensis probably deposited its eggs in closed nests near the shoreline to avoid desiccation. This is particularly pertinent with respect to the findings by Sexton and colleagues (2007) and on how the flexibility of the shell may provide clues to the evolutionary stages that led to the transformation from the amphibian eggs to the advanced amniotic egg. Perhaps, most surprising in this respect but providing further information on flexible egg shells in the fossil record from an unexpected animal group are findings by Lu et al. (2011) (Fig. 6.8), who described a sexually mature individual 6.2 The Amniotic Egg 137 of the pterosaur, Darwinopterus, preserved together with an egg. The material was recovered from Jurassic sedimentary rocks (160 MYA) in Liaoning Province, China. The egg preserved near the pelvic region of the pterosaur shows no trace of mineralized shell, cracking, or crazing. Instead the external features indicate that the egg had a relatively soft, parchment-like shell with possible pores. The authors suggest that presence of shell membranes that were sufficiently well developed to leave an impression were an indication that the egg had reached a late stage of development close to oviposition (laying). Surprisingly, the ratio of egg mass to adult mass is relatively low, unlike in birds but comparable to values for extant squamate reptiles. The evidence of a parchment-like eggshell points to burial and significant uptake of water after oviposition and that this allowed low parental investment (dinosaur and pterosaur eggs are discussed in Chap. 7).

6.3 The Reptile Integument

Life on land for the amphibians (a now outmoded term for animals leading to tetrapods but used for convenience) necessitated other changes. The epidermis in amphibians and more so in the reptiles takes over the functional role occupied by the dermis in fishes and their ancestors. The dermis we will see later plays a somewhat different but vital role in the mechano-structural function of the skin as a whole. However, the epidermis in living amphibians is thin, no more than 8-10 cell layers on average. To some extent to compensate for the relative vulnerability with respect to contact with the air, the epidermis has mucopolysaccharides that helps control desiccation (Duellman and Trueb 1986). Part of the reasons for the thinness of the epidermis is to facilitate cutaneous respiration which means that it must be kept moist and lubricated by secretions from abundant mucous glands located in the epidermis. Significantly, we see the beginnings of what will become the defining chemical material of the epidermis of all later amniotes—keratin. In extant amphibians, there is only a thin layer of dead keratinized cells, which may play some small part in coping against abrasion. The major breakthrough with respect to airproofing and resistance to abrasion by the epidermis occurs in reptiles. The adaptation involves a fundamental development of the skin with for the first time a distinctive epidermis that forms a complete body covering of scales, which are horny, tough extensions of the stratum corneum (see below and Chap. 7) (Fig. 6.9). It takes over the role of the dermis in fishes. The transition from an aquatic or semi-aquatic environment to a purely terrestrial one required an ever increasing number of new adaptations. The early reptilian integument had to adapt to the challenges of terrestrial life, developing a multilayered stratum corneum of the epidermis capable of providing mechanical protection, prevent desiccation, and provide ultraviolet protection, which together with the dermis provides a double layer of protection. The momentous development was in the composition of the scales, of an entirely new material—b-keratin. 138 6 Crucial Vertebrate Innovations

Fig. 6.9 Section of the skin and epidermis of a squamate reptile shortly before a molt. From Hildebrand (1995)

b-keratin is extremely tough and stable, and critically, extremely lightweight, unlike bone. b-keratin increases the mechanical resistance of the epidermis and protects the underlying softer a-keratin layer (Bragulla and Homberger 2009; Lingham-Soliar et al. 2010; Lingham-Soliar and Murugan 2013). It is also the main structural material of feathers, which will be discussed in greater detail in Volume 2 (it is probably safe to say that bird ﬂight would not have evolved were it not for this extremely tough and light material). For better mechanical protection, diverse reptilian scale types evolved. Reptiles solved the problem for ﬂexibility of the exoskeleton through folding the skin with a protruding outer layer of b-keratin and an underlying soft inner layer of connective tissue, containing a-keratin and perhaps some b-keratin, which may be much thinner and less rigid to form the ‘‘joint’’ or hinge (Fig. 6.10).

6.4 Scale Types

Three typical reptile scale types exist (Maderson 1965). In most reptiles the scales lie closely adjacent to the neighboring scales rather than overlapping them as in ﬁshes. The entire body and tail in snakes are covered with scales, which as we have seen are corniﬁed (converted to hard tissue) folds in the epidermal layers of the skin. These scales are usually arranged in longitudinal rows, the numbers and arrangement of which are characteristic of the species of snake. In overlapping scales the scale is asymmetric with the hinge region assigned to the posterior end. Tuberculate scales are found on the body of some lizards, like the gecko, which has a round surface without an anterior-posterior (A-P) axis (Maderson 1965, 1970). 6.4 Scale Types 139

Fig. 6.10 Schematic drawings showing different types of reptile scales. Scales (resting phase) are shown in multiple layers with names labeled in panel B. a Nonoverlapping tuberculate type scales. b Overlapping scales commonly seen in squamates. c Variations of microstructures from the Oberhäutchen layer illustrating short spines in (a, b) and long setaes in (c) (such as those in the adhesive pad lamellae in geckos). d Pits on the scales of anole, gecko, and iguana (mainly epidermal sensory organs). e Tactile sensory organ on the hinge side of a scale in Agama. Some follicle-like structures have clustered dermal cells associated to their base). f Scales with ridges are seen on the back of skink or the neck of anole. g Frills, or very elongated scales, are seen on the back of iguana. h The horn on the head of chameleon contains a bony element core (osteoderm). i Scales on the limb of crocodilians show only minor overlapping. j Keeled scales with a central, elevated corneous ridge are seen on the dorsal body of crocodilians and some armored agamid lizards (e.g., Australian spiny desert lizard or moloch). Legends: a, ﬁne ‘‘hair’’ on scales of anoles; b, Micro-ornamentation on scales of snakes; c, Toe pad of anole or gecko; dermal cells clustered at the base of sensory organs in Agama; AK a-keratin; BK b keratin; BP bone element. From Chang et al. (2009) 140 6 Crucial Vertebrate Innovations 6.4 Scale Types 141 b Fig. 6.11 Unusual scale types (a, b) Some scales have specialized surfaces to help them climb. a Toe pad of anole. (b) Longitudinal sections of digital pads shown in a. Note the hairy structures on the setae are variations of the Oberhäutchen layer. c, d Some scales have dorsal ridges (keels) which increase the protective properties of the scales. c Dorsal skink scale with three ridges. d Cross-section of an anole neck scale which exhibits one central ridge, e.g., some scales form pits, sensory organs with a simple structure at the dermal-epidermal junction. e Low power view of the distal edge of a scale. f Detail of the pit sensory organ in, e.g., Agama tactile sensory organ. It shows more of a complex epidermal structure with a nerve terminal in its base (arrow). Scale bars, 100 lm. Star arrow indicates the sensorial ﬁlament. From Chang et al. (2009)

Fig. 6.12 Reticulated scale pattern in a python. From Alexander and Marais (2007)

Some lizards, such as the iguana, also have an elongated scale (frill) on the dorsal region of the body (Chang et al. 2009) (Fig. 6.11). There are four extant orders of reptiles, Crocodilia (alligators and crocodiles), Chelonia (turtles and tortoises), Squamata (lizards and snakes), and Rhyncho- cephalia (tuatara) (Pough et al. 2004). Crocodilian scales show relatively few variations in gross morphology, are generally only a little overlapped and show a large surface composed mainly of hard b-keratinized, stratified epidermis (Alibardi and Thompson 2001). In squamates the most frequently occurring type of scale is the overlapping scale, which has distinct outer and inner surfaces, the most common scale type on the body of lizards and snakes. The overlapping scale is asymmetric, with the hinge region posteriorly. The outer surface is strongly cornified, which provides stiffness. During embryonic development, the morphogenesis of overlapping scales passes through the flat two-layered epidermis stage, the symmetric scale anlagen stage, the asymmetric scale anlagen stage, and the b-keratinizing asymmetric scale stage (Maderson 1985; Alibardi 1996, 1998). Specialization by cornification in the basic anapsid cotylosaurs involved a new form of corneous proteins, beta-keratins, from ancestral proteins. Snakes show perhaps the most varied shapes and sizes and patterns of their scales (Fig. 6.12). The scales in some species have sensory structures on the posterior margins called apical pits, and all scales have various micro-ornamen- tations, consisting of hair-like projections, holes, spinules (small spines), and other specializations visible only through an electron microscope. Scales on the ventral surface of the body are modified into broad plates in the majority of species and 142 6 Crucial Vertebrate Innovations are used in locomotion. In the anaconda the ventral scales are highly reduced in size, which appears to be a secondary reduction correlated with life in a predominantly aquatic environment where locomotion is achieved by swimming or movement in burrows. A unique feature of reptile skin is the exquisite arrangement of scales and pigment patterns, making them testable models for mechanisms of pattern formation (Chang et al. 2009). The colors of the scales may help in the snake’s camouflage in hiding its presence but in some species scales may be especially bright and colorful as a warning strategy against a possible unwitting encounter by another animal. Such warnings suggest some kind of previous learning or instincts by animals generally. Other harmless snakes may mimic coloration and patterns of some of the venomous species and consequently being avoided by potential predators. The epidermis of lepidosaurs is quite complex and interesting in other ways. In these reptiles an entire generation of the epidermis is sloughed off as a single unit at least several times a year, an activity that we are most familiar with in snakes. The process may be under hormonal control. Hildebrand (1995) described the process as it occurs from just after molting, i.e., the resting stage. The epidermis at this point consists of the stratum germinativum and an outer epidermal generation that characteristically has five layers. From the outside inwards there is first a thick, dead, acellular layer heavily keratinized by b-keratin. The surface of this layer is called the oberhautchen and has microscopic spicules (see below). Under this b-keratin layer is a thin mesos layer of unknown significance followed by a moderately thick layer of loose, dead, anucleate material comprised of a-keratin. Below this are two layers of living cells, an outer layer that will later be taken into the a-layer, and an inner layer that will later become clear and create the separation leading to sloughing (Fig. 6.10). At the end of the resting stage the germinal epithelium rapidly proliferates the various layers of an inner epidermal generation. As they mature they separate from the innermost layer of the outer epidermal generation, preceding sloughing. Recently, new investigations were done with respect to the mechanism of sloughing. Alibardi et al. (2009) have shown using microscopy and immunological and molecular studies that in the layer stratifica- tion of the epidermis there is a periodic reduction in the production of the cornification proteins that may be an important factor associated with the evolution in lepidosaurian reptiles of the cyclical shedding or sloughing of the outer part of the corneous layer. Furthermore, immunocytochemical, biochemical, and molecular biological studies have suggested that the cytological differences among the six main layers of lepidosaurian epidermis (oberhautchen, beta-, mesos-, alpha-, lacunar, and clear-layers) are derived from the genetic control of the production of their hard proteins during the sloughing cycle. How is scale renewal achieved in crocodiles and chelonians in which the keratinous plate on the outer surface is a large flat scale called a scute? Unlike in lepidosaurs, scutes of crocodiles and chelonians are not shed as a whole. Growth adds keratinous material over their entire inner surfaces, thus compensating for wear. Each growth wave extends beyond the previous margin to produce the familiar growth concentric rings on a turtle’s shell. However, in the individual 6.4 Scale Types 143 scutes of the carapace and plastron there must be some way of getting rid of old keratinous material. In some species of turtles (chelonians) there is a switch from thicker beta- into thinner alpha-keratin cells that determines the formation of a shedding layer, permitting a slow intra-corneous detachment of the more external layers with the sloughing of the superficial part of the corneous layer and formation of a new beta-layer; this process occurs only in shedding turtles but not in non-shedding turtle or in the tortoises where the superficial part of the corneous layer is gradually lost by natural wear (Alibardi et al. 2009), very much as we get rid of old keratinous skin, one of the main components of dust in the home. An interesting mechanical property of gekkonid lizard skin is worth brief mention. Bauer et al. (1989) investigated the skin of certain geckos. They found that the skin of Gekko gecko, behaves in ‘‘typical’’ vertebrate fashion but that of others, such as Ailuronyx seychellensis, exhibits unusual properties associated with identifiable morphological specializations. Under light and scanning electron microscopy they revealed that Ailuronyx dermis is functionally bilayered; the stratum compactum is divided into inner and outer layers by intervening loose connective tissue. The inner layer is strong and tough and does not differ significantly in its properties from that of the entire skin of G. gecko. However, its much thicker outer layer is only 1/20 as strong and 1/50 as tough as the inner layer and exhibits preformed zones of weakness. This peculiarity of the outer skin layer is used in nature as an antipredator escape mechanism by breaking away when seized. We have seen what may be regarded as typical scales in reptiles but scales may be modified for a variety of functions. In certain reptiles epidermal scales may overlie the bony scales of the dermis, e.g., the carapace (top) and plastron (bottom) of turtles’ shells, perhaps responsible for a certain confusion, for example, by a reporter a few years ago who suggested that these reptiles were unique in vertebrates in having their skeletons on the outside of the animal. The beak of turtles is composed of a modified epidermal scale covering the jawbone. Some of the most highly modified scales occur in certain species of gecko, in the family Geckonidae. In reptilian scales, layers of epidermis containing beta- keratin alternate with those containing alpha-keratin. A specific layer termed oberhautchen produces micro-ornamentation that interdigitates with those of the upper layer, termed clear layer, and forms the shedding complex. (Maderson 1965; Hiller 1970; Alibardi 1998; see Fig. 6.10). In these special scales, the 0.5–1.5 lm thick spinulae of the oberhautchen layer, that normally are 2–4 lm in length, grow into bristles or setae that can reach over 100 lm in length. Among the first workers to record, these highly modified scales of the footpads of certain species of gecko were Ruibal and Ernst (1965) and Stewart and Daniel (1972) (Fig. 6.13). Setae are distributed in arrays on approximately 20 leaf-like scansors of each. Each seta branches to form a nanoarray of hundreds of spatula structures that make intimate contact with the surface the gecko uses. Remarkably there are over a thousand species of gecko encompassing an impressive range of morphological variation to these structures. What is quite extraordinary is that these modified scales on the digital pads of geckos allow them to climb vertical surfaces and bear the entire weight of the animal suspended from a ceiling. It is achieved by forming a 144 6 Crucial Vertebrate Innovations

Fig. 6.13 Scanning electron micrographs of a scale from the plantar surface near the base of a hind toe of Gekko gecko. a Creased and indented protuberances bearing spinules of the Oberhautchen layer. (scale = 15 lm). b Magniﬁed view of a region near the center of (a). Arrow points to a line which may be a remnant of a boundary between presumptive Oberhautchen cells. (scale = 5 lm). After Ruibal and Ernst (1965)

‘‘frictional adhesive’’ on virtually all conceivable surfaces by producing an adhesion that enables either a tough bond or spontaneous detachment a highly complex biomechanical problem that workers such as Kellar Autumn and colleagues have been working on for many years (Autumn and Gravish 2008), showing ﬁne structure and biomechanics with important implications for technology (Fig. 6.14). Other workers such as (Alibardi et al. 2009; Alibardi and Toni 2006) have worked on the histological characteristics of setae by showing, for example, that the scales in gecko epidermis contain multiple forms of b-keratin (Toni et al. 2007). Setae apparently grow inside the cytoplasm of clear cells that form a cytoskeletal belt around the growing setae that probably molds their shape. At the beginning of setae formation the cytoplasm of both setae and clear cells is soft but progressively becomes corneous. Setae are mainly composed of beta- keratin of 12–18 kDa while the cytoskeleton of clear cells is made of other types of proteins, including cytokeratins. Partial primary sequence of some proteins of setae has shown that they share a common amino acid sequence with chick scale and feather keratin, which entrenches the basic ground plan keratin (keratin and collagen will be dealt with in greater detail in Volume 2). As mentioned above, the evolution of scales in reptiles marked the beginning of a chapter that would inﬂuence the development of all subsequent vertebrate life. Despite the fact that vertebrate skin appendages such as scales, feathers, hair, and teeth appear to be very different, studies on the Evo-Devo (evolution and development) of the amniote integuments show that they share a number of common developmental pathways (Sawyer and Knapp 2003; Alibardi and Toni 2006), such as the Hedgehog, BMP, and Wnt signaling pathways (Chang et al. 2009). 6.4 Scale Types 145

Fig. 6.14 Structural hierarchy of the gecko adhesive system. a Ventral view of a tokay gecko (Gekko gecko) climbing a vertical glass surface. b Ventral view of the foot of a tokay gecko, showing a mesoscale array of seta-bearing scansors (adhesive lamellae). c Microscale array of setae are arranged in a nearly grid-like pattern on the ventral surface of each scansor. In this scanning electron micrograph, each diamond-shaped structure is the branched end of a group of four setae clustered together in a tetrad. d Cryo-SEM image of a single gecko seta. Note individual keratin ﬁbrils comprising the setal shaft. e Nanoscale array of hundreds of spatular tips of a single gecko seta. f Synthetic spatula fabricated from polyimide at UC Berkeley in the laboratory of Ronald Fearing using nanomolding. From Autumn and Gravish (2008)

Developmental biologists, e.g., Ming Chuong (1998) have shown that variation and innovation in developmental processes may be a key mechanism of organ novelty. While the evolutionary origins and diversity of vertebrate integument appendages has long been of great interest a huge spur in the field has come from sensational finds of feathered dinosaurs in China (dealt with in greater detail in Volume 2). It has stimulated fresh and frequently different interpretations on the evolution of avian feathers by a number of workers such as Sawyer and Knapp (2003), Maderson (1972), Prum and Brush (2002), and Chuong and colleagues (2000, 2003) and toward heat conservation or endothermy (Wu et al. 2004). Despite the diverse nature of the integumental appendages among reptiles, birds and mammals, they share common developmental pathways. The neural crest and somatopleura cells interact with epithelium to form the skin and skin appendages. It has been proposed that hair originated by modification of scales (Alibardi 2003, 2004; Wu et al. 2004), however, there is no paleontologic evidence for intermediate forms. Evo-Devo on the other hand points to modulation and reorganization of gene regulatory networks that led to new forms of skin appendages with the evolution of feathers and hairs that possibly began from scales present in early reptiles (Chuong 1998). Perhaps most striking were findings by Eckhart et al. 146 6 Crucial Vertebrate Innovations

(2008) suggesting that last common ancestor of all extant amniotes contained cysteine-rich a-keratins, which served in the establishment of hard non-hair (scales?) epidermal structures and that these genes remained functional in sauropsids such as the green anole lizard, Anolis carolinenis, and were co-opted for a role in hair formation in mammals. The development of scales in Squamate reptiles begins with epidermal papillae, which are undulations of the epidermal surface producing symmetric dermo-epidermal elevations (Maderson 1965). The epidermis becomes undulated to form scale primordia due to differences in growth rate or mechanical forces between the epidermis and dermis. Four developing stages have been recognized by Alibardi (1996), including the flat bilayered epidermis stage, the symmetric scale anlagen stage, the asymmetric scale anlagen stage, and the b-keratinizing asymmetric scale stage. The asymmetric scale anlagen stage in the embryonic bearded dragon (Pogona vitticeps). Based on developmental anatomical data (Maderson and Alibardi 2000), the authors considered that a protofeather and its follicle are most easily derived from elongated reptilian scales. However, this is highly contentious topic in feather origins and will be considered in detail in Volume 2. Although we are familiar with the presence of reptilian or epidermal scales on the lower legs, feet, and spurs of birds; we are perhaps less so with their presence in the outer layer of the bills of birds. At the University of California at San Diego, materials scientists Seki et al. (2006) demonstrated the structure and biomechanics of the toucan, Ramphastos toco, bill. Because of the demands of flight, the beaks of birds have to be lightweight structures but also tough enough to deal with the frequently tough outer coverings of seeds, fruit, and invertebrates that may make up their diet. This is especially so in the toucan beak, which is about one-third the size of the entire bird whereas it masses it about one-twentieth. The bill possesses significant specific strength and thickness for food gathering, despite its extremely lightweight structure, and it is highly sensitive, being innervated by branches of the fifth cranial nerve. Seki and colleagues demonstrate how the strength and lightness is achieved. By SEM and schematic drawings, they show that the beak is formed of a sandwich composite with an exterior of tough keratin scales and a core composed of a lightweight fibrous network of closed cells made of collagen (Fig. 6.15). They follow this with mechanical tests that indicate that there is a synergistic effect between foam and the keratin scales—the foam stabilizes the deformation of the keratin shell by providing an internal support, which absorbs the stresses and increases its buckling load under compressive loading (Lingham-Soliar 2013; feather structure and mechanics will be discussed further in Volume 2).

6.5 Temporal Openings and Classiﬁcation of Reptiles

During the Mesozoic, reptiles underwent the most phenomenal radiation. As a consequence reptiles, i.e., the class Reptilia, are variously grouped in about 17-25 orders, only four of which survive to the present. Among the best and broadly- 6.5 Temporal Openings and Classiﬁcation of Reptiles 147

Fig. 6.15 Structure of Toco toucan (Ramphastos toco) beak. a Schematic cross-section. b Exterior of beak (keratin scales). c Interior of beak (foam). d, e Schematic drawings. d Keratin scale. e Closed cell of beak. From Seki et al. (2006) based clues to their classification are the number and positions of the temporal openings of the skull. The dorsal and lateral walls of the skull are covered by skull roofing bones, beneath which is the cranium which encloses the brain. These temporal openings in reptile skulls represent the space between the cranium and the bones of the temporal region. With the improvement in the jaw mechanism there followed an increase in the volume of jaw musculature. The only way in which to overcome the confinement of space to accommodate the jaw musculature was to open up the solid encasement at the sides of the skull to form temporal openings or to emarginate the roofing bones. The broad classifications of the reptiles based on the temporal openings seem to be universally accepted (Carroll 1988) (Fig. 6.16). 148 6 Crucial Vertebrate Innovations

Fig. 6.16 Phylogeny based on temporal openings among reptiles and their descendents. From Carroll (1988)

Turtles, living chelonians, are related to the anapsid stem reptiles, the Cotyl- osauria and like them have no temporal openings. However, some chelonians such as terrapins have the posterior and ventral edges of the temporal rooﬁng bones emarginated. They include the Captorhinida as a primitive stock. We will meet many of the other groups in later chapters. 6.5 Temporal Openings and Classiﬁcation of Reptiles 149

The ﬁrst group to diverge from the ancestral stock was the Synapsida. This group, also termed the mammal-like reptiles, includes the ancestor of mammals. They have a single pair of temporal openings situated low on the cheek and surrounded by the jugal, squamosal, and postorbital bones. Late in the Pennsylvanian a second major group diverged from the anapsids, the Diapsida. They are characterized by two pairs of temporal openings. One pair is located ventral to the postorbital, like that of the synapsids, and the second pair dorsal to the postorbital and squamosal and lateral to the parietal. The diapsids are so diverse as to contain two major subgroups, the lepidosauromorphs and the archosauromorphs, the latter giving rise to the major groups of amniotes, including the dinosaurs and pterosaurs. Lepidosauromorphs are represented by gigantic marine reptiles of the Late Cretaceous, the mosasaurs and by present-day sphen- odontids, lizards, and snakes. Two other major groups of marine reptiles, the ichthyosaurs and plesiosaurs (see Chap. 9 for all three groups) may have evolved from diapsids. Ichthyosaurs have an upper temporal opening like that of the diapsids but lack a clearly deﬁned lateral opening. The postorbital and squamosal form a wide cheek. The pattern has been termed parapsid or euryapsid and was thought to have arisen directly from an anapsid condition. The pattern in plesiosaurs may have been derived from that of early diapsids by elimination of the lower temporal bar and thickening of the postorbital and squamosal (For more detailed phylogenetic relationships the reader is referred to Estes and et al. (1988)).

References

Alexander G, Marais J (2007) A guide to the reptiles of Southern Africa. Struik Nat, Cape Town Alibardi L (1996) Scale morphogenesis during embryonic development in the lizard Anolis lineatopus. J Anat 188:713–725 Alibardi L (1998) Differentiation of the epidermis during scale formation in embryos of lizard. J Anat 192:173–186 Alibardi L (2003) Adaptation to the land: The skin of reptiles in comparison to that of amphibians and endotherm amniotes. J Exp Zoolog B Mol Dev Evol 2003; 298:12–41 Alibardi L (2004) Dermo-epidermal interactions in reptilian scales: speculations on the evolution of scales, feathers, and hairs. J Exp Zoolog B (Mol Dev Evol) 302:365–383 Alibardi L, Toni M (2006) Cytochemical, biochemical and molecular aspects of the process of keratinization in the epidermis of reptilian scales. Prog Histochem Cytochem 40:134–173 Alibardi L, Thompson MB (2001) Fine structure of the developing epidermis in the embryo of the American alligator (Alligator mississippiensis, Crocodlilia, Reptilia). J Anat 198:265–282 Alibardi L, Valle LD, Nardi A, Toni M (2009) Evolution of hard proteins in the sauropsid integument in relation to the corniﬁcation of skin derivatives in amniotes. J Anat 214:560–586 Autumn K, Gravish N (2008) Gecko adhesion: evolutionary nanotechnology. Phil Trans R Soc A366:1575–1590. doi: 10.1098/rsta.2007.2173 Bauer AM, Russell AP, Shadwick RE (1989) Mechanical properties and morphological correlates of fragile skin in gekkonid lizards. J Exp Biol 145:79–102 150 6 Crucial Vertebrate Innovations

Bragulla HH, Homberger DG (2009) Structure and functions of keratin proteins in simple, stratified, keratinized and cornified epithelia. J Anat 214:516–559. doi:10.1111/j.1469- 7580.2009.01066.x Carroll RL (1988) Vertebrate paleontology and evolution. WH Freeman, New York Chang C, Wu P, Baker RE, Maini PK, Alibardi L, Chuong C-M (2009) Reptile scale paradigm: Evo-Devo, pattern formation and regeneration. Int J Dev Biol 53:813–826. doi:10.1387/ ijdb.072556cc Chuong C-M (ed) (1998) Molecular basis of epithelial appendage morphogenesis. RG Landes Company, Austin Chuong C-M, Chodankar R, Widelitz RB, Jiang T-X (2000) EvoDevo of feathers and scales: building complex epithelial appendages. Curr Opin Genet Dev 10:449–456 Chuong C-M, Wu P, Zhang F-C, Xu X, Yu M, Widelitz RB, Jiang T-X, Hou L (2003) Adaptation to the sky: defining the feather with integument fossils from Mesozoic China and experimental evidence from molecular laboratories. J Exp Zool 298B:42–56 Deeming DC, Thompson MB (1991) Gas exchange across reptilian eggshells. In: Deeming DC, Ferguson MWJ (eds) Egg incubation: its effects on embyronic development in birds and reptiles. Cambridge University Press, Cambridge, pp 277–284 Deeming DC (2006) Ultrastructural and functional morphology of eggshells supports the idea that dinosaur eggs were incubated buried in a substrate. Palaeontology 49:171–185 DiMichele WA, Tabor, NJ, Chaney DS, Nelson WJ (2006) From wetlands to wetspots: environmenbtal tracking and the fate of Carboniferous elements In Permian tropical floras. In: Greb SF, Dimichele WA (eds) Wetlands through time. Special Paper 339, Geological Society of America, Boulder, Colorado, pp 223-248 Duellman WE, Trueb I (1986) Biology of amphibians. McGraw-Hill, New York Eckhart L, Valle LD, Jaeger K, Ballaun C, Szabo S et al (2008) Identification of reptilian genes encoding hair keratin-like proteins suggests a new scenario for the evolutionary origin of hair. Proc Natl Acad Sci USA 105(47):18419–18423. doi:10.1073/pnas.0805154105 Estes R, De Queiroz K, Gauthier J (1988) Phylogenetic relationships within Squamata. In: Estes R, Pregill G (eds) Phylogenetic Relationships of the Lizard Families. Stanford University Press, Stanford, pp 119–281 García RA (2007) An ‘Egg-Tooth’-like Structure in Titanosaurian Sauropod Embryos. J Vertebr Paleontol 27:247–252 Hildebrand M (1995) Analysis of vertebrate structure, 4th edn. Wiley, New York Hiller U (1970) Morpholgische Untersuchen der Haftborstenbildung und Hautung bei Tarentola mauritanica (Rept). Forma et Function 2:169–177 Hou L-H, Li P–P, Ksepka DT, Gao K-Q, Norell MA (2010) Implications of flexible-shelled eggs in a Cretaceous choristoderan reptile. Proc R Soc Lond 277:1235–1239 Lingham-Soliar T (2013) Feather structure, biomechanics and biomimetics: the incredible lightness of being. doi:10.1007/s10336-013-1038-0 Lingham-Soliar T, Bonser RHC, Wesley-Smith J (2010) Selective biodegradation of keratin matrix in feather rachis reveals classic bioengineering. Proc Roy Soc Lond B 277:1161–1168. doi:10.1098/rspb.2009.1980 Lingham-Soliar T, Murugan N (2013) A New Helical Crossed-Fibre Structure of b-Keratin in Flight Feathers and Its Biomechanical Implications. Plos One 8 (June, Issue 6): 1–12. e65849 Lu J, Unwin DM, Deeming DC, Jin X, Liu Y (2011) An Egg-Adult association, gender, and reproduction in Pterosaurs. Science 331:321–324 Maderson PFA (1965) Histological changes in the epidermis of snakes during the sloughing cycle. J Zool (Lond) 146:98–113 Maderson PFA (1970) Lizard hands and lizard glands: models for evolutionary study. Forma et functio 3:179–204 Maderson PFA (1972) On how an archosaurian scale might have given rise to an avian feather. Am Nat 176:424–428 Maderson PFA (1985) Some developmental problems of the reptilian integument. In: Maderson PFA, Gans C, Billett F (eds) Biology of the reptilia 14. Wiley, New York, pp 525–598 References 151

Maderson PFA, Alibardi L (2000) The development of the sauropod integument: a contribution to the problem of the origin and evolution of feathers. Amer Zool 40:513–529 Pough FH, Andrews RM, Cadle JE, Crump ML, Savitzky AH, Wells KD (2004) Herpetology. Prentice-Hall, New Jersey Pike DA, Andrews RM, Du W-G (2011) Eggshell morphology and gekkotan life-history evolution. Evol Ecol. doi:10.1007/s10682-011-9527-1 Prum RO, Brush AH (2002) The evolutionary origin and diversiﬁcation of feathers. Quart Rev Biol 77:261–295 Ruibal R, Ernst V (1965) The structure of the digital setae of lizards. J Morphol 117:271–294 Sawyer RH, Knapp LW (2003) Avian skin development and the evolutionary origin of feathers. J Exp Zool (Mol Dev Evol) 298B:57–72 Stewart GR, Daniel RS (1972) Scales of the Lizard Gekko gecko: surface structure examined with the scanning electron microscope source. Copeia 1972(2):252-257 Sexton OJ, Bramble JE, Heisler IL, Phillips CA, Cox DL et al (2007) Eggshell composition of squamate reptiles: relationship between eggshell permeability and amino Acid distribution. J Chem Ecol 31:2391–2401. doi:10.1007/s10886-005-7108-x Seki T, Kad B, Benson D, Meyers MA (2006) The toucan beak: structure and mechanical response. Mater Sci Eng C 26:1412–1420 Toni M, Valle LD, Alibardi L (2007) The epidermis of scales of gecko lizards contains multiple forms of beta-keratins including basic glycine-proline-serine-rich proteins. J Proteome Res 6:1792–1805 Wainwright SA, Biggs WD, Currey JD, Gosline JM (1976) Mechanical design in organisms. Edward Arnold, London Wu P, Hou L, Plikus M, Hughes M, Schenet J et al (2004) Evo-Devo of amniote integuments and appendages. Int J Dev Biol 48:249–270 Chapter 7 Dinosaur Integument

As we saw in Chap. 6, the epidermis in reptiles takes over the functional role occupied by the dermis in ﬁshes and their ancestors. This role is further examined in this chapter in the most exciting reptiles that lived on earth, the dinosaurs. Following that we look at the changing role of the dermis, which plays a somewhat different but vital role in the mechano-structural function of the skin as a whole and its impact on the lives of reptiles and mammals.

