Feeding Form,. Function, and Evolution in Tetrapod Vertebrates
This Page Intentionally Left Blank
Feeding Form, Function, and Evolution in Tetrapod Vertebrates Edited by
Kurt Schwenk Department of Ecology and Evolutionary Biology University of Connecticut Storrs, Connecticut
:#) San Diego
San Francisco
ACADEMIC PRESS A Harcourt Science and Technology Company
New York
Boston
London
Sydney
Tokyo
Cover photographs: (Inset) Harry W. Greene, © 1999. (Background) Nirvana Filoramo, © 2000. This book is printed on acid-free paper.
^^
Copyright © 2000 by ACADEMIC PRESS All lights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Requests for permission to make copies of any part of the work should be mailed to: Permissions Department, Harcourt Inc., 6277 Sea Harbor Drive, Orlando, Florida 32887-6777
Academic Press A Harcourt Science and Technology Company 525 B Street, Suite 1900, San Diego, California 92101-4495, USA http: / / www.academicpress.com
Academic Press Harcourt Place, 32 Jamestown Road, London NWl 7BY, UK http: / /www.academicpress.com Library of Congress Catalog Card Number: 99-63490 International Standard Book Number: 0-12-632590-1 PRINTED IN THE UNITED STATES OF AMERICA 00 01 02 03 04 05 EB 9 8 7 6 5
4
3
2
1
To my teachers: Warren F. Walker, Jr., James R. Stewart, and Marvalee H. Wake
This Page Intentionally Left Blank
Contents
Contributors Preface xiii
IV. Kinematics of Feeding: Feeding Stages V. Concluding Remarks 55 References 55
xi
49
S E C T I O N S E C T I O N
I
n
INTRODUCTION
AMPHIBIA
C H A P T E R C H A P T E R
Tetrapod Feeding in the Context of Vertebrate Morphology
Aquatic Feeding in Salamanders STEVEN M. DEBAN AND DAVID B. WAKE
KURT SCHWENK
I. II. III. IV. V.
I. Introduction 3 11. Approaches to the Study of Tetrapod Feeding 5 III. Concluding Comments 16 References 16
C H A P T E R
Introduction 65 Morphology 68 Function 82 Diversity and Evolution 88 Opportunities for Future Research References 92
92
C H A P T E R
An Introduction to Tetrapod Feeding
Terrestrial Feeding in Salamanders
KURT SCHWENK
DAVID B. WAKE AND STEPHEN M. DEBAN
I. Introduction 21 II. Morphology of the Feeding Apparatus III. Kinematics of Feeding:The Gape Cycle
I. Introduction 95 II. Morphology 97 III. Function 101
26 47
Vll
Contents
Vlll
IV. Diversity and Evolution 111 V. Opportunities for Further Research References 114
S E C T I O N
114
IV REPTILIA: LEPIDOSAURIA
C H A P T E R
C H A P T E R
Feeding in Frogs
8
KIISA C. NISHIKAWA
I. II. III. IV. V. VI. VII.
Introduction 117 Morphology of the Feeding Apparatus 119 Function of the Feeding Apparatus 124 Neural Control of Prey Capture 135 Evolution of the Feeding Apparatus 139 Conclusions 143 Current and Future Directions 144 References 144
C H A P T E R
Feeding in Lepidosaurs KURT SCHWENK
I. Introduction 175 II. Lepidosaurian Phylogeny and Classification 176 III. Natural History 178 IV. Morphology of the Feeding Apparatus 189 V. Feeding Function 220 VI. Specialized Feeding Systems 257 VII. Evolution of Feeding in Lepidosaurs 264 VIII. Future Directions 277 References 278
C H A P T E R
Feeding in Caecilians JAMES C. O'REILLY
I. 11. III. IV. V
Feeding in Snakes
Introduction 149 Morphology 150 Function 155 Evolution 161 The Future 163 References 164
DAVID CUNDALL AND HARRY W. GREENE
I. II. III. IV. V.
Introduction 293 Form and Function 301 Performance and Size 322 Evolution 322 Concluding Remarks 326 References 327
S E C T I O N
III REPTILIA: TESTUDINES
S E C T I O N
V REPTILIA: ARCHOSAURIA
C H A P T E R C H A P T E R
A Bibliography of Turtle Feeding KURT SCHWENK
I. Introduction II. Bibliography
169 169
10
Feeding in Crocodilians
JOHAN CLEUREN AND FRITS DE VREE
I. Introduction II. Morphology
337 340
Contents III. Function 347 IV. Evolution 354 References 357
IX
V. The Feeding Apparatus 421 VI. Feeding Function 439 VII. Control of Feeding Behaviors 444 References 444
C H A P T E R C H A P T E R
11 Feeding in Paleognathous Birds CAROLE A. BONGA TOMLINSON
14 The Ontogeny of Feeding in Mammals R. Z. GERMAN AND A. W. CROMPTON
I. 11. III. IV. V
Introduction 359 Materials and Methods 360 Morphology of the Hyolingual Apparatus 361 Function of the Hyolingual Apparatus 373 Evolution of the Feeding System 384 References 390
C H A P T E R
"12
I. II. III. IV. V.
Introduction 449 Morphology 449 Function and Mechanics of Suckling 450 Rhythmicity and Control of Suckling 453 Coordination of Swallowing and Respiration 455 VI. Transition from Suckling to Drinking at Weaning 455 VII. Evolutionary Considerations 456 References 456
Feeding in Birds: Approaches and Opportunities
C H A P T E R
MARGARET RUBEGA
I. Introduction 395 II. Patterns of Analysis III. Conclusion 406 References 406
396
S E C T I O N
VI MAMMALIA
15 Feeding in Myrmecophagous Mammals KAREN ZICHREISS
I. II. III. IV. V.
Introduction 459 Foraging Ecology 462 Morphology of the Feeding Apparatus 464 Functional Morphology 475 Evolution of Myrmecophagous Specializations 478 VI. Directions for Future Research 480 References 481
C H A P T E R
C H A P T E R
13
16
Feeding in Mammals
Feeding in Marine Mammals
KAREN M.HIIEMAE
ALEXANDER WERTH
I. Introduction 411 II. Mammalian Feeding System 414 III. The "Process Model" for Mammalian Feeding 416 IV. Mechanical Properties and Textures of Foods 419
I. Introduction 487 II. Feeding Strategies 492 III. Conclusions 521 References 521 Index
527
This Page Intentionally Left Blank
Contributors
Kiisa C. Nishikawa (117) Department of Biological Sciences, Northern Arizona University, Flagstaff, Arizona 86011. James C. O'Reilly (149) Program in Organismic and Evolutionary Biology, University of Massachusetts, Amherst, Massachusetts 01003. Karen Zich Reiss (459) Department of Biological Sciences, Humbolt State University, Areata, California 95521. Margaret Rubega (395) Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut 06269. Kurt Schwenk (3, 21,169,175) Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut 06269. Carole A. Bonga Tomlinson (359) Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138. David B. Wake (65,95) Museum of Vertebrate Zoology, University of California, Berkeley, California 94720. Alexander Werth (487) Department of Biology, Hampden-Sydney College, Farmville, Virginia 23901.
Numbers in parentheses indicate the pages on which the authors' contributions begin.
Johan Cleuren (337) Department of Biology, University of Antwerp, B-2610 Antwerp, Belgium. A. W. Crompton (449) Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138. David Cundall (293) Department of Biological Sciences, Lehigh University, Bethlehem, Pennsylvania 18015. Frits De Vree (337) Department of Biology, University of Antwerp, B-2610 Antwerp, Belgium. Stephen M. Deban (65, 95) Museum of Vertebrate Zoology, University of California, Berkeley, California 94720. Rebecca Z. German (449) Department of Biological Sciences, University of Cincinnati, Cincinnati, Ohio 45221. Harry W. Greene (293) Department of Ecology and Evolutionary Biology, Cornell University, Ithaca, New York 14853. Karen M. Hiiemae (411) Department of Bioengineering and Neuroscience, Institute for Sensory Research, Syracuse University, Syracuse, New York 13244.