7.1 More than Skin Deep

Modern reptiles, class Reptilia, are striking in many ways making up an extraordinarily diverse and rich fauna (globally there are *8,000 species). In South Africa alone there are a total of 517 described species (151 snakes, 338 lizards, 27 tortoises, and 1 crocodile; They are represented by three of the four orders, Squamata (lizards and snakes), Crocodylia (crocodiles), and Testudines (tortoises, terrapins, and turtles) i.e., all but Rhynchocephalia, with a number of new species descriptions proceeding at an unprecedented rate (Alexander and Marais 2007). Yet, despite the amazing array of modern-day reptiles they are nevertheless a tiny expression of their glorious ancestors and none more so than the dinosaurs. For the height of the amazing conquest of the land by the reptiles we need to go back hundreds of millions of years to two great orders of reptiles, the archosaurs popularly known as the ‘‘ruling reptiles and the lepidosaurs (which includes modern day snakes and lizards) . Here, the emphasis is on one group of archosaurs, the dinosaurs but later (Chap. 9) among these two major groups, after success on land, some turned their backs on it to return to the oceans where they grew to gigantic size and rose to the top of the food chain. They were the plesiosaurs, ichthyosaurs, and mosasaurs. Even after almost a century since it ﬁrst appeared, a book on dinosaurs written by WD Matthew in 1915 still conveys the air of fascination and mystery that these remarkable animals continue to evoke. Matthew achieves this by his ﬁne writing style and perceptive observations that are enhanced by astute comparisons with

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 153 DOI: 10.1007/978-3-642-53748-6_7, Ó Springer-Verlag Berlin Heidelberg 2014 154 7 Dinosaur Integument modern day animals, the latter skill unfortunately a dying art today, as remarked by Jarvik (1980), much to the field’s great loss. Matthew’s detailed descriptions of dinosaur skin are no exception and part of his descriptions is included here with little change. Matthew’s section on the carnivorous dinosaurs Allosaurus, Tyrannosaurus, and Ornitholestes is a little more speculative than that on the herbivores, although quite reasonable. He acknowledges that there is no exact knowledge but suggests that the skin was probably either naked or covered with horny scales as in lizards and snakes and that it was almost certain that dinosaurs such as Allosaurus were not armor- plated as in the crocodile because the remains of any such armor could not fail to be preserved with the skeletons, as it always is in fossil crocodiles or turtles (Fig. 7.1). Rather he considered that the skin might have been scaly like the skin of lizards and snakes because such horny scales of the body are not preserved in fossil skeletons of these reptiles. However, if so he thought that because in the lizard the scales of the head are ossified and preserved in the fossil their absence in the carnivorous dinosaurs was perplexing. What is amazing is how deeply he thought about dinosaur skin with careful weighing of the pros and cons before arriving at any decision. For instance, when he says we can exclude feathers from consideration in these large theropods, it is actually astonishing that he actually considered the possibility of feathers being part of the dinosaur integument at a time when perceived wisdom was that these were reptiles and by default scaly. Although he concluded in the end against feathers this must be viewed in the light of the state of the data on the integument at the time. He surmised that Allosaurus probably had skin similar to that of the duck-billed dinosaur (hadrosaur) i.e., a curiously patterned mosaic of tiny polygonal plates (not overlapping) that was thin and quite flexible. He suggested, rightly or wrongly, that these dinosaurs have no affinities to birds and that there was no evidence for feathers in any dinosaur known to him. It would not be difficult to contemplate if he was alive today the excitement with which he would have greeted the exciting finds of feathered non-avian dinosaurs from Laioning China and the controversies surrounding them. One thing is certain, he would have left no stone unturned in whatever conclusion he arrived at. Matthew was undoubtedly most excited about the specimen of the duck-billed dinosaur in the American Museum of Natural History (Fig. 7.2). While frequently the skeletons were found articulated with parts of the skin not uncommonly preserved, in one of these specimen so much of the skin is preserved that he referred to it as a ‘‘dinosaur mummy,’’ (Fig. 7.3) a term more recently used to describe a specimen of a hadrosaur whose skin was extensively investigated by Philip Manning and colleagues, which will be discussed later. Matthew’s ‘‘dinosaur mummy’’ was among a number of fine-mounted skeletons of the species on display at the American Museum of Natural History. The excitement that Matthew felt was clear in his words, ‘‘We all believe that the Dinosaurs existed. But to realize it is not so easy. Even with the help of the mounted skeletons and restorations, they are somewhat unreal and shadowy beings in the minds of most of us. But this ‘‘dinosaur mummy’’ sprawling on his back and covered with shrunken skin—a real specimen, not restored in any part—brings home the reality of this ancient world 7.1 More than Skin Deep 155

Fig. 7.1 Mounted skeleton of Allosaurus in the American Museum of natural history (top). After Osborn. Restoration of Allosaurus by C.R. Knight (Bottom). From Matthew (1915) even as the mummy of an ancient Egyptian brings home to us the reality of the world of the Pharaohs.’’ Henry Fairﬁeld Osborn’s ﬁne description of the specimen is repeated here from Matthew’s book (1915). ‘‘The reason the Sternberg specimen (Trachodon annectens [a hadrosaur, probably Corythosaurus]) may be known as a dinosaur ‘mummy’ is that in all the parts of the animal which are preserved (i.e., all except the hind limbs and the tail), the epidermis is shrunken around the limbs, tightly drawn along the bony surfaces, and contracted like a great curtain below the chest area. This condition of the epidermis suggests the following theory of the deposition and preservation of this wonderful specimen, namely: that after dying a natural death the animal was not attacked or preyed upon by its enemies, and the body lay exposed to the sun 156 7 Dinosaur Integument

Fig. 7.2 Two mounted skeletons of Trachodon (hadrosaur) in the American Museum. Height of standing skeleton 16 feet, 10 inches. From Matthew (1915) entirely undisturbed for a long time, perhaps upon a broad sand flat of a stream in the low-water stage; the muscles and viscera thus became completely dehydrated, or desiccated by the action of the sun, the epidermis shrank around the limbs, was tightly drawn down along all the bony surfaces, and became hardened and leathery, on the abdominal surfaces the epidermis was certainly drawn within the body cavity, while it was thrown into creases and folds along the sides of the body owing to the shrinkage of the tissues within. At the termination of a possible low- water season during which these processes of desiccation took place, the ‘mummy’ may have been caught in a sudden flood, carried down the stream and rapidly buried in a bed of fine river sand intermingled with sufficient elements of clay to take a perfect cast or mold of all the epidermal markings before any of the epidermal tissues had time to soften under the solvent action of the water. In this way the markings were indicated with absolute distinctness,… the visitor will be 7.1 More than Skin Deep 157

Fig. 7.3 The dinosaur mummy. Skeleton of a Trachodon [hadrosaur] preserving the skin stretched over a large part of the body. After Matthew (1915). The skeleton lies on its left side. The head can be seen (bottom left) as well as the right forelimb (top left). After Osborn. From Matthew (1915) able by the use of the hand glass to study even the finer details of the pattern, although of course there is no trace either of the epidermis itself, which has entirely disappeared, or of the pigmentation or coloring, if such existed. Although attaining a height of fifteen to sixteen feet the trachodons were not covered with scales or a bony protecting armature, but with dermal tubercles of relatively small size, which varied in shape and arrangement in different species, and not improbably associated with this varied epidermal pattern there was a varied color pattern. The theory of a color pattern is based chiefly upon the fact that the larger tubercles concentrate and become more numerous on all those portions of the body exposed to the sun, that is, on the outer surfaces of the fore and hind limbs, and appear to increase also along the sides of the body and to be more concentrated on the back. On the less exposed areas, the underside of the body and the inner sides of the limbs, the smaller tubercles are more numerous, the larger tubercles being reduced to small irregularly arranged patches. From analogy with existing lizards and snakes we may suppose, therefore, that the trachodons presented a darker appearance when seen from the back and a lighter appearance when seen from the front. The thin character of the epidermis as revealed by this specimen favors also the theory that these animals spent a large part of their time in the water, which theory is strengthened by the fact that the diminutive fore limb terminates not in claws or hoofs, but in a broad extension of the skin, reaching beyond the fingers and forming a kind of paddle. The marginal web which connects all the fingers with each other, together with the fact that the lower side of the fore limb is as delicate in its epidermal structure as the upper, certainly tends to support the theory of the swimming rather than the walking or terrestrial function of this fore paddle as indicated in the accompanying preliminary restoration that was made by Charles R. Knight working under the writer’s direction. One is drawn in the conventional bipedal or standing posture while the other is in a quadrupedal pose or walking 158 7 Dinosaur Integument

Fig. 7.4 The Dinosaur mummy. Detail of skin of underside of body. After Osborn. From Matthew (1915)

position, sustaining or balancing the fore part of the body on a muddy surface with its fore feet. In the distant water a large number of animals are disporting themselves.’’ Henry Fairfield Osborn’s attention to detail with respect to the dehydration/ desiccation process following the death of the dinosaur and its contribution to preservation of the carcass prior to fossilization is a testament to an understanding of preservation processes as a significant part of an organisms taphonomic history well ahead of its time, a field of research that is becoming increasingly neglected in recent years (Lingham-Soliar and Glab 2010) of which more will be said in Volume 2. Also, what was particularly enlightening was Matthew’s (1915) comparison of dinosaur scales with those found on extant reptiles, e.g., Heloderma (Figs. 7.4 and 7.5). Stephan Czerkas of the Dinosaur Museum in Utah is one of the leading experts on dinosaur skin and has done more than anyone else to show that to understand dinosaurs we need much more than knowledge of their skeletons. As Czerkas (1997) says, skeletal remains can be used to reconstruct an approximation of a dinosaur’s body shape but only fossilized skin impressions can provide tangible evidence to show what the animal looked like in life. Despite seminal studies by early workers such as WD Matthews and Henry Fairfield Osborn (1912), who used appropriate modern day analogs to such good effect, later studies for many years 7.1 More than Skin Deep 159

Fig. 7.5 Skin impression from the tail of a Trachodon (hadrosaur). The impressions appear to have been left by horny scutes or scales, not overlapping like the scales on the body of most modern reptiles, but more like the scutes on the head of a lizard. From Matthew (1915)

depicted dinosaurs with scaleless, relatively smooth skin like those of cetaceans and elephants a point that Czerkas drew attention to (1997). Indeed as he notes the underlying effect from this further exaggerates other popular beliefs that if dinosaurs looked so mammalian why should they not be considered endothermic? Czerkas alerts us to one important misrepresentation that dates back to Sternberg’s (1909) ‘‘dinosaur mummy’’ of Corythosaurus in the American Museum of Natural History described above by Henry Fairﬁeld Osborn. It concerns the widely held view that they were aquatic and that the right hand was paddle-like in that all the digits were connected by an integumentary web. This seemed unequivocal when another specimen discovered by Charles Sternberg was found. This specimen now in the Senckenberg Museum in Germany in which the right hand was similarly preserved and more complete apparently showed an even more paddle-like web of skin extending over the bones and beyond, appeared to unequivocally support an aquatic mode of life. However, the more rounded left hand does not support such a view and it appears that the ﬂattened web-like appearance of the right hand and those of the AMNH specimen were distortions brought about by desiccation. It seems rather that the hand must have been used in supporting the body and that it was a highly functional participant in terrestrial locomotion (Czerkas 1997). Stephan Czerkas ‘documentation of dinosaurs with skin largely involves the large herbivorous dinosaurs the sauropods, and in a few instances large theropods, e.g., Ceratosaurus, Carnotaurus, and a tyrannosaurid. However, as he notes skin impressions from small nonvolant theropods, or for that matter small ornithopods are unknown nor were there skin impressions that preserve the actual color of any dinosaur. The only ornithopod with preserved skin impressions mentioned by Czerkas (1997) is from the ankle of a Psittacosaurus. Despite more than 15 years since Czerkas made these points they by and large still hold true today. Never- theless, the few specimens that do exist are producing interesting information concerning the scales on smaller ornithopods and theropods and even color. For instance, with respect to Psittacosaurus fortunately more information is available now (Lingham-Soliar and Plodowski 2010). 160 7 Dinosaur Integument

Fig. 7.6 Psittacosaurus SMF R 4970 at the Senckenberg Museum, Germany. Top, Psittaco- saurus skeleton as preserved. Scale bar = 10 cm. After Lingham-Soliar and Plodowski (2010). Bottom, author’s life reconstruction of Psittacosaurus in the Jehol biota, Laioning, China. A predatory Sinosauropteryx lurks on the right

In 2002 Gerald Mayr and colleagues described interesting bristle-like filaments in the tail region of a small psittacosaurid dinosaur, Psittacosaurus. Psittacosaurus is, a small ceratopsian, herbivorous dinosaur about 1.5–2 m long, which may occasionally have adopted a bipedal stance (as frequently depicted in reconstructions) (Fig. 7.6). The bristle-like structures will be discussed later in the section on the highly charged controversies connected with the hypothesis of birds originating from dinosaurs (Volume 2). Here we concentrate on the scales. The Senckenberg Psittacosaurus came from the world famous Yixian Formation in Liaoning, China. It has had a fairly chequered, somewhat shady history before its rescue and purchase by the Forschungsinstitut Senckenberg in Germany, where it is now housed as specimen SMF R4970. Among the people connected with the specimen’s rescue and fine preparation and with providing an understanding of its lithology is Professor Gerhard Plodowski of the Institute. Because of the paucity of 7.1 More than Skin Deep 161 information on the skin of small dinosaurs the descriptions will be included here in some detail from the study by Lingham-Soliar and Plodowski (2010). The head and thoracic regions of the specimen of Psittacosaurus (up to the posterior limbs) are compressed dorsoventrally with only a trace of the latero- thorax visible. Only in the posterior part of the body (lumbar region) and in the tail are notable lateral and ventrolateral regions of the body visible. Of particular note, the epidermal preservations in Psittacosaurus represent actual scales in contrast to most cases of skin preservation in dinosaurs that are of a mold of the scales. The scales can be seen over most parts of the body including in a thin strip along the throat, left and right shoulders, right dorsal surface, dorsoventral caudal regions, and areas of the left anterior and posterior limbs. Three types of scales are preserved in Psittacosaurus: (1) large, round plate-like scales, (2) polygonal scales or tubercles, and (3) small-rounded pebble-like scales. The best preservation of the plate-like scales is in the shoulder region (Figs. 7.7 and 7.8). They vary in size, the largest approximately 10 mm in diameter, decreasing in size ventrad along the arm to approximately 4 mm diameter. Perhaps most interesting, we were able to detect, impressions of pigmentations over most of the skeletal elements among the most vivid were dark pigment impressions of three plate-like scales (black), surrounded by amber/brown impressions of tubercles, overlie the base of the left scapula (Fig. 7.7b) which will be discussed in greater detail below. The tubercles are numerous, small, and polygonal-shaped. They vary in size from approximately 0.75–1.5 mm in diameter and also vary in tone from light to dark brown. Well-preserved tubercles and plate-like scales occur in both shoulder regions (Figs. 7.7a and 7.8a). One of the most remarkable features connected with the scales in Psittacosaurus was the pigmented impressions left by the scales, which were found imprinted over virtually all the skeletal elements of the specimen. This was the first record of color associated with scales in a dinosaur. They give a very vivid impression of color and patterns and among the most informative were those at the base of the left scapula, probably comprising articular cartilage. It shows a well-defined ring-like pattern of amber tubercles surrounding the dark plate-like scales (Fig. 7.7b, white arrows). The ring has possibly survived better over a stiff surface such as bone or cartilage given that the skin as it contracts (as over a drum) is probably less likely to crease and distort as it might over muscle. The color too is generally brighter over a smooth surface, e.g., bone or cartilage—a brighter amber/brown and deeper black, presumably because the pigment has less chance to permeate bone or cartilage and dissipate as over soft tissue, analogous to the way color is brighter when printed on glossy rather than matt or absorbent paper. Dark tubercles are preserved at the boundary of the upper proximal part of the tail along the vertical midline and in the upper lumbar region. Small, light, pebble-like scales occur in a small section of skin in the region of the throat and in the ventrolateral regions of the posterior body and tail. These are assumed to have sparse pigmentation in life (Lingham-Soliar and Plodowski 2010). Of interest are the color patterns in Psittacosaurus, most notably in the right and left shoulder regions. The dark plate-like scales regularly dispersed among the amber brown tubercles would almost certainly have created an effective cryptic 162 7 Dinosaur Integument

Fig. 7.7 Preserved scales in Psittacosaurus SMF R 4970. a Left shoulder, details shows cryptic patterns created by light and darker brown tubercles, many in relief. b Dark plate-like scales surrounded by rings of amber/brown tubercles (white arrows) preserved over the base of left scapula (arrowheads). After Lingham-Soliar and Plodowski (2010)

pattern on the body of Psittacosaurus. Color from the scales has in places impressed itself onto the bone as in the humerus (Fig. 7.8e). Conceivably, differences in color tones of the tubercles themselves may even have contributed to the cryptic patterns (Figs. 7.7 and 7.8)asinAgama (Fig. 7.8f). Another form of crypsis, common in many extant animals, is also evident in Psittacosaurus— countershading, i.e., a dark dorsum and light ventrum (Fig. 7.6, bottom). 7.1 More than Skin Deep 163

Fig. 7.8 Left shoulder of Psittacosaurus SMF R 4970. a Tuberculate and large plate-like scales or grouping of smaller scales (arrow). b Light and dark brown tubercle imprints on bone; c Polygonal tubercles in relief showing traces of dark pigment; d Amber/brown and dark tubercle pigment imprints on the upper part of left humerus (lower humerus in view); e Tubercle pigment imprints (only) on left coracoid; f Skin of A. atricollis. Scale bar. a, e = 10 mm; b, c, d = 2 mm; f = 5 mm. After Lingham-Soliar and Plodowski (2010)

A distinctive ring of evenly sized tubercles surrounding the plate-like scales (Fig. 7.7b, white arrows (Lingham-Soliar and Plodowski 2010) is also evident in another ceratopsian dinosaur, Chasmosaurus. Also, comparable with Chasmo- saurus is a decrease in size of the large plate-like scales from the dorsal surface of the body ventrad. In the ceratopsian, Centrosaurus the skin tubercles are beautifully preserved but unlike in Chasmosaurus and Psittacosaurus, they are generally consistent in size, and the plate-like scales are disposed in rows (Lull 1933). Preservation of color in fossils such as Psittacosaurus is rare and largely restricted to invertebrates Subsequent reported expressions of colors in non-avian dinosaur and fossil birds concerns feathers (see Volume 2). In Psittacosaurus the patterns produced by the different hues, even if not the original colors, are so distinctive that there seems little doubt that they are genuine and responsible for some kind of cryptic patterns that occurred over the body. It appeared to involve two dominant colors that were tentatively identified, black and amber/brown. The most striking crypsis is a consequence of the large dark plate-like scales set amidst smaller, lighter, tubercles, most noticeable in the shoulder region of Psittacosaurus (Figs. 7.7a and 7.8a). The pattern created by the tubercles and plate-like scales are reminiscent of that in the lizard Agama atricollis), i.e., a precise pattern of large dark spiny scales surrounded by smaller and lighter tubercles. Bleaching and 164 7 Dinosaur Integument weathering may have affected the intensity of the color, but a combination of data from color overlying bone and soft tissue allows a more reliable assessment. Next, we come to the recently contentious question of scales in theropod dinosaurs of which we knew very little, until now. In 2006 Ursula Gohlich and Luis Chiappe described an exquisitely preserved small Late Jurassic theropod dinosaur that they named Juravenator starki. The importance of this little dinosaur and others in the group known as coelurosaurs is that from among them many paleontologists believe are species that gave rise to birds. This rare fossil came from Schamhaupten and is only the second non-avian theropod found in the laminated limestones of the Late Jurassic Solnhofen reef archipelago of Bavaria, after the discovery of the celebrated Compsognathus nearly 150 years ago. It represents the best-preserved predatory, non-avian dinosaur in Europe. Soft tissue is preserved along the tibiae, and particularly between the 8th and the 22nd caudal vertebrae, where it defines the outline of the tail. The latter region allows observation of the skin surface and other soft parts (Fig. 7.9). The integument of Juravenator is formed of uniformly sized, smooth tubercles (about 15 tubercles per 25 mm of preserved tissue) similar in appearance to the small, conical, and non-imbricated tubercles of other non-avian dinosaurs. More sensationally, an array of feathers has been reported among non-avian coelurosaurs. However, the absence of either feathers or skin follicles associated with the preserved integument of Juravenator indicates, as the authors state, that at least the central portion of the tail of this coelurosaur was devoid of feathers. They also add that the discovery of Juravenator sheds light on a poorly known segment of coelurosaur history and that it indicates that the evolution of feathers in theropods might have been more complex than previously envisaged. When the theropod dinosaur Sinosauropteryx was unveiled to the public, fuzz along the dorsal surface was interpreted as feathers. The euphoria extended to Sinosauropteryx being the only dinosaur to make the front page of The New York Times. Soon after it was formerly proposed in the journal Nature that integumental structures preserved along the dorsal surface of Sinosauropteryx were the remains of feather precursors or protofeathers (Chen et al. 1998), which was reiterated in a follow-up paper (Currie and Chen 2001). However, the hypothesis of protofeathers was subsequently challenged by Lingham-Soliar et al. (2007) who proposed an alternative explanation, namely, that the integumental structures were degraded collagen fibers. The authors based the new hypothesis on data from a previously undescribed specimen of Sinosauropteryx, IVPP V12415 (Fig. 7.10), as well as from the holotype and referred specimen and, significantly, on taphonomic experiments showing a striking resemblance between the alleged protofeathers with decaying collagen fibers in a decomposing dolphin (Lingham-Soliar 2003a). However, more recently Zhang and colleagues (2010) dropped a potential bombshell when they claimed they had irrefutable evidence that the filaments in Sinosauropteryx contained melanosomes, color organelles known to be present solely in bird feathers. This was subsequently challenged and demonstrated that the claim had little scientific merit (Lingham-Soliar 2011; this will be discussed further in Volume 2). Although increasing evidence showing that the filaments 7.1 More than Skin Deep 165

Fig. 7.9 Integument of Juravenator starki. a Specimen photographed under ultraviolet light. b Specimen photographed under normal light. Abbreviations: c9, c11, and c13, caudal vertebrae 9, 11, and 13. Courtesy of Zhang et al. (2010). Copyright Macmillan Publishers, Ltd discovered in Sinosauropteryx are collagen structural fibers supporting an external dorsal frill, the original idea that the filaments are protofeathers is still clung to, if a little less tenaciously, by many workers in the story of feather origins. In Sinosauropteryx IVPP V12415, most of the tail was displaced upward as a consequence of postmortem opisthotonus (strong recurvature of the head and neck backward and the tail forward, over the back). In the terminal part of the tail a small patch of tissue is preserved on its surface. It comprises 4–5 circular or scale- like structures which, importantly, overlie the filamentous structures (see Lingham-Soliar 2013), considered by many workers to be protofeathers (Fig. 7.11). Given that they underlie the scale-like structures of the epidermis, this is highly improbable. The surfaces of the scale-like structures are clearly papulose (see circle inset). The papilla radiate from a central point, a feature reminiscent of papilla observed in extant reptile scales (Matthew 1915; Lingham-Soliar 2013, Fig. 7.5). Between the displaced tail segments a larger patch of similar epidermal structures is preserved. These too, show the distinctive papulose surface. Pro- truding from beneath the patch of scales are fibers that presumably acted to anchor it to the dermis (Fig. 7.11, small arrows, and rectangular inset). As with Jurave- nator, the overwhelming evidence is that these basal theropods were covered in scales i.e., the skin of Sinosauropteryx was typically reptilian and not covered in feathers (Fig. 7.12). 166 7 Dinosaur Integument

Fig. 7.10 Sinosauropteryx IVPP V12415. The specimen shows the displacement of parts of the tail from the main caudal region (during postmortem opisthotonus). Near the tail base the chevrons become increasingly separated from the tail as its curvature increases. Near the distal end the separation of the terminal part of the tails leaves traces of scales between the displaced sectors (detail in rectangle in Fig. 7.11). After Lingham-Soliar (2013) 7.2 Organic Analysis of the Dinosaur Integument 167

Fig. 7.11 Sinosauropteryx IVPP V12415. Displacement of caudal vertebrae (approx. no. 39) during postmortem recurvature. Inset 1 shows reconstruction and perfect match of the displaced tail sections even with respect to a slight kink in one of the structural fibers (interpreted by others as protofeathers) of the dorsal crest (arrow). Circle shows patch of approx. four scales, showing a papulose surface with papilla radiating around a central point. Inset 2 shows scales with central papilla and protruding from under the scales are attachment fibers that connect them to the dermis. In the bottom left of the main picture can be seen fibers from the opposite side of the tail comprising the dermis. After Lingham-Soliar (2013)

7.2 Organic Analysis of the Dinosaur Integument

Skin impressions observed in dinosaur ‘mummies’ such as the hadrosaurs at the AMNH (above) and Senckenberg Museum are interpreted as trace fossils and are rare. However, the mummiﬁed appearance of another hadrosaur dinosaur, Edmontosaurus sp. (MRF-03), from the Upper Cretaceous, Hell Creek Formation, North Dakota is also characterized by large areas of uncollapsed skin forming an 168 7 Dinosaur Integument

Fig. 7.12 THe author’s life restoration of Sinosauropteryx in the Jehol of Laioning, China. In the background is an active volcano, known to be part of the environment. After Lingham-Soliar (2012)

‘envelope’ preserved through early mineralization. The preservation includes more than just the usual mineralized tissue but traces of organic material. It is believed that Edmontosaurus was preserved through rapid burial and by numerous factors that came together at the same time. Sediment enclosing the specimen was found to be mainly composed of fine sand-sized grains of quartz, feldspars, and rock fragments with some higher plant-derived organic material. This was closely associated with the pore waters which enabled a complex series of chemical changes and the formation of a number of chemicals and chemical compounds that ultimately were available to replace the dinosaur soft tissue with carbonate minerals. Crucially, supply of iron to the rotting dinosaur was maintained by iron-rich ground waters 7.2 Organic Analysis of the Dinosaur Integument 169 that flowed through the higher porosity sand (Manning et al. 2009; see Volume 2 for more extensive discussions on the taphonomy and preservation of fossils). The intricate processes involved in soft tissue preservation have invariably meant that many of the clues to the fossilization process have invariably been locked in the fossilized material, which because of its irreplaceable nature does not avail itself to usual destructive chemical investigations. However, more recently leading workers in the field (Manning et al. 2009; Edwards et al. 2011) have developed methods involving the use of the latest technology and analytical techniques that includes small angle X-ray scattering, Synchrotron Rapid Scanning X-ray Fluorescence (SRS-XRF Fourier transform infrared spectroscopy (FTIR), gas chromatography mass spectrometry (Py-GCMS), and amino acid analyses, which, importantly is either nondestructive to the fossil material or uses minute samples for analysis. Using such methods, Manning et al. (2009) showed that the amino acid composition of the mineralized skin envelope of Edmontosaurus MRF- 03 clearly differed from the surrounding matrix. Although intact proteins could not be obtained they were able to show that microstructures were strongly reminiscent of soft tissue. Among factors they proposed for such remarkably high fidelity preservation of the integument were as mentioned, rapid burial, a reduced oxygen environment and rapid replacement of the soft tissue with carbonate minerals—the latter outpacing microbial decay. It seems that it also ensured that some breakdown products of organic molecules at the point of burial, whether endogenous to MRF- 03 or of microbial origin, were preserved. Edwards et al. (2011) used nondestructive Fourier Transform Infrared (FTIR) spectroscopy in their work on the mapping of ancient fossil organic material without damage to the fossil specimens. Their results in a study of an Eocene specimen (50 MYR old) of fossilized reptile skin demonstrated that not only was it possible but the material was not a simple impression, mineralized replacement, or an amorphous organic carbon film, but that it contained a partial remnant of the living organism’s original chemistry, in this case derived from the animal’s pro- teinaceous skin. The work is ongoing and promises information from trace fossil material that was previously thought impossible. The work of Manning and colleagues involves some of the best, sincere and effective use of the latest technology rather than simply technology for technology’s sake or to put it another way the mistaken belief that there is as much science in a subject as there is technology in it.

7.3 Dinosaur Eggs and Embryos

As we saw in Chap. 6, a momentous innovation in reptiles was the evolution of the stiff-shelled amniotic egg, the embryo’s own barrier against the environment, which would irrevocably mark the conquest of land. Here, we will look at how in recent years dinosaur eggs are providing an increasing source of information on the biology and lifestyles of these fascinating reptiles. 170 7 Dinosaur Integument

Fig. 7.13 Sauropod trace fossils. a Transmitted light microscopy image of skin fragment from sauropod embryo. Intricate scale pattern is clearly visible. b Scanning electron micrograph (SEM) of sauropod shell in tangential section, showing external tubercles (T), shell units membrane (SM), and organic cores (OC). Scale bar, 200 mm. After Schweitzer et al. (2005)

Schweitzer et al. (2005) investigated the possibility of molecular preservation in Late Cretaceous titanosaurid sauropod dinosaur eggshells. The quality of morphological preservation of the fossil shells was demonstrated using transmitted light and scanning electron microscopy. The external shell morphology revealed closely spaced knobby projections and (Fig. 7.13) pores for gas exchange similar to modern avian eggshells but differing from the flexible shell of lepidosaurians (Sexton et al. 2005). Electron microscopic examination revealed organic cores immediately internal to the mineralized shell. Based upon location, structure, and comparison with extant taxa, the authors hypothesized that this material might represent remnants of preserved organic molecules, permineralized by calcite precipitation during fossilization. Significantly, Schweitzer and colleagues tested the eggshell using immunological analyses. While acknowledging skepticism with respect to the ability for organic molecules to preserve over such vast periods of time as well as problems with respect to contamination, the authors suggest that their results are viable and that they demonstrate organic compounds and antigenic structures similar to those found in extant eggshells. If this is so it would be interesting to see whether or not future studies confirm the amino acid content between dinosaur and extant avian eggs and even to establish possible differences between rigid dinosaurian and flexible lepidosaurian eggs. Studies on the structures of dinosaur eggs and comparisons with those of modern reptiles and birds have extended to equally fascinating research on embryos found within dinosaur eggs. For instance discovery of a remarkable nesting site near Auca Mahuevo in the Patagonian badlands of Argentina was strewn with thousands of eggs, dozens of which still containing dinosaur embryos inside (Chiappe et al. 1998). The nesting site dates from the late Cretaceous and is approximately 70-90 million years old. In addition to tiny embryonic bones, 7.3 Dinosaur Eggs and Embryos 171

Fig. 7.14 Bone morphology of the sauropod embryos from Auca Mahuevo. a Cluster of teeth of specimen PVPH-112. b Left postorbital in dorsolateral view (PVPH-112). c Right orbital region in lateral view (PVPH-112). a to b and c show maximum width and length, respectively. d Natural cast of the integument of the sauropod embryos from Auca Mahuevo. d Large patch of skin (PVPH-126), showing a triple row of large scales. Abbreviations. La, lacrimal; sc, sclerotic plates. PVPH, Paleontologia de Vertebrados, Museo Municipal ‘‘Carmen Funes,’’ Plaza Huincul. Reprinted after Chiappe et al. (1998). Courtesy of Macmillan Publishers, Ltd many of the eggs contain patches of delicate fossilized skin, providing the first glimpse of the soft tissue covering embryonic dinosaurs (Fig. 7.14). Petrographic analysis of the Auca Mahuevo specimens showed the presence of clay, quartz, and a carbonate mineral, indicating that the preserved skin is the cast of the original integument. The general skin pattern consists of round, nonoverlapping, tubercle- like scales of subequal size (300 lm in diameter). One of the fossils has a distinct stripe of larger scales near its upper surface, which probably extended along the animal’s back (Fig. 7.14d). Scale patterns on the embryonic skin of the Auca Mahuevo specimens are reminiscent of the mosaic pattern of small osteoderms associated with the skeleton of the titanosaur Saltasaurus loricatus, also from the Late Cretaceous of Argentina. The authors were unable to identify with certainty the sauropod species that had laid the eggs at Auca Maheuvo, but the discovery of tiny teeth in the eggs provides 172 7 Dinosaur Integument an intriguing clue (Fig. 7.14a). One embryo alone has at least 32 individual pencil- shaped teeth, each small enough to fit easily into the capital ‘‘O’’ at the beginning of this sentence. The only sauropod dinosaurs alive at the end of the Cretaceous period with teeth this shape were sauropods known as titanosaurs. The remains of these dinosaurs are common near Auca Maheuvo, making it very likely that the embryos belong to this group. Unlike any other known sauropods, titanosaurs had bony, armored plates embedded in their skin. The embryo’s skin, however, does not show any signs of armored plates, indicating that these grew only after the dinosaurs had hatched. This growth pattern mirrors that seen in modern armored lizards and crocodiles, the juveniles of which lack the bony patches in the skin that are present in adults (Chiappe et al. 1998). Descriptions of the embryonic skin of the Auca Maheuvo fossils (Coria and Chiappe 2004) are among the most remarkable of dinosaur scales regardless of their embryonic status. The descriptions are therefore reproduced here virtually unchanged. The findings showed six different patterns of integument, recognized on the basis of the shape, size, and arrangement of individual tubercles:

1. Ground tubercles: This is the prevailing pattern in all specimens. These tubercles are subequal, pebble-like, smooth, and approximately 300 lmin diameter. They are densely arranged (12–14 tubercles/mm2) and uniformly distributed. 2. Large, elongated tubercles: Large tubercles, almost 3 mm long, surrounded by ground tubercles. In the only known example (MCF-PVPH-135), more than 20 ground tubercles surround a large tubercle. The surface of the latter is somewhat inﬂated, with at least four slightly elevated areas. This condition, and the presence of ground tubercles that partially coalesce with the edges of the large tubercle, suggests that the latter consists of at least four-fused tubercles, which are larger than a typical ground tubercle. 3. Parallel rows of large tubercles: They form stripe-like arrangements formed by a series of parallel rows of tubercles, with the largest tubercles located in the central row. Some variation notwithstanding this pattern is essentially symmetrical along its main axis. Central tubercles are roughly polygonal (pentagonal or hexagonal); their main axes are perpendicular to the row and the contact with the adjacent tubercles is along their sides. The adjacent rows of tubercles on each side of the central row consist of tubercles that are approximately half the size of the central ones. Each of these smaller tubercles is wedged between every two large central tubercles. In MCF-PVPH-126, the central row contains tubercles that are approximately 800 lm across. On each side of this row, there is a row of tubercles approximately 400 lm in diameter. In CFPVPH-686, one row of large tubercles is also ﬂanked by parallel rows of smaller tubercles. 4. Rosette-like arrangement: A common pattern consisting of a large, oval, or slightly polygonal (pentagonal to hexagonal) tubercle surrounded by smaller tubercles. These peripheral tubercles are similar to, but slightly larger than, ground tubercles. The number of peripheral tubercles varies from specimen to specimen. In MCF-PVPH-130, the central tubercle (800 lm) is encircled by 11 7.3 Dinosaur Eggs and Embryos 173

tubercles. This central element is surrounded by seven tubercles in MCF- PVPH-140 and MCF-PVPH-682, and 13 or 14 in MCFPVPH-135. In addition, the large tubercles of two rosettes in MCF-PVPH-680 are encircled by 10 and 13 smaller tubercles. 5. Flower-like arrangement: Seven to eight tear-shaped tubercles converging on a minute, central tubercle. While the central tubercle is much smaller than the average ground tubercle, the converging tear-shaped tubercles are 1.5–2 times the size of typical ground tubercles. In MCF-PVPH-126, at least four of these structures are observed associated with rosette-like pattern. Other specimens with ﬂower-like arrangements include MCF-PVPH-147, 680, 681, and 686. 6. Striate-like rows: Parallel rows of minute, elongated tubercles arranged in a tight formation. This pattern gives the skin a conspicuous striated appearance. These tubercles are the smallest observed among all available specimens. Rows are separated by relatively deep and constricted grooves. The size of the tubercles may vary along the same row and sometimes tubercles appear to be fused to each other, forming long multitubercle structures. Over 20 parallel rows of these tubercles are preserved in MCF-PVPH-142.