XI
This Page Intentionally Left Blank
Preface
This book addresses the first of these challenges. It examines in depth and breadth the myriad solutions to the essential problem of feeding in one clade of animals. It summarizes and synthesizes for the first time in 15 years our burgeoning knowledge of tetrapod feeding systems and how they have evolved. It explores the "variations on a theme" in this system and in so doing provides grist for evolutionary theorists. However, its proximate goals are more modest. The book is intended to instruct novice morphologists and others interested in animal form and function. It is both an introduction to the field and a presentation of the "state of the art." It is aimed at advanced undergraduate and gradutate students, as well as experts in the field who wish to delve outside their taxonomic bounds. It provides an accessible entree into an exploding literature and showcases our impressive knowledge—but it highlights also our great ignorance. As Rubega points out in Chapter 12, there are many dissertations on tetrapod feeding left to be written. The greatest possible outcome I can imagine for this book is that it will stimulate and provoke the next generation of morphologists to fill in the gaps and shoot down the dogma. In an effort to promote its utility to students, the book begins with two introductory chapters that establish the conceptual, historical, and factual contexts within which the empirical chapters can be interpreted. The empirical chapters provide a more-or-less phylogenetic survey of tetrapod vertebrate feeding systems. Although a phylogenetic approach is emphasized throughout, there are some cases in which I judged other criteria to be more useful in organizing current knowledge. Hence, some chapters are not limited to a monophyletic taxon, but are based on functional types (e.g., "marine mammals," Chapter 16), dietary types (e.g., "myrmecophagous mammals," Chapter 15), or the medium in which feeding occurs (e.g., "aquatic
Vertebrate morphology stands accused of failing to contribute meaningfully to the neo-Darwinian evolutionary synthesis. Although I strongly doubt this (see Chapter 1), I nonetheless take it as a challenge for the future. Our evolutionary theory is at a cusp—the power and efficacy of reductionism are undeniable, but equally so is its failure to deal effectively with intrinsic, organismal attributes. Despite leaps and bounds in our understanding of genetic- and population-level evolutionary phenomena, we remain almost embarrassingly ignorant about the fundaments of phenotypic evolution. Answers to the most basic questions are beyond our grasp: Why do some lineages evolve rapidly while others remain static? How do complex systems full of interacting characters evolve? And once evolved, how can they change? Our approaches to these questions are often simplistic and too facile. We point to "phylogenetic constraint," for example, as if it explains the failure of a lineage to evolve in some expected way when, in fact, it does little more than describe our ignorance. What is the mechanistic basis of phylogenetic constraint? Is it all just genetic background, or are there phenotypic processes that interfere with diversification and adaptive evolution or that facilitate it? Organisms are multihierarchical, complex systems, and as in other such systems, each level expresses emergent properties that are unpredictable, even unknowable, from the vantage point of other levels. If we want to understand the principles governing the evolution of phenotypes, it is logical, indeed, necessary to study the phenotype directly. Who better to do this than morphologists? The challenge to vertebrate morphology in the next century is therefore twofold: to develop the empirical database and conceptual tools needed to create a phenotypebased evolutionary theory, and to forge a new evolutionary synthesis by integrating this theory with the gene-based, neo-Darwinian paradigm.
Xlll
XIV
Preface
feeding in salamanders/' Chapter 3). Each chapter is authored by an expert or experts on the group, including both veteran and younger workers. I am very pleased to be able to include chapters on little-known groups, such as caecilian amphibians (Chapter 6), crocodilians (Chapter 10), paleognathous birds (Chapter 11), myrmecophagous mammals (Chapter 15), and marine mammals (Chapter 16). However, my goal of complete taxonomic coverage of all tetrapods was not quite achieved. Owing to many factors, a chapter on turtle feeding could not be completed. To mitigate this taxonomic breach, I have prepared a brief bibliography of turtle feeding to serve as an entree into the literature (Chapter 7). This book has a long and tortured history—even longer and more tortured than most edited works! It was inspired by a symposium on the ecology and evolution of feeding systems in lower vertebrates presented at the annual meeting of the American Society of Ichthyologists and Herpetologists in Austin, Texas, in 1993, to which I was a contributor. The symposium was organized by Drs. Peter Wainwright and Kiisa Nishikawa. Dr. Charles Crumly, a systematic herpetologist and editor at Academic Press, was in attendance. I had been toying with the idea of editing a book in the area of feeding, so when Chuck approached me with the idea I was thrilled to take it on. By the end of the meeting several authors were already lined up. That was the easy part! The project ebbed and flowed over the years as the author roster grew and shifted. Consequently, there is a large span of time over which chapters were completed and submitted. Although I have tried to update the literature where necessary, some chapters are inevitably not as current as others. Thus, authors who worked most dilligently to complete their manuscripts in time for early deadlines should not be held to blame for editorial shortcomings. There are many people to thank for their contributions, direct and indirect, to this project. I must begin by expressing my deep gratitude for the inspiration of my teachers, Warren R Walker, Jr., James R. Stewart, and Marvalee H. Wake, to whom this book is dedicated. Warren Walker first taught me vertebrate biology and comparative anatomy as a junior at Oberlin College and it was his deep knowledge and masterful teaching that led me to embark on a career in vertebrate morphology. I remain in awe of his knowledge of comparative anatomy; his course serves as the benchmark from which I measure my own feeble attempts. Warren had the poor taste to take a sabbatical leave my senior year at Oberlin, but this sad event (for me) had a positive side—James Stewart was hired to replace him that year. Jim came to Oberlin fresh out of Berkeley with a
new set of experiences and ideas. I watched firsthand as he put together his own terrific course on comparative anatomy and I was given the opportunity to assist teaching in the lab. Jim supervised my senior thesis research (on lizard feeding!) and became a friend as well as a mentor. His calm, philosophical, and scholarly approach to both life and science deeply impressed me and continues to inspire me now. At Oberlin, Jim regaled me with stories of Berkeley, the Museum of Vertebrate Zoology, and the "Herp Lab,'' so after a short hiatus as a zookeeper at the Bronx Zoo, I was thrilled to be accepted into Marvalee Wake's lab at Berkeley for graduate study. Marvalee, to me, is the quintessential vertebrate morphologist—painstaking, detailed, thorough, and a scholar of the highest order. She was also the perfect advisor. She knew unerringly when to leave me on my own and when to push me. She supported my work and my psyche. Most important, she set a high standard in the lab and maintained it by example. The depth and breadth of her work on caecilian amphibians are a model of achievement and a personal source of inspiration. I have depended on Marvalee's wisdom for the last 20 years and still turn to her when I am in need of counsel. I am profoundly grateful to each of these people who have contributed so critically to my professional, intellectual, and personal development—often in ways they cannot imagine. Whatever strengths my work has shown since are owed to their mentor ship. I thank my friend and editor at Academic Press, Chuck Crumly, for seeing this project through from the beginnning and for alternately holding my hand and kicking my butt, as required. Donna James and Joanna Dinsmore at AP provided much-needed help in the final stages of manuscript preparation, for which I am very grateful. Mary Jane Spring not only prepared some wonderful original artwork for my chapters, but also slaved over a hot scanner to produce many composite plates and other figures for reproduction. My father, George Schwenk, generously produced the penand-ink illustrations that introduce each section of the book. A number of people critically read chapters in whole or in part, offered comments, checked facts, and/or helped with bibliographic sources: William E. Bemis, A. W. Crompton, Nirvana I. Filoramo, Leo J. Fleishman, Harry W. Greene, Susan W. Herring, Dominique G. Homberger, Parish A. Jenkins, Jr., Kenneth V. Kardong, Nate Kley, John H. Larsen, Matthias Ott, Margaret Rubega, Carl D. Schlichting, Adam Summers, Carole Tomlinson, Giinter P. Wagner, Marvalee H. Wake, and Kentwood D. Wells. David Cundall, Harry Greene, Carl Schlichting, and Giinter Wagner supported this effort with their friend-
Preface ship, beer, and a high tolerance for whining. I thank my graduate students. Nirvana Filoramo and Charles Smith, for their forbearance in dealing with a busy and distracted advisor. My family—George Schwenk, Elizabeth Schwenk, Deborah Schwenk, John Schwenk, and Natalia Schwenk—have always been there for me and don't even seem to mind when I lapse into soliloquies about tongues and lizards. Finally, I thank my wife, Sandford von Eicken, and my son, Colton Schwenk, for their love and incredible patience.
XV
Work in my lab and preparation of the manuscript were made possible by grants from the University of Connecticut Research Foundation and the National Science Foundation (IBN-9601173) whose financial support is gratefully acknowledged. Kurt Schwenk Storrs, Connecticut April 2000
This Page Intentionally Left Blank
S E C T I O N
I
ff^^y^/^^y^,'
This Page Intentionally Left Blank
C H A P T E R
1 Tetrapod Feeding in the Context of Vertebrate Morphology KURT SCHWENK Department of Ecology and Evolutionary Biology University of Connecticut Storrs, Connecticut 06269
I. INTRODUCTION A. Why Study Feeding? B. Delimiting the Topic II. APPROACHES TO THE STUDY OF TETRAPOD FEEDING A. Themes in Vertebrate Morphology B. Schools of Vertebrate Morphology C Techniques of Vertebrate Morphology III. CONCLUDING COMMENTS References
L INTRODUCTION A. Why Study Feeding? Despite huge strides in our understanding of gene structure and function, we remain largely ignorant about how phenotypes evolve. This apparent paradox arises from the fact that phenotypes embody emergent properties not directly codified in the genes. Fience, these properties are unknowable even with an exact and complete knowledge of the genetic system. It is possible to deduce them only from direct study of the phenotype itself. Knowledge of such properties is essential because it is the phenotype, after all, and not the genotype, that performs in an environmental context and it is this phenotype-environment interaction which determines lifetime reproductive success (fitness). A comprehensive view of organismal evolu-
FEEDING (K.SchwenKed.)
tion and diversity must, therefore, integrate genebased (bottom-up) and phenotype-based (top-down) approaches. Because morphology and function are the direct objects of selection, their study can contribute uniquely to the formulation of such a comprehensive view. Indeed, morphological studies have, historically, spawned many of the fundamental concepts of comparative biology: homology, analogy, adaptation, constraint, and Bauplan, for example. Furthermore, morphological principles might be critical to our understanding of how lineages navigate "phenotype space" through evolutionary time. Why do some systems remain static while others diversify? Do certain phenotypes confer intrinsically stable configurations that resist modification? How are segmental or modular body plans functionally integrated and how does the degree of their integration affect their ability to evolve? Such questions are best addressed through phenotypic analysis and it is in this context that a deep and detailed knowledge of form and function gains its greatest value. General principles of form-function evolution can be approached through a process of induction from specific systems. Tetrapod feeding is attractive in this regard because it offers several attributes that enhance its utility as a model system. First, tetrapod feeding systems are hugely variable, ranging from the edentulous jaws and extraordinarily protrusible tongues of ant- and termite-feeding mammals that feed frequently on many minute prey, to the syringe-like fangs and
Copyright © 2000 by Academic Press. All rights of reproductiori in any form reserved.