In contrast to the studies above involving among the latest sauropods known, another study on dinosaur eggs takes us back to sauropod origins. It involves a remarkable study on fossil eggs that had lain in a museum repository for over 30 years. Robert Reisz and colleagues (2010) describe in one of the eggs a skillfully prepared embryonic skeleton of the sauropodomorph dinosaur Massospondylus from the Lower Jurassic of South Africa i.e., about 190 million years old, predating the sauropod eggs described by Chiappe et al. (1998) by about 100 million years. Massospondylus belongs to a group of dinosaurs known as prosauropods, widely believed to be the ancestor of sauropod dinosaurs. They grew to large size, were four-legged dinosaurs, and possessed long necks. Thus the eggs contain the oldest known embryos of any dinosaur known or for that matter of any land-dwelling vertebrate. The preservation of the embryos is exquisite, permitting a complete reconstruction of the skeleton and detailed interpretations of the anatomy (Fig. 7.15). The level of ossiﬁcation—how much of the skeleton has turned to bone—reveals that the embryos were close to hatching. The fossils also reveal that the future hatchlings would have been oddly proportioned and would have looked very different from the adults of the species. The 20 cm (8 inch) embryos were quadrupedal (walked on all four legs), with relatively long front limbs and dis- proportionately large heads. In contrast, the 5 m (16.5 foot) long adults had relatively tiny heads and long necks; they mostly likely were bipedal, given that their forelimbs are much shorter than their hind limbs. This implies that as the dinosaurs matured, their necks and hind limbs grew much faster than their forelimbs and head. Schweitzer et al. (2005) note that the preservation of embryonic soft tissues is highly signiﬁcant because it indicates unique depositional and geochemical conditions that resulted in rapid and complete mineralization of these very labile soft tissues before degradation could occur. These taphonomic conditions extend to the eggshell, and as a result, both eggs and their contents represent a unique 174 7 Dinosaur Integument

Fig. 7.15 Massospondylus embryo. a Drawing of the embryo skeleton. b Photograph of skeleton of the ﬁrst embryo of Massospondylus. White materials represent eggshell that remained around the embryonic skeleton after they had been exposed. Reprinted from Reisz et al. (2010). Courtesy AAAS opportunity to elucidate taphonomic conditions resulting in exceptional preservation. These eggs containing such fragile and labile elements as embryonic bone and fragments of embryonic skin attest to unusual taphonomic and diagenetic conditions and provide an opportunity to expand the correlation between unusual morphological preservation and the presence of endogenous molecules. 7.3 Dinosaur Eggs and Embryos 175

Fig. 7.16 Eggshell microstructure of the Phu Phok eggs. a SEM of the nodular surﬁcial ornamentation. Note the bimodal distribution of the height of the nodes. This type of ornamentation has previously been reported in elongatoolithid eggs, a parataxonomic family that has been traditionally associated with non-avian theropods. b Thin section of eggshell radial section. As for the SEM picture, note the presence of the three structural eggshell layers and also of a black diagenetic line (dl) here interpreted as bacterially mediated micrite. After Buffetaut et al. (2005)

Minute theropod eggs and embryos from the Lower Cretaceous of Thailand (Buffetaut et al. 2005) show many important features that the authors propose were highly important in the dinosaur-bird transition. The shell thickness, without the surficial ornamentation, averages 354 lm, a value very similar to that of the much more voluminous Gallus gallus (domestic chicken) eggs and 30 % more than that of an ornithothoracine bird from the Late Cretaceous of Argentina (Schweitzer et al. 2002). The nodular surface ornamentation compares with that of many non- avian dinosaur eggs (Mikhailov 1997; Grellet-Tinner et al. 2004). The taller nodes average 183 lm, while the smaller and more numerous ones are twice smaller (92 m). Pore-like fissures at the level of the taller nodes are likely the result of compression during a late diagenetic phase rather than being of biological origin. The authors consider that diagenetic replacement at the ultrastructural level did not alter the eggshell microstructure or the authenticity of the three prismatic structural layers, a synapomorphy of Mesozoic Ornithothoraces and modern birds, as opposed to the two-layered condition seen in non-avian theropod eggs (Grellet- Tinner and Chiappe 2004) (Fig. 7.16). However, it is now believed (Eric Buffetaut, personal communication 2013) that the eggs (Buffetaut et al. 2005) belong to a lizard and not a theropod dinosaur thus changing the interpretations on eggshell structure with respect to birds and non-avian dinosaurs. It does not, however, diminish the contribution of the fossil eggs and embryos to our understanding of developmental stages in reptiles. An important question in dinosaur behavior concerns egg-laying and incubation, which many authors have proposed show close correlations with extant birds and will be considered below. Charles Deeming, one of the world’s leading experts on egg-structure and related studies investigated the ultrastructural characteristics of dinosaur eggshells in order to calculate water vapor content so as to assess their 176 7 Dinosaur Integument conductance and thereby obtain information on the dinosaur’s nesting environment (Deeming and Thompson 1991; Deeming 2006). A variety of shell parameters were used against predicted egg mass and comparisons made with allometric equations for bird eggs. He found that shell thickness was generally larger than seen for extant birds. Deeming’s findings were very interesting in his novel approach to understanding one of the fundamental questions of dinosaur ‘‘nesting’’ behavior. He found that by calculating the water vapor conductance of the eggshell there was an allometric increase with egg mass, parallel to the bird values, but for any given egg mass values for dinosaurs were an order of magnitude higher. Mass-specific water vapor conductance was unaffected by egg mass but was an order of magnitude higher than the bird values. Water vapor conductance per pore showed an allometric decrease with egg mass but again the predicted values were an order of magnitude higher than for bird eggs. The ultrastructural characteristics of dinosaur eggshells indicated that the nesting environment had to be saturated with water vapor. The high-conductance eggshell precludes incubation in an environment that is much less than 100 % saturated with water vapor. It is simply not possible for dinosaur eggs to lie partially exposed to the elements. Physical evidence of water vapor conductance shows that dehydration of the egg contents would be rapid. Even under humidity conditions of over 99 %. A dinosaur egg only half buried in a substrate would lose water rapidly to the air. Such physical conditions mean that most of the eggs considered by Deeming, including those of hadrosaurs, titanosaurs, troodontids, and oviraptorids and the Jurassic allosauroid theropod Lourinhanosaurus antunesi could not have been left exposed or even incubated by an adult dinosaur. Deeming also gave thought to the brooding-type incubation (with contact between the eggs and adults), akin to that of birds, recently attributed to Oviraptor philoceratops (Dong and Currie 1996) and Troodon formosus (Horner 2000; Varricchio et al. 1997; see Deeming 2006 for references). However, Deeming indicates that the theropod eggs including the elongatoolithid eggs studied, which are identical to those known to belong to oviraptorids, have an eggshell porosity over five times higher than the shell of bird eggs of an equivalent mass. Nevertheless, Deeming tries to reconcile this apparent conflict with the long-held view that some theropods sat on their nests in a brood-like manner. He proposes that the Oviraptor specimen may have been sitting on top of a nest mound rather than contact-incubating the eggs, a compromise that more resembled the nesting behavior of some modern crocodilians (Deeming 2002). Significantly, contrary to popular belief of similar nesting behaviors, dinosaur eggs resemble more those of modern reptiles, in particular crocodiles (also producing hard-shelled eggs), than those of birds and maintenance of incubation conditions would have depended on the prevailing environment.

7.4 The Dermis in Dinosaurs

Most studies with respect to the dinosaur integument have involved the epidermis and its derivative structures. The biomechanically important role of protection and/ 7.4 The Dermis in Dinosaurs 177 or support for the enclosed body mass by the dermis was relatively understudied. However, in the past 30 years the mechanical properties of the dermis have been rigorously studied in many animal groups, mainly large marine animals. Such studies have shown exactly how the skin provides both a rigid framework to support the body contents and a flexible covering to allow changes in shape associated with mobility and locomotion (Wainwright et al. 1978; Pabst 1996; Lingham-Soliar 2005a, b and references therein). The knowledge that emerged is that the dermis in many modern marine vertebrates is characterized by collagen fibers oppositely oriented in left- and right-handed helices in multiple layers that may extend the entire depth of the dermis. Remarkably, a similar architecture was found in just one fossil marine vertebrate, the ichthyosaur (Lingham-Soliar 1999, 2001; 2010; see Chap. 9 and Volume 2), enabled by the excellence of soft tissue preservations in the Posidonia Shale of Baden Württemberg (Toarcian), southern Germany. On the other hand, dinosaur integumental preservations were rather sparse, that is, until the explosion of finds from the Lower Cretaceous Yixian Formation in Liaoning Province, China in the late 1990s. Yet, even then the emphasis was on deciphering epidermal structures that were interpreted as protofeathers (Volume 2). A major breakthrough was made, however, with the discovery of Psittacosaurus sp. MV53 from the Lower Cretaceous Yixian Formation in Liaoning Province, China (part of the famous Jehol biota). It is housed in the Palaeontological Museum, Nanjing (Nanjing Institute of Geology and Palaeontology) and is the first dinosaur to provide detailed information on the collagen fiber structure of the dermis. A cross-section of preserved soft tissue, directly adjacent to the surface integumental fibers, lies on the ventrolateral part of the body near the last three thoracic ribs and anterior to the left femur (Fig. 7.17). The fine details of the preserved tissue in Psittacosaurus MV53 are probably a consequence of rapid burial and mineralization (Seilacher et al. 1985; see taphonomic section in Volume 2). Given that preservation of integumental structures in Liaoning dinosaurs, e.g., Sinosaurop- teryx (Chen et al. 1998; Currie and Chen 2001; Lingham-Soliar et al. 2007) are described as ‘blurry or foggy’ (Fucheng et al. 2006) perhaps attributable to brit- tleness that frequently occurs with pyritized soft tissue (Leng and Yang 2003), the soft tissue preservation in Psittacosaurus is remarkably good. The entire cross-section lies beneath surface integumental filaments (oppositely oriented) and extends from a height of approximately 10–15 mm deep (Fig. 7.17a, b) and width of approximately 35–40 mm. The vertical face in one part of the cross- section is more unevenly fractured. As a consequence the fibers can be seen extending in a few layers in opposing directions (Fig. 7.14b, vertical arrow). At the base of the cross-section, a tangential section is exposed (parallel to the surface layer) and shows fibers perfectly oriented in opposite directions (Fig. 7.17c). The fiber layers are closely comparable to those of sharks (Lingham-Soliar 2005a, b) (Fig. 7.17d). Remarkably, an excess of 25 fiber layers were clearly identified in the upper two-thirds of the section, ignoring gaps arising from degradation that would clearly have represented further layers. In total there are an estimated 40 fiber layers in the dermis of Psittacosaurus, the most thus far recorded in any species of animal with only a few species of sharks showing a comparable number (Motta 1977). 178 7 Dinosaur Integument

Fig. 7.17 A cross-section of the skin (dermis) of Psittacosaurus MV 53 (Nanjing Museum of Geology and Palaeontology, China). a Left posterolateral side of the body shows surface and a vertical fracture, i.e., a cross-section, in the fossil (extending from a variable height of approx. 10–15 mm (see b; the fossil shows fibers extending in opposing directions (surface, black arrows, and base (see c, inset); described in detail in Feduccia et al. 2005). b the vertical section area is fractured and as a consequence the cross-section shows fibers in two different planes (yellow arrows) and the fiber bundles are clearly visible (also in inset). c Opposing layers of fibers in tangential view, i.e., parallel to the surface fibers, exposed at the base of the section; arrows show respective fiber angles; inset, also in tangential view, shows detail of the surface. fibers from a oriented in opposing directions. d, e Dermal fibers in the white shark C. carcharias. d Tangential view of fiber bundles (left-handed); owing to the thinness of the histological section only a trace of the right-handed fibers are present (bottom right). e Cross-section of fiber bundles from skin from the body of the shark. Scale bars, a 20 mm, b 1.5 mm, and c, d 1 mm. After Lingham-Soliar (2008))

The ﬁbers are present as parallel arrays, with one layer in left-handed orientation and the other in right-handed orientation, at approximately 40° to the animal’s long axis. The orientations closely match those on the surface of the animal as well as at other locations in the same specimen suggesting a uniformity of the ﬁber orientations. 7.4 The Dermis in Dinosaurs 179

To date, all integumental structures described in dinosaurs, whether interpreted as ‘protofeathers’ or structural fibers, occur on the surface of the animal or on adjacent substrate. These findings with respect to the dermis of Psittacosaurus represented a significant development in so far as it was the first cross-sectional exposure to reveal a complex architecture of layers of structural fibers in the dermis of a dinosaur. The unique preservation of integumental structures in Psittacosaurus MV53 provides a depth of information that virtually matches that of living animals, i.e., of multiple fiber layers through the height of the dermis in which the fibers are oriented in the left- and right-handed directions in alternate layers and identifiable with that of many vertebrates thus far studied, such as sharks (Motta 1977; Lingham-Soliar 2005a, b), dolphins (e.g., Pabst 1996), modern reptiles (Feduccia et al. 2005), and the extinct ichthyosaurs (e.g., Lingham-Soliar and Plodowski 2007; Lingham-Soliar and Wesley-Smith 2008). In helical fibers that are free from the skeleton, fiber angle plays a critical role with respect to strain due to bending of the body. Fiber angles are greater at higher strains than lower, e.g., the skin of an animal is usually under tension even when it is not being deformed by some movement (Wainwright et al. 1978). In Psittacosaurus, a fiber system that would enable changes in fiber angles in response to different levels of tension would reflect the mechanical function in the ventrolateral and ventral regions of the body where there would be stress variations connected with expansion of the stomach and gut during feeding. However, the role of collagen fibers is not merely to stiffen the tissue at high strains or enable mobility. Numerous layers of fibers also contribute very much to its toughness and high work of fracture. A thick, highly reinforced dermis in dinosaurs, particularly in prey animals such as Psittacosaurus, probably evolved as a means of mechanical protection that would include to a lesser extent predatory dinosaurs (Lingham-Soliar 2008) (the helical fiber architecture and mechanics are discussed in detail in Volume 2). Finally, how did this remarkably fortuitous section in Psittacosaurus come about, was it purely a serendipitous fracture by natural forces occurring at some point during the taphonomy (death to fossilization) of the animal or something more definable? As it turns out against all odds it was possible to pinpoint the likely source of the damage. The fracture was almost certainly caused by a predator/scavenger’s biting on the abdomen of Psittacosaurus. There are a number of lines of circumstantial evidence that in totality produce a highly plausible argument that may give a glimpse of the life and death struggles of extinct animals such as Psittacosaurus of the Jehol biota: (1) The concave shape of the fracture (Fig. 7.17b) matches the convex outward tooth face of a predator or scavenger. (2) Trauma around the possible tooth tips, inferred from fibers in the vicinity emerging as it were out of the section and a general raggedness. (3) The exposure of fiber layers on the tangential plane at the base of the section, is suggestive of the ‘cut and pull’ feeding technique—the tear at the base extending along the plane of least resistance to expose a ‘clean’ tangential section. (4) Perhaps most striking is a deep indentation or excavation near the base of the cross-section and another smaller one alongside, which apparently represent the tips of tooth impressions, the smaller probably made by a 180 7 Dinosaur Integument slightly shorter tooth. Both impressions, despite a difference in size, show apparent radiating stress fractures emanating from the tooth tip that are strikingly similar in shape. The soft tissue information on the dermis gleaned from Psittacosaurus is quite unique among terrestrial fossil animals and only equalled in the marine ichthyosaurs (see Chap. 9 and Volume 2). Hence, it is instructive, particularly with respect to how mechanical strength and protection is achieved in the smaller more agile dinosaurs without compromising flexibility and movement.

7.5 Dinosaur Skin Modiﬁed as a Sail in Temperature Control?

The importance of the skin (principally the dermis) was taken to another level as we shall see in the discussion on sailback pelycosaurs (Chap. 8), namely in a highly specialized application involving temperature control. However, the idea of dorsal sails in temperature control was not restricted to pelycosaurs. Elongated neural spines in the Gondwana dinosaurs Ouranosaurus, a large iguanodont, and Spinosaurus, a gigantic theropod (Sereno et al. 1994), were popularly interpreted as support struts for high crests or sails which, as in pelycosaurs, functioned in thermoregulation and enabled these reptiles to either absorb or dissipate heat by altering the orientation of the sail with respect to the sun through different basking strategies. Jack Bowman Bailey (1997) tested the prevailing theory by examining the dorsal neural spines of Ouranosaurus, Spinosaurus, and several other long-spined dinosaurs, including the purported sailbacks, Acrocanthosaurus and Hypacrosaurus. Bailey proposed that the elongated sacral, caudal, and sometimes dorsal vertebrae in dinosaurs rather than the popular view of bearing crests or sails, were support for humps, which he considered were a relatively common dinosaurian traits that served as thermobiological buffers as well as fat storage depots to support egg production and a migratory lifestyle. Bailey’s study (1997) included comparisons with a number of modern-day functional analogs to help resolve the sail-versus-hump question, among which were mammals such as the Pleistocene Bison antiquus antiquus (Leidy), and agamid and iguanid lizards. He had, however, first to determine whether or not the neural spines in dinosaurs were in any way comparable with those of the sailback pelycosaurs before more specific comparisons with respect to presence of a sail could even be contemplated. In Dimetrodon the spines were shown to be very tall, gracile and narrowly cylindrical, distally tapering to a point, and that morphologically they resemble spines on the backs or tails of a few modem iguanid and agamid lizards (Basiliscus, Lophura, and Hydrosaurus). These slender, delicate, tapering spines were found to be biomechanically incapable of coping with the usual anteroposterior stresses of locomotion associated with anchoring points of muscles and ligaments as in for example a hump. Bailey concluded that the functional morphology of Dimetrodon spines is thoroughly consistent with the sail 7.5 Dinosaur Skin Modified as a Sail in Temperature Control? 181

Fig. 7.18 A tetrapod with elongated neural spines (also see ﬁgure of Dimetrodon in Chap. 8). Scale bar = 1m. Ouranosaurus nigeriensis Taquet, (GDF 300), an iguanodont dinosaur, Early Cretaceous of Africa; the anteroposteriorly expanded neural spines and ossiﬁed tendon lattice suggest attachment of a thick pad or hump rather than a sail. After Bailey (1997)

hypothesis. Indeed, direct confirmation was provided by incorrectly healed fractured (fracture non-union (Lingham-Soliar 2004)) spines showing that in life the spines of Dimetrodon were encased in a fleshy membrane of skin. Using large ungulate mammals as analogs, Bailey found that the elongated neural spines in a number of large dinosaurs as typified by Ouranosaurus were not consistent with the sail hypothesis. Rather he established that they were functionally involved in structural support and leverage of the anterior and posterior extremities of the body. Biomechanically the long spines located near the sacrum acted, as in some dinosaurs, in weight-bearing analogous to a cantilever bridge. However, in Ouranosaurus, the tallest and most erect are placed well in front of the hind legs, marking them as important centers of support more comparable with the parabolic bowstring-bridge paradigm. Ultimately, in dinosaurs such as Our- anosaurus the added leverage of the tall spines placed mid-dorsally, i.e., with the mass of the associated hump placed well forward of the larger hind limb, tensile energy loading suggests a mostly quadrupedal gait that includes the ability to gallop consistent with modern day analogs such as the rhinoceros. Another advantage proposed is that analogous with the bison, forward placement of the tall and erect neural spines in Ouranosaurus is reminiscent of those bison species with very low-slung heads, almost touching the ground, an adaptation for browsing on short vegetation (Bailey 1997 and references therein). Thus in the sail versus the hump hypothesis in dinosaurs, the hump appears to have been convincingly supported (Fig. 7.18).

7.6 Dinosaur Tracks and Traces

Dinosaur tracks are trace fossils also known as ichnofossils. They have been found on every continent of the world except Antarctica, but compared to dinosaur skeletons they are still relatively rare. Among the early intensive studies were those 182 7 Dinosaur Integument of fossil trackways (ichnofacies) of the Connecticut Valley, by professor Edward Hitchcock, president of Amherst College. Hitchcock systematically excavated, described, and classiﬁed thousands of tracks in remarkable detail, culminating in a major work (Hitchcock 1858). Studies of dinosaur tracks were subsequently neglected but the 1980s saw a renaissance with an explosion of interest and research. Papers presented at a major symposium on dinosaur ichnofossils were subsequently published in a book entitled Dinosaur Tracks and Traces (Gillette and Lockley 1989). Dinosaur tracks can be described using a number of features:

1. Number of toes 2. Size of footprint 3. Shape of footprint and toeprints 4. Relative arrangement of digits 5. Claw marks 6. Heel marks 7. Interdigital webbing 8. Skin impressions

A book that had an important influence on dinosaur tracks and traces was Tony Thulborn’s (1990) Dinosaur Tracks. The book provided a rich source of dinosaur tracks from around the world including important finds from Australia and was of especial significance in the levels of scientific information that can be obtained from dinosaur tracks such as the interpretation of dinosaur behavior and locomotion, with greater potential in this context than skeletal fossils. For example, estimates of normal dinosaur walking and running speeds can be calculated on the basis of the analysis of trackways (below) and in sauropods in particular, group behavior (Ostrom 1972). In dinosaur fossilized trackways (ichnofacies) the epidermal structures may play an important, even indispensable part, in the identification of the trackmaker. The claws which are hardened (keratinized) modification of the epidermis are frequently preserved as a negative impression and are particularly marked in carnivorous dinosaurs. Individual prints provide data on the size and shape of the trackmaker’s foot, and the number the toes. Clear prints can even reveal details of the soft anatomy of the foot, including the pattern of pads and muscles on the feet, and the flexibility of the digits. These track features, combined with trackway patterns, reveal important clues about the identity of the trackmaker. Studies by Lingham-Soliar and colleagues concern the tracks of both theropods and sauropods in the Chewore region of Zimbabwe in which Middle Jurassic strata are exposed (Lingham-Soliar et al. 2003, 2004). The quality of the Chewore dinosaur tracks are remarkably good and even show skin creases and claws (Figs. 7.19 and 7.20). Among the most intriguing trackways discovered were several in close associations, which comprised a total of 88 tracks of large bipedal dinosaurs (Fig. 7.21). Several features of the tracks show with reasonable certainty that they belong to theropods: 7.6 Dinosaur Tracks and Traces 183

Fig. 7.19 The Chewore (Zimbabwe) dinosaur footprints. The longest series of nine Allosaur-like footprints extends parallel to the river bed before turning into the right bank. Scale = 25 cm. After Lingham-Soliar et al. (2003)

(1) they are tridactyl and signiﬁcantly longer than wide (ratio of foot length:foot width is 4:3); (2) distinctive curved claws are present on a number of; (3) they have tapering digits, and (4) curvature of digits II and III; (5) there is a narrow gauge of the trackways, that is, right–left bipedal progression along a narrow midline. (6) Tracks are on average 5 cm deep in the anterior toe region, about a third deeper than in the posterior part of the foot (excluding the metatarsal impression), and most show a strong medial curvature and angulation of digits II 184 7 Dinosaur Integument

Fig. 7.20 Chewore dinosaur footprint. One of the well- preserved tracks west– northwest in a narrow strip (from Fig. 7.19);white arrow shows distinctive slender claw; digit pads are evident including skin creases. After Lingham-Soliar et al. (2003)

and III, suggesting a species trait and shows the suggested walking posture and average size of the dinosaur. To assess the size of the theropods a measurement of 4.5 9 foot length was used to calculate hip height (e.g., Alexander 1976) and the greatest hip height was found to be 211.5 cm. While it is commonly thought that theropods and ornithopods habitually walked in a digitigrade (toe-walking) manner work, Kuban (1989) showed that trackmakers in Paluxy Riverbed of Texas occasionally walked in a plantigrade or plantigrade- like manner, impressing their soles and heels as they walked–thus making elongate tracks. In the Chewore tracks (Lingham-Soliar et al. 2003) occasionally long slip marks are associated with metatarsal impressions (Figs. 7.21 a, b). The plantigrade stance may be associated with slippery conditions as inferred from sediment collapse, also noted in the digit impressions in some tracks (Fig. 7.20, arrowed) (Kuban 1989) and by the dynamic interaction between foot and substrate (Thulborn 1990). The dinosaurs of the Chewore region in Zimbabwe are important for a number of reasons. These were large theropods, probably averaging an estimated total height of about 3 meters. The multiple trackways of a large theropod dinosaur taxon all in close associations indicates a highly reasonable possibility that the dinosaurs were gregarious (Lingham-Soliar et al. 2003). This phenomenon is rarely recorded in theropod dinosaurs with the possible exception of trackways in the Morrison Formation in the USA. Perhaps most fascinating is the reasonable speculation that these theropods were pack-hunting, which may add to our understanding of theropod feeding behavior. Lingham-Soliar et al. (2003) proposed a number of points in support of the view that the tracks were made by theropod dinosaurs traveling as a group: 7.6 Dinosaur Tracks and Traces 185

Fig. 7.21 Jurassic theropod trackways in Zimbabwe. a number of trackways cross the dry river bed (white arrows show three double tracks). Scale = 50 cm. b The main area of site 5 containing 30 tracks (based on tracing of a detailed photo). Five trackways are identiﬁed (I–V). The hatched area to the left represents a lower bed. Zig-zag patterns area schematic representation of some areas in which ripples are preserved within the track site. Three of the double tracks are indicated. c Diagram of footprint in Fig. 7.20, includes measurements. d Silhouette of a large theropod trackmaker with suggested posture (ca. 3 m tall). After Lingham- Soliar et al. (2003)

1) All the footprints belong to large theropods of apparently a single species. (2) In a rapid-drying environment the substrate should show footprints of different preservational types, for example, if they were laid down with longer time intervals separating them. This is apparently not the case since most of the tracks have remarkably similar shapes and depths. It is pertinent too that the tracks forming double imprints are equally sharply defined (Fig. 7.17). This is unlikely if there was a significant time interval between the first and second imprints since desiccation, for example, would likely result in differences in track quality and increased likelihood of destruction of the earlier formed track by the later. On the contrary the track of a succeeding dinosaur cuts precisely into that of the preceding 186 7 Dinosaur Integument dinosaur, with little interference to the latter’s shape or quality. (3) Tracks are aligned in one general direction, that is, there are no returning or random tracks or those of other species. The most parsimonious explanation is that individuals of the same taxon made the tracks within a relatively short space of time. (4) Three sets of double imprints in a relatively small number of tracks suggest animals traveling in relatively close association. Crossing over of the trackways, as opposed to the side by side alignment of sauropod (herbivore) trackways (Lingham-Soliar 2003b), is considered reasonable and inevitable in predatory pack-hunting animals (Lingham-Soliar, personal observation of wolves in northern Europe and numerous carnivores in Africa), in which hierarchies (e.g., in wolves denoted by alpha, beta, and omega males/females) prompt ‘safe’ distances being maintained between pack members. The sauropod tracks described by Lingham-Soliar et al. (2004) are the first to have been discovered in sub-Saharan Africa. The presence of very large sauropod tracks in close vicinity to the theropod tracks and in the same geological horizon may lend further credibility to the hypothesis of pack hunting and a predator/prey interaction between these two groups of dinosaurs (Fig. 7.22). The present dinosaur associations in Zimbabwe are interesting in the context of the African dinosaur fauna. Brachiosaurus was first discovered in Africa in Tendaguru, Tan- zania (e.g., see Fastovsky and Weishampel 1996) although there was no record of association with large theropods, in contrast to the theropod/sauropod associations of the Morrison Formation in the United States (Prince and Lockley 1989). Thus the Ntumbe region in Zimbabwe, as with the Morrison Formation, speaks to a theropod fauna with associated giant sauropods. Smaller theropods coexisted with large Ntumbe theropods described above. Their tracks were no more than 7 cm long. Short trackways showed that in one the theropod traveled at a walking pace and in the other at a trot (calculated from formula in Thulborn 1990). Interestingly even smaller tracks between 4.5 and 5 cm were also recorded (Lingham-Soliar and Broderick (2000). The trackway showed impressions of both the hands (manus) and the feet (pes) indicating that the animal was on all fours. If it was a theropod dinosaur this is incongruous with their normal gait. On the other hand, a theropod walking on all fours strongly pointed to one possibility, that the dinosaur was immature. However, trying to establish that the footprints belonged to an immature (‘juvenile’) dinosaur and not a small adult species was a real problem. The small size was a firm beginning. The tridactyl footprint shape and narrowness of the tracks clearly indicated this was a theropod dinosaur. However, at complete odds, the unusually strong outward rotation of the feet was not consistent with a theropod dinosaur. On the other hand in immature animals and even in human babies taking their first steps, balance is helped by rotating the feet outwards. There are also signs that the dinosaur had slipped and fallen toward the end of the trackway (indicated by a very deep extended hand print and a skidmark of the foot), conditions more consistent with an ‘inexperienced’ animal. The combined evidence strongly pointed to our trackmaker being an immature dinosaur making possibly its first steps. 7.6 Dinosaur Tracks and Traces 187

Fig. 7.22 One of the best preserved sauropod tracks found close to the therepod tracks described here. a Distinct claw impressions are seen. Uniquely a high ridge or mound surrounds the track which reveals the impact of the foot on the mud and that drying of the mud was rapid, i.e., not allowing time for the mound to collapse. Scale bar = 1m.b A revisit to the remote site, two years later in 2003 shows that it was almost completely destroyed. Only the claw region is identiﬁable (white rectangle). The smooth bank (top) suggests elephants rubbing themselves against it and that they were the most likely cause of the destruction. After Lingham-Soliar et al. (2004)

Fig. 7.23 Base Camp. Being armed (photo of the author on the ﬁrst expedition to the remote Chewore region in Zimbabwe in 1990) was necessary because of the daily danger from rhino poachers who were invariably better armed with Kalashnikovs 188 7 Dinosaur Integument

Fig. 7.24 The smallest dinosaur footprints recorded in Africa (Chewore, Zimbabwe) and possibly in the world. Details (track nos. 5 and 8) show manus and corresponding pes tracks. From a cast made by the author as described in the text. Scale bar = 4 cm. From Lingham-Soliar and Broderick (2000)

The notable paleontologist Charles Sternberg captured vividly in his classic memoir The life of a fossil hunter (1909) some of his astonishing adventures. He brilliantly captures the hardships and dangers of fossil collecting during the bygone age of the Wild West, perhaps most memorable was his quite nonchalant description of searching for fossils in the Black Hills of Dakota—only weeks after Custer’s defeat at Little Big Horn with the Indian danger still very much in the air. Perhaps it is the lure of fossils that make one oblivious to the dangers as seen in 7.7 Pterosaurs 189 perhaps a unique set of experiences connected with the Chewore tracks (Lingham- Soliar and Broderick 2000). These dinosaur tracks are located in the most remote part of Zimbabwe, situated on the Zambian and Mozambican borders. The meandering river valleys and escarpments make the region one of nature’s great unspoilt places with a variety of wildlife that includes elephants, lions, buffalo, and crocodiles. It was also the home of the largest population of black rhino in the world and tragically, the rewards for rhino horn are huge. Consequently, we were faced with the daily danger of raiders armed with Kalashnikovs, for whom life was cheap and who outgunned us armed with just AK 47 s (Fig. 7.23) (today the rhinos have been relocated further south where they are better protected). As recorded in the above publication the dangers from the wildlife were also ever present and it was not the place to be lulled into a false sense of security. On one occasion a lone male lion broke from the undergrowth on the bank of the dry riverbed and charged at the author, stopping at the last moment before returning to the bank. Never- theless, the danger from this irate lion did not deter our team from making a latex mold of the ‘juvenile’ dinosaur’s trackway (Lingham-Soliar and Broderick 2000) (Fig. 7.24) despite the continued presence of the lion (probably recently ousted from its pride), which we were reminded of by its constant deep throaty growls a few meters away in the adjacent bush.

7.7 Pterosaurs

Although pterosaurs are not dinosaurs, both belong to the archosaurs, a group to which birds and crocodiles also belong. Ancestors of pterosaurs, like those of birds, tended toward a bipedal gait, which thus freed the forelimbs for other uses, which evolved into wings in birds and pterosaurs. Unlike in birds, instead of feathers, pterosaurs developed a wing surface formed by a membrane of skin similar to that of bats. However, in bats all of the fingers except the thumb support the membrane. In pterosaurs, the membrane was attached solely to the elongated fourth finger (there was no fifth finger). The first three fingers were slender, clawed, clutching structures. The flight membrane will be discussed in the section on the evolution of flight in Volume 2.

References

Alexander G, Marais JM (2007) A guide to the reptiles of southern Africa. Struik Nature, Cape Town Alexander RM (1976) Estimates of speeds of dinosaurs. Nature 262:129–130 Bailey JB (1997) Neural spine Elongation in Dinosaurs: Sailbacks or Buffalo-Backs? J Paleontol 71:1124–1146 190 7 Dinosaur Integument

Buffetaut E, Grellet-Tinner G, Suteethorn G, Cuny G, Tong H et al (2005) Minute theropod eggs and embryo from the Lower Cretaceous of Thailand and the dinosaur-bird transition. Naturwissenschaften. doi:10.1007/s00114-005-0022-9 Chen P-J, Dong ZM, Zheng SN (1998) An exceptionally well preserved theropod dinosaur from the Yixian Formation of China. Nature 391:147–152 Chiappe LM, Coria RA, Dingus L, Jackson F, Chinsamy A, Fox M (1998) Sauropod dinosaur embryos from the LateCretaceousof Patagonia. Nature 396:258–261 Coria RA, Chiappe LM (2004) Piel embrionaria de sauro0podos (Dinosauria) del Creta0cico Tardı0o de Auca Mahuevo, Patagonia. Argentina. Ameghiniana 41:9R Currie PJ, Chen P-J (2001) Anatomy of Sinosauropteryx prima from Liaoning, northeastern China. Can J Earth Sci 38:1705–1727 Czerkas SA (1997) Skin. In: Currie PJ, Padian K (eds) Encyclopedia of Dinosaurs. Academic Press, San Diego, pp p669–p675 Deeming DC (2002) Importance and evolution of incubation in avian reproduction. In: Deeming DC (ed). Avian incubation: behaviour, environment and evolution. Oxford University Press, Oxford, pp 1–7 Deeming DC, Thompson MB (1991) Gas exchange across reptilian eggshells. In: Deeming DC, Ferguson MWJ (eds) Egg incubation: its effects on embyronic development in birds and reptiles. Cambridge University Press, Cambridge, pp 277–284 Deeming DC (2006) Ultrastructural and functional morphology of eggshells supports the idea that dinosaur eggs were incubated buried in a substrate. Palaeontology 49:171–185 Dong Z-M, Currie PJ (1996) On the discovery of an oviraptorid skeleton on a nest of eggs at Bayan Mandahu, Inner Mongolia, People’s Republic of China. Can J Earth Sci 33:631–636 Edwards NP, Barden HE, van Dongen BE, Manning PL, Larson PL et al (2011) Infrared mapping resolves soft tissue preservation in 50 million year-old reptile skin. Proc R Soc B 209:3218. doi:10.1098/rspb.2011.0135 Fastovsky DE, Weishampel DB (1996) The evolution and extinction of the dinosaurs. Cambridge University Press, Cambridge Feduccia A, Lingham-Soliar T, Hinchcliffe JR (2005) Do Feathered Dinosaurs Exist? Testing the hypothesis on neontological and paleontological evidence. J Morphol 266:125–166 Fucheng Z, Zhou Z, Dyke G (2006) Feathers and ‘feather-like’ integumentary structures in Liaoning birds and dinosaurs. Geol J 41:395–404. doi:10.1002/gj.1057 Gohlich UB, Chiappe LM (2006) A new carnivorous dinosaur from the Late Jurassic Solnhofen archipelago. Nature 440:329–332. doi:10.1038/nature04579 Grellet-Tinner G, Chiappe LM (2004) Dinosaur eggs and nesting:implications for understanding the origin of birds. In: Currie PJ, Koppelhus EB, Shugar MA, Wright JL (eds) Feathered dragons: studies on the transition from dinosaurs to birds. Indiana University Press, Indiana, pp 185–214 Grellet-Tinner G, Chiappe LM, Coria RA (2004) Eggs of titanosaurid sauropods from the Upper Cretaceous of Auca Mahuevo (Argentina). Can J Vertebrate Paleontol 41:949–960 Gillette DD, Lockley MG (eds) (1989) Dinosaur tracks and traces. Cambridge University Press, Cambridge Hitchcock E (1858) Ichnology of New England. A Report on the Sandstone of the Connecticut Valley, Especially its Fossil Footmarks. W White, Boston. Reprinted 1974 by Arno Press, New York Jarvik E (1980) Basic structure and evolution of vertebrates. Academic Press, London Kuban GJ (1989) Color distinctions and other curious features of dinosaur tracks near Glen Rose, Texas. In: Gillette DD, Lockley MG (eds) Dinosaur tracks and traces. Cambridge University Press, Cambridge, pp 426–440 Leng Q, Yang H (2003) Pyrite framboids associated with the Mesozoic Jehol biota in northeastern China: implications for microenvironment during early fossilization. Prog Nat Sci 13:206–212 Lingham-Soliar T (1999) Rare soft tissue preservation showing ﬁbrous structures in an ichthyosaur from the Lower Lias (Jurassic) of England. Proc Roy Soc Lond B266:2367–2373 References 191

Lingham-Soliar T (2001) The ichthyosaur integument: A means for a stiff, smooth skin. Lethaia 34:287–302 Lingham-Soliar T (2003a) The dinosaurian origin of feathers: Perspectives from dolphin (Cetacea) collagen fibers. Naturwissenschaften 90:563–567 Lingham-Soliar T (2003b) A functional and biomechanical perspective of locomotion in a giant sauropod dinosaur – or why sauropods don’t waddle. N Jb Geol Pälaont Abh 229:19–30 Lingham-Soliar T (2004) Palaeopathology and injury in the extinct mosasaurs (Lepidosauro- morpha, Squamata) and implications for modern reptiles. Lethaia 37:255–262 Lingham-Soliar T (2005a) The Dorsal Fin in the White Shark, Carcharodon carcharias:A Dynamic Stabilizer for Fast Swimming. J Morphol 263:1–11 Lingham-Soliar T (2005b) Caudal fin in the white shark Carcharodon carcharias (Lamnidae): A dynamic propeller for fast, efficient swimming. J Morphol 264:233–252 Lingham-Soliar T (2008) A unique cross-section through the skin of the dinosaur Psittacosaurus from China showing a complex fibre architecture. Proc Roy Soc B 275:775–780. doi:10.1098/ rspb.2007.1342 Lingham-Soliar T (2012) The evolution of the feather: Sinosauropteryx, life, death and preservation of an alleged feathered dinosaur. J Ornithol 153:699–711. doi:10.1007/ s10336-011-0787-x Lingham-Soliar T (2013) The evolution of the feather: scales on the tail of Sinosauropteryx and an interpretation of the dinosaur’s opisthotonic posture. J Ornithol 154:455–463. doi:10.1007/ s10336-012-0910-7 Lingham-Soliar T, Broderick T (2000) An enigmatic Early Mesozoic dinosaur trackway from Zimbabwe. Ichnos 7:135–148 Lingham-Soliar T, Plodowski G (2007) Taphonomic evidence for high-speed adapted fins in thunniform ichthyosaurs. Naturwissenschaften 94:65–70. doi:10.1007/s00114-006-0160-8 Lingham-Soliar T, Wesley-Smith J (2008) First Investigation of the collagen D-band ultrastructure in fossilized vertebrate integument. Proc R Soc B 275:2207–2212. doi:10. 1098/rspb.2008.0489 Lingham-Soliar T, Glab J (2010) Dehydration: A mechanism for the preservation of fine detail in fossilised soft tissue of ancient terrestrial animals. Palaeogeogr Palaeoclimatol Palaeoecol 291:481–487. doi:10.1016/j.palaeo.2010.03.019 Lingham-Soliar, T, Plodowski G (2010) The integument of Psittacosaurus from Liaoning Province, China: taphonomy, epidermal patterns and color of a ceratopsian dinosaur. Naturwissenschaften 97:479–486. DOI 10.1007/s00114-010-0661-3 Lingham-Soliar T, Broderick T, Ahmed A-K (2003) Closely associated theropod trackways from the Jurassic of Zimbabwe. Naturwissenschaften 90:572–576 Lingham-Soliar T, Ahmed A-K, Broderick T (2004) Giant sauropod tracks from the Mid-Late Jurassic of Zimbabwe in close association with theropod tracks. Lethaia 37:467–470 Lingham-Soliar T, Feduccia A, Wang X (2007) A new Chinese specimen indicates that ‘protofeathers’ in the Early Cretaceous theropod dinosaur Sinosauropteryx are degraded collagen fibres. Proc R Soc B 274:1799–1805. doi:10.1098/rspb.2007.0352 Lull RS (1933) A revision of the ceratopsia or horned dinosaurs. Memoirs Peabody Mus Nat Hist 3:1–175 Manning PL, Morris PM, McMahon A, Jones E, Gize A et al (2009) Mineralized soft-tissue structure and chemistry in a mummified hadrosaur from the Hell Creek Formation, North Dakota (USA). Proc R Soc Lond B 276:3429–3437. doi:10.1098/rspb.2009.0812 Matthew WD (1915) Dinosaurs with special reference to the American Museum Collections. New York American Museum of Natural History. By permission of Project Gutenberg http:// www.gutenberg.org Mayr G, Peters DS, Plodowski G, Vogel O (2002) Bristle-like integumentary structures at the tail of the horned dinosaur Psittacosaurus. Naturwissenschaften 89:361–365 Mikhailov KE (1997) Fossil and recent eggshell in amniotic vertebrates: fine structure, comparative morphology and classification. Palaeontology 56:1–80 192 7 Dinosaur Integument

Motta PJ (1977) Anatomy and functional morphology of dermal collagen fibres in sharks. Copeia 1977:454–464 Osborn HF (1912) Integument of the iguanodon dinosaur Trachodon. Am Mus Nat Hist Mem 1:33–54 Ostrom JH (1972) Were some dinosaurs gregarious? Palaeogeogr Palaeoclim Palaeoecol 11:287–301 Pabst DA (1996) Morphology of the subdermal connective sheath of dolphins: a new fiber- wound, thin-walled, pressurized cylinder model for swimming vertebrates. J Zool Lond 238:35–52 Prince NK, Lockley MG (1989) The sedimentology of the Purgatoire Tracksite Region, Morrison Formation of Southeastern Colorado. In: Gillette DD, Lockley MG (eds) Dinosaur tracks and traces. Cambridge University Press, Cambridge, pp 155–164 Reisz RR, Evans DC, Sues H-D, Scott D (2010) Embryonic skeletal anatomy of the sauropodomorph dinosaur Massospondylus from the Lower Jurassic of South Africa. J Vertebrate Paleontol 30(6):1653–1665 Sereno PC, Wilson JA, Larsson HCE, Dutheil DB, Sues H-D (1994) Early Cretaceous dinosaurs from the Sahara. Science 266:267–270 Sternberg CH (1909) The life of a Fossil Hunter. Indiana University Press (1990) p 286 Sexton OJ, Bramble JE, Heisler IL, Phillips CA, Cox DL (2005) Eggshell composition of squamate reptiles: relationship between Eggshell permeability and Amino acid distribution. J Chem Ecol 31: 2391–2401. DOI:10.1007/s10886-005-7108-x Schweitzer MH, Hill CL, Asara JM, Lane WS, Pincus SH (2002) Identification of immunore- active material in mammoth fossils. J Mol Evol 55:696–705 Schweitzer MH, Chiappe L, Garrido AC, Lowenstein JM, Pincus SH (2005) Molecular preservation in Late Cretaceoussauropod dinosaur eggshells. Proc R Soc B 272:775–784. doi:10.1098/rspb.2004.2876 Thulborn RA (1990) Dinosaur tracks. Chapman and Hall, London Wainwright SA, Vosburgh F, Hebrank JH (1978) Shark skin: function in locomotion. Science 202:747–749 Zhang F, Kearns SL, Orr PJ, Benton MJ, Zhou Z et al (2010) Fossilized melanosomes and the colour of Cretaceous dinosaurs and birds. Nature 463:1075–1078. doi:10.1038/nature08740 Chapter 8 Mammal-Like Reptiles

8.1 In the Shadow of the Dinosaurs

When the thecodonts (‘‘socket-toothed’’), the group thought to include the ancestral stock of all other archosaurs, including birds, dinosaurs, pterosaurs and crocodiles, appeared on the evolutionary scene, another group of reptiles were also evolving. They were the mammal-like reptiles or therapsids. The group had a checkered history having survived the Permian boundary crisis and becoming virtually extinct by the end of the Triassic—possibly a consequence of competition from more efficient predators within the thecodonts. Their near complete demise would allow the dramatic evolution of the archosaurs, the pinnacle of which were the dinosaurs but it would also leave a small window of opportunity for what would later become the most highly evolved species on the planet—mammals. It is interesting to conjecture what would have happened if some of the early mammal- like reptiles, bristling with so much potential, had not suffered such major extinctions and depletions early in their history and if so how their descendants, presumably fully fledged mammals, might have developed from some of these magnificent mammalian-reptile experiments (and not as it turned out much later from the small shrew-like survivors)—and indeed how our own human species might have evolved if at all? The mammal-like reptiles evolved from captorhinomorphs such as Eocapto- rhinus (Fig. 8.1) during the Carboniferous and can be distinguished by the lower or synapsid temporal opening (see Chap. 6). They can be divided into the more primitive order Pelycosauria, of the Upper Carboniferous and Lower Permian (most notably from the Texas Red Beds) and from the more advanced Therapsida of the Upper Permian and Triassic (most notably from the South African Karoo and Russia).

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 193 DOI: 10.1007/978-3-642-53748-6_8, Ó Springer-Verlag Berlin Heidelberg 2014 194 8 Mammal-Like Reptiles

Fig. 8.1 Representative of the family Captorhinidae. a Reconstruction of the skeleton of Eocaptorhinus. b Occipital view of skull. Note the massive stapes and large posttemporal fossa. c Palate of Captorhinus showing multiple tooth rows. The ectopterygoid bone is missing; its position is replaced by a medial process of the jugal. From Carroll (1988)

8.2 Pelycosaurs

Pelycosaurs can be most readily recognized by their sprawling gait and undiffer- entiated teeth, conditions strongly reminiscent of true amphibians and reptiles, but the presence of lateral temporal openings shows unequivocally that they are related to the ancestry of mammals. However, one member belonging to the most advanced family, the Sphenacodontidae, was perhaps the most notable and noticeable of the pelycosaurs. This was Dimetrodon, a carnivorous pelycosaur that reached lengths of around 3 m (Figs. 8.2 and 8.3). The long neural spines gave Dimetrodon its unmistakable appearance and it is thought they performed the function of temperature control. If this was so then it gives further credence to the view that pelycosaurs were ectotherms rather than endotherms and that they probably lacked a tendency to a high metabolic rate or other physiological traits that characterize mammals. The 8.2 Pelycosaurs 195

Fig. 8.2 Dimetrodon. Skeleton of the carnivorous pelycosaur from the Permian of the Lower Permian of Texas, 3 m long. Experiments (see text) have shown that the long neural spines that supported a membrane acted to increase the rate of heat exchange with the environment. From Romer and Price (1966)

Fig. 8.3 Dimetrodon, skull and lower jaws. a, b Lateral view. c Medial view of lower jaw. Abbreviations. a angular; art articular, cor coronoid; d dentary; sa surangular; f frontal; j jugal; l lacrimal; m maxilla; n nasal; sph sphenoid; part prearticular; po postorbital; pm premaxilla; prf prefrontal; ps parasphenoid; pt pterygoid, q quadrate; sm septomaxilla; sp splenial; sph sphenethmoid; sph sphenoid; sq squamosal; st stapes. From Romer and Price (1966) and Carroll (1988) 196 8 Mammal-Like Reptiles

Fig. 8.4 Edaphosaurus an herbivorous pelycosaur from the late Pennsylvanian and early Permian. a Skeleton 3 m long. Unlike in Dimetrodon, the spines bear transverse processes. b, c, d Skull in lateral, occipital, and palatal views. After Romer (1966) and Carroll (1988) spines were most studied in Dimetrodon and are thought to have supported a single sheet of tissue and to have accommodated a network of blood vessels. An ample blood vascular supply would permit the sail to rapidly absorb heat and transfer it to the rest of the body. From experiments conducted by Bramwell and Fellgett (1973) they were able to calculate that a 200 kg Dimetrodon would take 205 min to warm up to 26–32 °C. However, with the sail they calculated that it would take 80 min, i.e., well under half that time. This would have given them a head start over other predators and prey by allowing them to be active earlier in the day (see Chap. 7 for discussion on sailbacks in dinosaurs). Dimetrodon was not alone in the possession of greatly elongated neural spines as they were also strikingly present in the herbivorous pelycosaur, Edaphosaurus (Fig. 8.4), from the late Pennsylvanian and early Permian (Kemp 1982; Carroll 1988) (Fig. 8.5). 8.3 Therapsids 197

Fig. 8.5 A pictorial family tree of the pelycosaurs. From Romer (1966)

8.3 Therapsids

One pelycosaur family is generally considered to have given rise to the more advanced therapsids. Primitive therapsids are present as fossils in certain Middle Permian deposits; later forms are known from every continent except Australia but are commonest in the Late Permian and Early Triassic of South Africa. But from the beginning of their history these reptiles follow evolutionary trends that led to functional specializations very different from those reached by any other reptiles. The limbs and limb girdles were modified for quadrupedal locomotion. The gait in carnivorous therapsids such as Thrinaxodon, Cynognathus, and Lycaenops becomes mammalian, with the legs drawn under the body in a near-vertical stance, a departure from the sprawling posture of their predecessors the pelycosaurs. In addition the vertebral column becomes flexed to enable raising of the head (Fig. 8.6). The dentition is differentiated compared to simple conical teeth of other reptiles. In advanced therapsids a high level of development is attained with sharply contrasted incisors, canines, and cheek teeth or molars. Many changes occurred in the skull, e.g., the quadrate and quadratojugal bones were reduced to very small elements, often loosely connected to the skull (in true mammals they would become the ear ossicles). In more advanced therapsids a secondary palate develops below the original reptilian palate, which served to separate the nasal passage from the mouth thereby increasing the efficiency of breathing particularly while feeding. Also, it is among the more advanced cynodonts that hearing is 198 8 Mammal-Like Reptiles

Fig. 8.6 Skeleton, showing wolf-like, mammalian gait of the Gorgonopsid, Lycaenops (1 m long). From Colbert (1955) and Carroll (1988) thought to have developed along the line of the modern mammalian condition, which may merely be a culmination of the cynodont condition, in which the angular, articular, and quadrate bones were already involved in sound conduction. The sound transmitting tympanic membrane is according to this hypothesis, thought to have existed and to have enclosed an air space within a recess created by processes of the angular and articular bones (Allin 1975) (Figs. 8.7 and 8.8). During the Permian there was an overwhelming array of therapsids, dino- cephalians, dicynodonts, and theriodonts in Russia and particularly in South Africa, in the Karoo area. The sediments in South Africa form the Karoo series, one of the classic stratigraphic sequences of earth history, uniquely covers the complete extent of both the Permian and Triassic periods. Among advanced carnivorous therapsids, two groups of theriodonts ﬂourished, the gorgonopsians and therocephalians (Huttenlocker 2009; Huttenlocker et al. 2011) (Figs. 8.9 and 8.10). Gorgonopsids were the dominant carnivores at the time. The canine teeth are enlarged and the cheek teeth are reduced. The postcranial skeleton is best known in the genus Lycaenops. Tom Kemp (1982) one of the leading experts on the mammal-like reptiles, structure and functional biology, showed that while the postcranial skeleton of gorgonopsids resembles that of modern cursorial mammals the posture of the forelimb retains primitive features with the humerus, being held essentially horizontally. He proposed that the structure and mechanics of the femur in one member of the group, Lycaenops, was more comparable with that of crocodiles than modern cursorial mammals. It was capable of two styles of locomotion, one that adopted a sprawling posture and the other in which the femur was held at about 45° to the body, which enabled a semi-parasagittal gait. Whereas gorgonopsids did not survive after the end of the Permian and are not considered directly related to the more advanced group of carnivorous therapsids, the therocephalians, and cynodonts had a particularly interesting feature that may have passed on to later therapsids (Fig. 8.9). 8.3 Therapsids 199

Fig. 8.7 Skull of the primitive ictidorhinid gorgonopsian Rubidgina from the Upper Permian of South Africa. a Dorsal view. b Lateral view. Skull length about 10 cm. The orbits are much larger than the temporal openings in contrast with the more advanced gorgonopsians such as Lycaenops (Fig. 8.6)

The therocephalians were a predominant group of carnivorous therapsids, some with a superﬁcial resemblance to gorgonopsids (Huttenlocker 2009; Ivakhnenko 2011). However, they were much more diverse than the gorgonopsians and included small insectivorous forms and large carnivores and one late group adopted a herbivorous diet but in a manner quite different to the dicynodonts (Kemp 1982). Some of the larger therocephalians competed with the large gorgonopsians as carnivores of the late Permian. The earlier therocephalians were in many respects as primitive as the gorgonopsids, but they did show certain advanced features. Although they differed with gorgonopsians with respect to the temporal opening for broader jaw adductor muscle attachment and in development of an incipient secondary palate, they shared certain similarities such as large canines and loss of the cheek teeth. Kemp (1982) has shown from an almost complete skeleton of a therocephalian that some members of the group had very 200 8 Mammal-Like Reptiles

Fig. 8.8 Skull of the Lower Triassic cynodont Thrinaxodon. a Dorsal b palatal c lateral. Skull length about 8 cm. Abbreviations as in Figs. 8.1 and 8.3. From Carroll (1988) mammalian proportions with relatively long ribs, attenuated lumbar ribs, and a highly reduced tail. Furthermore, the discovery of maxilloturbinal ridges in forms such as Glanosuchus, suggests that at least some therocephalians may have been warm-blooded. This is based on the knowledge that animals with respiratory turbinates can maintain a high rate of breathing without the adverse effects of dehydrating the nasal passages and lungs, conditions indirectly supporting a high metabolic rate and endothermy. Unfortunately, these bones are very delicate and therefore have not yet been found in fossils. But rudimentary ridges like those that support respiratory turbinates have been found in advanced therocephalians as noted above (Ruben and Jones 2000) (Figs. 8.10 and 8.11). 8.4 The Evolution of Hair 201

Fig. 8.9 A pictorial family tree of the therapsids. From Romer (1966)

8.4 The Evolution of Hair

Hair characterizes mammals and one of the problems that faces paleontologists is the thorny question of its evolution. Many smaller members of mammal-like reptiles have numerous small foramina or pits on their snouts that may have supported vibrissae or whiskers. Could the various hypotheses on foramina or pits in the snout of certain mammal-like reptiles provide clues to the ﬁrst appearance of hair in mammals? Perhaps no one has done more to attempt to understand the presence of such small infraorbital pits on the snout of small mammal-like reptiles than the eminent Russian palaeontologist, Leonid B Tatarinov. Among the theories he proposed for the presence of such pits was that in some forms they may have been part of an electroreception system in aquatic therocephalians (Tatarinov 1994; see Chap. 4). In other forms such as Theriognathus we see an extraordinarily large number of infraorbital pits. The distribution of these pits occurs over a large area of the antero-lateral surface of the snout as seen in a specimen housed in the Geology Museum of Tuebingen University (Lingham-Soliar unpublished data, Fig. 8.11). Among the hypotheses for the function of similar pits is that they may have housed sensory hairs or vibrissae (see Kemp 1982). However, rostral foramina were demonstrated in the snout of at least one modern-day reptile whose scaly skin lacks both vibrissae and hair, namely in the lizard Tupinambis, considered similar in shape, number, and distribution to those of the cynodont Thri- naxodon (Ruben and Jones 2000). Kemp (1982) observed that much of the external surfaces of the bones of the snout and dentary are ﬁnely sculptured with minute foramina often associated with tiny grooves and that this type of surface is typical 202 8 Mammal-Like Reptiles

Fig. 8.10 Cranial restorations of selected therocephalian taxa in left lateral aspect. a The akidnognathid eutherocephalian Moschorhinus kitchingi (modified from Durand 1991); b akidnognathid Promoschorhynchus (based on SAM-PK-K10014 and RC 116); c hofmeyriid Mirotenthes digitipes; d whaitsiid Theriognathus microps; e basal ‘‘ictidosuchid’’ baurioid Ictidosuchoides longiceps (modified from Hopson 1994); f derived bauriid Bauria cynops (modified from Brink 1963). Skulls not to scale. From Huttenlocker et al. (2011) of modern reptiles where the skin is tightly applied to the bone surface. If this is so then there was clearly absence of a fleshy snout in Theriognathus and consequently, if hair was present, absence of whisking movements (see below). If the idea that the pits housed functional sensory vibrissae then given the absence of whisking musculature two conditions it seems should be fulfilled: (1) the vibrissae would have to have been located over a wide area of the snout and (2) they would have to have been fixed deeply at widely differing angles, i.e., radiating in a virtual circle to capture as many distance signals as possible (analogous to a large fixed-dish satellite as opposed to a small rotating-dish satellite). This is where The- riognathus comes into its own—in sheer density of the pits and area covered (from beneath the eye socket to snout tip) and, significantly, radiation of the infraorbital pits from a central point as demonstrated by inserting bristles into the deep pits (Fig. 8.11b), they are unlike those in any other mammal-like reptile. It is not inconceivable that they may have housed vibrissae. If the pits housed vibrissae then their radiation would be functionally viable only if the bony pits enabled precise and varied orientations. This is comparable, e.g., to the precise orientation 8.4 The Evolution of Hair 203

Fig. 8.11 Skull of Dixeya (Theriognathus) showing rostral foramina. a The foramina extend on the dorso-lateral surfaces as far as the anterior rim of the arbits. b Artificial hairs inserted into the deep foramina without manipulation of angles of feathers by being lodged in bony pits in the wing tips of birds where there is barely any musculature. Alternative hypotheses such as the pits being associated with nerves or blood vessels are hardly likely to have required such a drastic modification of the bony structure of the snout (a few pits would suffice as in modern mammals) (Author’s unpublished results). How does the above hypothesis of vibrissal pits fit in, if at all, with the way in which vibrissae are thought to have evolved in modern mammals? In contrast to the snout of reptiles, the snout in modern mammals is fleshy and muscular. Because the vibrissae are directly connected with a neuro-muscular system they are capable of highly complex movements and function. Brecht and colleagues (1997) documented vibrissal architecture in a series of mammals to identify evolutionary conserved features of vibrissal organization. As a result of this analysis they 204 8 Mammal-Like Reptiles distinguish between a frontal microvibrissal system and macrovibrissal system of the mystacial pad. The latter is invariably comprised of whiskers aligned in regular rows and may be of particular significance when trying to understand the suborbital pits in mammal-like reptiles. They found that in each row, whiskers were oriented perpendicular to the animal’s rostrocaudal axis; all shared a specific dorsoventral orientation. In all species, progressing from rostral to caudal in any vibrissal row, there was a precisely exponential increase in whisker length. They demonstrate that there are two distinctive separations of function, as a distance detector and as a high-resolution tactile sensor. Mystacial macrovibrissae were critically involved in spatial tasks, the lateral orientation of the mystacial whiskers maximizes the search space for distance information. Each whisker appeared to act as a lever-like transducer, providing information as to whether or not—but not where—an individual vibrissa had been deflected. Patterns of whisking movements are highly complex (achieved by finely controlled muscular activity), for instance, the authors suggest that the rostrocaudal whisking movement might be considered a distance scanning behavior, as it allows a row of whiskers—all of different length—to move through an overlapping segment of space. This distance detector model they propose is functionally very different from traditional concepts of whisker function, in which the mystacial whiskers were hypothesized to provide a tactile surface for recognizing objects. In contrast, it is the function of the microvibrissae located at the rostral tip and about 40 times denser than the mystacial vibrissae that appear to form a high-resolution tactile sensor that operate in a more touch-like manner. The importance of vibrissae in modern mammals may provide clues to the possible presence of sensory hairs in mammal-like reptiles and the view that a sensory function might reflect the primitive condition that appeared long before an insulative pelage had evolved. Indeed, Ruben and Jones (2000 and references therein) draw attention to very recent data that suggest that a complete insulatory covering of hair or fur may not have existed until the appearance of the earliest mammals (see below). Modern reptiles with their scaly skin may be considered a poor candidate for questions on the origin of hair. But are they? Eckhart and colleagues (2008) investigated the evolutionary history of the main components of mammalian hair, the important structural proteins cysteine-rich type I and type II keratins, also known as hard a-keratins or ‘‘hair keratins.’’ They did this by comparing the genomic loci of the human hair keratin genes with the homologous loci of the chicken, Gallus gallus, and of the green anole lizard, Anolis carolinenis. What they found was rather surprising. The genome of the chicken contained one type II hair keratin-like gene, and perhaps even more surprising the lizard genome contained two type I and four type II hair keratin-like genes. The lizard hair keratin-like genes were expressed most strongly in the digits, indicating a role in claw formation. Perhaps most significantly their data showed that cysteine-rich a-keratins are not restricted to mammals. The identification of hard keratin genes in reptiles and in birds strongly argues against the concept that hair keratins were an evolutionary innovation that occurred in the mammalian lineage after the divergence from sauropsids. The authors propose that the evolution of mammalian hair 8.4 The Evolution of Hair 205 involved the co-option of preexisting structural proteins. They suggest the following revision of hair evolution at the molecular level: The last common ancestor of all extant amniotes contained cysteine-rich a-keratins, which served in the establishment of hard non-hair epidermal structures. These genes remained functional in sauropsids, such as A. carolinensis, and were co-opted for a role in hair formation in mammals.

8.5 Cynodonts and the Origin of Modern-Day Mammals

As we saw above the mammal-like reptiles that can be said to be nearest in line to modern mammals occur in the theriodonts (Huttenlocker 2009; Huttenlocker et al. 2011). The most advanced group was the cynodonts. In the Triassic they sup- planted two other members of the theriodonts, the therocephalians, and gorgonopsians (Fig. 8.12). Cynodonts show the clearest approach to the mammalian condition. The majority were small to medium-sized. One species, Cynognathus, was large and is diagnostic of the Cynognathus Zone in the upper part of the Lower Triassic in southern Africa. With a massive skull, large canines and broad adductor chamber, all the indications are that Cynognathus would have had a powerful bite and would have been among the most formidable of predators in the early Triassic. This view is reinforced by other characteristics of the skull. The maxillary bone was expanded laterally while the mandibles were so large that they formed almost the entire lower law. Among the mammal-like reptiles, the teeth are among the most highly specialized and differentiated. Beverly Halstead spent a significant part of his early career at the Royal Dental Hospital in London and Universities of London and Oxford in research on teeth and their evolutionary history. His work has given paleontologists a much better insight into teeth and their evolutionary development in the mammal-like reptiles (see Chap. 1). The teeth in mammal-like reptiles are specialized by regions on the jaws; incisors in front for gripping and pulling, followed by stabbing tearing canines, and then a row of cheek teeth (premolars and molars) diversely functioning for seizing, crushing, grinding, or cutting food. This latter tooth structure would reach its pinnacle of development in mammals, in the tribosphenic molar (below). As we have seen the mammalian characters were gradually appearing over many millions of years in a number of impressive looking mammal-like reptiles that culminated with advanced forms among therocephalians and cynognathids. The therocephalians show a progressive evolution from featuring many primitive or ‘‘reptilian’’ traits to having increasingly advanced ‘‘mammalian’’ morphology. However, they are regarded as the sister-group of cynodonts and any similarity between the advanced therocephalians and cynodonts should probably be ascribed to convergent evolution (Sigurdsen 2006). While therocephalians went extinct, cynodonts underwent rapid diversification. Yet, if size is anything to go by, the arrival of the first true mammals may be considered somewhat anticlimactic. From the diversification of cynodonts stemmed a distinct lineage of small carnivorous or 206 8 Mammal-Like Reptiles

Fig. 8.12 Hypotheses of the cladistic relationships of eutheriodont therapsids. Strict consensus of 12 MPTs supporting a monophyletic Therocephalia (bold numbers at nodes indicate bootstrap values [50 %; consensus indices are provided next to bootstrap values). From Huttenlocker et al. (2011)

insectivorous mammals in the Late Triassic and early Jurassic. Their fossilized remains have been collected from a bone bed in Great Britain dating from the Rhaetian Stage at the end of the Triassic as well as from China. Unfortunately, the evolutionary transition from therapsid reptiles to mammals at the close of the Triassic is nowhere clearly demonstrated by well-preserved fossils. 8.6 The First Mammals 207

Fig. 8.13 The early mammal Morganucodon. a Skeletal reconstruction. From Jenkins and Parrington (1976). b Life restoration From Crompton (1968). Also see Carroll (1988)

8.6 The First Mammals

The therapsids suffered pulses of extinctions in the Late Permian but became virtually extinct by the end of the Triassic, possibly because of competition from more efficient predators, such as the thecodonts. The first true mammals, which were very small and shrew-like, appeared in the Late Triassic. The best known is Morganucodon, whose fossilized remains have been collected from a bone bed in Great Britain dating from the Rhaetian Stage at the end of the Triassic (Fig. 8.13). Kermack and colleagues (1981) described the skull of two species of Morgan- ucodon from the lower Jurassic, M. oehleri from China, and M. watsoni from Wales. The skull was no more than 3 cm long but was large relative to the body. Despite its small size, Morganucodon achieved a relatively higher brain size than any of the mammal-like reptiles and in all probability a higher metabolic rate than them too. Carroll (1988) pertinently notes that many features of the skeleton show a more or less continuous change through the advanced mammal-like reptiles and the early mammals but that the establishment of a complex suite of dental characters is coincident with what is generally considered the origin of mammals. Part of these changes exemplified in the dental characters of Morganucodon was the 208 8 Mammal-Like Reptiles development of a particularly important cheek tooth type, the tribosphenic or multifunctional molar with several points, cusps, and valleys on each tooth to perform seizing, crushing, and grinding at once. A herbivorous diet however, necessitated special development. Plant material had to be crushed to break down the tough cellulose walls and silica that causes excessive wear. Hence, the tribosphenic molar was highly modified in herbivores. The shear component is reduced in favor of excessive apposition and the crowns of the teeth develop ridges to give a more effective grinding surface to the teeth. While tooth replacement is of less of a problem to mammals but largely concerns the lower vertebrates, workers such as Kermack (1956) and Crompton (1963) have shown alternation in, e.g., therapsid dentitions and a degree of alternation in mammals. This may be an appropriate time therefore to discuss the problem of tooth replacement in reptiles before the advent of true mammals. In reptiles there is a continual succession of teeth throughout their lives with waves of replacement of teeth passing from the back of the jaws to the front as opposed to strict alternative tooth replacement. In the advanced mammal-like reptiles and primitive mammals the wave of replacement passing from front to back and the number of generations were severely reduced. So that at any one time there were no great gaps in the dentition or at the other extreme unwanted development of teeth (resulting in irregular clumps of fused teeth randomly scattered through the jaws). In contrast, in nonmammals the sequence of tooth replacement is highly complex (Edmund 1960). Edmund (1960) demonstrated that in the dentitions of almost all nonmammalian vertebrates, teeth are replaced in waves which regularly sweep through alternate tooth positions. He explained the ontogeny of these patterns of tooth replacement in terms of biological units called Zahnreihen. This resulting model of tooth replacement patterns in lower vertebrates has been accepted by nearly all workers studying tooth replacement. However, it was challenged by Osborn (1971), in particular with respect to Edmund’s explanation of the ontogeny of the patterns of tooth replacement in terms of biological units called Zahnreihen. Osborn (1971) argued that there is no unequivocal evidence, either during development or in adult animals, that Zahnreihen have any biological significance. The significance of Osborn’s findings (1971) are presented below. With respect to Zahnreihen, Osborn (1971) concluded that it had no significance in Lacerta. Rudimentary teeth were produced with varying frequency in positions 3, 5, 6, 8, 10, and 13. Contrary to the predictions of all previous theories explaining the ontogeny of tooth development in reptiles it was in these apparently random positions that the first teeth were produced. Furthermore, apart from during the first few days of embryonic dental development, it was clear that the development of a row of alternating teeth was initiated in sequence from the back to the front of the jaw to be followed by a similar sequence of development of the intervening teeth. The extraordinary phylogenetic and ontogenetic stability of the phenomenon of tooth alternation indicates that it is related to stable and therefore probably simple processes. Furthermore, because it seems to have existed in the very earliest vertebrate dentitions that have been studied, it is possible that 8.6 The First Mammals 209 alternation is achieved by a process which may be inseparable from the development of rows of teeth. On the basis of this evidence Osborn proposed a new model to explain the sequence of tooth initiation in reptiles. Based on investigations in Lacerta vivipara, Osborn (1971) was able to demonstrate the following observations on the dental development. (a) Ectomesen- chymal cells migrate anteriorly through the developing jaws initiating a reaction from the oral ectoderm. (b) The oral ectoderm develops competence to react to the ecto-mesenchyme in three stages. First, it generates abortive clumps of ectodermal cells; second, it becomes capable of inducing the adjacent ectomesenchymal cells to form dentine; and third, it becomes capable of laying down enamel. (c) At all times the dental lamina has the potential of taking part in tooth development according to the regional competence achieved. (d) Developing tooth germs produce a condition which inhibits tooth development around them. This can be argued as follows. It has been speculated above that, once it has achieved regional competence, the dental lamina is at all times capable of taking part in tooth development but that developing teeth are surrounded by a region which tempo- rarily inhibits the initiation of further tooth development. This ensures that the teeth are spaced apart, for without the inhibition irregular clumps of fused teeth would be randomly scattered through the jaws. Osborn adds that this argument can be extended to include the possible phylogenetic precursors of teeth, the odontodes covering the carapaces of ostracoderms, and the scales of fish. In other words, the separation between the elements contained on surfaces which produce either odontodes, scales, or teeth is ensured by the same embryological process. If this is true then the alternating replacement of such mineralized bodies when arranged in single rows along a narrow gill arch (as in dentitions) is inevitable and concomitant with the evolution of such rows. Thus, the first tooth row is initiated along an embryological gradient related to the migration of ectomesenchymal cells. The teeth are separated because each is surrounded by a sphere of inhibition. Due to this same inhibition phenomenon, which may be related to all equivalent mineralized bodies, the second tooth row develops in the interspaces of the first tooth row and the alternation has been established. It is not surprising that some degree of alternation persists in all dentitions in which teeth are replaced. Using these assumptions, Osborn (1971) proposed that it was possible to explain all stages in the development of the wave replacement of alternate teeth in Lacerta vivipara and that it was also possible to explain previous observations on the ontogeny of reptilian dentitions. The sphere of inhibition which surrounds developing teeth is particularly important because it ensures that developing teeth are evenly spaced through the jaw. He further argued that the wave replacement of alternate teeth is an automatic sequel to this and is of only secondary functional significance. 210 8 Mammal-Like Reptiles