Kurt Schwenk venom glands of some caenophidian snakes that feed infrequently on few, very large prey. Within this array of systems, however, there has been the repeated acquisition of certain types. Lingual prehension of food, for example, is found in frogs, salamanders, turtles, lizards, birds, and mammals, and truly projectile tongues have evolved independently within salamanders, frogs, and lizards. Such phenotypic diversity allows us to address basic Darwinian questions regarding pattern and process of evolution and biodiversity; why are there so many types of feeding system? Conversely, the repeated acquisition of functionally analogous systems permits us to examine the evolutionary dynamic between extrinsic factors (e.g., environmental selection, adaptation) and intrinsic factors (e.g., developmental and functional constraint) in determining form. Despite extreme variation in form and function, tetrapod feeding systems are amenable to comparative analysis because they represent modifications of the same basic apparatus, comprising, for the most part, a set of unequivocally homologous parts. Many of the relevant skeletal components, for example, derive from the ancestral splanchnocranium (visceral skeleton), a defining set of vertebrate characters significant in the origin of the group and present in the tetrapod head under every variety of form and function (Hanken and Hall, 1993). Indeed, splanchnocranial elements of the feeding apparatus in one clade are modified to function as part of the auditory system in another, or the chemosensory system in a third. Such phenotypic recycling well illustrates the evolutionary truism that novelties are most often realized through the modification of preexisting forms. Finally, the relative functionality of the feeding system has, without a doubt, a large impact on individual survival and hence lifetime reproductive success. It is reasonable, therefore, to presume that feeding systems are under strong selection and that variations in feeding performance will have significant fitness consequences. Such a presumption allows us to apply optimality criteria and engineering principles in analyses of feeding system design and to identify trade-offs and constraints on design modification because very little about the feeding system is likely to be a result of random processes (e.g., fixation through drift). As such, the feeding system is well suited to analyses that address the relative contributions of adaptation and historical contingency to the phenotype. The importance of the feeding system to survival and fitness is underscored by its large impact on the tetrapod body plan, primarily through its influence on cranial form. The concentration of both feeding and
sensory components onto one portion of the body reflects a general, historical trend of increasing cephalization in vertebrate evolution. This, in turn, has set the stage for a complex integration and coevolution of feeding and sensory systems in a number of tetrapod clades through competition of these systems for limited cranial space and their shared use of certain anatomical parts. The characteristics of tetrapod feeding systems outlined earlier—ample variation, homology of parts, reasonable presumption of fitness consequences for variation, large impact on body form, and integration with other systems—promote the utility of feeding systems for comparative, evolutionary studies. Specific insights will emerge from detailed analyses of cladespecific patterns and cross-clade comparisons, as considered in individual chapters of this book. B. D e l i m i t i n g the Topic The decision to limit the coverage of this book to tetrapod vertebrates (as opposed to all vertebrates, including fish) was driven both pragmatically and conceptually. Pragmatically, issues of length and the tradeoff of depth versus breadth had to be considered. As editor, I felt strongly that tetrapod feeding systems were in need of a more in-depth treatment than they had been accorded heretofore. Although various aspects of feeding in fishes have been treated to several overviews (e.g., Lauder, 1982b, 1985; Bemis, 1986; Schaefer and Lauder, 1986; Motta, 1988; Westneat, 1990,1995a; Sanderson et al, 1991; Aerts and De Vree, 1993; Lauder and Shaffer, 1993; Wainwright and Lauder, 1992; Sanderson and Wassersug, 1993; Gerking, 1994; Frazzetta, 1994; Stouder et al, 1994; Vandewalle et al, 1994; Wu, 1994; Gosline, 1996; Drost et al, 1998), the rapidly growing literature in tetrapod feeding has barely been harnessed since the seminal papers of Bramble and Wake (1985) and Hiiemae and Crompton (1985). Only two major overviews of tetrapod feeding have appeared (Smith, 1993; Bels et al, 1994), but their coverage is too broad to achieve the depth for tetrapods aimed for here. One book provides in-depth coverage of "eating" in the context of human biology (Linden, 1998). Perhaps the most important reason to exclude fish from consideration, however, is because tetrapod feeding represents a real and significant departure from fish feeding—so much so that research in each area largely operates within a different paradigm. This departure arises from the simple fact that tetrapods, by definition, evolved to feed on land from ancestors that fed exclusively within water. The radically different
1. Tetrapod Feeding in Vertebrate Morphology physical properties of air versus water required an equally radical remodeling of the feeding system. The impact of the medium (air vs water) is so great that feeding system phenotype, overall, is widely viewed as "medium dependent" (Lauder, 1985; Bramble and Wake, 1985; Liem, 1990; Denny, 1990). Fish (presumably including those ancestral to tetrapods) rely almost universally on either suspension or suction mechanisms for feeding (e.g., Sanderson and Wassersug, 1993; Lauder and Shaffer, 1993), which exploit the high density of food particles in water, their buoyancy, and the frictional forces generated by the high viscosity of water. Faced with a low viscosity medium capable of producing only nominal frictional forces, a relatively much lower density of prey, and the need to lift and support the weight of any food item actually obtained, tetrapod ancestors could rely neither on suction nor on suspension feeding and thus required an entirely new approach to food acquisition and manipulation. They achieved this through invention of a true evolutionary novelty: a mobile, muscular tongue. A second area largely neglected in the present volume is the postcranial contribution to feeding function. Obviously, once captured, processed, and transported to the esophagus, food is further digested, assimilated, and reduced to excreta in the remainder of the gut, and these latter processes can have profound effects on body form. The reason for this neglect is largely historical; postcranial digestion and processing are mostly considered within the province of comparative physiology, whereas this book is rooted in the traditions of functional and evolutionary morphology (see later) (DuUemeijer, 1994). From this perspective, most feeding function occurs within the mouth and pharynx, hence the chapters herein tend to be limited to this region. Nonetheless, it is worth remembering that postcranial feeding adaptations can be significant and must be tightly integrated with cranial specializations for particular diets or feeding modes. An obvious example of such integration is mammalian herbivory, which evinces striking adaptations of the gut, including complex stomachs and enlarged caecae associated with the presence of a symbiotic microfauna capable of digesting cellulose (no vertebrate produces its own cellulase enzyme). A less well-known example is the relationship between feeding mode and digestive physiology in snakes. Most "advanced" (macrostomatan) snakes eat only rarely, but when they do they are likely to eat extremely large prey relative to their body size (Chapter 9). Due to a previous evolutionary commitment to an elongate, narrow-diameter body and small head [probably driven by a period of fossoriality and locomotory adaptation (e.g., Greene, 1997)], there was a
considerable mechanical challenge to be met before large prey could be engulfed. The cranial and mandibular apparatus of such snakes was radically modified for production of a wide gape, thus permitting passage of 'large prey through a small-diameter head into an extensible gut. The gut in such snakes is capable of very rapid and large-scale physiological upregulation when the occasional food item presents itself—for example, the intestine actually doubles in mass, mostly through mucosal growth (Secor and Diamond, 1998). Return to an atrophied, or quiescent, condition during long intervals between meals is energetically advantageous to these ectothermic animals. Thus in snakes, body form and the cranial apparatus are tightly integrated, and these aspects of the phenotype are correlated with less obvious, but equally dramatic adaptations of the intestine and its physiology. This example serves to remind us that feeding form and function are more than head deep, a fact the reader is urged to bear in mind.
II. APPROACHES TO THE STUDY OF TETRAPOD FEEDING As for any aspect of the phenotype, the feeding system can be studied in several complementary ways. Although the authors of the various chapters in this book were asked to cover certain general areas, differences in their contributions reflect not only peculiarities of the taxa they treat, but also differences in philosophy, technique, and approach. There is no agreed upon schema for these different approaches and most studies and investigators defy characterization by simple, typological labels. However, it is worth considering, in a broad sense, the different philosophical bases and "schools of thought" characterizing vertebrate morphology as an intellectual realm. It is this realm that has provided the context for this book and its individual contributions (see also Liem and Wake, 1985).