8.7 Mammals Emerge into the Light

Why did the mammal-like reptiles despite suffering a series of serious extinctions and competitive setbacks against the dominant dinosaurs in particular (see below) succeed to become ultimately, in Homo sapiens, the most highly evolved and successful vertebrate species on the planet? Simpson (1950) sums up in his book the reasons for the success of the mammals during their long history as well as the place of humans in this incredible evolutionary journey. Mammals had developed numerous interrelated anatomical and physiological characters not present in the other reptile groups that in the course of time were the basis for the development of new ways of life never achieved by the reptiles but such potent features nevertheless began in the latter group. The mammal-like reptiles exploited unusual possibilities for diversification, the outcome of which was that they would overshadow the achievements of all the other reptiles put together. Among the many developments within the reptile-mammal line, live birth and parental care must be given high place. As Simpson eloquently put it, eggs were no longer deposited and left at the mercy of an egg-hungry world nor even given such lesser care as external (as in birds) or internal (as in some reptiles) incubation. The developing embryo was protected within the mother’s body and nourished during the incubation as well as for a time after by milk from the mothers body (the milk-producing mammary glands being yet another evolutionary novelty). The novelties that arose are connected to one vital change in mammals, endothermy, which would necessitate a series of interrelated developments that would define the mammalian condition. We will briefly look at how these dramatic and momentous changes occurred. Mammals were able to sustain higher and more sustained levels of activity by maintaining a constant body temperature related to physiological means such as a constant level of metabolism. Related to this, the bones of the skeleton grow in a way that maintain firm, bony joints even while they are growing and ultimately knit firmly at size characteristics specific for each kind of mammal. These arrangements provide mechanical strength greater than, e.g., in reptiles which led to quadrupedalism and allowing the legs to be drawn directly under the body and for the body to be held well off the ground with minimal energy expenditure. This was a major advance from the typical sprawling gait of basic reptiles. Sustained activity and a constant metabolism require a high and regular food intake and efficient utilization. Early mammals such as the very small, shrew-like Morgan- ucodon resembled modern insectivores, which suggests that their activities and physiology were also similar. Their small size suggests that they had a fairly high metabolic rate (Carroll 1988). To conserve energy that would have been rapidly lost because of their small size they would probably have had to been covered by an insulative coat of hair (below). These characteristics meant they were able to avoid competition with the dominant carnivores at the time, the dinosaurs, by in all probability being nocturnal and secreting themselves away from the dinosaurs during the day. Ironically, it was the unchallengeable magnificence of the 8.7 Mammals Emerge into the Light 211 dinosaurs that led to the success of mammals. Unlike, their mammal-like ancestors, who came close to extinction on a number of occasions, they had learnt to avoid the dinosaurs at all costs. One of the problems associated with small size and high metabolic rate is rapid heat loss. Small extant mammals are only able to maintain their body heat by effective insulation. Unfortunately, no traces of hair have been found associated with morganucodontid fossilized material and the search for evidence involves indirect evidence. For instance, there is strong circumstantial evidence implicit in numerous specializations in Morganucodon mentioned above and particularly connected with the heterodont dentition and in improvements in the mechanics of locomotion such as by bringing the plane of action of the limbs close to the trunk. It is clear therefore that the mammalian condition evolved as a complex, interrelated system. Near-endothermy, i.e., the occurrence of a high metabolic rate, may have been achieved in the late therapsids (Brink 1956), and may provide clues to when full endothermy was achieved and to the evolution of hair in mammals. Above we touched upon why the presence of respiratory turbinates in fossil animals may be important in establishing endothermy? Ruben and Jones (2000) proposed that the presence of incipient respiratory turbinates in Permian therocephalians provides an early and essential roadmap for understanding the evolution of endothermy and hair in mammalian ancestors. The authors consider that respiratory turbinates alone have a strong functional association with endothermy and override previously adduced lines of evidence. As Ruben and Jones (2000 and references therein) indicate respiratory turbinates occur in greater than 99 % of all extant birds and mammals. They facilitate an intermittent countercurrent exchange of respiratory heat and water between respired air and the moist, epithelial linings of the turbinates, thereby significantly reducing respiratory water and heat loss that would otherwise be associated with the high rates of lung ventilation associated with mammalian and avian endothermy. The authors show that rudimentary anterolateral rostral ridges for support of respiratory turbinates first appear in some large, late Paleozoic therocephalian therapsids, e.g., Glanosuchus, a wolf-like pristerognathid, some 40–50 million years prior to the origin of the Mammalia. Many therocephalian therapsids appear to have been active, dog-, bear-, or lion- size (20–100 kg) carnivores that inhabited regions with subtropical to tropical climates and are considered to have probably been inertial homeotherms that were unlikely to have required a hairy or furry covering for thermoregulatory purposes. On the other hand, Lower-Middle Triassic cynodont therapsids, e.g., Thrinaxodon as well as the earliest mammals, e.g., Morganucodon, appear to have possessed respiratory turbinate development similar to that of extant mammals and are hypothesized to have possessed an insulative covering. Yet other recent data (Thiessen 1992) suggest that a complete insulatory covering of hair or fur may not have existed until the appearance of the earliest mammals. This is inferred from fossil evidence indicating that Harderian glands, structures associated with grooming and maintenance of insulatory pelage in extant mammals, only occurred in the earliest very small mammals such as Morganucodon. As Crompton et al. 212 8 Mammal-Like Reptiles

(1978) proposed an insulative fur covering might have become a necessity only when taxa in the therapsid-mammal lineage became extremely small (i.e., total length of 50 mm) and nocturnal. Morganucodontids have been thoroughly studied and considered by many authors over the years as representatives of the earliest mammals (Kielan-Jaw- orowska 2004). However, it is not without controversy, principally whether or not to classify Morganucodon as a mammal at all. Some authors limit the term mammal to the crown group, Mammalia, which would not include Morganucodon. Mammalia would comprise the most recent common ancestor of living monotremes (echidnas and platypuses) and therian mammals (marsupials and placentals) and all descendants of that ancestor (Rowe 1988), a decision based on ancestry. To accommodate Morganucodon and some related taxa falling outside his crown group, Rowe defined the Mammaliaformes as comprising the last common ancestor of the Morganucodontidae and Mammalia and all its descendants. However, most workers have accepted that mammals, as a group, are defined by the possession of a special, secondarily evolved jaw joint between the dentary and the squamosal bones (albeit in morganucodontids showing transitional stages of both the dentary-squamosal and articular-quadrate jaw joints), which has replaced the primitive reptilian one between the articular and quadrate bones in all modern mammalian groups, changes that mark the evolution of the mammalian ear that commenced in certain advanced cynodonts although the complete separation of the middle ear occurs later in the Cretaceous (see below). Another vital character in this classification scheme is possession of the tribosphenic or multicusped molar (discussed above) and its multifunctional role in mammalian evolution (Kielan-Jaworowska 1981, 2004). One point is worth noting with respect to dental characteristics. Diagnostic dental characters are not consistent in all morganucodontids, which some authors suggest warrant recognition of separate species and point to a diverse assemblage of morganucodontids forming part of the base of a prototherian radiation. Regardless of whether morganucodontids are classified as mammaliformes or mammals, numerous characters of the group suggest that their appearance was almost certainly before the eutherian-metatherium split (among placentals, marsupials, and monotremes). Among the principle features that characterizes the mammalian infraclass Eutheria, to which our own species belongs, is the presence of a placenta, which facilitates exchange of nutrients and wastes between the blood of the mother and that of the fetus, which separates it from the marsupials and monotremes. However, when exactly the split occurred was very much a matter of speculation, until recently. Since the late 1990s, Chinese fossils have been revo- lutionizing our views on evolutionary morphological development and lineage, perhaps none more so than on the question of dinosaurs and the origin of birds (Volume 2). To some extent mammals took a backseat but that changed in 2002. Ji and colleagues (2002) reported discovery of the skeleton of a eutherian mammal from the Lower Cretaceous Yixian Formation of northeastern China (the same horizon that has yielded diverse fossil vertebrates, invertebrates, and plants). 8.7 Mammals Emerge into the Light 213

Fig. 8.14 Eomaia scansoria (Chinese Academy of Geological Sciences (CAGS) 01-IG-1a, b; holotype). a Fur halo preserved around the skeleton (01-IG-1a, many structures not represented on this slab are preserved on the counter-part 01-IG-1b, not illustrated). b Identification of major skeletal structures of Eomaia. c Reconstruction of Eomaia as an agile animal, capable of climbing on uneven substrates and branch walking. as astragalus; c canine; c1–c7 cervical vertebrae 1–7; ca1–ca25 caudal vertebrae 1–25; ch chevron (caudal haemal arch); cl clavicle; cm calcaneum; cp1–9 carpals 1–9; cr1–11 costal cartilages 1–11; dn dentary; dpc deltopectoral crest; en entocuneiform; ep epipubis; f frontal; fe femur; fi fibula; hu humerus; I1–4 lower incisors 1–4; il ilium; im ischium; is infraspinous fossa of scapula; j jugal; la lacrimal; lb lambdoidal crest; L1– L6 lumbar vertebrae 1–6; mb manubrium sterni; mp1–5 metacarpals 1–5; ms masseteric fossa; mt1–mt5, metatarsals 1–5; mx maxillary; n nasal; p parietal; pa ossified patella; pb pubis; phi intermediate phalanges; php1–5 proximal phalanges 1–5; px premaxillary; ra radius; r1–r13 thoracic ribs 1–13; s1, s2, sacral vertebrae 1 and 2; sa sagittal crest; sc scapula; sq squamosal; ss supraspinous fossa of scapula; stb1–5 sternebrae 1–5 (sternebra 5 is the xiphoid); ti tibia; t1–t13 thoracic vertebrae 1–13; ug1–5 ungual claws 1–5; ul ulna. Reprinted permission of Ji et al. (2002). Courtesy of Macmillan Publishers, Ltd 214 8 Mammal-Like Reptiles

Fig. 8.15 Eomaia scansoria dentition and mandible (composite reconstruction). a Lower P3–M3 (right, lingual view). b Lower P3–M3 (right, labial view). c Upper dentition (incomplete) and mandible (right, labial view). d Mandible (left, lingual view). ap angular process; C and c upper and lower canine; co coronoid process of mandible; dc dentary condyle (articular process); etd entoconid; f cuspule f (anterolabial cingulid cuspule for interlocking); hﬂ hypoﬂexid; hyd hypoconid; hyld hypoconulid; I1–5 and I1–4 upper and lower incisors; M1–3 and M1–3 upper and lower molars; med metaconid; mf posterior (internal) foramen of mandibular canal; mks Meckel’s sulcus; ms masseteric fossa; P1–5 and P1–5 upper and lower premolars; pad paraconid; prd protoconid; ptf ptergygoid muscle fossa; sym mandibular symphysis. Reprinted permission of Ji et al. (2002). Courtesy of Macmillan Publishers, Ltd 8.7 Mammals Emerge into the Light 215

Fig. 8.16 Holotype specimen of Juramaia sinensis, Beijing Museum of Natural History (BMNH) PM1343B. a, b Specimen photograph and morphological identiﬁcation. c Restoration of the partly preserved skeleton and skull. d Restoration of hand (ventral view; alignment of incomplete and scattered carpals is conjectural). Abbreviations. ac acromion (scapula); ag angular process (dentary); C, c upper or lower canine; ca carpals; cl clavicle; cod coronoid (dentary); cos coracoid process (scapula); cv1–7, cervical vertebrae 1–7; dc dentary condyle; ecc ectepicondyle; enf entepicondylar foramen; hh humeral head; ht humeral trochlea; I1–5 upper incisors 1–5; Ju jugal; M, m upper or lower molar; manus, hand; mc1–5 metacarpals 1–5; oc occipital condyles; ol olecranon process; P1–5 upper premolars 1–5; ph phalanges; r1–13; thoracic ribs 1–13; ra radius; sn semilunar notch (ulna); sp scapular spine; tv1–13, thoracic vertebrae 1–13; ul ulna. Reprinted permission of Luo et al. (2011). Courtesy of Macmillan Publishers, Ltd 216 8 Mammal-Like Reptiles

Fig. 8.17 The increased enechelon postvallum shearing of upper molars in the earliest eutherians in contrast to metatherians that lack a strongly developed postvallum shearing by metacingulum, except for the Late Cretaceous Pediomys. Nodes 1 Cladotheria 2 Boreosphenida 3 crown Theria 4 Eutheria (including Placentalia) and 5 Metatheria (including Marsupialia). Reprinted permission of Luo et al. (2011). Courtesy of Macmillan Publishers, Ltd

The fossil, Eomaia, is dated to 125 million years ago and is 10 cm in length and virtually complete. It possesses numerous skeletal evolutionarily advanced features that distinguish Eomaia from currently known eutherians, the earliest-known metatherians (including marsupials), and nontribosphenic therians. Hairs are preserved as carbonized filaments and impressions around most of the body, although their traces are thin on the tail (Fig. 8.14). At the time the paper was written it was the first record of fossilized hairs in the Cretaceous or earlier, the previously earliest reported were in Tertiary placentals and multituberculate mammals. The pelage appears to have both guard hairs and a denser layer of under hairs close to the body surface. On the basis of hundreds of characters sampled from all major Mesozoic mammal clades and principal eutherian families of the Cretaceous, Eomaia was placed at the root of the eutherian tree with Murtoilestes and Pro- kennalestes, all three taxa considered closer to living placentals than to living marsupials. Eomaia is placed in Eutheria by numerous advanced characters of the wrist and ankle (Fig. 8.14) and the dentition (Fig. 8.15). It also is the first to show 8.7 Mammals Emerge into the Light 217

Fig. 8.18 Dentition and mandible of Maotherium asiaticus (HGM 41H-III-0321; holotype) and interpretation of its middle ear based on homoplastic evolution of the mammalian middle ear by paedomorphic retention of ossified Meckel’s cartilage among extinct clades of mammaliaforms. a The preserved Meckel’s cartilage (ossified) and its attachment to the mandible. b Interpretive restoration of the middle ear (dashed line) and its attachment to the mandible. Red arrow, mid- length curvature of Meckel’s cartilage (identical to eutriconodonts); blue arrow, mediolateral separation; the ectotympanic and malleus are conjectural, based on the middle ears of multituberculates and Yanoconodon. After Ji et al. (2009) limb and foot features that are known only from climbing and tree-living (scansorial) extant mammals, in contrast to the terrestrial or cursorial features of other Cretaceous eutherians (Clemens 1973). It is another example of early eutherian diversification potentials with respect to locomotory adaptations. It is always difficult when a fossil such as Eomaia, which incidentally is Greek for dawn mother, should be toppled from its place as the oldest or mother of all eutherian mammals, especially given that the primacy is explicit in the name. The problem came home when Lou et al. (2011) reported the discovery of a new eutherian from the Jurassic of China, 160 MYA, Juramaia sinensis, which extends the first appearance of the eutherian–placental clade by a considerable gap of about 35 Myr from the previous record of Eomaia. Phylogenetically, Juramaia sinensis is one of the basal-most eutherians and it is currently the earliest-known eutherian (Fig. 8.16). It provides crucial inferences on the ancestral features of all eutherians including that of the dentition. Juramaia weighed about 15–17 g, was covered in hair, and its tooth morphology indicates it was an insectivore. It shows an identical dental formula with Eomaia (I5–C1–P5–M3/I4–C1–P5–M3 and an increased postvallum shear surface consistent with the earliest eutherians, in contrast to metatheriums (Fig. 8.17). Compared with fossil mammals of the Early Cretaceous Yixian formation, phalangeal indices of Juramaia are between Eomaia, a scansorial mammal, and the eutriconodont Jeholodens jenkinsi, which is interpreted to be terrestrial. However, the authors consider that Juramaia’s habitat preferences would have been similar to that of Eomaia and Cretaceous and Early Cenozoic 218 8 Mammal-Like Reptiles metatherians and to living scansorial or arboreal didelphids. Other important evolutionary developments were also occurring in this critical period of earth’s history. Maotherium asiaticus from the early Cretaceous Yixian Formation of Laioning, China shows a number of ancestral skeletal features of living therians, among the most important of which is the disconnection of the middle ear (ossicles) from the mandible by reabsorption of Meckel’s cartilage, which sheds light on the evolution of the definitive mammalian middle ear (Ji et al. 2009) (Fig. 8.18). Fastovsky and Weishample (1996) note that these tiny, insect-eating creatures, the earliest mammals, were a far cry from the large, carnivorous, ground-dwelling and diurnal dinosaurs and that they do not seem to have been competitive with the earliest dinosaurs but that rather each group was behaving in different ways. Thus, the small size of these early mammals, although anticlimactic from their large wolf-, bear-, and lion-sized forebears, may have been what ultimately led to their survival against the dinosaurs. As mentioned above discretion seems to have been the better part of valor.

References

Allin FF (1975) Evolution of the mammalian middle ear. J Morphol 147:403–438 Bramwell CD, Fellgett PB (1973) Thermal regulation in sail lizards. Nature 242:203–205 Brecht M, Preilowski B, Merzenich MM (1997) Functional architecture of the mystacial vibrissae. Behav Brain Res 84:81–97 Brink AS (1963) On Aneugomphius ictidoceps Broom and Robinson. Palaeontol Afr 1956:97–115 Carroll RL (1988) Vertebrate paleontology and evolution. WH Freeman, New York Clemens WA Jr (1973) Fossil mammals of the type lance formation wyoming—part III. Eutheria and summary. Univ Calif Publ Geol Sci 94:1–102 Colbert EH (1955) Evolution of the vertebrates. John Wiley & Sons, Inc., New York Crompton AW (1963) Tooth replacement in the cynodont Thrinaxodon liorhinus seeley. Ann S Afr Mus 46: 479–521 Crompton AW, Taylor CR, Jagger JA (1978) Evolution of homeothermy in mammals. Nature 272:333–336 Eckhart L, Valle LD, Jaeger K, Ballaun C, Szabo S et al (2008) Identiﬁcation of reptilian genes encoding hair keratin-like proteins suggests a new scenario for the evolutionary origin of hair. Proc Natl Acad Sci 105:18419–18423. doi:10.1073/pnas.0805154105 Edmund AG (1960) Tooth replacement phenomena in the lower vertebrates. Contrib Life Sci Div Roy Ontario Mus 52: 1–90 Fastovsky DE, Weishhampel DB (1996) The evolution and extinction of the dinosaurs. Cambridge University Press, Cambridge Huttenlocker AK (2009) An investigation into the cladistic relationships and monophyly of therocephalian therapsids (Amniota: Synapsida). Zool J Linn Soc 157:865–891. doi:10.1111/j. 1096-3642.2009.00538.x Huttenlocker AK, Sidor CA, Smith RMH (2011) A new specimen of Promoschorhynchus (Therapsida: Therocephalia: Akidnognathidae) from the lower Triassic of South Africa and its implications for theriodont survivorship across the permo-triassic boundary. J Vertebr Paleontol 31(2):405–421 References 219

Ivakhnenko MF (2011) Permian and triassic therocephals (eutherapsida) of eastern Europe. Paleontol J 45: 981–1144. doi: 10.1134/S0031030111090012 Ji Q, Luo Z-X, Yuan C-X, Wible JR, Zhang J-P, Georgi JA (2002) The earliest known eutherian mammal. Nature 416:816–822 Ji Q, Luo Z-X, Zhang X, Yuan C-X, Xu L (2009) Evolutionary development of the middle ear in mesozoic therian mammals. Science 326:278–281 Kemp TS (1982) Mammal-like reptiles and the origin of mammals. Academic Press, London Kermack KA (1956) Tooth replacement in mammal-like reptiles of the suborders gorgonopsia and therocephalia. Philos Trans R Soc London B 240:95–133 Kermack KA, Mussett F, Rigney HW (1981) The skull of Morganucodon. Zool J Linn Soc 71:1–158. doi:10.1111/j.1096-3642.1981.tb01127.x Kielan-Jaworowska Z (1981) Evolution of the therian mammals in the Late Cretaceous of Asia. Part IV. Skull structure in Kennalestes and Asioryctes. Palaeontol Pol 42:25–78 Kielan-Jaworowska Z, Cifelli RL, Luo Z-X (2004) Mammals from the age of dinosaurs: origins, evolution, and structure. Columbia University Press, New York Luo Z-X, Yuan C-X, Meng Q-J, Ji Q (2011) A Jurassic eutherian mammal and divergence of marsupials and placentals. Nature 476:442–445 Osborn JW (1971) The ontogeny of tooth succession in Lacerta vivipara Jacquin (1787). Proc B R Soc Lond B 179:261–289 Romer AS (1966) Vertebrate paleontology, 3rd edn. University of Chicago Press, Chicago Rowe T (1988) Deﬁnition, diagnosis and origin of Mammalia. J Vertebr Paleontol 8:241–264 Ruben JA, Jones TD (2000) Selective factors associated with the origin of fur and feathers. Am Zool 40:585–596 Simpsom GG (1950) The meaning of evolution. Yale University Press, New Haven Tatarinov LP (1994) On the preservation of rudimentary rostral tubular complex of crossopterygians in theriodonts and on possible development of the electroreceptor systems in some members of this group. Doklady Akademii Nauk 338:278–281 Thiessen DD (1992) Function of the Harderian glands in the Mongolian gerbil (Meriones unguiculatus). In: Webb SM, Hoffmann RA, Puig-Domingo ML, Reiter RJ (eds) Harderian glands. Springer, Berlin, pp 127–140 Sigurdsen T (2006) New features of the snout and orbit of a therocephalian therapsid from South Africa. Acta Palaeontol Pol 51:63–75 Chapter 9 Reptiles Return to the Sea

9.1 Problems of Developing Fresh Sea-Legs

At the time that the mammal-like reptiles were making their mark on land and the thecodonts were giving rise to novel forms culminating in the aptly named ‘‘ruling reptiles,’’ the dinosaurs, other advanced groups of terrestrial reptiles did a complete about-turn with an evolutionary journey back to the sea. Some of the proposals for this dramatic occurrence are severe competition on land for food resources on the one hand and the lure of a plentiful supply of fish in the sea on the other (Lingham-Soliar 2003). We have seen earlier in numerous publications by Jennifer Clack, Per Ahlberg, and others how in the first terrestrial pioneers fleshy fins had to change into robust limbs to make the transition from water to land possible (Hall 2007). Now, some 200 million years later when some of the descendants of these early colonizers chose to return to the sea, the problems were reversed particularly with respect to locomotion—robust legs had to be converted back into fleshy fins. With differences in mechanical loading bone elongation had to be reduced, muscle insertion points changed, and rounded cross-sections, particularly in the digits, had to become highly flattened. These were daunting challenges that had to be met if this life-changing swop was to occur and for these vertebrates to successfully recol- onize the oceans and compete with animals for which the seas had been home for hundreds of millions of years. And meet the challenges they did, with the greatest style and efficiency to become among the most amazing animals to inhabit the planet. They would be grouped together under the term of ‘‘secondarily adapted marine vertebrates.’’ The main reptilian invasion of the seas occurred during the Triassic and extended into the Cretaceous. It included three major groups of marine reptiles of the Mesozoic era. Two groups of marine reptiles during the Triassic and Jurassic periods of the Mesozoic have captured the imagination of the public and scientists alike. They were the ichthyosaurs and plesiosaurs. Another group, less known to the public but dominating the close of the Mesozoic when the former two groups had either become extinct or were fading, were the mosasaurs, which grew to gigantic proportions

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 221 DOI: 10.1007/978-3-642-53748-6_9, Ó Springer-Verlag Berlin Heidelberg 2014 222 9 Reptiles Return to the Sea

Fig. 9.1 In his work The Book of the Great Sea-Dragons, Hawkins (1840) entertained a very idiosyncratic view of the marine reptiles, seeing them as monstrous creations of the devil

(Lingham-Soliar 1995a). One way or another they excited the imagination of the public, and less frequently regarded as monsters created by the devil (Fig. 9.1). Reptilian legs had to transform back into ﬁns or paddles for steering and balance in the water, and in one group for active locomotion. The tail was readapted to locomotion—with the exception of plesiosaurs in which they were highly reduced. Quite remarkably, each of the three groups of marine reptiles would be identiﬁed by highly distinctive forms of locomotion that would only much later be emulated by extant groups of vertebrates. In a general sense, these three groups of secondarily adapted marine reptiles provide a window to the genetic depth of the ancient reptiles. We will start with the plesiosaurs, a group with not even a remote equivalent among present day vertebrates.

9.2 A Serpent Threaded Through a Turtle

Plesiosaur ancestry is still rather poorly known. Among nearest semiaquatic ancestor are the nothosaurs, with closely related forms such as Pistosaurus and Ceresiosaurus (Lingham-Soliar 1995b) (Figs. 9.2 and 9.3). Plesiosaurs included two major groups, the plesiosaurids, which possessed a small head atop of a long neck that arose from a bulky ovoid body terminating in a short, fat tail and the pliosaurids, which differed from plesiosaurs essentially by a massive head armed 9.2 A Serpent Threaded Through a Turtle 223

Fig. 9.2 Cast of the fossil of the nothosaur Ceresiosaurus in the geological museum, University of Tubingen. The original is at the University of Zurich. The genus reached 4 m in length. Note, the ‘‘juveniles’’ or perhaps embryos close to the specimen. Embryos, as in ichthyosaurs, would have anticipated the fully marine lifestyle of nothosaur descendent, the plesiosaurs

Fig. 9.3 Phylogeny of Sauropterygia, nothosaurus, and plesiosaurs. From Lingham-Soliar (1995b) 224 9 Reptiles Return to the Sea

Fig. 9.4 Four specially constructed experimental models by the author in the Geological and Palaeontological Museum, Tubingen (GPIT) based on Plesiosaurus brachypterygius GPIT 477/1/ 1, Peloneustes philarchus GPIT 1754/3, Cryptocleidus eurymerus GPIT 1754/1, and Liopleur- odon ferox GPIT 1754/2 (Lingham-Soliar 2000) with a battery of formidable teeth, at the end of a short stocky neck, and a barrel- shaped body—a group that included some of the largest carnivores ever to inhabit the oceans. Among the strangest-shaped plesiosaurids were the elasmosaurs. The head was especially small and the neck extremely long and snake-like emerging from a bulky somewhat dorso-ventrally ﬂattened body that terminated in a short tail. So strange was the shape of the plesiosaurids that the notable natural historian of the nineteenth century, the Reverend W. D. Conybeare, described Plesiosaurus as akin to a serpent threaded through a turtle and noted that ‘‘in its motion this animal must have resembled the turtles more than any other’’ (Conybeare 1824). In fact he probably did not realize how close to the truth he had come with respect to the animal’s locomotion, which would only be realized some 135 layers later. Pliosaurids showed the same distinctive body shape among the different species (Figs. 9.4 and 9.5). They reached lengths of at least 15 m and weights in excess of 50 tons. The limbs became greatly elongated, compressed, and tapered while the tail was highly reduced and clearly ineffective as a propeller. The latter feature in both plesiosaurids and pliosaurids presented a puzzle as to how they propelled themselves through the waters, which has been the subject of much debate even to the present day. 9.2 A Serpent Threaded Through a Turtle 225

Fig. 9.5 Three mounted plesiosaur skeletons (GPIT). a Liopleurodon ferox. b Peloneustes philarchus. c Cryptocleidus eurymerus. Note, in the latter the limbs are incorrectly mounted (back to front) and should be reversed

By the mid-1920s paleontologists believed they had cracked the problem of plesiosaur locomotion. In 1924, DMS Watson published a classic paper on the functional anatomy of the elasmosaurid shoulder girdle and forelimb. He was able to identify the muscles from tell-tale scars on the bones and to document the changes that took place during the evolution of the bones of the shoulder girdle. In the elasmosaurs, he showed that the scapula developed into a large flat bone that had much the same area as the coracoids from which he was able to show that the muscles anterior and posterior to the glenoid articulation of the forelimb were comparable. He proposed that the two pairs of flippers were moved back and forth like oars and by engaging the paddles on just one side complex twisting and turning movements could be achieved. Watson’s hypothesis prevailed for three decades until Beverly Halstead (Tarlo 1958) proposed an innovative new theory that revolutionized how the locomotion of these animals would forever be perceived. On the basis of an unusual bone 226 9 Reptiles Return to the Sea identified as a scapula, Halstead reconstructed the pectoral musculature of a giant pliosaur. From the evidence of skin impressions, he considered that the limbs were hydrofoils and not oars. He described their movements thus, ‘‘the most effective swimming stroke is one in which the forelimb is driven downwards through the water with its dorsal surface facing forwards and slightly upwards, from a position in which the limb is somewhat above the horizontal and a little forward from the glenoid. At the end of this stroke the limb is adducted backwards by the coraco- brachialis muscle, and drawn forwards ready for the next swimming stroke by the deltoid, scapulohumeralis anterior, and subscapularis muscles.’’ The motion inferred is that of underwater flight and was groundbreaking. Although Halstead worked out the wing-like action from his detailed analysis of the forelimbs, he still viewed the hind limbs in the traditional rowing model, believing that they were adducted into the body to produce the main thrust of the swimming stroke. It paved the way for Jane Robinson (1975) who published the first comprehensive account of the locomotion of plesiosaurs, establishing a strong case for subaqueous flying among plesiosaurs. In contrast to Halstead, she proposed that there was no evidence to consider that the fore and hind limbs functioned differently. In a follow-up paper (Robinson 1977), she carefully discussed the functional differences between paddles, oars, and hydrofoils. She envisaged a propulsive force on both the upstroke as well as the downstroke. However, one of the main problems associated with this interpretation was that the musculature for the upstroke was relatively poorly developed. This became known as the four wing problem, which was specifically considered by (Frey and Riess 1987). They proposed rather ingeniously that ‘‘the wing beat cycle needs an alternation of up- and-down stroke of the front and hind limbs; one pair moves ventrally in the downstroke providing thrust: the other pair is supinated into the waterflow with a positive angle of attack. This pair is swept up passively, reaching the dorsal extreme of the wingbeat cycle when the other wing pair is at the ventral extreme: the downstroke is then initiated by pronation of the wing.’’ While this apparently accounted for the reduced musculature for the recovery stroke in particular for the posterior limbs, Michael Taylor (in a personal communication in Halstead 1989) suggested that vortices from the forelimbs would create turbulence under same phase relations thereby increasing drag on the hindlimbs. He proposed a modified view of phase differences between the fore- and hind-limbs (see below). Godfrey (1984) likened plesiosaur swimming to that of sea lions. He envisaged that the limbs were involved in equal flight and rowing strokes, a hypothesis that was favored by Lingham-Soliar (2000). BH Newman filmed sea lions swimming at London Zoo and came up with what he saw as a solution to the four wing problem that would get around the problems inherent in the hypotheses of (Tarsitano and Riess 1982) and (Frey and Riess 1987). Newman (see Halstead 1989) proposed that there was a general assumption that motion in a fluid by an organism occurred in a straight line. However, he countered this assumption by proposing that in most instances aquatic tetrapods move in a vertically undulating or wave-like pathway. 9.2 A Serpent Threaded Through a Turtle 227