A. T h e m e s in Vertebrate Morphology Animal anatomy is certainly one of the oldest, if not the oldest, biological science, tracing its origins to preAristotelian times (Cole, 1944; Singer, 1957); however, the term "morphology" was only coined in 1800 (Nyhart, 1995). It was during the 19th century that morphology, as a discipline, achieved its heyday, becoming a dominant area of biological research intellectually, if not institutionally (Nyhart, 1995). Certainly, Darwin's
Kurt Schwenk (1859) formulation of the theory of natural selection grew out of an intellectual milieu in which the extensive and painstaking documentation of morphological variation within and among species figured prominently. It is therefore ironic that morphology is seen as having contributed little to the 20th century evolutionary synthesis, commonly referred to as neo-Darwinism (Ghiselin, 1980; Mayr, 1980; but see Waisbren, 1988). While this may be true, it is equally true that neoDarwinism has failed to incorporate many of the lessons of morphology and there is growing dissatisfaction among some morphologists with the ability of neo-Darwinian theory to deal adequately with the totality of phenotypic evolution (see later). In any case, since Darwin, few fields have embraced evolutionary theory as enthusiastically as morphology {contra Ghiselin, 1980), for it is in evolution that the morphological concepts of homology, analogy, Bauplan, and "unity of type," for example, become sensible and elevated in significance. Indeed, the entire field of systematics as currently practiced is based on the character concept, a pre-Darwinian morphological construct melded with neo-Darwinian evolutionary theory. One cannot overstate the centrality of morphology to the development of modern biological thought. What follows is a subjective and largely overlapping list of dichotomies that attempt to clarify and situate the discipline and subdisciplines of vertebrate morphology. The dynamic and protean nature of the field precludes even the pretense of total agreement in this formalization, but one hopes it might be of some heuristic value to new students of vertebrate form and function. 1. Morphology vs
Anatomy
Unlike botanists, who make a clear and formal distinction between plant morphology and plant anatomy (the former referring to whole-plant form and the organization of parts, the latter to the fine structure, or histology, of parts), zoologists are uncharacteristically vague in their usage of these terms. Morphology "deals with the form of living organisms, and with relationships between their structures" (from the Greek stem morpho), whereas anatomy is "the science of the structure of the bodies of humans, animals, and plants" (derived from the Greek stems ana- and -tomy, meaning "repeated cutting") (Oxford English Dictionary; Brown, 1993). Although these definitions would appear to be more-or-less synonymous, in current zoological usage they connote somewhat different things, the sense of which is hinted at in the etymology. Morphology is the study of "form," which can be generalized to all hierarchical levels, from organelle to whole
organism. It is also concerned with the relationships among structures, hence it includes emergent features of form such as relative size, allometry, and even function and physiology for some (see later). Thus, morphology is a more expansive term, subsuming anatomy (as toad is to frog, so anatomy is to morphology). Anatomy is largely limited to the hierarchical level of organs, or body parts, i.e., those elements of form revealed by dissection, and historically has been associated with human beings (Owen, 1866; Singer, 1957). For example, Gegenbauer (1878), epitomizing the late 19th century view, divided anatomy ("the doctrine of structure") into anthropotomy and zootomy, the dissection of humans and nonhuman species, respectively. According to Gegenbauer, anatomy is not a science because it is restricted to the empirical generation of descriptive data. However, these data achieve the status of science when they lead to synthesis and abstraction, only possible by comparing anatomical data among species, i.e., "comparative anatomy," or what Owen (1866) called "homological" and "zoological anatomy." Gegenbauer (1878) included anatomy within the larger field of morphology such that all of biology could be divided into physiology and morphology (see later) and morphology, in turn, into anatomy and embryology. Thus, while some treat the terms anatomy and morphology equivalently, most modern usage continues to reflect the 19th century view in which anatomy is restrictive, descriptive, based on dissection of body parts, and primarily (but not exclusively) anthropocentric, whereas morphology is holistic, multihierarchical, often synthetic and concerned not only with the structure of the parts, themselves, but relationships among the parts, and of parts to function. Although Owen (1866: vii) did not use the word morphology, his notion of "zoological anatomy" captures nicely the essence of modern animal morphology as "that which investigates the structure of an animal in its totality, with the view of learning how the form or state of one part or organ is necessitated by its functional connections with another, and how the co-ordination of organs is adapted to the habits and sphere of life of the species." The history of these ideas and disciplines is reviewed admirably by Russell (1916), Cole (1944), Singer (1957), and Nyhart (1995), among others. Perhaps the most important distinction between anatomy and morphology as formal disciplines is that anatomy is mostly unconcerned with the origin of structure, whereas the origin and generation of form are at the philosophical core of morphology (e.g., Gegenbauer, 1878; His, 1888; Russell, 1916; Thompson, 1942; Davis, 1960; Nyhart, 1995; Webster and Goodwin, 1996). Indeed, since the early 19th century mor-
1. Tetrapod Feeding in Vertebrate Morphology phologists have endeavored to discern "rules of form" that might underlie ontogenetic and phylogenetic morphological transformations. Interest in rules of transformation predates an evolutionary world view (e.g., Driesch, 1908; Nyhart, 1995). Elucidation of such rules was the goal of Geoffroy Saint-Hilaire (1818; in Russell, 1916) and the "transcendental" morphologists, and explored by Owen (e.g., 1848,1849) in the context of "archetypes." However, this fundamental interest in morphological rules of transformation is clearly evident in the modern structuralist movement (see Piaget, 1970; Rieppel, 1990), which, although it takes many forms, has at its core the notion that phenotypic hierarchies manifest emergent properties that certainly influence, if not dictate, directions of further phenotypic evolution (e.g., Russell, 1916; Whyte, 1965; DuUemeijer, 1974, 1980; Riedl, 1978; Lauder, 1982a; Ho and Saunders, 1979, 1984; Roth and Wake, 1985; Rieppel, 1986; Wagner, 1986; Wake and Larson, 1987; Wake and Roth, 1989; D. Wake, 1991; Smith, 1992; van der Weele, 1993; Schwenk, 1995, 2000; Amundson, 1996; Hall, 1996, 1998; Raff, 1996; Webster and Goodwin, 1996; Arthur, 1997; Wagner and Schwenk, 1999; Schwenk and Wagner, in preparation). Indeed, radical "process structuralists" (see Smith, 1992) go so far as to suggest that random variation, natural selection, and phylogenetic history (the tenets of Darwinism) are secondary players in the generation of hierarchically organized phenotypic diversity. They seek instead a "rational," predictive science of form based on rules of selforganization and organismal development (e.g.. Ho and Saunders, 1979,1984; Goodwin, 1989; Webster and Goodwin, 1996). Whether radical or moderate, inherent in the structuralist, morphological view is the sense that the atomistic, neo-Darwinian, gene-based paradigm of phenotypic evolution is incomplete and that the organism creates the context for its own further evolution, thus setting the stage for a top-down chain of evolutionary causality in phenotypic evolution (Whyte, 1965; Wagner and Schwenk, 1999). 2. Form vs Function The conceptual dichotomy of form and function is ancient, at least as old as Aristotle (Russell, 1916; Lauder, 1982a; Padian, 1995). It was identified by Russell (1916) as a major theme in the history of morphology, as evident in the title of his classic book {Form and Function), and it has continued to be a central theme in the study of phenotype throughout this century (e.g., Woodger, 1929; Bock and von Wahlert, 1965; Lauder, 1981,1982a; Gans, 1969,1988; Wake, 1992; Lauder et a/., 1995; Padian, 1995; Weibel, 2000). Historically, this dichotomy was discussed in terms of primacy (e.g., Rus-
sell, 1916; Woodger, 1929; Lauder, 1982a, 1995, 1996; Appel, 1987; Amundson, 1996): is it form that determines function or function that determines form? In the early 19th century, George Cuvier, for example, believed in the primacy of function, positing that form was imposed by the functional demands of the environment and that similar forms reflected similar functional "adaptation." He believed that form and function were so tightly bound in this causal sense that one could, through the "principle of correlation," predict the whole form of an animal from a single part. In contrast, his contemporary, Etienne Geoffroy St. Hilaire, sought a "pure morphology of organic forms" wherein function emerged as a mere by-product of structure. Indeed, Geoffroy eschewed any consideration of function in comparative anatomy and believed that changes in function followed secondarily upon primary changes in structure (Russell, 1916). Although the debate between Cuvier and Geoffroy epitomizes the contention in the field regarding form and function early in the 19th century, the dichotomy was formalized on a grander scale later in the century when the whole of biology was cleaved into physiology and morphology (e.g., Owen, 1866; Gegenbauer, 1878; see Russell, 1916; Nyhart, 1995). Physiology comprised all aspects of organismal function, whereas "the investigation of the material substratum of those functions, and accordingly of the phaenomena of form of the body and its parts, as well as the explanation of the phaenomena of form by reference to function, is the business of Morphology" (Gegenbauer, 1878:1). Physiology already had a long history separate from morphology and the two fields continued through the 19th century and into the 20th to develop in isolation (Russell, 1916; Woodger, 1929). They have remained distinct traditions until quite recently and even now a complete rapprochement remains elusive (e.g., Weibel et al, 1998; Weibel, 2000). Although there is obvious overlap in domain, comparative physiologists, for the most part, continue to study function of a different sort and in a different context than functional morphologists. Woodger (1929) regarded the splitting of form and function as an abstraction not manifest in nature. The "antithesis" results from employing "different modes of apprehension" and from artificially separating space and time (anatomy focusing on the former, physiology the latter) which are simultaneous attributes of living structures. Certainly, most modern evolutionary morphologists would consider the question of formfunction primacy moot, recognizing the chicken-andegg nature of the dichotomy in a historical context. Indeed, evolutionists do not acknowledge a distinction at all, capturing the seamlessness of form and function in
8
Kurt Schwenk
the inclusive term "phenotype." Nonetheless, vestiges of the dichotomy persist within morphology. This is most apparent in the question of how one studies function: must one study function directly or is it possible to infer function from structure? Certainly pure, descriptive, anatomical studies are not much interested in function and are therefore not relevant here, but there is, at the same time, a long tradition in vertebrate morphology of inferring animal function from static form. This is most apparent in the 20th century paleontological literature where functional speculation is commonplace (e.g., Rainger, 1989; Hopson and Radinsky, 1980; Thomason, 1995), but it is also rife in studies of living species. Indeed, the ability to predict function from form is at the conceptual core of "ecomorphology" (see later). The presumption that function can be inferred from form stems more or less directly from Cuvier's principle of correlation (see earlier discussion). It depends absolutely on the degree to which form and function are integrated so that each can stand as a proxy for the other. However, Lauder (1995, 1996) has shown that structure and function are not always tightly matched in a predictive sense, thus falsifying the assumption upon which the principle of correlation is based. He advocates a direct, experimental approach to function based on analysis of living specimens and emphasizes the need for quantification of functional data. Experimental, quantitative approaches to functional morphology are relatively recent. Przibram (1931:14), for example, noted, "Whereas in Physics, Chemistry or even Physiology nobody would nowadays try to state laws without experimental evidence, the general statements in animal Morphology have been mostly based on speculations. Only of late has the experimental method been gaining ground in Biology and research is being carried on in a large scale at institutions adapted to the p u r p o s e . . . Again I would refer to Physics and Chemistry as brilliant examples of what may be achieved in the way of unraveling the laws of nature by quantitative experiment and mathematical formulation based thereon." Thompson (1942:2) was, likewise, intent on a more quantitative approach to form: "But the zoologist or morphologist has been slow, where the physiologist has long been eager, to invoke the aid of the physical or mathematical sciences." Thus, while the inseparability of form and function may be intellectually acknowledged by most practicing morphologists, the ancient dichotomy remains manifest in investigators' approach to function: inferential and qualitative or experimental and quantitative. These different approaches are elaborated further. Although the dangers of inferring function from static form seem clear, seldom, if ever, is concern ex-
pressed for the converse: can the detailed study of function in the absence of equally detailed structural data be misinformed? Unquestionably (see also Cans, 1986). I am frequently impressed by the near total lack of morphology evident in some studies of putative "functional morphology" in which quantitative aspects of function are analyzed in a structural vacuum, almost wholly uninformed by knowledge of the relevant anatomy. At the very least, morphological data can, in such cases, eliminate from consideration alternative mechanistic hypotheses. Further, functional conclusions can be revealed as false, or even absurd, when held in the light of form. The dichotomy of form and function is an undeniably useful, heuristic tool. However, we must acknowledge that this dichotomous view is a philosophical construct, simplifying for us the complex notion oiphenotype in which form and function are interwoven dynamically and infrangibly (Woodger, 1929). As Ruffini (1925) noted, "form is the plastic image of function" (in DiDio, 1986:197). There can be no argument for primacy, only relationship. Therefore, all approaches to the study of phenotype are valid, if not complete, so long as conclusions follow from data and speculation is labeled as such. If only for pragmatic reasons, most individual studies will continue to emphasize either form or function, with syntheses relatively rare. Nonetheless, form and function, experiment and description, qualitative and quantitative data must be held as equally important, complementary, and ideally, "reciprocally illuminating" elements in the study of morphology [Lombard (1991) and Lauder (1995) offer related perspectives on these issues]. 3. Idiographic vs
Nomothetic
Idiographic studies are those that characterize a specific case without regard to its typicality or generality. Studies which treat the morphology or function of a single species are of this type. An organism, or part thereof, can be treated purely as mechanism and one can ask, simply, what does it look like? what are its component parts? and how does it function? without regard to whether the results are applicable to a larger group or peculiar to the case at hand. Such studies can have high intrinsic interest and are valuable in proportion to the quality of their data, but they do not inform us about the next system we might study, nor how one form evolves into another. Pattern, causality, and prediction derive, instead, from nomothetic studies, which strive to elucidate higher principles, rules, or, in the strict sense, scientific laws. Comparative morphology, like Owen's (1866) zoological anatomy (see earlier discussion), falls within
1. Tetrapod Feeding in Vertebrate Morphology this realm because it seeks generalities beyond the specific (Gegenbauer, 1878). However, nomothetic, comparative, vertebrate morphology does not require an evolutionary world view. As noted earlier, early 19th century morphologists, such as Cuvier and Geoffroy, were very much interested in general rules of form without reference to historical descent. Indeed, some modern structuralists (see earlier discussion) are also interested in an organism-based theory of form independent of evolution, arguing that * 'evolution provides only limited insight into the problem of form as regards both the causal explanation of form and the relations between forms . . . what is required is the development of a specific causal-explanatory theory of form, a theory of morphogenesis in the most comprehensive sense . . . such a theory will be as fundamental to biology, if not more so, at least as the theory of evolution" (Webster and Goodwin, 1996 :ix; see also His, 1888). Nonetheless, most modern, comparative studies in vertebrate morphology are rooted in the neo-Darwinian tradition, but framed in an explicit cladistic, phylogenetic context (e.g., Lauder, 1981, 1990; Greene, 1986a), whereas still others seek some middle ground (e.g.. Wake and Larson, 1987; Wagner and Schwenk, 1999). Whether organism centered or phylogenetic, such studies are part of a long, nomothetic tradition in vertebrate morphology that has contributed fundamentally to the principles of modern biology. 4. Laboratory vs Field Most vertebrate morphology must, of necessity, be undertaken within the laboratory. However, any consideration of function, whether it is inferential or experimental, is at risk if it fails to consider the actual behavior of organisms in the field (Greene, 1986b, 1994). For the inference of function from form, natural history data can circumscribe the universe of possibilities. For example, one need not waste time speculating about the role of certain skull attributes in processing flesh if it is known from observation and stomach content analyses that the organism in question is an insect specialist. Likewise, a treadmill study of locomotion in gibbons would be largely uninformative about the evolution of its form, whereas initial field observations of its natural locomotory mode (brachiation) would suggest a different experimental approach. These examples are obvious and might appear silly, but there is no doubt that functional and evolutionary morphology can be led astray in failing to consider the natural behavior of animals. A classic case is that of monitor lizard (Varanus) morphology and diet. A persistent theme in functional-morphological (and other) studies is that varanid lizards are adaptively special-
ized for carnivory and the ingestion of large prey (e.g., Rieppel, 1979; Smith, 1982). As such, experimental studies of feeding function have analyzed monitors during feeding on rodents (e.g.. Smith, 1982; Condon, 1987). However, a comparative study of diet in monitor lizards revealed that nearly all species are insectivorous, eating many small prey items rather than a few large ones (Losos and Greene, 1988). Varanus exanthematicus, the most commonly used species in liveanimal studies (e.g.. Smith, 1982), is one of the most extreme insect specialists within the family. Indeed, it is in many other respects unusual and derived— neither a typical lizard nor even a typical varanid. Thus, the experimental conditions of these studies were unnatural, or at least, atypical, and therefore the generality of their results is suspect. Certainly the interpretation of derived varanid cranial attributes as adaptations for eating large, vertebrate prey is without foundation. Rather, the typical monitor lizard cranial apparatus seems to represent a specialized pincer system for effectively nabbing elusive prey with the tips of the jaws (Frazzetta, 1983; see Chapter 8). B. Schools of Vertebrate Morphology The simple 19th century division of biology into physiology and morphology is no longer tenable. Experimentalism, technological advancement, new techniques, new theory, and increasing specialization have driven fragmentation of biology, generally, and vertebrate morphology, specifically. Because no investigator can master all possible approaches to morphology, individual studies tend to emphasize one approach, technique, or school of thought. The following list of approaches to vertebrate morphology is hardly exhaustive, but it represents a cross section of the field as currently practiced. Elements of these different approaches are evident in the chapters of this book. 1. Descriptive
Morphology
Description remains the foundation upon which morphological studies must ultimately rest. Good descriptive morphology is as rare as it is beautiful, virtually an art. Unfortunately, it may be a dying art because it is little valued within the context of modern science. This is tragically short-sighted because one thing is clear about high-quality, well-documented descriptive anatomy: it may represent the only truly hard, objective data in morphological research, as free of fashion and interpretation as possible. It is, therefore, timeless. This timelessness is critical because it means that descriptive data can be used in the service of other studies and new theories, including those not conceived at
10
Kurt Schwenk
the time the data were generated. As such, descriptive data are of value to future generations of biologists— it is common, for example, for 19th century and even older literature to be cited in modern morphological studies, often in contexts, such as cladistic, systematic analysis, never imagined by the original authors. For how many other fields, or even other approaches to morphology, can this be said? Of what value now, for example, are systematic analyses based on the once widespread and "cutting edge" method of phenetics? How many molecular studies more than 10 years old remain in currency? In contrast, a morphological treatise, such as D. Dwight Davis' (1964) classic study of the giant panda, is unceasingly relevant. Indeed, its value only increases with time as the availability of rare species declines. It is a telling fact that Davis' (1964) conceptual, evolutionary analysis of his descriptive work is dated and of dubious value in today's milieu of cladism and molecular genetics, but the 300 pages of painstaking, descriptive anatomy that precedes it remain—untainted, unfiltered, uninterpreted, and fully available for reanalysis in another context, now or in the future. Descriptive morphology can apply to any hierarchical level, but it is done most commonly at the anatomical level (see earlier discussion). Anatomies can be regional, such as Oelrich's (1954) Anatomy of the Head of Ctenosaura pectinata, or they can be systemic, as in Romer's (1956) Osteology of the Reptiles. Tissue-level morphology is usually referred to as histology, loosely limited to those aspects of morphology resolvable by light microscopy. Cellular and subcellular morphology studied by means of scanning and transmission electron microscopy is referred to as ultrastructure. In the parlance of vertebrate morphology, structure usually refers to organs and organ systems, although it can be in reference to any hierarchical level. However, in the biomedical community, "structural biology" is purely molecular. 2. Evolutionary
Morphology
Evolutionary morphology defies precise characterization because it is the most inclusive school of vertebrate morphology (D. Wake, 1982; M. Wake, 1991, 1992). It can be experimental, functional, or purely descriptive, but it has at its core an evolutionary intent. By definition, therefore, it is comparative. As such, it most clearly manifests the tradition of 19th century comparative anatomy in that it continues to seek higher level understanding of the generation and transformation of form (see earlier discussion; Davis, 1960). Evolutionary morphology is distinct from morpho-
logical systematics in that the latter uses morphology in the service of phylogeny, whereas the former uses phylogeny in the service of morphology. As such, the historical relationships of organisms are of secondary interest to the evolutionary morphologist whose main concern is the history of the characters themselves, i.e. character analysis in the purest sense. Whereas the systematist uses morphological characters as a matter of routine, in evolutionary morphology the character concept is, itself, a subject of study (Wagner, 2000). Furthermore, studies in evolutionary morphology may attempt not only to generate patterns of character evolution, but to address ultimate issues of causality, as such, the processes of phenotypic evolution (e.g., Galis, 1996; Wagner and Schwenk, 1999). 3. Functional