Fig. 9.6 Short-necked plesiosaur (pliosaur) swimming cycle. a beginning of swimming stroke, shape of hydrofoil provides lift; b–d hydrofoil pulled downwards, with dorsal surface facing forward and upward lift translated into forward and upward thrust, body moves forward and upward, flow of water oblique across body (shown by long arrows); e at end of power strome limb rotates to provide less resistance, vortices weaken; f–h hydrofoil rotates into horizontal plate, vortices washed off; i–k body sinks through water, hydrofoil raised passively in recovery stroke, pulled forward by action of dorsal girdle musculature to correct position for next power stroke (modified from a sketch by B.H. Newman). Note: The body axis retains its horizontal orientation in relation to the flow of water. From Halstead (1989)

It made all calculations easier in such a case because friction of water particles is never along the entire length but is oblique. This he proposed is the most efficient way to move through water, i.e., with the main downward power stroke of the hydrofoil limbs, the body will move forward and up through the water. At the end of the power stroke, the vortices weaken and as the vertical movement of the body lessens, the flow of water automatically rotates the limb into the horizontal plane and the vortices are washed off. The animal then begins to sink so that the water flow passively lifts the limb into the correct position for the following swimming power stroke. At no point does the leeward turbulence of the forelimbs interfere with the action of the hind limbs. Thus, as Halstead (1989) suggested, it seemed that Newman’s hypothesis would eliminate the problems in Frey and Riess’ model (Fig. 9.6). Nevertheless, Halstead (1989) pointed out that the various hypotheses still left unexplained the disparity in size between the fore and hind limbs. The long-necked 228 9 Reptiles Return to the Sea plesiosaurids have longer forelimbs and the short-necked pliosaurids longer hindlimbs. While I was on Royal Society of London post-doctoral fellowships to the University of Tubingen (1993–1995), which houses some of the finest plesiosaur specimens in the world, it seemed inevitable that the problem of plesiosaur underwater flight would rear its head (Lingham-Soliar 2000). Flight depends on lift, the same principle that helps propel birds in air and penguins, marine turtles, and sea lions in water. Lift is gained from the natural aerofoil shape of the wing: long, broad but tapering at the end. In cross-section, the wing is cambered above and concave below. The air/water traveling along the wing’s upper surface has to follow a longer path to the trailing edge than that traveling along the lower surface. The faster the fluid travels, the lower the pressure. So the flow below the wing exerts a greater pressure on the wing than the flow above, producing lift. Animals may use these characteristics as well as certain natural conditions such as thermals to glide through the air without expending much energy. However, in most other situations in order to remain air borne and travel from one point to another they do this by flapping flight. This is achieved by powerful adductor muscles and less powerful abductor muscles that control the wings. As we have seen, the theory of four-winged flight in plesiosaurs has been dogged by the undisputed less developed musculature for the downward power stroke and even less-developed for upward recovery stroke of the posterior limbs. This was at the heart of complex hydrodynamic proposals over the years. The traditional treatment of all four plesiosaur limbs as identical functional structures in the four wing flight hypothesis was challenged by Lingham-Soliar (2000), upon which much of the discussion below is based. One of the problems for animals that fly or swim is to avoid falling or sinking. Although the problem is greater in air because denser water is capable of contributing to an animal’s buoyancy, the problem of staying afloat is still serious in water particularly in large animals. For instance, tuna have to swim continuously to stay afloat and thunniform sharks such as Carcharodon carcarias need enormous low-density livers (Chap. 4), and large air-breathing ichthyosaurs had large flattened limbs that served as hydrofoils. What of the plesiosaurs? Contrary to previous studies, Lingham-Soliar (2000) showed that there were significant differences between the dynamic morphology of anterior and posterior limbs in plesiosaurs (Figs. 9.7 and 9.8). The anterior limbs are swept back as in swallow’s wings—the posterior ones are relatively straighter. Research has shown that the swept-back or crescent shape is dynamically more efficient than the straight wing for flight. Unlike other underwater fliers (penguins, marine turtles, sea lions), in plesiosaurs there is no elbow or wrist-joint in the limbs. Hence, simulta- neous delicate changes in pitch, direction, etc., during flight was severely limited. Of further significance is the way in which marine animals such as sharks rise or descend in the water by elevating or lowering the leading edge of the limbs. Lift in a hydrofoil works best when it is inclined at a precise angle to the water flow—known as the angle of attack. Hence, the wing’s capacity to act simultaneously as a rudder and active propeller is impractical. In aeroplanes engineers have dealt with precisely this problem. The propeller at the front of the plane provided both thrust and lift by ingenious design involving the blade angle. The major source of lift, however, comes 9.2 A Serpent Threaded Through a Turtle 229

Fig. 9.7 Major increase in sweepback of anterior limbs a–f of six plesiosaur species compared to the posterior limbs g–l. From Lingham-Soliar (2000) from the wings. In other words, flight functions are shared not just in animals but in machines. In underwater fliers such as penguins and sea lions, the posterior limbs are not involved in flight but have another important function—because of their greater distance from the centre of balance they are ideal for rotating and maneuvring and critically important in abrupt and rapid orientation of their bodies during the pursuit of prey or escape from predators. A detailed examination of plesiosaurid and pliosaurid osteology, musculature, and hydrodynamics confirmed the hypothesis that the posterior limbs were predominantly involved as a rudder and in buoyancy. Why then would the posterior 230 9 Reptiles Return to the Sea

Fig. 9.8 The pliosaur Rhomaleosaurus victor (Stuttgart Museum, Germany). The specimen is preserved in situ and dramatically shows the impressive sweepback of the anterior limbs compared to the posterior, giving the most visual clues to different functions of the two pairs of limbs. Note, also the strong curvature of the humerus compared to the femur

limbs be so streamlined as to give the impression that it was an active ﬂight organ if it wasn’t? Penguin, turtle, and sea lion posterior limbs aren’t. The main reason is the massive size of most plesiosaurs and especially the pliosaurs. There evidently was only one solution regardless of whether the limbs were involved in active underwater ﬂight, in rowing, as a rudder or as a static structure in achieving passive lift (as the wings of an aeroplane—they would have to involve a hydrodynamic design that would reduce drag. The hydrodynamic shape and design of the anterior and posterior limbs of plesiosaurs are testament to this. In rowing, e.g., the limbs could not be lifted out of the water during recovery to reduce drag hence a hydrodynamic shape would be useful. Furthermore, it is unlikely that plesiosaurs ventured onto land, unlike penguins and sea lions and even marine turtles (to lay eggs), and consequently there was no reason for a compromise in structure, i.e., hind limbs that would also accommodate terrestrial locomotion. Furthermore, in the absence of evidence for other forms of buoyancy, the posterior hydrofoils would provide passive lift that would help prevent the animal from sinking, just as 9.2 A Serpent Threaded Through a Turtle 231

Fig. 9.9 Plesiosaur underwater flight. a Hydrofoil lift dynamics. b Anterior limb of a plesiosaur in a cycle of flight (lift) and rowing (drag based) phases. After Lingham-Soliar (2000) the wings in an airplane provide lift and are separate from the power source. All the anatomical and morphological evidence in plesiosaurs support an effective division of functions between the anterior and posterior limbs. The anterior limbs would be employed with a partial rowing phase as in sea lions (Lingham-Soliar 2000) (Fig. 9.9). The question, therefore, is why was the idea of 4-wing flight in plesiosaurs conceived and perpetuated in the literature by a number of notable workers despite numerous obstacles if it was not a realistic model for plesiosaur locomotion. The short answer is that it was not so much a realistic answer as an exciting conundrum to solve. Most beguiling was the fact that there are two pairs of ostensibly similar- looking hydrofoils that on face value looked as though they performed identical functions. The obstacles against such an interpretation such as reduced musculature necessitated the complex solutions that followed over the years. The concept of all four limbs being used as flight organs in underwater swimming was not only both novel and exciting, but was irresistible to the scientific psyche. In purely theoretical terms on paper, they might well have worked—but for a fatal flaw. These were living animals not robotic machines travelling in a straight line. Ple- siosaurs were highly predacious animals and their highest speeds would clearly have had to have occurred during pursuit of prey. One of the avoidance tactics employed by prey being pursued by a fast predator is rapid change in direction. Hence, a successful predator would have to evolve pursuit skills to deal with abrupt changes in direction while still maintaining high speeds. Here the idea of 232 9 Reptiles Return to the Sea water flow from the anterior wings aiding the flight motions of the posterior wings, staggered or not, comes apart. Because plesiosaur wings are relatively inflexible paddles (lacking elbow and wrist joints), the ability to initiate flexural changes (involving changes in direction, height, etc.), as for example in a bird wing, while at the same time maintaining active forward thrust was mechanically impossible. The assistance that the posterior wings might gain from the water flow from the anterior limbs might theoretically work for an animal traveling in a straight line but the limbs lack complex mechanisms as in, e.g., penguins for turning movements. However, a recent study proposes a solution to this problem. Carpenter et al. (2010) proposed what they say is a previously undiscussed hypothesis, i.e., that the flight movements were semi-synchronous, meaning that there was a slight lag between the anterior and posterior limbs. However, that idea was actually first proposed by Michael A. Taylor as a personal communication in Halstead (1989), as stated above. Taylor concluded that ‘‘the phase differences [between the flippers] are adjusted to promote synergy between the vortices produced by the forelimb and the action of the hindlimb’’ based as Halstead states on the assumption that the vortices from the forelimbs will create turbulence under same phase relations and increase drag on the hindlimbs. Carpenter et al. (2010) nevertheless were the first to consider that as formidable pursuit predators, plesiosaurs would need to be highly maneuverable and capable of making rapid turns. It may, however, be the undoing of their hypothesis on four-wing flight. They suggest that that plesiosaurs had anhedral flippers (negatively angled in the neutral position) similar to those of cetaceans, which facilitated a lift-based control of yaw and roll in maneuvering. In a roll, the relatively much larger limbs compared to those of cetaceans would require maximum abduction (against the body) which seems unfeasable given that the authors state earlier that posterior rotation at the glenoid is severely restricted. Even if the comparison with cetaceans and fishes were to work it would be restricted to diving pursuit or free fall because the theory would come apart in horizontal pursuit. The obvious failure lies in the fact that the power source and control surfaces, i.e., flukes and flippers, respectively, are completely separated in cetaceans. In the theory of active 4-wing swimming in plesiosaurs they are not, hence cetaceans are a poor analog for plesiosaurs. This is exacerbated by the fact that success in the theory is dependent on precise synergy between anterior and posterior limbs, i.e., a small lag between the pairs. Just as in cetaceans, the only way for direction changes to be effected in plesiosaurs is by a separation of functions—power from the anterior limbs and direction control by the posterior limbs, which is aided by their considerable distance from the centre of gravity, enabling large directional moments because of the long lever arm. Probably the most effective turning movements would be achieved, e.g., by altering the angle of attack of one or other of the posterior wings (dependent on whether a left or right turn was needed) while forward speed was effected by the powerful anterior limbs, i.e., the last thing needed was synchrony between the anterior and posterior limbs and indeed between the left and right posterior limbs. Given that the pursuit of prey and even avoiding other predators (in the case of smaller plesiosaurs) is likely to have taken up a significant part of plesiosaur’s 9.2 A Serpent Threaded Through a Turtle 233 lifestyle, when greatest speed and maneuvrability were imperative synergies between the anterior and posterior limbs would be disrupted at each abrupt turning movement. Furthermore, using right or left anterior and posterior limbs to effect sharp turns has two major disadvantages—first the anterior limb has a short relatively ineffective lever arm for turning compared to the posterior and second it means using one of the powerful anterior limbs for slowing down one side of the animal to aid turning makes no sense at all. The sensible thing is to use the powerful, anatomically and morphologically superior anterior limbs as the power source (as in penguins, marine turtles, and sea-lions) and the beautifully long- lever-armed posterior limbs to effect direct changes effortlessly. Carpenter et al. (2010) suggest a synchronous movement between the fore- and hindlimbs was supported by computer-generated animation showing semi-synchronous movements. The reservation about computer-generated animations is that frequently beauty lies in the eye of the program loader. It really depends on how many improbabilities are loaded in comparison with the probabilities: ‘‘The magic of Disney’s animations made Dumbo fly effortlessly through the air but in the words of the ‘streetwise’ crows ‘Did you ever see an elephant fly’’’(Lingham-Soliar 2000). Carpenter et al.’s criticisms of other studies seem rather ungenerous. In their abstract, they (Carpenter et al. 2010) state with respect to previous studies ‘‘almost without exception these models were based on limited analyses and limited data that failed to adequately examine joint morphology, joint kinematics or test the hypotheses.’’ While there are inevitably gaps in information, science invariably proceeds by small steps and by the accumulation of knowledge. Among the enormous contributions, warts and all, as we have seen above, were the superb anatomical studies of the musculature of plesiosaur limbs by Watson in 1924, the seminal idea that the front limbs were used in flight by Halstead (Tarlo 1958), complex and innovative anatomical and functional analyses by Robinson (1975, 1977), Godfrey (1984), Feldkamp (1987), and novel interpretations of the four- wing problem by Tarsitano and Riess (1982) and Frey and Reiss (1987). It is only with hindsight knowledge that more and more comprehensive studies can be made. With time one expects ideas to become more complete and exhaustive and gaps filled and hopefully, increasingly instructive. Perhaps the remark by Sir Isaac Newton, by no means a modest or generous man himself with respect to the achievements of many contemporaries, yet writing to his rival Robert Hooke said, ‘‘If I have seen further it is by standing on the shoulders of giants.’’ Whether or not this was a snide innuendo with respect to Robert Hooke’s diminutive height, hopefully without being too cynical, is another possibility.

9.3 The Fastest of Them All

Ichthyosaurs lived at about the same time as the plesiosaurs. They were members of an extinct group of marine reptiles that had captured the imagination of the public more than any other group of marine animals of the past. Ichthyosaurs had a 234 9 Reptiles Return to the Sea

Fig. 9.10 Life reconstruction of a Jurassic ichthyosaur, Ichthyosaurus. Reproduced from a cover illustration (modiﬁed here from the original by the author). Courtesy of the Royal Society, London

wide geographic distribution and they spanned almost the entire Mesozoic Era. They are represented by a variety of forms (Motani 1999) with sizes ranging from about 1 m to 15 m long while one gigantic species Shastasaurus sikanniensis, found by Canadian Paleontologist Elizabeth Nicholls, was 21 m long. It was in the Jurassic, however, that ichthyosaurs achieved their classic fish-like body shape, giving rise to their name (literally ‘‘fish-lizards,’’ from Greek ikhthus ‘‘fish’’ ? sauros ‘‘lizard’’). From their body shape there is no question of the appropriateness of the term fish-lizard for Jurassic ichthyosaurs given their outward similarity to fish such as tuna and swordfish. Yet part of their great fascination was that ichthyosaurs also look strikingly like dolphins (Figs. 1.4 and 9.10). Also remarkable is the fact that Jurassic ichthyosaurs despite being descended from egg-laying reptilian ancestors produced young via live birth. Several specimens found in the excellent fine-grained early Jurassic shale of southern Germany, particularly around Holzmaden, show numerous embryos within the body outline and some apparently in the process of giving birth. Some of the earliest and remarkable finds of ichthyosaurs came from southern England. These include the excellent ichthyosaur fossil finds made by the young fossil collector Mary Anning, which she made along the Dorset coast during the early nineteenth century. Among the earliest notable studies were those made by the Reverend William Buckland and Sir Richard Owen. Many Jurassic ichthyosaurs were about 3–4 m in length and what identified them most with fishes were their deep spindle- or tear-dropped body shape and a large crescent-shaped tail fin. The skull and jaws were long and contained numerous sharp teeth. The eyes were very large and the nostrils were positioned far back on the top of the skull (another specialized adaptation to an aquatic existence). These forms probably fed largely upon fish and cephalopods and have been studied extensively by a number of specialist workers (McGowan 1992; McGowan and Motani 2003). 9.3 The Fastest of Them All 235

Jurassic ichthyosaurs are considered to have been the fastest swimmers of the Mesozoic seas. This idea emerged when the notable British mathematician Sir James Lighthill classified ichthyosaurs along with the fastest of today’s swimmers, tuna, dolphins and lamnid sharks, based on a striking similarity of body and caudal fin shape. The three groups of living marine vertebrates belong to a special category of high-speed swimmers defined as thunniform and by default it included ichthyosaurs. Many workers following Lighthill’s (1975) informal classification have generally accepted the idea of ichthyosaur high-speed swimming potential based on its external characteristics of a teardrop body shape and high-aspect ratio tail (Massare 1988; McGowan 1992). The signs were certainly good enough to reasonably establish that ichthyosaurs were fast swimmers but whether or not they would truly compare with the fastest swimmers in the ocean would require much more information (below). We need to jump forward to an important study by Philip Motta (1977) of the University of Florida. He showed that the skin, in particular the dermis, of sharks, comprised layers of collagen fibers arranged in left- and right-handed helices around the body. Shortly after, Wainwright and colleagues (1978) demonstrated the impact of the cross-fiber dermal structure in the swimming dynamics of sharks. Other workers showed similar fiber systems in fast swimming tuna and dolphins (see Volume 2). It became clear that high-speed swimming in the extant thunniform groups involved more than body and tail shape and that the biomechanical characteristics were more complex than previously thought. Hence, if ichthyosaurs were to maintain their status in the thunniform category of swimming it would be necessary reassess them and, to start with, to know if there was any evidence of a similar fiber system to that of the living thunniform swimmers given its functional ramifications. Although there are a number of ichthyosaur specimens with fossilized soft tissue, notably from the Lower Jurassic Posidonia Shale of Southern Germany, so- called skin preservation is in fact often decayed and transformed soft tissue. This material may outline the body as a black film (Martill 1993). While such specimens provide valuable information on the hydrodynamic body and tail shapes of ichthyosaurs such as Stenopterygius (Fig. 9.11), information on actual skin structures is rare (e.g., Martill 1995; Lingham-Soliar and Reif 1998). Subsequent studies involved a search for structural fibers in the ichthyosaur integument (Lingham-Soliar 1999a, 2001, Lingham-Soliar and Plodowski 2007). As a result an alternating system of oppositely oriented fibers was conclusively shown in the dermis of Jurassic ichthyosaurs such as Stenopterygius quadricissus and Ichthy- osaurus sp. (Figs. 9.12 and 9.13), which was similar to the dermal fiber architecture of extant thunniform swimmers. Furthermore, a subsequent study established beyond doubt on the basis of the D-banding molecular structure that that the fibers were collagenous (Fig. 9.14), filling in the last part of the puzzle, i.e., that the fibers had to be stiff and inextensible, both characteristics essential if the cross-fiber system were to work comparable to that of the extant thunniform swimmers (Lingham-Soliar and Wesley-Smith 2008). The biomechanics of the 236 9 Reptiles Return to the Sea

Fig. 9.11 Stenopterygius quadricissus SMF 457 (Senckenberg Museum, Germany), approximately 2.3 m long, showing a large dorsal ﬁn and high-aspect-ratio caudal ﬁn, reminiscent of the thunniform. After Lingham-Soliar and Plodowski (2007)

Fig. 9.12 The ichthyosaur ? Ichthyosaurus GLAHM V1180 from England. A cross-fiber architecture is seen in two layers of the dermis preserved over the jaws of the ichthyosaur. The fine fibers of one layer shows about 40 strands. The surface layer (whitish and hazy) is considerably decomposed but on close scrutiny shows fibers oppositely oriented to the deeper layer. Scale bar = 0.5 cm. After Lingham-Soliar (1999a). Courtesy of the Royal Society, London dermal architecture of Jurassic ichthyosaurs as well as of a number of other living and extinct animals will be discussed in detail in Volume 2.

9.4 Last of the Great Marine Reptiles

Of all the major groups of marine reptiles mosasaurs had the shortest geological record, conﬁned entirely to the Upper Cretaceous period. Yet, in the short time of approximately 25 million years (about a tenth that of the ichthyosaurs and 9.4 Last of the Great Marine Reptiles 237

Fig. 9.13 High-tensile fibers near the fin base in 3–4 layers in the Jurassic ichthyosaur Stenopterygius quadricissus SMF 457; arrows show two sets of fibers oriented at high opposing angles and one of fibers at more acute angles, i.e., three different and opposing orientations. The outermost layer is degraded providing a less than pristine appearance which ironically, it is as a consequence of such degradations that information on deeper layers are obtained. After Lingham- Soliar and Plodowski (2007)

Fig. 9.14 Collagen ﬁbrils in the ichthyosaur ?Ichthyosaurus GLAHM V1180 The ﬁbrils show the repeating beaded molecular structure of 67 nm, i.e., the D-bands that conclusively characterizes collagen. Scale bar = 0.5 lm. Lingham- Soliar and Wesley-Smith (2008). Courtesy of the Royal Society, London

plesiosaurs) they achieved global diversity, with remains found even in Antarctica, gigantic size and domination of the marine fauna when the ichthyosaurs and plesiosaurs were rapidly diminishing or had become extinct. Their demise is shrouded in mystery and one possibility is that it may have been due to the end of Cretaceous catastrophic event, which also coincided with the disappearance of the 238 9 Reptiles Return to the Sea

Fig. 9.15 The ‘‘grand animal de Maastricht.’’ Top etching from faujas st. Font (1799) of the removal of the ‘‘grand animal de Maastricht’’ in 1780 from the mine shaft. Bottom the cast of the fossil skull housed in the Natural History Museum, London. It was presented by the French anatomist Baron Georges Cuvier to Gideon Mantell (discoverer of the ﬁrst dinosaur) in 1825 dinosaurs. However, the best place to start is in the discovery of their ﬁrst remains, which was nothing short of sensational. In 1766 during mining excavations, strange enormous jaw bones of an unknown animal were unearthed in St. Pieter’s Mountain, Maastricht, The Netherlands. Four years later in 1780, in the same vicinity, a more sensational discovery followed, that of the better preserved, incomplete skull of what the discoverers called the ‘‘Grand animal de Maestricht’’ and removed to the keeping of a Doctor Hoffman (Fig. 9.15). The skull was about 1.2 m long and most investigators of the time, most prominently Faujas de Saint-Fond (1799), believed that it belonged to a giant crocodile. On the other hand the anatomist Pieter Camper (1786) was con- vinced that it belonged to a whale. Nevertheless, he correctly pointed out many characters that were not consistent with those of crocodiles. The true identity of the skull was left to Adriaan Gilles Camper (1800), Pieter Camper’s son, who 9.4 Last of the Great Marine Reptiles 239

Fig. 9.16 Possibly ancestry of the mosasaurs. The terrestrial ancestor was a reptile similar to Varanus komodoensis (right about 3.4 m long), followed by a semi-aquatic reptile, Aigialosaurus (centre, silhouetted, about a meter long) and the fully aquatic Mosasaurus hoffmanni (left, about 17 m long). below each animal is shown the counterpart forelimb, demonstrating the transition from a terrestrial to aquatic lifestyle. From Lingham-Soliar (1999a, b) demonstrated that the fossil was neither a crocodile nor a whale but a giant lizard closely related to the modern-day Varanus niloticus (A. G. Camper’s Lacerta dracaena) (Fig. 9.16). This lizard status (classified today as lepidosaurs) of mosasaurs was first communicated by A. G. Camper in letters to Cuvier in 1800 and followed in several later publications (e.g., 1812). Hence, the early literature, including the first record in the Transactions of the Royal Society by Peter Camper (1786), was primarily concerned with its identity and it was only much later in 1822 that it was named Mosasaurus (after the Meuse River near which the fossil was found) by the English naturalist the Reverend William Conybeare (Lingham- Soliar 1995). Surprisingly little more was said about this remarkable fossil, besides its discovery, until just over 200 years after its discovery when it was first fully scien- tifically described (Lingham-Soliar 1995), coincidentally also in the Transactions of the Royal Society (see below). Mosasaurus hoJfmanni (the species named after the first owner of the ‘‘Grand animal de Maestricht’’ Doctor Hoffman) is one of the most advanced of all known mosasaur species. Although Hainosaurus bernardi (Lingham-Soliar 1992a)at approximately 15 m long (the head alone about 1.5 m long; Fig. 9.17) is usually regarded as the largest mosasaur, an enormous almost complete dentary of Mosasaurus hoffmanni, in the Natural History Museum, Maastricht was measured at one meter. Based on proportion of dentary to the entire lower jaw in other specimens the entire lower jaw was calculated to be 1.6 m long. Further calculations showed 240 9 Reptiles Return to the Sea

Fig. 9.17 A life reconstruction by the author of the gigantic skull of one of the largest mosasaurs known, Hainosaurus bernardi, housed in the Royal Institution of Brussels, Belgium. The skull is approximately one meter long. From Lingham-Soliar (1999a, b)

that the whole animal would have been close to 17.6 m long, making it the largest marine reptile known (Lingham-Soliar 1995a). Besides presenting an opportunity to study its morphology and anatomy, its enormous size was also important in investigations on head mobility and mechanics involved in the feeding on large prey and in the evolution of the lifestyle of this gigantic mosasaur (Fig. 9.18). As we saw in Chap. 2, teeth are modified descendants of bony dermal plates that formed the armour of ancestral fishes and consistent with the theme of the book we will stick to the skin and its derivatives. However, in doing so we will hopefully come to understand some of the fascination that surrounds these long vanished sea dragons, as they are often referred to in the popular literature. Mosasaurs possessed teeth set in deep sockets known as the thecodont (‘‘socket- toothed’’) form of dentition, similar to that of the archosaurs. Tooth bases are cemented in deep pits and there was a continuous form of tooth replacement (Edmund 1960). Successional teeth emerge from the posteromedial region of each tooth, occurring in waves as described earlier. The replacement waves appear to have been very successful in mosasaurs because examinations in numerous specimens show that there were no glaringly dangerous gaps in the dentition or patches of small incompletely developed teeth. Rather, teeth are uniform in size along the length of the tooth rows with only the first two and last two teeth showing any significant reduction in size (Fig. 9.18). This excludes the premaxilla in which the teeth are typically for mosasaurs very small barring one species, Goronyosaurus, in which they were especially large (Lingham-Soliar 2002). The most distinguishing feature of the teeth is the pronounced development of external tooth facets or prisms, perhaps more so than in any fossil species examined. In most specimens, the tooth crowns possess two prisms at the slightly convex buccal surface and five along the deeply convex or U-shaped lingual surface. The U-shaped tooth bases are strongest anteriorly, becoming less pronounced 9.4 Last of the Great Marine Reptiles 241

Fig. 9.18 Top, partial skull of Mosasaurus hofmanniI RSNB R26 (frontal and parietal absent). Bottom, Mosasaurus hoffmanni restored skull, lateral view based on the holotype (cast) and IRSNB R26. Abbreviations: fmag foramen magnum; for foramen; gl intermandibular articulation; iam internal auditory meatus; in internal naris; inpi incisura piriformis; intbar internarial bar; istp infrastapedial process of quadrate; j jugal; jo aperture for Jacobson’s organ; I lachrymal; mands mandibular symphysis or syndesmosis; mpals foramen for median palatine sinus; meckca Meckelian canal; meckfo Meckelian fossa; mx maxilla or maxillary suture; o orbit; of olfactory lobe; op opisthotic; ot otosphenoidal crest of basisphenoid; p parietal; or suture for parietal; paf parietal foramen; pal palatine; ptpalu sutural union between pterygoid and palatine; pcr posterior carina on tooth crown; pmx premaxilla; pofex postorbitofrontal excavation; pof postorbitofrontal; or suture for postorbital frontal; popr paroccipital process of opisthotic; pra prearticular; prf prefrontal; or suture for prefrontal; prf ala prefrontal alar; ps parasphenoid; pt pterygoid; ptte pterygoid teeth; pvp postero-ventral process of jugal; q quadrate; quap quadratic process of pterygoid; qcond quadratic condyle; respit resorption pit on tooth base; ret retroarticular process; reto replacement tooth; ro roughened area (e.g., pitted or striated); ros rostrum; rosforfrn roughened area for sutural contact between frontal and premaxilla; sstp suprastapedial process of quadrate; sa surangular; sep septomaxilla; soc supraoccipital; sm branch of M. depressor mandibulae; sp splenial; spit stapedial pit; spl splint supporting premaxilla and maxilla; sq squamosal; st supratemporal; tcav tooth cavity. Lingham-Soliar (1995a). Courtesy of the Royal Society, London posteriorly with the external surface changing from ﬂat to slightly convex (changing earlier in the tooth row of the maxilla). The lingual surface is nevertheless always greater than the buccal. Associated with the changes in shape and number of prisms, the posterior carina (pcr in Fig. 9.18) shifts from a somewhat lateral position in the anterior teeth to a posterior position further along the tooth row. In the lingual surfaces of the maxillary teeth the striae are not clearly dis- cernible until the sixth or seventh tooth crowns. 242 9 Reptiles Return to the Sea

The teeth of Mosasaurus hoffmanni may quite fairly be labeled as lethal killing structures for literally anything that moved in the oceans. We will look at the ferocious and lethal use they were put to. Perhaps blind aggression may not be too strong a term in the context of tooth impressions found on the enormous carapace (over a meter in diameter) of a giant extinct marine turtle, Liopleurodon hoffmanni. In a study the usual suspects were whittled down to Mosasaurus hoffmanni.A functional analysis (Lingham-Soliar 1991a) showed that so great was the impact from the teeth on the massive carapace that around the tooth impressions were a series of widely radiating stress fractures. It was a clear indication of enormous bite forces Mosasaurus hoffmanni was capable of inﬂicting on prey. However, the consequences of their aggressions were not always just to the prey as was seen in a paleopathological study on Mosasaurus hoffmanni (Lingham-Soliar 2004) described below. Before we do that, it is necessary to say a few words on mosasaur swimming. Mosasaurs used much of the posterior part of the body in swimming in a mode referred to as axial subundulatory (Lingham-Soliar 1991b). Because it is capable of producing rapid starts from a relatively stationary position it is ideal for ambush predation, a lifestyle that mosasaurs are thought to have followed. However, one form of mosasaur, Plioplatecarpus marshi, shows some important morphological characteristics, which include a massive scapula (hence increased muscular insertion area), that suggests incipient underwater ﬂight (Lingham-Soliar 1992b), although the long tail was probably still used as the main form of locomotion (Fig. 9.19).

9.4.1 Disease, Injuries, and Bone Repair in Mosasaurus Hoffmanni and Their Wider Implications

Fractures in a number of dentaries (anterior lower jaw) of Mosasaurus hoffmanni bear testimony to the violence of mosasaur predation. At the Royal Institute of Belgium, there are in the collections a number of large jaw bones that had apparently been split into two as though they were twigs (Lingham-Soliar 2004). We know this was not postmortem breaks because of the presence of the deﬁnitive tell-tale bony callus or haematoma in otherwise perfectly preserved dentaries (Fig. 9.20). This is part of the bone healing process which commences at the very moment that a fracture occurs. Most of our knowledge of bone repair comes from living mammals. The specimens of M. hoffmanni show some remarkable details that provide insights into the bone healing process in extinct gigantic reptiles that is rarely observed even in modern reptiles. As a squamate reptile, it provides data that in many ways cannot be distinguished from those in mammals. For instance, a number of deep pits bear witness to a series of continuous sinus drainage canal from the internal part of the fracture (seen in lateral and transverse view of the dentary; Fig. 9.21c, 9.4 Last of the Great Marine Reptiles 243

Fig. 9.19 Anterior proﬁles of three contemporary mosasaurs all showing relative proportions of the scapula (hatched). a Plioplatecarpus marshi. b Platecarpus. c Plotorosaurus. After Lingham-Soliar (1992a, b)

d), which are vital to invasion by cellular elements of the fracture hematoma, which form organic matrix and mineral into bone (the callus) as a bridge to the fracture. One of the problems associated with the bone-healing process in modern day mammals is bacterial infection. Two specimens of Mosasaurus hoffmanni show deﬁcient healing associated with microbial infection of the bone (microbial erosion around the tooth alveoli) a form of osteomyelitis, that may have been associated with an open fracture that permitted direct access for disease-causing organisms (Fig. 9.20). Indeed, interestingly we can tell that mosasaurs were no strangers to tooth ache with insight from specimens at the Royal Institution in Brussels (ISRNB) on disease and pain that can rarely be obtained from a fossil. Pathogenic bacteria may have gained access to deeper tissue from secondary or post-traumatic infections resulting from the fracture and/or from defects in the teeth (caries). Infections affect both hard and soft tissues and are present in fossils as a cavity with a distinct margin (Bricknell 1987; Sawyer and Erickson 1984). In one mosasaur specimen (IRSNB R25), the alveolar margin of a missing tooth nearest 244 9 Reptiles Return to the Sea

Fig. 9.20 Mosasaurus hoffmanni IRSNB R25, left dentary. a. Shows the fracture and bony callus associated with infection, in buccal view. b. (Diagrammatic representation of a the single arrow shows apparent bacterial erosion around the tooth alveoli. The series of arrows show bacterial erosion on the buccal surface. Two scratches can be seen to the right of the fracture (see text). c Buccal-occlusal view showing callus over-growing much of the sixth alveolus (arrow 1), large area of bacterial erosion (arrow 2), osteolytic canal (arrow 3), ﬁfth cranial mandibular canal (arrow 4) and fragment of sixth tooth base (arrow 5). Scale bar (a, c) = 5 cm. Lingham-Soliar (2004) the fracture (Fig. 9.21b, arrow) is signiﬁcantly widened and probably represents bacterial erosion that started at the base of the tooth. The papulose surface of the bony callus also suggests considerable bacterial activity. Dental caries may develop as a result of defects in the teeth and the inability to masticate properly, etc. (Bricknell 1987). This may be exacerbated by injury, as in the mosasaur specimens described here. The smooth margin of the cavity indicates an abscess rather than post mortem damage (Rothschild and Martin 1993; Sawyer and Erickson 1984). Four other cavities (Fig. 9.20b, radiating arrows) may also represent bacterial erosion, which may continue post mortem for substantial periods of time. Although the buccal surface in IRSNB R27 shows no obvious signs of pustular (pyogenic) cavities and disease, infection is evident on the lingual surface, in the form of a large pit near the dorsal part of the fracture. It represents the outlet of an abscess canal, running from the internal part of the fracture (seen in the transverse view; Fig. 9.21d). Because of the excellent surface preservation of this specimen, it adds to the unlikelihood that the cavities and state of the tissue around the fractures resulted from post mortem damage rather than from disease. The 9.4 Last of the Great Marine Reptiles 245

Fig. 9.21 Mosasaurus hoffmanni IRSNB R27, right dentary. a Shows a well-formed bony callus surrounding a fracture non-union (arrow 1) in buccal view. Adjacent to this is an associated stress fracture (arrow 2); toward the middle of the figure is an almost completely healed fracture running round the dentary and completely fused in places (arrow 3). Arrow 4 shows a circular pit anteriorly and arrow 5 a long scratch mark. b Shows best the distinctive bony callus in a wide angled lingual view (arrow shows an emerging replacement tooth). Note the pits near the fracture and the enormous tooth bases. The splenial is absent. c The detailed lingual view of the dentary shows the abscess pit on the bony callus and the almost completely healed old fracture extending from the fourth tooth base. The arrow points to both the fracture and the slight convex swelling on the upper dental margin that indicates the last trace of the bony callus associated with the fracture. d Transverse section of the anterior part of the broken dentary shows the continuation of the osteolytic pit from the lingual side (arrow 1). A shallow osteolytic pit is also evident (arrow 2). The large pit (arrow 3) represents the canal for cranial nerve V. O = occlusal, L = lingual, B = buccal. Scale bars = 4 cm. From Lingham-Soliar (2004) 246 9 Reptiles Return to the Sea lifestyle of Mosasaurus hoffmanni, in particular unbridled aggression no doubt played an important part in its tooth-decay problems with wounds and injuries to the gums being invaded by bacteria that eventually spread to the teeth and bone. One of the problems in bone healing is when a fracture unites slowly or not at all because of deficient blood supply to one or more of the bone fragments or separation of the fragments by distention or interposition of a tendon or ligament or skin, or excessive motion at the fracture site. This condition is evident in a few dentaries of Mosasaurus hoffmanni despite the presence of distinct calluses. It is, however, possible that the bone broke post mortem along the incompletely healed fracture although judging from the strength of the earlier fractures in IRSNB R27 with minor traces of the healed fracture, this seems unlikely. Fracture nonunion has been noted in the postcranial skeletons of a number of animals, e.g., birds, lizards, and mammals that have survived in the wild (Harris 1978) and as pseu- doarthrosis (false joint) in the ribs of the extinct ichthyosaurs (reptiles) (personal observation, Staatliches Museum, Stuttgart, Germany). As noted in earlier chapters, the skull, mandibles (lower jaws) and clavicles or collar bones in reptiles, birds, and mammals are membrane bones of dermal origin, the remaining vestiges of dermal plates that protected the body of ancient jawless fishes (ca. 450 million years ago). In birds and mammals within 1 to 2 weeks, after injury, a provisional bony callus containing secondary fibrocartilage is formed around the ends of the fracture, uniting the fragments and enveloping the fracture site. However, Irwin and Ferguson (1986) demonstrated that fractured reptilian dermal bones, as in amphibians, do not form secondary cartilage. It is clear therefore that the fossilized bony calluses in the present specimens are not provisional cartilaginous precursors probably formed within a few days of the injury (cartilage is rarely preserved in fossils and markedly different from fossilized bone). Endosteal new bone, on the other hand, forms slowly and may be noted as early as the third week (Irwin and Ferguson 1986; Rothschild and Martin 1993). This gives an estimate of the minimum age of the injury in the present specimens. The bony callus is then resorbed over a period of ca. 16 weeks (Rothschild and Martin 1993), giving an estimate of the upper age of the injury. Thus, the survival after injury of the individuals described here is probably about 16 weeks. This period is noted in parenthesis and is probably a minimum time since bony callus formation and resorption are quite variable and dependent on the animal’s general health, e.g., chronic osteomyelitis, poor nutrition, etc., could greatly slow the time of bony callus resorption. An obvious question concerns how effective was the rehealing in Mosasaurus hoffmanni. It is clear that the old fracture at the narrow end of the dentary in IRSNB R27 had knitted effectively enough to withstand obvious stresses occurring at the time of the latest fracture nonunion, given that the latter occurred at a much thicker part of the dentary. Another problem sometimes associated with severe fractures is the loss of vascular supply resulting in the devitalized bone becoming necrotic (Rothschild and Martin 1993). However, the condition of the dentaries anterior to the haematomas in specimens IRSNB R25 and R27 is particularly good (note, the presence of one or two alveoli is a normal condition in mosasaurs 9.4 Last of the Great Marine Reptiles 247 because of the system of continuous tooth replacement in these reptiles; the tip of a large replacement tooth is seen emerging from one). The mandibular artery and fifth mandibular or trigeminal nerve, housed in the mandibular canal (Lingham- Soliar 1995a) and extending the length of the dentary, were evidently not damaged. Immobilization of the fracture would probably have helped prevent damage to vital nerves and blood supply to the dentary and will be considered next. How did Mosasaurus hoffmanni survive such serious fractures to their vital feeding apparatus given that they underwent in so many cases the time-consuming repair processes. In mammals, there have been few studies on such injuries to animals in the wild but the few instance e.g., of foxes with broken dentaries have shown that an inability to feed almost certainly resulted in the animal’s death. Mosasaurs as reptiles have slower metabolic rates compared to mammals but given their instinctively aggressive nature that would not entirely account for the reahe- aling. The answer may lie in the morphology of mosasaurs as reptiles as distinct from mammals. The skull and jaws of mosasaurs were highly kinetic being formed of numerous bones, somewhat resembling the condition in snakes (Lingham-Soliar 1995a; Lee et al. 1999). Each of the anterior lower jaws unlike in mammals comprises two bones the dentary and the splenial that are ligamentously united. An explanation for perfect alignment of the fractured dentaries may lie in this peculiar construction of this mosasaur lower jaw. The dentary is a laterally broad girder of bone that houses the teeth. The splenial, on the other hand, is a laterally flattened flange of bone, deep posteriorly and tapering anteriorly, lying medial to the dentary and supporting it ventrally. The splenial alone forms the articulation with the posterior jaw unit. This is well preserved in a number of specimens (Lingham-Soliar 1995a). It extends anteriorly to approximately the second dentary tooth and overlaps the Meckelian Fossa. This peculiar splenio-dentary association may have played a crucial role in the healing of mosasaur dentaries. During predation the dentary and teeth would take the major part of the force when it closed upon prey. This would only be indirectly transmitted to the splenial via connective tissue, much of it presumably absorbed by the latter and by the mobile splenio-angular joint. It is therefore highly likely that the splenial did not fracture and remained ligamentously bound to the dentary. The intact splenial, bound by connective tissue to the dentary, would have acted as a natural splint that held the fractured segments of the dentary in place (veterinarian surgeons in treating broken jaws may introduce an internal splint to immobilize the jaw; Fig. 9.22). The splenial, together with suppression of aggressive predatory behavior, mentioned above, seems the most parsimonious explanation for the excellent alignment during healing of the jaws. The teeth of Mosasaurus hoffmanni combined sharp cutting edges with robust size. Penetration of the skin of the prey is rendered by the sharp anterior and posterior carina while the unique cutting edges or suction-breaking grooves of the prisms on the buccal surfaces of the teeth help to further tear the tissue. This is unequaled in other marine reptiles including other mosasaurs. The teeth are moderately recurved posteromedially and the tips are slightly blunt, features that show an adaptation to great loadings (described in detail in Lingham-Soliar 1995a). 248 9 Reptiles Return to the Sea

Fig. 9.22 A 1.5 m Varanus niloticus specimen that had suffered several injuries to the head, including a fractured right dentary, in a collision with a vehicle. A rib splint supports the fracture Scale bar = 6 cm. From Lingham- Soliar (2004)

9.4.2 Mosasaur Integument

As we have already seen, a remarkable amount of data have been obtained on the integument of fossil vertebrates, in particular dinosaurs and ichthyosaurs. However, with respect to mosasaurs despite the enormous quantities of fossil remains found all over the world, there has been very little material on the epidermal scales and indeed any soft tissue structures either from the epidermis or dermis. One of the earliest descriptions was by FH Snow in 1878 who wrote ‘‘But, so far as I am able to learn, nothing has been hitherto known of the general covering of the saurian body in any genus, and nothing whatever of the dermal covering in the genus Liodon [Tylosaurus]. It might have been expected that this covering would be found to consist of larger plates, like those of the alligator and crocodile. On the contrary, it is composed of small scales, much resembling in size, shape and arrangement, the scales of living Ophidians.’’ In a photograph (Fig. 9.23), Snow showed the impressions of about 3000 scales and noted that there were about 90 scales per inch, being somewhat smaller than those of a large rattlesnake (80 scales per inch). The scales are diamond shaped (3.3 mm 9 2.5 mm), with a raised ridge (carina) on the long axis. Mike Everhart notes in his webpage devoted to mosasaurs (2010) that scales of a similar size and shape, but lacking the central ridge, are also known from a specimen of Ectenosaurus clidastoides (FHSM VP-401) in the collection of the Sternberg Museum of Natural History, of which more will be said later. However, from the very beginning, the need to discover what mosasaurs looked like on the outside was dogged by sparse or poor preservation and errors were inevitable. For instance, Samuel Williston (1899) wrote, ‘‘It is certain that none of the Kansas forms of this order were covered with bony scutes, as described by Marsh, the bones so described being, undoubtedly, sclerotic plates.’’ Ironically, Wiilliston himself was similarly misled with respect to the integument. His error almost certainly led to an error in one of the otherwise ﬁnest life reconstructions of mosasaurs, i.e., KR Knight’s painting of a tylosaur that hangs in the American Museum of Natural History (AMNH). 9.4 Last of the Great Marine Reptiles 249

Fig. 9.23 A mosasaur from the upper Santonian-lower Campanian of Tylosaurus (KUVP-1075) from Cove County, Kansas showing scales that were ﬁrst described by FH Snow in 1875

Williston (1899) had unfortunately interpreted soft tissue remains associated with a tylosaur specimen as part of a nuchal fringe, which as it happens very much adds to the drama of KR Knights dynamic reconstruction of the mosasaur. In truth, however, the remains represented the splayed out and compressed cartilaginous tracheal rings (Lingham-Soliar 1991b), which Williston, himself retracted in a later, albeit rather obscure publication. If it had turned out to be true, Williston’s description would have had serious taxonomic implications given that among present day reptiles a nuchal fringe is known only in iguanid lizards, i.e., a unique phylogenetic branch of reptiles unrelated to mosasaurs. The paucity of mosasaur epidermal scales may be explained by the type of scale patterns, in reptiles, i.e., they lie predominantly adjacent to each other, unlike the strongly overlapping condition in ﬁshes. This leaves the interconnecting membranes, which is made up of the much softer a-keratin compared to the harder b- keratin of the scales and considerably more exposed and hence more susceptible to scavenging by invertebrates (Lingham-Soliar 2012). In this context, it is pertinent that the a-keratin is highly nutritious. The scales also overlie a thick collagenous layer comprising a-keratin. In decomposition experiments epidermal scales were found to be among the ﬁrst structures to be dislodged from reptile carcasses on land by ants and beetles feeding on the soft keratin and in fungal-assisted delin- eation of feather keratin the keratin matrix, thought to be composed of a-keratin, was selectively degraded in preference to b-keratin (Lingham-Soliar et al. 2010, Lingham-Soliar and Murugan 2013). Numerous marine invertebrates would have performed a similar role. Sparse data on the integument of mosasaurs was redressed to a large extent in a recent description of the integument of the mosasaur, Ectenosaurus clidastoides by Lindgren and colleagues (2011). E. clidastoides was a long-snouted mosasaur from the Santonian (Upper Cretaceous) part of the Smoky Hill Chalk Member of the 250 9 Reptiles Return to the Sea

Fig. 9.24 Epidermal and dermal structures in mosasaurs. a External view of imbricating and obliquely arrayed scales showing keeled surface in the epidermis. b Transverse and tangential sections through helically arranged ﬁber bundles in the dermis. Scale bars = 2 mm. From Lindgren et al. (2011). Courtesy of Plos One, Open Access

Niobrara Formation in western Kansas, USA. This mosasaur at a stroke provides data not just of the epidermal scales but also of the underlying dermis (Fig. 9.24). The authors discovered an elaborate system of multiple-layered fiber bundles in Ectenosaurus which they rightly described as a significant development in so far as it represents the first unambiguous record of deeper soft-tissue structures in the skin of a mosasaur. The uniform shape and diminutive proportions suggest that they originated from the neck and/or trunk of the animal (Fig. 9.24a). The hydrodynamic aspects of keeled body scales had been dealt with elsewhere (Lindgren et al. 2010) but the supportive sculpturing on the underside of the scales was reported for the first time in the study reporting dermal fibers (Lindgren et al. 2011). What was the function of the keeled scales? The authors proposed that the ‘‘multiple keels’’ on the body scales of Plotosaurus may in fact be supportive sculpturing as, e.g., in the osteoderms and scales in certain nonavian dinosaurs and in many extant lizards and notably, two parallel, longitudinal crests occur on the larger body scales of some monitor lizards. In Ectenosaurus, on the other hand, they suggested that the longitudinal ridges might act as attachment sites for underlying ligaments or connective tissue and that they may serve a function as anchors, thereby providing strength to the skin. The scales of Ectenosaurus are considerably smaller in size (2.7 9 2.0 mm) than are those of a Platecarpus specimen (3.8 9 4.4 mm) studied by the Lindgren and colleagues (2011), despite a comparable estimated total body length of the two animals (5.9 vs 5.7 m). Additionally, they note that in a specimen of Tylosaurus the scales are 3.3 mm, firmly in between the scale size in Ectenosaurus and Platecarpus. Based on skeletal dimensions, the authors estimated the length of the tylosaur to be approximately 5 m in overall body length and that size discrimi- nations with Platecarpus may also suggest that differences in scale size between 9.4 Last of the Great Marine Reptiles 251 the three species might be age related. However, based on the fact that extant lizards generally hatch with a fixed number of scales which then grow in size with each molt (resulting in larger scales in older individuals), they suggest that the discrepancy in scale size may be the result of age differences between the three individuals mentioned and the possibility that Ectenosaurus represents a younger individual than Platecarpus, and that Ectenosaurus might reach body lengths well beyond those of Platecarpus. Also interesting, was the discovery by Lindgren et al. (2011) of layers of pale, strand-like structures, subjacent to the scales, exposed in both transverse (i.e., at right angles to the skin surface) and tangential (i.e., parallel to the skin surface) sections. Some of the strands are flattened (Fig. 9.19b) others retain much of their presumed original three-dimensional form (Fig. 9.24b). In extant vertebrates, similar strands represent the fossilized remains of structural fiber bundles from the dermis. The thickest fiber bundles are located deepest in the skin (i.e., approximately 2 mm below the skin surface in its present, somewhat compressed state) and the thinnest are the outermost. The fiber bundles are either straight, densely spaced and oriented at acute angles (i.e., almost parallel) to the long axis of the animal, or are arranged in tightly stacked layers with alternating left- and right-handed orientations (a crossed helical architecture; Fig. 9.23b) with a minimum of eight layers in which fiber angles are in the range of 20–70° to the long axis of the animal (predominantly 40–55°). As functional explanations for the integument the authors propose that the characteristics of the scales covering the body, i.e., small-sized, firmly anchored and keeled may have contributed to an anterior-posterior channelling of the water flow, thereby reducing frictional drag arising during increased speeds such as in the pursuit of prey. They add that surface deformation (and consequent frictional drag), may have been further reduced by the cross woven helical fiber bundles in the subjacent dermis. As the authors note, the functional significance may be interpreted from findings of a similar system of fibers that has been shown to be involved in counteracting fluid drag by retaining a smooth body surface, enabling stiffness, and counteracting torsional stresses. The authors propose that the combination of small-sized, firmly anchored body scales and a complex meshwork of alternating crossed-helical and longitudinal fiber bundles was evidence that the anterior part of the body of Ectenosaurus was stiffened so that it was held somewhat rigid during locomotion while the side–side thrust-producing flexure was restricted to the posterior trunk and tail. Accordingly, they proposed that Ectenosaurus probably utilized the faster subcarangiform rather than anguilliform mode of swimming as employed by different groups of fishes (discussed in Lingham-Soliar 1991b).

9.4.3 Extinction

The disappearance of mosasaurs is an enigma of the K-T extinction. Unlike plesiosaurs and ichthyosaurs, they were at the point of an enormous radiation. From a handful of species in the Turonian, they expanded to a total of 252 9 Reptiles Return to the Sea

Fig. 9.25 Based on evidence of a large meteorite crater at Yukaton dated around the K-T extinction, one theory is that the aftereffects of the meteorite impact caused the extinction of major vertebrate groups on earth, which included the dinosaurs and mosasaurs. The Baringer crater shown in the Canyon Diablo region of Arizona is similar but smaller than that at Yukaton. From Lingham-Soliar (1999a, b) approximately 70 plus species worldwide during the course of their evolution. The global stratigraphic record at this point is also good. Hence, the indications are that the demise of the mosasaurs was sudden and unexpected and perhaps due to a possible catastrophic event such as the aftermath effects following a large meteorite impact on earth (Lingham- Soliar 1994, 1999b (Fig. 9.25). Remarkably, like the dinosaurs, they were replaced by mammals and equally remarkably, as with the reptiles, a group of fully terrestrial mammals after enjoying success on land for millions of years would return to the ocean. This event occurred according to the fossil record during the Early Eocene (about 50 million years ago). To add insult to injury these pioneers, this time mammals, the ﬁrst fully marine form being Ba- silosaurus, would even take on the shape of mosasaurs and be described as mo- sasauriform. They, however, would evolve to become among the most successful vertebrates on the planet. They are the dolphins and their relatives, which make up the Cetacea (discussed in the biomechanics of high-speed swimming vertebrates in Volume 2).

References

Bricknell I (1987) Palaeopathology of pleistocene proboscideans in Britain. Modern Geol 2:295–309 Camper AG (1800) Lettre de AG Camper- G Cuvier sur le ossements fossiles de la montagne de St. Pierre, Maestricht. J Phys Chim D’hist Nat 51:278–291 References 253

Camper AG (1812) Memoire sur quelques parties moins connues du squelette des sauriens fossiles de Maestricht. Annls Mus Hist nat 19:215–214 Camper P (1786) Conjectures relative to the petrifactions found in St Peters mountain near Maastricht. Phil Trans R Soc Lond B 76:443–456 Carpenter K, Sanders F, Reed B, Reed J, Larson P (2010) Plesiosaur swimming as interpreted from skeletal analysis and experimental results. Trans Kansas Acad Sci 113:1–34. doi: http:// dx.doi.org/10.1660/062.113.0201 Conybeare WD (1824) Additional notices on the fossil genera Icthyosaurus and Plesiosaurus. Trans Geol Soc Lond 2:103–123 Edmund AG (I960) Tooth replacement phenomena in the lower vertebrates. Contr Life Sci Div R Ont Mus 52:1–90 Faujas de Saint-Fond B (1799) Histoire naturelle de la montagne de Saint-Pierre d Maestricht. Paris Feldkamp D (1987) Foreflipper propulsioni n the Californian sea lion Zalophus californianus. J Zool Lond 2l2:43–57 Frey E, Riess J (1987) Consideration concerning plesiosaur locomotion. Neues Jahrbuch für Geologie und Paliiontobgie Abhandlungen 164:193–194 Godfrey SI (1984) Plesiosaur subaqueous locomotion: a reappraisal. Neues Jahrbuch für Geologie und Palaontologie Mh 1984:661–672 Hall BK (ed) (2007) Fins into limbs: evolution, development and transformation. Chicago University Press, Chicago Harris S (1978) Injuries to foxes (Vulpes vulpes) in suburban London. J Zool Lond 186:567–572 Halstead LB (1989) Plesiosaur locomotion. J Geol Soc 146:37–40. doi: 10.1144/gsjgs.146.1.0037 Hawkins TH (1840) The Book of the great sea-dragons, ichthyosauri and plesiosauri gedolim taninum of moses. Extinct Monsters of the Ancient Earth, London Irwin CR, Ferguson MWJ (1986) Fracture repair of reptilian dermal bones: can reptiles form secondary cartilage. J Anat 146:53–64 Lee SY, Bell GL Jr, Caldwell MW (1999) The origin of snake feeding. Nature 400:655–659 Lindgren J, Caldwell MW, Konishi T, Chiappe LM (2010) Convergent evolution in aquatic tetrapods: insights from an exceptional fossil mosasaur. PLoS ONE 5(8):e11998. doi:10.1371/ journal.pone.0011998 Lindgren J, Everhart MJ, Caldwell MW (2011) Three-dimensionally preserved integument reveals hydrodynamic adaptations in the extinct marine lizard Ectenosaurus (Reptilia, Mosasauridae). PLoS ONE 6(11):e27343. doi:10.1371/journal.pone.0027343 Lighthill J (1975) Mathematical biofluid dynamics. Society of Industrial and Applied Mathematics, Philadelphia Lingham-Soliar T (1991a) Predation in mosasaurs—a functional approach. Natural structures, principles, strategies and models in architecture. Proc SFB 230:169–177 Lingham-Soliar T (1991b) Locomotion in mosasaurs. Modern Geol 16:229–248 Lingham-Soliar T (1992a) The tylosaurine mosasaurs of Europe and Africa. Bull Inst R Sci nat Belg 62:171–194 Lingham-Soliar T (1992b) A new mode of locomotion in mosasaurs: subaqueous flying in Plioplatecarpus marshi. J Vertebr Paleontol 12:405–421 Lingham-Soliar T (1994) The cretaceous-tertiary extinction—a catastrophe. Biologist 41:215–218 Lingham-Soliar T (1995a) Anatomy and functional morphology of the largest marine reptile known, Mosasaurus hoffmanni (Mosasauridae, Reptilia) from the upper Cretaceous, upper Maastrichtian of the Netherlands. Phil Trans R Soc Lond B 347:155–180 Lingham-soliar T (1995b) Nothosaurier und die evolution der plesiosaurier. In: Liebau A, Westphal F (eds.) Schwimmsaurier. Ausstellungskataloge der Universtat Tubingen 24:19–26 Lingham-Soliar T (1999a) Rare soft tissue preservation showing fibrous structures in an ichthyosaur from the Lower Lias (Jurassic) of England. Proc R Soc B 266:2367–2373 Lingham-Soliar T (1999b) What happened 65 million years ago. Science Spectra 17:20–29 254 9 Reptiles Return to the Sea

Lingham-Soliar T (2000) Is the four-wing problem real or merely an atheoretical exercise? N Jb Geol Paläont Abh 217:45–87 Lingham-Soliar T (2001) The ichthyosaur integument: skin fibers, a means for a strong, flexible and smooth skin. Lethaia 34:287–302 Lingham-Soliar T (2002) The first occurrence of premaxillary caniniform teeth in the Varanoidea: Presence in the extinct mosasaur Goronyosaurus (Squamata: Mosasauridae) and its paleoecological implications. Lethaia 35:187–190 Lingham-Soliar T (2003) Extinction of ichthyosaurs: a catastrophic or evolutionary paradigm? N Jb Geol Paläont Abh 228:421–452 Lingham-Soliar T (2004) Palaeopathology and injury in the extinct mosasaurs (Lepidosauro- morpha, Squamata) and implications for modern reptiles. Lethaia 37:1–8 Lingham-Soliar T (2012) The evolution of the feather: Sinosauropteryx, life, death and preservation of an alleged feathered dinosaur. J Ornithol 153:699–711. doi:10.1007/ s10336-011-0787-x Lingham-Soliar T, Plodowski G (2007) Taphonomic evidence for high-speed adapted fins in thunniform ichthyosaurs. Naturwissenschaften 94:65–70 Lingham-Soliar T, Reif WE (1998) Taphonomic evidence for fast tuna-like swimming in Jurassic and Cretaceous ichthyosaurs. N Jb Geol Paläont Abh 207:171–183 Lingham-Soliar T, Wesley-Smith J (2008) First investigation of the collagen D-band ultrastructure in fossilized vertebrate integument. Proc R Soc Lond B 275:2207–2212. doi:10.1098/rspb.2008.0489 Lingham-Soliar T, Murugan N (2013) A new helical crossed-fibre structure of b-keratin in flight feathers and its biomechanical implications. Plos One 8(6):e65849 Martill DM (1993) Soupy substrates: a medium for the exceptional preservation of ichthyosaurs of the Posidonia shale (Lower Jurassic) of Germany. Kaupia 2:77–97 Martill DM (1995) An ichthyosaur with preserved soft tissue from the Sinemurian of southern England. Palaeontology 38:897–903 Massare JA (1988) Swimming capabilities of Mesozoic marine reptiles: implications for method of predation. Paleobiology 14:187–205 McGowan C (1992) Dinosaurs, spitfires and sea dragons. Harvard University Press, Cambridge McGowan C, Motani R (2003) Ichthyopterygia, handbook of paleoherpetology 8. Verlag Dr Friedrich Pfeil, München Motani R (1999) Phylogeny of the ichthyopterygia. J Vertebr Paleontol 19:472–495. http://www. ucmp.berkeley.edu/people/motani/ichthyo/phylogeny.html Motta PJ (1977) Anatomy and functional morphology of dermal collagen fibres in sharks. Copeia 1977:454–464 Robinson JA (1975) The locomotion of plesiosaurs. N Jb Geol Paläont Abh 149:286–332 Robinson JA (1977) Intracorporal force transmission in plesiosaurs. N Jb Geol Paläont Abh 153:86–128 Rothschild BM, Martin LD (1993) Paleopathology. Disease in the fossil record. CRC Press, Boca Raton Sawyer GT, Erickson BR (1984) Injury and diseases in fossil animals. Bull Field Mus Nat Hist 58:20–25 Snow FH (1878) On the dermal covering of a mosasauroid reptile. Trans Kansas Acad Sci 6:54–58 Tarlo LB (1958) The scapula of Pliosauncs macrornem Phillips. Palaeontology1:193–199 Tarsitano S, Riess J (1982) Plesiosaur locomotion—underwater flight versus rowing. Neues Jahrbuch für Geologie und Paläontologie Abh 164:188–192 Wainwright SA, Vosburgh F, Hebrank JH (1978) Shark skin: function in locomotion. Science 202:747–749 Watson DMS (1924) The elasmosaurid shoulder-girdle and fore-limb. Proc Zool Soc Lond 2:885–917 Williston SW (1899) Some additional characters of mosasaurs. Kans Univ Q 8:39–41 Index

A Adductor muscles, 199, 228 Acanthodes, 39, 67f Agama, 139f, 162 Acanthodians, 33, 34f, 54, 67f tactile sensory organ, 140-141f development, at different grades of, 70f Agama atricollis, 163 gill arch structures, 66 plate-like scales, 163 hyoid arch, 34 Age of Fishes, 99 mandibular arch, 33 Age of Reptiles, 5 pectoral and pelvic fins, 37 Agnathans, 3, 11, 19, 52f relations bony armor, 3 with bony vertebrates, 39 denticle whorl of, 50f to teleostomes or osteichthyian fishes, Devonian shift, 39 36–37 gill arches, 35 and sharks, 67 modern agnathans, 23 tooth replacement system, 63 olfactory ducts of, 28 Acanthostega, 99, 100, 101f, 103f, 108, 110f, vertebrate status, 21 112f, 114, 123, 126 Aigialosaurus, 239f digits details, 118 Ailuronyx seychellensis, 143 fins to feet, transformation of, 119f Albumen, 133 forelimb of, 102f Allosaurus, 154 and Ichthyostega, 113, 114, 115 mounted skeleton of, 155f left hind limb of, 113f, 121 Alpha-keratin (a-keratin), 138, 139f, 142, 143, pectoral and pelvic fins, 121f 249 Acanthostega gunnari, 123 and b-keratin, 249 Aceraspis (Silurian-Devonian), 24f cysteine-rich, 146, 205 Achoania jarvikii, 81 hair keratin, 204 Acrocanthosaurus, 180 Ameloblastin (AMBN) gene, 52 Actinistia (coelacanths), 81 Amelogenin (AMEL) gene, 52 Actinopterygians, 54, 56, 59, 69 Amelotin (AMTN) gene, 52 actinopterygian-sarcopterygian split, 74 American Museum of Natural History development, grades of, 70f (AMNH), 102, 154, 158, 159, 248 euteleostome fishes (Osteichthyes), Allosaurus in, mounted skeleton of, 155f relationship among, 79f Corythosaurus in, 159 extant, 68f skin impressions, 167 ganoid scales, 80 Amino acid analyses, 169 in fish evolution, 71f, 72 Amnion, 131, 133 phylogenetic relationships among, 78f Amniotic eggs, 5, 129, 131–137 tail morphologies of living and albumen, 133 fossil forms, 73-74f amnion, 133

T. Lingham-Soliar, The Vertebrate Integument Volume 1, 255 DOI: 10.1007/978-3-642-53748-6, Ó Springer-Verlag Berlin Heidelberg 2014 256 Index

Amniotic eggs (cont.) Asterolepis ornate, 47f conductivity of water vapor, 136 Astraspis, 20f, 22, 23 cross-section of, 131f Astraspis desiderata, imperfect dorsal Darwinopterus egg, 136f shield of, 21f eggshell, 134f of geckos, 134f flexible-shelled eggs, 135 B Hyphalosaurus baitaigouensis, eggs and Basiliscus, 180 hatchling, 135f Basilosaurus, 252 Ichthyostega, 133 Bauriacynops, 202f K-strategy of reproduction, 133 Ben Nevis Formation, Spitzbergen, 17f proline, 134 Beta-keratin (b-keratin), 5, 137, 138, 141, 142, of reptiles, 130 144 titanosaurs, premaxillary embryonic mor- and a-keratin, 249 phology of, 132f Bison antiquus antiques, 180 Amphibians, 114, 123, 124, 126, 129 Blacktip shark. See Carcharhinus limbatus eggs, 129 BM-40/osteonectin, 50 extant, 108 BMP, 144 modern, skin of, 130f Bones and reptiles, 137, 194, 246 aspidin, 43–46. See also Aspidin tadpole stage of, 133 calcium phosphate, 16 transformation of amniotic egg, 133, 134, carapace, 21–22 136 cartilaginous endoskeleton, 19 Amphistylic jaw, 61 dentine. , 46–48. See also Dentine Ampullae of Lorenzini, 90, 91 dermal plates, ornamentation of, 43f of Carcharodon carcharias (white shark), during food shortage, 18 92f oldest vertebrates, evolutionary signifi- Ampullary receptor organs, 91 cance of, 42–43 for prey detection, 91 dermal plates, 42 Anaconda, 142 phosphate store, 16 Anaspids, 11, 23, 26f protective armor, 19 Angle of attack, 226, 228, 232 Bony fishes, 37, 59, 67, 88, 104. See also Anolis carolinenis (green anole lizard), 146, Osteichthyans; Osteichthyes 204 and four-legged vertebrates, 109f Anterior limbs, 228, 229f, 230f, 231, 232, 233 neuromasts of, 87 Apteronotus, 94 scale armour of, 123 Arandaspis, 22 swim bladder in, 123 Archaeopteryx, 6f Bony integument, 3, 13 Aspidin, 22, 43 The Book of the Great Sea-Dragons aspidinocytes, 46 (Hawkins), 222f cell processes of, 44–45 Brachiosaurus, 186 lacunae of, 44 Breathing on land, problems of, 123–126 and bone, 43 aerial respiration, 123 collagenous fiber bundles, 44 air-breathing fishes and amphibians, 124 earlier heterostracans, outer surface of, 46 lungs vs. swim bladder, 123 forming cells, 46 pectoral fins on tetrapod stem, 122f lamellar aspidin, 44f Polypterus sp., diagrammatic view of, 125f Sharpey’s fibers, 44, 45f Buoyancy, 123, 228, 229, 230 spongy aspidin, 44f in fishes, 94–95 trabecular dentine, 46 in sharks, 95 uncalcified fiber bundles, 44 hydrofoils, 95 Asterolepis, 47f, 48 slow swimming, 95 Index 257

C Cladistics, 100–105 Canal neuromasts (CNs) aquatic and terrestrial existence, dichotomy in detecting transient currents, 87–88 between, 100, 101f Captorhinus, palate of, 194f Phylogenetic Systematics, 101 Carboniferous Cladoselache, 61, 62f Devonian-Carboniferous boundary, 110 Clarius, 93 mammal-like reptiles, 5–6, 193 Cleidoic eggs. See Amniotic eggs new tetrapods from, 115 Climatius, 67f reconstruction of closing stages of, 130f Coelocanths sharks, 61 Latimeria, 93 stegocephalians, phylogeny of, 103f subclass Coelacanthomorpha, 74 Carcharhinusfalciformis (silky shark), 81 Coelurosaurs, 164 Carcharhinus limbatus (blacktip shark), 81, 83 Collagen fibers, 7, 44, 124, 126, 164, 176, 177, Carcharhinus melanopterus, tooth sets, 50f 179, 235 Carcharhinusplumbeus (sandbar shark), 91 Compsognathus, 164 Carcharodon carcarias (white shark), 123, Corvaspis (Downtonian heterostracan), 42 228 Crocodilia, 141 Carcharodon carcharias (white shark), 61, Crocodylia, 153 62f, 95 Cryptocleidus eurymerus, 224f ampullae of Lorenzini in, 92f mounted skeletons, 225f Carnotaurus, 159 Ctenacanthus, 63f Carassius auratus (goldfish), 87 Cut and pull feeding technique, 179 Cementum, 50 Cyathaspid, 20f Cephalaspid, 3, 4f, 23, 25f Cynodonts, 197, 198, 212 ostracoderms, 24f life reconstruction, 7f Cephalaspis, 25f and modern-day mammals, 205–207 Cephalaspis signata, 25f diversification of, 205–206 Ceratopsian dinosaur (Chasmosaurus), 163 Cynognathids, 6, 205 Ceratosaurus, 159 Cynognathus, 6, 197, 205 Ceresiosaurus, 222 Cytokeratins, 144 cast of fossil, 223f Chelonia, 141 Chewing motions, 84 D Chewore dinosaurs Darwinopterus, 136f, 137 importance of, 184 Jurassic sedimentary rocks, 137 tracks, 186 Delphinus capensis, 95 footprints in, 182, 183f, 184f, 188f Dentine, 22, 50, 59, 81 Chicken. SeeGallus gallus armor hypothesis, 48 Chondrichthyans, 37, 39, 54, 56, 59, 61, 66, aspidin to, 46 67, 69, 79 in Astraspis, 46 phylogenetic relationships, 65f and fine tubules, 46 scalation of, 77–78 in Ganosteus, 48 Chondrichthyes (cartilaginous fishes), 59, 61 in heterostracan armor, 44f early Paleozoic, 61 heterostracan dentine, 45f subclass Elasmobranchii, 61–65 in Lacerta vivipara, 209 Cladoselache (Cladoselachimorpha), of placoderms, 77 61 in Poraspis sp., 43f jaw suspension types, 64f psammosteid microstructure, 42–43 Xenacanthus (Xenacanthimorpha), 61 sensitivity, 46 subclass Holocephali, 65–66 in Tartuosteus, 46 elasmobranchs, differences from, 65–66 Dermal denticles, 3, 77 Clade Elpistostegalia + Tetrapoda, ancestral Dermatocranium, 4 characters for, 107f Dermis, 44, 74, 79, 146, 235, 248, 250f, 251 258 Index