Morphology
The focus of functional morphology, transparently, is function. As discussed earlier, there is contention about whether studies that are limited to the inference of function from structure can be considered "functional." Nonetheless, purely inferential papers continue to appear with "functional morphology," or sometimes "functional anatomy" in their titles. Although the latter term may be an attempt at truth in advertising (i.e., it flags the paper as inferential), "functional anatomy" is to be discouraged because it is, in some sense, an oxymoron (see discussion earlier). Thus, in most recent and I would say preferred usage, functional morphology refers specifically to analyses based on the direct measurement of function in live, behaving animals. Functional morphology may be evolutionary; it is often purely mechanistic and idiographic (D. Wake, 1982; M. Wake, 1992). Like good descriptive morphology, good functional studies of single species can be of great intrinsic interest, and to the extent that data are objective, they are available to other investigators for meta-analyses. It is worth noting here that while most modern functional morphology is "experimental," there are two quite different usages of this word in the field. The first is appropriate and consistent with its use in traditional, reductionist scientific method in referring to hypothesis testing. Such a hypothetico-deductive approach to function usually requires manipulations of experimental subjects and careful controls, thus in functional morphology such studies often require invasive techniques and the use of complex technology. Nonetheless, it is important to note that hypothesis testing, and therefore experimentation sensu stricto, is not restricted to functional studies. It is possible to test certain types of hypotheses, even functional hypotheses, by reference to morphology alone. However, such descriptive
11
1. Tetrapod Feeding in Vertebrate Morphology analyses are never labeled "experimental." Thus, in current usage, the application of hypothetico-deductive analysis is insufficient to qualify a study as experimental. Nor would the use of control and manipulated subjects seem sufficient. Rather, the principal criterion appears to be the application of technology itself. This has led to a second, diluted, and much more common (if inappropriate) use of the term to mean any analysis employing technologically based methods. For example, many studies characterize muscle activity patterns during normal behavior using electromyography (EMG) and are labeled "experimental" even though they are purely descriptive—they show X muscle to be active during Y behavior. There need be no control, manipulation of the conditions, nor testing of functional hypotheses. However, not all use of technology seems to merit an experimental epithet. For example, use of binoculars to describe the same behavior, Y, in the field would not be considered experimental, although it may serve to test a functional hypothesis and scientifically it may be the more valuable of the two studies! Therefore, the currency of the technology employed, or perhaps its degree of invasiveness, seems to be a deciding factor. I see no obvious solution to these terminological paradoxes, but highlight them here as a reminder to students of form that scientific merit is a quality independent of technique or titular fashion. 4.
Biomechanics
Biomechanics might be regarded as a subdiscipline of functional morphology—certainly many functional morphological studies contain elements of biomechanical analysis (e.g., Gans, 1976; Rayner and Wootton, 1991). However, biomechanics is, in the strict sense, directly inspired by various fields of engineering and is far more mathematical and less anatomical than is typical of functional morphology (e.g., Fung, 1993). Indeed, biomechanics quickly blends into theory and modeling (see later). In addition, its more-or-less pure design approach to form (see Lauder, 1996) means that biomechanical studies generally eschew comparative or evolutionary issues. In fact, most of the biomechanics literature is biomedical and sports oriented. This can be contrasted with transformation morphology (see later) which, though often highly biomechanical, is concerned with understanding ontogenetic and phylogenetic transformations in form and function. Biomechanics in (nonbiomedical) morphology is especially concerned with the physics of biological materials and surrounding media, and their consequences for organismal form, function, and, occasionally, evolution (e.g., Wainwright et al, 1976; Alexander, 1985; Wainwright, 1988; Vogel, 1988, 1994; Vincent, 1990;
Denny, 1993). For example, it was noted at the outset of this chapter that the tetrapod feeding apparatus is, to a large extent, medium dependent in the sense that, depending on whether the organism lives in air or in water, the physical properties of the medium will impose a limited realm of phenotypic solutions to the problem of ingesting, processing, and swallowing food. Thus the convergence of feeding mechanisms in fish and secondarily aquatic tetrapods could be predicted from biomechanical first principles, i.e., the viscosity and fluid dynamics of water. Indeed, it is often found that certain aspects of the phenotype conform to the expectations of mechanical optimization in some parameter or other in a way that clarifies their functional and evolutionary significance. For example, although the tongue of squamate reptiles was well known to be a chemical sampling device, the functional significance of its forked form in snakes and some other species was not understood. An engineering approach in a comparative context, however, suggested that the fork provides a two-point sampling device for the detection of chemical gradients useful for following pheromonal trails (Schwenk, 1994). Fluiddynamic theory further suggested that the rapid oscillation of the tongue in the air characteristic of these species is an adaptation to enhance the molecular diffusion of environmental chemicals into the fluid on the tongue's surface, thereby amplifying the chemical signal carried into the mouth (Schwenk, 1996). Thus, while biomechanics can be pursued idiographically without reference to evolution, it can powerfully inform questions about the evolution of form and function (Lauder, 1991; M. Wake, 1992). 5. Developmental
Morphology
Like evolutionary morphology, developmental morphology is a broad field that grew out of the 19th century tradition of embryology (see earlier discussion). It implies no particular technique or approach, but is concerned with the ontogenetic transformation of form. Most often such studies are limited to the embryonic period, but they may treat any life stage so long as their concern is ontogenetic transformation. Because the focus of developmental morphology is on intrinsic attributes of individual organisms, it is the part of vertebrate morphology most clearly identified with biological structuralism (e.g., Wagner, 1988; Hall, 1994; D. Wake, 1991; Webster and Goodwin, 1996). Developmental studies have revealed principals of selforganization and pattern formation at the organismic level that are likely to be relevant to the directions and dynamics of phenotypic change in lineages through evolutionary time. Yet neo-Darwinian theory springs
12
Kurt Schwenk
from population-level phenomena, hence it has neglected the role of the organism in directing patterns of phenotypic change. To a large extent, the modern notion of "evolutionary constraint" has emerged from the field of developmental morphology (e.g., Maynard Smith et al, 1985; Wagner, 1986, 1988; D. Wake, 1991; Schwenk, 1995; Schwenk and Wagner, in preparation). 6. Ecological
Morphology
Ecological morphology (or ecomorphology) is a rather amorphous hybrid of several morphological schools (as suggested in the title of one review; Wainwright, 1991). Nominally it owes its origin to van der Klaauw (1948), but conceptually it is rooted in the ancient observation of the "fit" between organisms and their environments. As such, the prevailing definition of ecomorphology is "the study of the relationship between the morphology of the organism and its environment" (Wainwright and Reilly, 1994a: 3). Indeed, it is difficult to say how the basic premise of ecomorphology differs from Cuvier's "principle of correlation"—both assume that organismal form is tightly correlated to environmental conditions. Like Cuvier, ecomorphology aspires to predict form from ecology and ecology from form (e.g., Emerson, 1991; Motta and Kotrschal, 1992; Norton and Brainerd, 1993; Wainwright and Richard, 1995; Losos et al, 1998). However, unlike Cuvier, modern ecomorphologists invoke natural selection and adaptation to account for the formenvironment fit. Thus, they espouse the need for integrating historical approaches into ecomorphological studies (e.g., Motta and Kotrschal, 1992; Wainwright and Reilly, 1994a; Losos and Miles, 1994; Westneat, 1995b). Wainwright and Reilly (1994b) provide an entree into the ecomorphological literature. As they noted, ecomorphological studies usually are either mostly ecological or mostly morphological (Wainwright and Reilly, 1994a). However, at its best, ecomorphology attempts a true synthesis by addressing questions that are fundamentally ecological through the phenotypic analysis of organisms. The key point is that the study of phenotype is pursued at the population level. I think it is only here that ecomorphology clearly distinguishes itself. Otherwise it tends to suffer one of three fates: it is simply ecology with a few superficial measurements (e.g., limb length) thrown in; it is simply good functional or evolutionary morphology, which should, after all, incorporate ecological and natural history data into its analyses (see earlier discussion); or, most commonly and worst of all, it suffers from naive adaptationism {sensu Gould and Lewontin,
1979), which finds that organisms are, indeed, adapted to their environments, e.g., animals with big mouths can, in fact, eat big prey. Although such studies are not without merit, they evince little conceptual advancement over "natural theology" (e.g., Amundson, 1996)—things are as they must be (see also Liem, 1993). Rather than using the form-environment fit as a point of departure, ecomorphology can and should probe the limits of this assumption by incorporating notions of constraint and other potentially limiting factors into its paradigm (e.g., Barel et ah, 1989; Liem, 1993; Losos and Miles, 1994). Certainly the relationship between the organismal phenotype and the environment is far more complex than usually assumed (e.g., Simpson, 1953; Whyte, 1965; Greene, 1982; Wagner and Schwenk, 1999). 7. Transformation
Morphology
Transformation morphology is related to evolutionary and developmental morphology, but it is concerned specifically with the process of ontogenetic and phylogenetic phenotypic transformation (Barel, 1993; Galis, 1996). Its focus is less on the pattern of transformation (e.g., documentation of form-function complexes in comparative, or ontogenetic series) than on the rules of transformation underlying observed patterns. According to Galis (1996:128), these rules can be elucidated "by constructing biomechanically feasible transformation schemes, by studying key structural changes that break important constraints enabling a cascade of changes, and by studying the mechanisms that preserve the match between form and function during ontogenetic and evolutionary change." Transformation morphology has grown out of the Leiden school of morphology (e.g., van der Klaauw, 1945; DuUemeijer, 1959, 1974, 1980, 1989) and, as such, its emphasis is functional. Furthermore, as suggested by Galis' (1996) quote, modeling and engineering approaches to evolutionary transformation (e.g., Zweers, 1991; Zweers and Vanden Berge, 1997; van Leeuwen and Spoor, 1992; Galis, 1992,1993; Galis and Drucker, 1996; see later) are preferred to the comparative, phylogenetic methods typical of North American workers (e.g., Lauder, 1981,1990; Lauder et al, 1995; Larson and Losos, 1996). 8, Constructional
Morphology
Like transformation morphology, with which it overlaps, constructional morphology is largely a European tradition. It was formalized by Seilacher in the 1970s (e.g., 1970, 1973, 1979), but has earlier roots in Leiden (see references given earlier) (Reif et al., 1985).
1. Tetrapod Feeding in Vertebrate Morphology Constructional morphology deals explicitly with constraints on adaptive evolution arising from the physical properties of materials and the mechanisms of their deposition and growth, although it is occasionally more broadly construed as pertaining to any sort of structural constraint (e.g., Reif et ah, 1985; Barel et at, 1989; see also papers in Schmidt-Kittler and Vogel, 1991). Whereas ecomorphology takes as its starting point the adaptive nature of organismal form, constructional morphology focuses on nonadaptive aspects of form; it therejfore disputes the notion that all morphological variation reflects adaptive responses to different environments. Roth (1989), for example, showed that differences in dental form among nominal species of fossil elephant are attributable to "fabricational noise," reflecting individual differences in masticatory stresses during postnatal development of the teeth. As such, quite large differences in the form of the elephantid dental battery are a consequence of the uniquely retarded growth of elephant teeth as compared to other mammals. While the phenotypic plasticity that permits such individual responses to juvenile stresses might, itself, be an adaptation (e.g., Schlichting and Pigliucci, 1998), the particular morphcJiogical variants that arise are the result of nonadaptive processes stemming from mechanistic aspects of tooth development and growth. It is these processes, and their effects on the evolution of phenotype, that are the focus of constructional morphology. 9.
Morphometries
Morphometries is a catch-all term for a variety of methods that seek to capture and quantify complex shapes based on external measurements. It grew out of the emerging field of biometry in the late 19th century (Bookstein, 1994) and later, the pioneering work of D'Arcy Thompson (1917, revised and expanded in 1942) who sought a mathematization of zoology and morphology. Although best known for his application of Cartesian coordinates to the study of shape transformation, Thompson (1917) actually dealt more extensively with other quantitative aspects of form, especially patterns of growth. This work strongly influenced the subsequent, formal study of allometry, as evident in the dedication of Huxley's (1932) classic book to Thompson. Allometry, therefore, might be considered a part of the morphometric tradition (Huxley, 1932; Gould, 1966). In light of his stated goal, it is ironic that Thompson's (1942) use of deformed Cartesian coordinate planes to describe shape change was largely intuitive and not particularly quantitative. However, a num-
13
ber of methods have been described that accomplish Thompson's (1942) intent with greater rigor (e.g., Rohlf and Bookstein, 1990; Bookstein, 1991; Marcus et al, 1996; McLellan and Endler, 1998). However, these are limited in application to two-dimensional shapes, or two-dimensional projections of three-dimensional forms, which diminishes their utility to morphologists interested in complex anatomy (however, some threedimensional morphometric methods are being developed, e.g.. Roth, 1993). Perhaps of greater concern is the question of using coordinate-based, landmark data in evolutionary character analysis. As noted previously, the issue of atomizing organisms and delimiting characters is an area of active empirical and conceptual work (e.g., Wagner and Schwenk, 1999; Wagner, 2000; Schwenk, 2000). It is by no means clear that landmark data fulfill the criteria of homology necessary to establish them as "characters" in the sense of semiautonomous units of phenotypic evolution. Bookstein (1994), for example, has argued that biometrical shape characters are not commensurate with traditional, phenotypic characters as used in systematic analyses, whereas others suggest that, in the proper context, morphometric characters can be used this way (e.g., Zelditchefa/.,1992,1993). 10, Theory and Modeling Although most morphological theory is conceptual (e.g., DuUemeijer, 1974; Gould and Lewontin, 1979; Wake et al, 1983; Wagner and Schwenk, 1999), it can also be highly mathematical (e.g.. Burger, 1986; Van Leeuwen, 1991, 1997). As for other biological disciplines (e.g., ecology and evolution), theory in morphology is most effective in reciprocity with empiricism. The case of chameleon tongue projection offers an excellent example of this synergism. The mechanistic basis of the explosive, ballistic projection of the chameleon tongue defied understanding for centuries (see Chapter 8). However, an incisive, experimental study by Wainwright and Bennett (1992) implicated the use of hydrostatic elongation of the lingual accelerator muscle in this function. The Wainwright-Bennett model, based on an in vitro study, was beautifully supported by a theoretical model of the biomechanics of the accelerator muscle, which accurately predicted details of hyolingual form and function (Van Leeuwen, 1997). The combined results of these studies have, in turn, been corroborated and extended by an in vivo study using high-speed X-ray movies (cineradiography) of chameleon tongue projection (Schwenk et al, in preparation). As a result of this interplay among descriptive anatomy, functional inference, in vitro and
14
Kurt Schwenk
in vivo experimentation, and mathematical modeling, the mechanism of chameleon tongue projection is now reasonably well understood. Because many functional-morphological studies focus on musculoskeletal systems, it is no surprise that most of the inspiration for functional models comes fron\ mechanical engineering and materials science (e.g., Gans, 1974; Wainwright et ah, 1976; Alexander, 1983) (see also Section II, B, 4). Frazzetta's (1962) application of a "quadric-crank" model to the amphikinetic lizard skull is an early example of this approach, but mechanical modeling approaches are varied (see references given earlier; other examples include de Jongh et a/., 1989; Otten, 1989; Galis, 1992, 1993; Russell and Thomason, 1993; Weishampel, 1993, 1995; Greaves, 1995; Herrel et al, 1998). Models are especially powerful when they make predictions that can be tested empirically or with natural history data. For example, Galis' (1992) model of bite forces in a cichlid fish showed that ontogenetic changes in the pharyngeal jaws accounted for the inability of small fish to pierce the integument of certain prey types and the absence of such prey in the natural diets of the fish until they achieved a certain size (GaUs, 1993). The model accurately predicted the size at which fish started to include the harder prey type in their diets. Galis (1993) suggested that such patterns establish causality, as opposed to mere correlation, although this assertion is open to debate. C. Techniques of Vertebrate Morphology One of the great beauties of vertebrate morphology is that meaningful results can be obtained with minimal expense and laboratory sophistication. Many studies begin with dissection and, as argued earlier, highquality descriptive anatomy remains at the core of any morphological analysis. Beyond dissection and description, a variety of techniques are used to reveal increasingly reductionist hierarchical levels of anatomy and to record, measure, and quantify components of function. Many of these techniques are revealed in the individual chapters of this book and it is worth reviewing some of them here. I. Anatomical
Techniques
The art of dissection is little changed since the time of Aristotle, with the exception that tissues and whole organisms can now be fixed and preserved indefinitely for later examination. While vivisection is rarely necessary in morphological studies, occasionally fresh, unfixed tissue is required, as in some histochemical
procedures. Otherwise it is desirable to "stiffen" the tissue through fixation, most often in 10% formalin (formaldehyde solution), which not only allows ample time for one's study, but also facilitates the manipulation of the organs to be dissected, as in the separation of muscles along facial planes. The latter process can be made easier through differential staining of muscle and connective tissue with a topically applied stain (Bock and Shear, 1972). Very fine-scale dissection {microdissection) is possible with finely machined tools and a dissecting scope. Skeletal anatomy is often a central element in morphological studies. It is most often examined from dried skeletons, usually held in the collections of major research museums. Although they can be prepared manually, skeletons, especially those of small, delicate species, are best prepared with colonies of dermestid beetles, which conveniently consume any remaining, dried flesh. If the skeleton is retrieved from the beetles at the appropriate time, a clean, but articulated skeleton results. Study of skeletal anatomy can be enhanced through preparation of dried hone-ligament or honemuscle preparations (Hildebrand, 1968). The skeleton in situ can be studied indirectly with radiography (X-ray pictures). This is most useful for simple measurements such as long bone length and width. Stereo radiographs can be made to study more complex structures, such as the skull. Another very powerful technique for studying skeletons in situ is clearing and staining (Wassersug, 1976; Hanken and Wassersug, 1981). In this technique the surrounding flesh is enzymatically macerated and rendered transparent while the bone and cartilage of the skeleton are differentially stained red and blue, respectively. This method can be extended to include simultaneous staining and visualization of the nervous system (Filipski and Wilson, 1984,1985; Bloot et al., 1985) and the circulatory system (Russell et al., 1988). Minute elements of anatomy are revealed through light and electron microscopy. These techniques are used so widely that they fall into the realm of "standard technique" (e.g., Presnell and Schreibman, 1997), although to do them well usually requires skill learned through long experience. Most light microscope histology is carried out with tissues embedded in paraffin or paraffin-plastic polymers and sectioned on the order of 5 to 20 ^tm. Embedding in harder n\edia, such as epoxy resin, allows for thinner sections: 0.5 to 2 /xm for light microscopy and 0.01 to 0.05 yam for transmission electron microscopy (TEM). Scanning electron microscopy (SEM) reveals only surface features of cells and organs and therefore involves coating and examining whole, three-dimensional forms rather than sections.