Dermis (cont.) E in Ailuronyx, 143 Early Paleozoic, 18, 61 crossed fiber architecture in, 124f Chondrichthyes (cartilaginous fishes), collagen fibers in, 126 origin, 61 in dinosaurs, 176–180 Ectenosaurus, 250, 251 of extant vertebrates, 7 spines, transverse processes, 196f in fish, 5, 7, 137, 153 Ectenosaurus clidastoides, 248, 249 to scales, 84f Edmontosaurus sp. (MRF-03), 167, 169 Dermoskeleton, 11 Eglonaspis, 20f Diatoms, 18f Eifelian, 5, 41f, 111 Digits, 100, 113, 115, 118, 121, 141f, 159, 204, Eigenmannia (gymnotiform fish), 89f 221 Elasmobranchs , 36f, 37, 61, 65, 66, 77, 90, flexibility of, 182 91, 94. See also Sharks, Dimetrodon, 180 infraclasses, 64 skeleton of, 195f Elasmosaurus, 224 skull and lower jaws, 195f Electric fishes, 90, 91, 93 spines in, 194–195 Electric organ discharge (EOD), 90, 93 Dinosaur mummy, 154, 157f Electrocommunication, 90, 93, 94 of Corythosaurus, 158, 159 Electrolocation, 93, 94 Heloderma, 158 active, 90, 92, 94 skin underside, 158f Elpistostegalia, 77, 106, 108 Dinosaur Museum in Utah, 158 ancestral characters for, 107f Dinosaur trace fossils Embryonic bearded dragon. SeePogona epidermal structures, 182 vitticeps footprints, smallest, 186, 188f Enamel, 46, 49, 50, 52, 80, 81, 83f, 209 latex mold of, 189 Enamel in (ENAM) gene, 52 Jurassic theropod trackways, 185f Enamel matrix proteins (EMPs), 52 sauropod tracks, 185–186 Enamel-like structures, 22, 79 claw impressions, 187f Enameloid, 49, 52, 77, 81 tracks, features, 181–182 Eocaptorhinus, 193 and theropods, 182–183 reconstruction of skeleton of, 194f Dinosaurs, 6, 153 Eocene, 7, 70f, 169, 252 dehydration/desiccation process, 158 Eolates, 70f dermis in, 176–180 Eomaia, 216, 217 collagen fibers, 177 Eomaia scansoria, 213f Psittacosaurus sp. MV53, 177–178, dentition and mandible, 214f 178f Euphanerops (=Endeiolepis), 24 Sinosauropteryx, 177 Euryapsid, 149 eggs and embryos, 169–176 Euselachii, 63f Massospondylus embryo, 174f order Ctenacanthiformes, 64 preservation of soft tissues, 173 order Hybodontiformes, 64 water vapor conductance, 175, 176 Eusthenopteron, 99, 102f, 106, 107, 107f, 108, integument, organic analysis, 167–169 110, 110f, 118, 119f nesting behavior, 176 life reconstruction of, 100f skin, in temperature control, 180–181 pectoral and pelvic fins of, 121f tracks and traces, 181–189. See also Eusthenopteron foordi, 77, 80f, 83f, 104 Dinosaur trace fossils Euthacanthus, 67f Diplacanthus, 67f Eutheriodont therapsids, cladistic relationships Dipnomorpha (lungfish), 81 of, 206f Domestic chicken. SeeGallus gallus Evo-Devo (evolution and development), 144, Doryaspis, 20f 145 Downtonian heterostracan. SeeCorvaspis Evolution of jaws, 3, 4, 33–37, 49 Drepanaspis, 20f primitive visceral skeleton, 36f Duck-billed dinosaur (hadrosaur), 154 External fertilization, 129 Index 259

External mechanical stimuli reception, G in fishes, 85–86 Galeaspids electroreception, 88–89 freshwater colonizations, 13 acousticolateralis system, 90 radiation of, from South China, 15f ampullary or lateral-line organs, 90 Gallus gallus (domestic chicken), 133, 204 ampullary receptors, 91 eggs, 174 electroreceptor organ of, 89f eggshell, 134f function of, 93–94 Ganoin, 79 mormyromasts, 90 Ganoine, 80, 81 in non-teleost taxa, 88 Ganosteus, 48 small pit organs, 90 Ganosteus stellatus, 45f tuberous organs, 91 Gas chromatography mass spectrometry mechanoreception in (Py-GCMS), 169 inner ear, 85 Geckos, 139f, 143 lateral line (LL) system, 85, 86, 87–88 adhesive system, 145f calcareous shells, 134f Gekko gecko, 143 F frictional adhesive, 144, 144f Feeding, in fishes, 84–85 scanning electron micrographs of challenges, 84 scale, 144f teleost feeding, changes, 84–85 structural hierarchy of, 145f jaw expansion and protrusion, 86f Geologic time scale, 2f Fins to limbs, 117–122 Geological and Palaeontological Museum, Acanthostega, 118, 121 Tubingen (GPIT), experimental cladogram of transformation of, 119f models of, 224f Elginerpeton, 123 Gill arches, 4, 33–35, 49, 110 Eusthenopteron, 118 Givetian, 5, 41f, 54f, 111 Ichthyostega, 117, 119 Glanosuchus, 200, 211 limbs of, 119 Global environmental changes, 99–100 Panderichthyids, 118 Glomerular kidney, 13 pectoral and pelvic fins, comparison ion-regulatory function, 15 of, 121f in osmoregulation, 13 pectoral and pelvic girdles, 122 Gnathostomes, 3 Rhizodontids, 118 genus level diversity curves, 54f Tulerpeton, 119 peri-Hangenberg extinction First mammals, 207–209 selectivity, 54 First tetrapods, 114–115 post-Hangenberg configuration of Upper Devonian to the Triassic, 114 vertebrate biodiversity, 56 Fish evolution, morphological (and ecological) species-level faunal composition, convergence in, 71f 55-56f tail shape, improvement in, 72 Gnathostomes, scales, 77–81. See also Scales Fish-tetrapod transition, 4–5 cosmoid scale, 81 amniotic eggs, 5 cosmine, 81 reptilian integument, 5 Eusthenopteron foordi, 80f Fitness, Darwinian concept of, 1 ganoid scales, 79 Forelimbs vs. hind limbs, 227, 228 ganoine, 80–81 synchronous movement, 233 Gogonasus, 105, 107f, 110 terrestrial locomotion, 230 Goldfish. See Carassius auratus, 87 Four wing flight hypothesis, 228 Gondwana dinosaurs. SeeOuranosaurus Four wing problem, 226 Gooloogongia, 105 Fourier transform infrared spectroscopy Gorgonopsian, 198, 199 (FTIR), 169 Rubidgina, skull of, 199f Frey and Riess’ model, 227 Gorgonopsid, 199 260 Index

Lycaenops, 198f I Goronyosaurus, 240 Ichnofossils, 181. See also Dinosaur trace Grand animal de Maastricht, 238, 238f fossils Green anole lizard. SeeAnolis carolinenis Ichthyosaurs, 149, 233 Gymnotiform fish. See Eigenmannia, 89f fastest swimmers, 235 GLAHM V1180, 236f life reconstruction of, 234f H skin preservation, 235 Hainosaurus bernardi, 239 Ichthyostega, 103f, 108, 110f, 112f, 114, 115, life reconstruction of skull of, 240f 123 Hair, evolution of, 201–205 external fertilization, 133 functional sensory vibrissae, 202 hind limb skeleton of, 113f neuro-muscular system, 203 reconstructions, 116, 117f hair keratins, 204 transformation of fins to feet, 119f microvibrissal and macrovibrissal Ictalurus, 93 systems, 204 Ictidosuchoides longiceps, 202f rostral foramina, skull of Dixeya, 203f Immunological analyses, 170 structural proteins, co-option, 205 ‘‘Inside-out’’ hypothesis, 4, 49, 51f vibrissae, importance of, 204 Invertebrates, 146, 163, 212, 249 vibrissal organization, 203 marine invertebrates, 249 whisker, 204 Iron-rich ground waters, 168 Hammerhead shark. SeeSphyrna lewini Ischnacanthus, 67f Harderian glands, 211 Isurus oxyrinchus (shortfin mako), 81, 82 Hedgehog, 144 scales in, 84f Hell Creek Formation, North Dakota, 167 Hemicyclaspis, 24f Herbivorous diet, 208 J alternative tooth replacement, 208 Jamoytiiformes, 25 tribosphenic molar, 208 Jamoytius, 24, 26f Heterostraci, 19–22 body parts and topological interpretation of arandaspids, 19 holotype, 28f astraspids, 19 phylogenetic analysis, 29f eriptychids, 19 Jamoytius kerwoodi, 23, 27f heterostracans Jarvik’s hypothesis, 105 importance of, 22 Jawless placoderms, 4 oral plates of, 16, 17f Jaws, evolution of, 33–37 ostracoderms, 20f ossifications, 33 Hiemenia ensis, 82f chondrocranium, 33 Higher teleosts, 72 primitive jawed vertebrates Holocephalans, 65 acanthodians (class Acanthodii), 33 modern, 66 placoderms (class Placodermi). See Homalacanthus, 67f Placoderms Homo sapiens, 210 primitive visceral skeleton, 36f Hybodus, 63f sharks, 33 Hydrodynamics, in fishes, 94–95 Jeholodens jenkinsi, 217 Hydrofoils, 94, 95, 226, 227f, 228, 230, Jewels of the seas. See Diatoms 231, 231f Juramaia sinensis, 217 Hydrosaurus, 180 holotype specimen of, 215f Hypacrosaurus, 180 Jurassic allosauroid theropod (Lourinhano- Hyphalosaurus baitaigouensis, 135 saurus antunesi), 176 eggs and hatchlings, 135f Jurassic ichthyosaurs, 8, 234, 235, 236 Index 261

hypothesized life reconstruction of, 8f, 234f Liopleurodon hoffmanni, 242 Jurassic period Lissamphibians, 108 Chewore region of Zimbabwe, 182 extant, 100, 101 Darwinopterus, 137 Liver, 95 diverse elasmobranch group in, 63f in Charcarodon carcarias, 96f, 123, 228 insectivorous mammals, 206 Living and fossil craniates, interrelationships Juramaia. SeeJuramaia sinensis of, 60f Juravenator. SeeJuravenator Loganellia (Thelodonti), 50f Lourinhanosaurus antunesi, 176 Lophura, 180 Massospondylus. SeeMassospondylus Lower Carboniferous stegocephalians, 103f modern sharks, 61 Lower Cretaceous, theropod eggs and embryos Morganucodon skull, 207 from, 175f tail mechanics, 72 Lower Cretaceous Yixian Formation, 177, 212 theropod trackways in Zimbabwe, 185f Lower Devonian, 14f, 17f Juravenator, 164, 165 cephalaspid, 4f Juravenator starki, integument of, 165f Cephalaspis signata, 25f Drepanaspis, 20f Hiemenia sp., 82f K Radotina, dorsal view of skull, 38f Kenichthys, 105, 106f Lower Eocene Keratin, 137, 144, 249 Eolates, 70f fibrils, 145f Lower Jurassic of South Africa, 173 human hair keratin genes, 204 Lower Jurassic Posidonia Shale of Southern -like gene, 204 Germany, 235 scales, 146, 147f Lower Ordovician rocks of Russia, 14 shell, 146 Lower Permian Kiearaspis, cephalic head shield, 4f Dimetrodon, 195f K-T extinction, 251 mammal-like reptiles from, 193 large meteorite crater at Yukaton, 252f Lower Triassic Cynognathus Zone, 205 Thrinaxodon, 211 L skull of, 200f Lacerta, 208 Lungfishes (Protopterus), 123 Lacertadracaena, 238–239 subclass Dipnoi, 74 Lacerta vivipara, 209 Lungs, 123 dental development, 209 lung ventilation, 124 Late Jurassic theropod dinosaur. See Juravenator Late Permian M larger therocephalians, 199 Mammal-like reptiles, 6, 149, 193 primitive therapsids in, 197 Cynodonts, 205–207 therapsids extinction, 207 from captorhinomorphs, 193 Lateral line (LL) system, 85, 86, 87 hair, evolution of, 201–203 lateral line (LL) organ, 88f life reconstruction, 7f neuromasts, 87 Pelycosaurs, 194–197 canal neuromasts (CNs), 87 therapsids, 197–201 cephalic lateral line canals, 87 Mammals, emergence, 210–218 superficial neuromasts (SNs), 87 anatomical and physiological characters, Lepidosauromorpha (lizards), 134 210 Lepidosaurs, 153, 239 constant body temperature, 210 epidermis of, 142 eutherian-metatherium split, 212 The life of a fossil hunter (Sternberg), 186 mammalian middle ear, 218 Liopleurodon ferox, 224f metabolic rate, 211 mounted skeletons, 156f near-endothermy, 211 262 Index

upper molars, enechelon postvallum Middle Ordovician shearing of, 216f of America, 19 Mandible, 39, 61, 136f, 205, 246 of Spitsbergen and Australia, 19 in Eomaia scansoria, 214f Middle Ordovician Harding Sandstone of of Maotherium asiaticus. SeeMaotherium Colorado, 13 asiaticus Middle Permian, primitive therapsids, 197 Meckel’s cartilage, reabsorption of, 218 Middle Triassic cynodont therapsids, 211 replacement bones of, 36f Mimia, 70f Maotherium asiaticus, 218 Mirotenthes digitipes, 202f dentition and mandible of, 217f Modern fishes, evolution of, 59–61 Marine reptiles, 149, 221, 222f, 236–242 buoyancy. See Buoyancy Massospondylus (sauropodomorph dinosaur), Chondrichthyes. See Chondrichthyes (car- 173 tilaginous fishes) Matthew, WD, 153, 154, 155 external mechanical stimuli. See External MCF-PVPH-126, 172, 173 mechanical stimuli reception, in MCF-PVPH-130, 172 fishes MCF-PVPH-135, 172 feeding. See Feeding, in fishes MCF-PVPH-140, 172 hydrodynamics, 94–95 MCF-PVPH-147, 173 Osteichthyes. See Osteichthyes (bony MCF-PVPH-639, 132 fishes) MCF-PVPH-680, 173 scales. See Scales MCF-PVPH-681, 173 Morganucodon, 6, 207, 207f, 210, 211, 212 MCF-PVPH-682, 172 Morganucodon oehleri, 207 MCF-PVPH-686, 172, 173 Morganucodon watsoni, 207 The Meaning of Evolution (Simpson), 6, 19 Morganucodontids, 211, 212 Medoevia, 105, 107f Morrison Formation, United States, 184, 186 Meemannia, skull roof, 83f Mosasaurs Meemannia eos, 81 disease, injuries, and bone repair in, Melanosomes, 164 242–248 Mesozoic, 61, 216 anterior profiles of contemporary mo- Chondrichthyes (cartilaginous fishes), sasaurs, 243f radiation, 61 bone healing process, 242 Ichthyosaurs, 233–234 chronic osteomyelitis, 246 Jurassic ichthyosaurs, 235 deficient healing and infection, 243 marine reptiles of, 221 distinct calluses, 246 modern chimeras, 66 Meckelian Fossa, 247 Ornithothoraces and modern birds, syna- poor nutrition, 246 pomorphy of, 175 secondary fibrocartilage, 246 reptile radiation, 146–147 surviving fracture, 247 Middle Cambrian teeth of, 247 Pikaia, 15 violence of predation, 242 Middle-Cambrian Burgess Shale, 15 extinction, 251–252 Middle Devonian integument of, 248–251 Ctenurella, 38f epidermal and dermal structures in, Eifelian, 111 250f Hiemenia sp., 82f multiple keels, 250 Panderichthys, 108 Upper Cretaceous period, 236 Pycnosteus, 20f from upper Santonian-lower Campanian, Tiktaalik, 110f 249f trackways, 116 Mosasaurus hoffmanni, 239f transformation of scales, 83f Mosasaurus hoffmanni IRSNB R27, 245f, 246 Middle Jurassic, Chewore region of Zimba- Mosasaurus hofmanni IRSNB R25, 243, 244f, bwe, 182 246 Index 263

Mosasaurus hofmanni IRSNB R26, 241f lobe-finned fishes, 108 axial subundulatory, 242 radiation in, 108 teeth of, 242 Osteostraci, 23–30 Moschorhinus kitchingi, 202f Anaspids (Anaspida), 23, 26f Moythomasia, 70f Arandispis, 23 Multicusped molar, 212 Astraspis, 23 Murtoilestes, 216 cephalaspids, 23, 24f benthic mode, 23 cephalic shield, 23 N orbital fenestrae, 23 Nectaspis areolata, 25f ostracoderms, 24f Newman hypothesis, 227 pectoral fins, 23 Niobrara Formation in western Kansas, USA, gnathostomes, 23 250 early Silurian Jamoytius kerwoodi, 23 Nothosaurs, 222 olfactory ducts of, 28 cast of fossil, 223f stem-gnathostomes, 25 phylogeny of, 223f nasohypophysial complex, 26 Ostracoderms, 1, 3, 11–12 anatomy of, 3 O bone, 16–17 Odontogenic ameloblast associated (ODAM) feeding, 16 gene, 52 fossilized impression, 12f Ordovician, 1, 11, 15, 49 heterostracans, 16, 17f. See also Acanthodians or spiny sharks, 66 Heterostraci Astraspis, 20f importance of, 22 fine tubules in, 46 Osteostraci, 23–30 bony armor in, fragments of, 19 Anaspids (Anaspida), 23, 26f environment of early vertebrates in, 13 Cephalaspids, 23, 24f Ornitholestes, 154 gnathostomes, 23 Orthodentine, 79, 80 shapes and sizes, 12f Osborn, Henry Fairfield, 158 Ouranosaurus (Gondwana dinosaurs), 180, Osteichthyans (bony fishes), 37, 39, 74, 79 181 Osteichthyes (bony fishes), 59, 66 Ouranosaurus nigeriensis, 181f class Acanthodii, 66–67 ‘‘Outside-in’’ hypothesis, 4, 49, 51f class Actinopterygii, 68 Oviraptor philoceratops, 176 subclass Cladista, 69 Oviraptorids, 176 subclass Chondrostei, 69–70 order Palaeonisciformes, 69 subclass Neopterygii, 72–74 P dorsal fins, in evolution of, 74, 75f Paired fins, 23, 33, 72 tail morphologies of living and fossil evolution of, 37 forms, 73-74f Paleozoic, 18 subclass Sarcopterygii, 74–77 Chondrichthyes, 61 actinopterygian-sarcopterygian split, 74 Glanosuchus, 211 osteolepiforms, 77 jawless vertebrates, 50 tetrapodomorphs, 77 palaeoniscoids, 69 Osteolepidiformes, 77 sharks Osteolepiform-porolepiform-panderichthyid body form diversity, 62f (subclass Tetrapodomorpha), 74 infraclass Euselachii, allied Osteolepiforms, 104, 105, 107, 113 with, 63f Eusthenopteron, 110f, 118 Panderichthys, 108, 109, 110, 110f, 111 fin rays in, 118 pectoral and pelvic fins, 121f in interrelationships of bony fishes and tail reconstructed from, 112f four-legged vertebrates, 109f transformation of fins to feet with, 119f 264 Index

Panderichthys rhombolepis, 121f Pleistocene, 180 Papilla, 49 Plesiosaurs, 149, 222 epidermal papillae, 146 active 4-wing swimming, 232 in extant reptile scales, 165 anterior limbs papulose surface with, 167f as power stores, 233 Parapsid, 149 sweepback of, 229f Parasemionotus, 70f 4-wing flight in, 231 Pectoral apparatus, 4 limb musculature of, 233 Pectoral fins, 23, 37, 61, 62, 67f, 74, 77f, 91, phylogeny of, 223f 95, 106, 107f, 109, 112f, 118 plesiosaurids, 222 cladogram of, 122 swimming to sea lions, 226 and pelvic fins, 121f underwater flight, 231f of Tiktaalik, reconstruction of, 111f Plesiosaurus brachypterygius, 224, 224f Pectoral girdles, 77f, 107f, 122 Plioplatecarpus marshi, 242, 243f Pectoral musculature, 226 Pliosaurids, 224, 228, 229 Peloneustes philarchus Plotorosaurus, 243f experimental model of, 224f Plotosaurus, 250 mounted skeletons, 225f Plotosus, 90 Pelvic claspers, 66 Pompous oration, 107 Pelvic fins, 37, 61, 62, 67f, 74, 77f, 99, Pogona vitticeps (embryonic bearded dragon), 118, 121f 146 and pectoral fins, 121f Posterior limbs, 122, 161, 226, 228, 229, 230, Pelvic girdles, 77f, 121f, 122 231, 232, 233 Pelycosaurs, 194–197 Pre-gnathostomes, 34 Dimetrodon. See Dimetrodon Primitive visceral skeleton, 36f Edaphosaurus. See Edaphosaurus Probelesodon, 6 pictorial family tree of, 197f Prokennalestes, 216 Sphenacodontidae, 194 Promoschorhynchus, 202f Permian, 6 Proto-bird Archeopteryx, 110 boundary crisis, 193 Psammosteus, 42, 43 carnivorous pelycosaur from, 195f Psammosteus megalopteryx, 45f deposits, 197 Psarolepis romeri, 81 Edaphosaurus in, 196f Psittacosaurus (small psittacosaurid dinosaur), modern holocephalans, 66 159 Rubidgina from, 199f head and thoracic regions of, 161 therapsids scales array of, 198 pigmented impressions of, 161 extinctions in, 207 types of scales preserved, 161, 162f therocephalians, 211 SMF R4970, 160, 160f -Triassic boundary, 6 left shoulder of, 163f Pharyngolepis, 26f preserved scales of, 162f Phu Phok eggs, eggshell microstructure of, Psittacosaurus sp. MV53, 177–178, 178f 175f dermis Phylogenetic Systematics, 100, 101 fiber angle, 179 Phytoplanktons, 17, 18 multiple fiber layers, 179 Placoderms, 35f life and death struggles, 179 Arthrodires, 38f Pteraspis, 20f, 42 Devonian shift, 39 Pteroniscus, 85f dominant vertebrates of Devonian, 37–42 Pterosaurs, 189 gnathostomes, 39 Ptomacanthus, 39 functional mandibular disparity among, Ptomacanthus anglicus, 40-41f 41f Ptomocanth anglicus, 39 and ostracoderms, 39 Pulse-like discharge, weak, 90 Platecarpus, 243f, 250, 251 Pycnosteus, 20f Index 265

R striate-like rows, 173 Rainbow trout (Oncorhynchus mykiss), 88 trace fossils, 170f Rana esculenta, 126 Sauropterygia, phylogeny of, 223f Ray-finned fishes, jaw evolution in, 85f Scale colour, camouflage, 142 Reptiles Scale types, 138–146, 139f egg shell, 136 elongated scale (frill), 141 eggtooth or eggbreaker, 131 exquisite arrangement, 142 integument, 137–138 and hormonal control, 142 cutaneous respiration, 137 oberhautchen, 143 keratin, 137 overlapping scales, 138, 141 return to sea. See Mosasaurs reticulated scale pattern, 141f bone elongation, reduction in, 221 scale renewal, in crocodiles, 142 robust legs to fleshy fins, 221, 222 setae, 143 temporal openings and classification of, tuberculate scales, 138 146–149 unusual types, 140-141f Captorhinida, 148 Scales Cotylosauria, 148 cosmoid scale, 81 phylogeny based on, 148f cosmine, 81 Reptilia, 153 ganoid scales, 79 reptilian integument, 5 ganoine, 80–81 Rhaetian Stage, 206 and swimming hydrodynamics, 81–84 Rhizodontiformes, 77 Hiemenia sp., 82f Rhomaleosaurus victor, 230f transformation of, 83f Rhynchocephalia, 141, 153 Scanning electron microscopy, 170, 170f Ruling reptiles, 153, 221 Sebastes, 85f Semidentine, 77 Senckenberg Museum in Germany, 159 S Sharks, 59, 61, 67, 77, 88 Salmo, 85f ampullary receptors, 91 Salmon, lungfish, and cow, 104 behavioral experiments, 90 relationship of, 105f infraclass Euselachii, allied with, 63f Sandbar shark. SeeCarcharhinus plumbeus lamnid sharks, 95, 235 Sarcopterygian fish. SeeTiktaalik roseae locomotion of, 81 Sarcopterygian fishes, 4–5 marine sharks, 93 descendants for, 108–114 negatively buoyant, 95 epistostegalian-tetrapod divergence, Paleozoic sharks, body form diversity, 62f 113 spiny sharks, 66 fishes and land vertebrates, 110 swimming sharks, 81 foot morphologies, 113f thunniform sharks, 228 osteolepiform fishes, 113 tooth replacement system, 63 right pectoral fin of 111f white sharks, 95 tetrapod trackways, 112f Shastasaurus sikanniensis, 234, 235 Tiktaalik. SeeTiktaalik skull and jaws, 234 evolution of tetrapods from, 4–5 Shortfin mako. SeeIsurus oxyrinchus Sauropodomorph dinosaur. Shuyu zhejiangensis, 30f SeeMassospondylus Silky shark. SeeCarcharhinus falciformis Sauropods Simpson, George Gaylord, 6 embryos, bone morphology of, 171f Sinosauropteryx, 164, 165 patterns of integument, 172 IVPP V12415, 164, 166f, 167f flower-like arrangement, 172–173 postmortem opisthotonus, 165 ground tubercles, 172 life restoration of, 168f large, elongated tubercles, 172 melanosomes, 164 parallel rows of large tubercles, 172 Skeletal mineralization, evolution rosette-like arrangement, 172 of, 53-54f 266 Index

Skull, 34, 72, 107, 115, 136f, 196f, 234, 246, Synapsida, 149 247 Synchrotron Rapid Scanning X-ray Fluores- in Cynognathus, 205 cence (SRS-XRF), 169 dermal skull bones, 107f in Dimetrodon, 195f of Dixeya, 203f T of Hainosaurus bernardi, 240f Tartuosteus maximus, 45f of Juramaia sinensis, 215f Tatarinov, Leonid B, 201 of Lower Triassic cynodont Thrinaxodon, Teeth, origin of, 49–50 200f canonical view, 49 Meemannia skull roof, 83f dental lamina, 49 of Morganucodon, 207 genes and, 50–54 of Mosasaurus hofmanni IRSNB R26, 241f gene duplication, 51 occipital view of, 194f SPARCL1 (Sparc-like 1), 50–51 in Natural History Museum, London, 239, ‘‘inside-out’’ hypothesis, 49 238f odontode evolution of placoderms, 77 inside and out gene regulatory hypoth- in Radotina, dorsal view of, 38f esis for, 52f replacement bones of, 36f theories of, 49–50, 51f of Rubidgina, 199f oral vs. dermal tooth-like units, evolution of selected therocephalian taxa, 202f of, 49 skull reconstruction, of titanosaurian ‘‘outside-in’’ hypothesis, 49 embryos, 132f second generation tooth, 49 skull roofs, 110f Teleostomi, 37 in Acanthostega, 114 Teleosts, 54f, 66, 88, 93 structure of, and elasmobranch infraclass- higher, 72 es, 64 diversification of dorsal fin in, 75f temporal openings of, 147 paired fin locations, phylogeny in tetrapodamorpha, 77 of, 76-77f in Therapsids, changes in, 197 modern, 44, 72 Small angle X-ray scattering, 169 palaeoniscoids, 71f Small psittacosaurid dinosaur. See Terrestrial vertebrate locomotion, evolution of, Psittacosaurus 115–117 Soft-shelled turtles. SeeTrionyx Acanthostega, 115 SPARC (secreted protein, acidic, and rich in Genoa River tracks, 116f cysteine), 50 Ichthyostega, 115, 117f SPARCL1 (Sparc-like 1), 50–51 Testudines, 153 secretory calcium-binding phosphoprotein Tetrapodamorpha, 77 (SCPP) gene family, 50 Tetrapodomorph stem group Sphyrna lewini (hammerhead shark), 91 new phylogeny of, 106f Spinosaurus, 180 Tetrapodomorpha, 74, 104, 107 Spiny sharks, 39 Tetrapods, 100–105. See also Breathing on Squalus, 63f land, problems of; Fins to limbs; Squamata, 141, 153 Sarcopterygian fishes Stenohaline, 13 aquatic and terrestrial existence, dichotomy Stenopterygius quadricissus SMF 457, 236f between, 100, 101f collagen fibrils in, 237f definition of, 100 high-tensile fibers near the fin base, 237f phylogenetic definition of, 101 Stensio, Erik, 3 Devonian and Lower Carboniferous Styloichthys changae, 81 stegocephalians, 103f Suction feeding, 84 elongated neural spines, 181f Superficial neuromasts (SNs) oxygenating the body, 100 orientation (rheotaxis), 87 Sarcopterygian limbs, 102f Synapomorphies, 104 Thecodont (socket-toothed), 193, 207, 221, 240 Index 267

Thelodonts insectivorous mammals in, 206 faunal provinces and palaeogeography, 14f mammal-like reptiles, 193 in variations in salinity, 13 neopterygians, 72 Therapsids, 193, 197–201 Parasemionotus, 70f advanced, 197 reptilian invasion of seas, 221 carnivorous, 197 therapsids extinctions, 207 canine and cheek teeth, 198 therocephalians, 205 dentition, 197 Thoracopterus, 71f gorgonopsians, 198 Thrinaxodon, skull of, 200f therocephalians, 198, 199 Tribosphenic molar, 212 cranial restorations of, 202f Trionyx, 48 pictorial family tree of, 201f Trionyx sinensis, 47f primitive forms, 197, 199 Troodon formosus, 176 Theriognathus, 201, 202 Tuberous electroreceptor organs, 91 Theriognathus microps, 202f active electrolocation, 92 Thoracopterus, 71f electric signals, 94 Thrinaxodon, 197, 201, 211 for encoding timing of electric organ skull of, 200f discharge, 93 Thunniform shark (Carcharodon carcharias), for encoding stimulus amplitude, 93 95, 96f Tulerpeton, 103f, 119 ecomorphological interpretations, 96 Tursiops aduncus, 95 feeding niche partitioning, 95 Tyrannosaurid, 159 Thyestes (Silurian), 24f Tyrannosaurus, 154 Tiaojishan Formation of Liaoning Province, China, 136f Tiktaalik, 110f U fishes and land vertebrates, 110 Underwater acoustics, 85 right pectoral fin of, 111f Upper Cambrian, 15 Tiktaalik roseae (sarcopterygian fish), 109, Upper Carboniferous, 193 110 Upper Cretaceous, 236, 249 Titanosaurs, 132f, 176 Edmontosaurus sp. (MRF-03), 167 skeleton of Saltasaurus loricatus, 171 Upper Devonian Toucan, 146 Acanthostega gunnari, 123 Ramphastos toco, 146 Ctenacanthus, 63f structure of, 147f Eustenopteron foordi, 80f Trace fossils, 167 Hiemenia sp. scales, 82f dinosaur tracks, 181 labyrinthodonts, 114 sauropod trace fossils, 170f palaeoniscoid fishes from, 70f Trachodon (hadrosaur) Sauripterus, 118 mounted skeleton of, 156f Tulerpeton from, 119 mummy, skeleton of, 157f Upper Permian, 193 skin impression from tail of, 159f Rubidgina from, 199f theory of color pattern, 157 swimming, 157 V Transmitted light microscopy, 170, 170f Varanus komodoensis, 239f Tremataspis (Silurian), 24f Varanus niloticus, 239, 248f Triassic, 6, 114, 198, 207 Vertebrates, 22, 33, 123, 179, 222 cynodonts amniotic eggs, 133 life reconstruction of, 7f bone. See Bone, evolutionary significance therapsids, 211 of oldest vertebrates elasmobranch group in, 63f bony vertebrates, 39 gorgonopsians, 205 dermis in, 7 268 Index

extant vertebrates, 7 tetrapoda, 100, 106f, 109f dominant vertebrates of Devonian. See Vitellin, 18 Placoderms early evolution of, 15 early vertebrates, 3, 23 W phosphate store in, 16–17, 18 Watson’s hypothesis, 225 extinct vertebrates, 3, 53-54f Wave-type discharge, continuous, 90 first vertebrates, 1, 4, 18 Weigeltaspis goclmani bone-like nature of armor of, 42 Sharpey’s fibers in basal aspidin, 45f fossil vertebrates, 212, 248 White shark. SeeCarcharodon carcarias; freshwater origin of, 14, 18 Carcharodon carcharias jawed vertebrates, 3, 28, 49, 51 Wnt signaling pathways, 144 earliest. See Gnathostomes evolution of jaws in, 49 jawless vertebrates, 3, 4, 19, 48, 49, 50, 52 X land vertebrates, 39, 110 Xenocanthus, 61, 62f lung ventilation in, 124 Xenopus laevis, 126 marine vertebrates, 7, 13, 177, 221, 235 modern vertebrates, 1, 3 non-fish vertebrates, 103 Y nonmammalian vertebrates, 208 Yixian Formation in Liaoning, China, 160, Ordovician period, 11 177, 212, 218 environment of, 13 phylogeny of, 53-54f proto vertebrates, 18 Z terrestrial vertebrates, 16 Zahnreihen, 208 bone, as calcium store, 16