1. Tetrapod Feeding in Vertebrate Morphology Both Ught microscopy and TEM require differential staining of cells and organelles in order to visualize the structure of interest. Both techniques dissolve, denature, or otherwise destroy some chemical constituents of the cells or tissues that one might like to reveal. Thus, instead of fixing the tissue chemically, it can be hardened and stabilized through freezing and sectioned on a microtome encased in a freezing compartment (a cryostat). In this way the chemical inclusions of the tissue are more or less preserved for visualization by treating frozen sections with stains that react with the desired compounds (histochemistry). It should be noted that all histological staining is histochemical in this sense, but usually histochemistry refers to frozen tissue techniques, which offer a greater variety and, more important, specificity of staining. 2. Functional
Techniques
Perhaps the most basic, and underrated, functional technique is observation. As His (1888) noted, "observation, though generally well marked in children, is more and more neglected, or even suppressed, by the usual school education." Careful observation of unrestricted animals performing natural behaviors is the first step of any functional analysis. Ideally, such observations are made in the field as well as in the laboratory, although this is remarkably rare in functional studies. Many functional techniques represent little more than enhancements of our powers of observation. First among these is standard photography, which captures and freezes for observation a rapid behavior, particularly if short shutter speeds and stroboscopic illumination are used; 35-mm and larger format film provides high-resolution images. Of course, photography has the disadvantage of being unable to record an entire behavioral sequence (the most rapid 35-mm motor drives are usually capable of no more than three to five frames per second). High-speed cinematography (cine film), in contrast, can capture a complete, rapid behavior (frame rates of 500 per second and higher are possible, although 50 to 250 fps is typical), but the rapid speed of the film through the camera and the small, 8- or 16-mm format reduces the resolution possible for individual frame analysis (few biologists have access to commercial 35-mm cine equipment). Furthermore, short shutter speeds and the chemistry of film emulsions require high light levels for proper exposure. Synchronized, stroboscopic illumination again can greatly enhance the resolution of individual cine frames and reduce the heat generated by photo floods. Cine is only infrequently used now that high-speed videography systems are available. Video has many ad-
15
vantages over film, including very high frame rates (up to 1000 fields per second), long recording times, high photosensitivity (only moderate light levels are necessary), and, in some cases, digital image capture and storage, but it still lacks the single-frame resolution of cine and so is not appropriate for some applications. It also has the virtue of being inexpensive to operate, although initial equipment cost is very high, whereas cine equipment is relatively less expensive, but has a high operational cost. Images generated by photography can be subjected to computer-based image analysis. Many commercial and publicly available (e.g., NIH Image) software packages permit numerous types of measurement (e.g., distance, area, image density) directly from imported images. Similarly, motion or kinematic analysis can be performed on sequential images obtained through cinematography or videography. Most often, the X-Y coordinates of relevant points on sequential images are obtained by digitizing them and these data are used for kinematic analysis. Many examples of kinematic plots are found in the chapters of this book. They show relative motion of desired points (e.g., the tips of the jaws) relative to time so that movements of various parts can be shown in synchrony and quantified throughout an entire behavioral sequence. Kinematic analyses can be coupled to simultaneous electromyography (Loeb and Cans, 1986; Cans, 1992) for information about muscle activity. Most often bipolar electrodes are used and these are inserted through a needle into the muscle of interest. The electrode measures the electrical potential across the dipole and when the muscle is active this voltage spikes. Simultaneous recording of activity in many muscles is possible; in some cases a reference muscle that exhibits rhythmic or otherwise predictable behavior is used to synchronize muscle activity patterns in separate groups of muscles measured in different experiments. Muscle activity patterns (motor patterns) are often illustrated in a summary bar diagram that shows relative onset and offset, and relative activity level in all muscles measured in conjunction with relevant kinematic plots. A variety of electronic transducers are now used to measure various functional parameters. Strain gages convert deformations along their axes into microvoltages, which can be resolved into compression and tension. Other transducers measure force, pressure, flow, etc. A unique, analogue method of goniometry for measuring minute flexion across limited-motion joints was employed by Condon (1987). In this technique a pointer and a protractor were glued to two bones
16
Kurt Schwenk
across a putative kinetic joint in a lizard skull. The technique was able to measure deflections as small as T, far better resolution than that provided by cineradiographic and strain gage techniques used in previous studies. This technique deserves to be exploited more widely. Lauder has introduced the use of Digital Particle Image Velocimetry in the analysis of fish locomotion (Lauder et al, 1996; Drucker and Lauder, 1997). This method is a great advancement over traditional flow visualization techniques, exploiting the use of laser light to illuminate a precise plane of particles in a flow tank, which are recorded on high-speed video. Software permits a three-dimensional reconstruction of fluid direction and velocity over time. This review has touched on only some of the anatomical and functional techniques employed by modern morphologists. There are many others, particularly if one includes specialized forms of light and electron microscopy and neurological techniques; however, this brief overview should provide an adequate introduction to the field and subsequent chapters of this book. III. C O N C L U D I N G C O M M E N T S If nothing else, the preceding overview should reveal that modern vertebrate morphology is a diverse and intellectually vibrant field. It is unique among the biological sciences in its combination of ancient knowledge with cutting-edge technological and conceptual advances. At its best it is the most integrative of sciences, moving fluidly among hierarchical levels and drawing insight from the interplay of disciplines as divergent as molecular genetics and community ecology. It is hard to imagine a more synthetic field, nor one more fundamentally relevant to the the cornerstones of comparative biology: systematics, evolutionary biology, and ecology. Nonetheless, as a discipline it has suffered its share of indignities, periods of professional and institutional stagnation, and even the sneering disregard of its neophytic, reductionist cousins. We have seen the early, impudent promises of molecular genetics to obviate morphology by "answering" the ultimate questions of phenotypic evolution fade as the emergent complexity of genetic, developmental, and functional systems becomes ever more apparent. I therefore conclude this chapter as I began it—by proclaiming that phenotypic approaches to comparative biology are not only deeply interesting, they are essential. If nothing else, the property of emergence requires that this is so. It is fruitless and petty to argue for primacy of one hierarchical level over another. Morphology shows us clearly that organisms are webs of interaction
and integration, not linear chains of cause and effect. Our task is to connect hierarchies, to hlur their boundaries, not to separate them. Morphologists are uniquely poised to contribute in this regard. Dramatic advances will emerge from explicit attempts to integrate topdown and bottom-up approaches in the study of form and function. Acknowledgments I am grateful to Nirvana Filoramo, Ken Kardong, Carl Schlichting, and Giinter Wagner for commenting on the manuscript. Willy Bemis pointed out some useful references. Preparation of the manuscript was supported by grants from the University of Connecticut Research Foundation and the National Science Foundation (NSF IBN-9601173) to the author.
References Aerts, P., and F. DeVree (1993) Feeding performance and muscular constraints in fish. J. Exp. Biol. 177:129-147. Alexander, R. M. (1983) Animal Mechanics, 2nd Ed. Blackwell, Oxford. Alexander, R. M. (1985) The ideal and the feasible: physical constraints on evolution. Biol. J. Linn. Soc. 26:345-358. Amundson, R. (1996) Historical development of the concept of adaptation. Pp. 11-53. In: Adaptation. M. R. Rose and G. V. Lauder (eds.). Academic Press, San Diego. Appel, T. A. (1987) The Cuvier-Geoffroy Debate: French Biology in the Decades Before Darwin. Oxford Univ. Press, Oxford. Arthur, W. (1997) The Origin of Animal Body Plans: A Study in Evolutionary Developmental Biology. Cambridge Univ. Press, Cambridge. Barel, C. D. N. (1993) Concepts of an architectonic approach to transformation morphology. Acta Biotheor. 41:345-381. Barel, C. D. N., G. C. Anker, F. Witte, R. J. C. Hoogerhoud, and T. Goldschmidt (1989) Constructional constraint and its ecomorphological implications. Acta Morphol. Neerl.-Scand. 27:83-109. Bels, V. L., M. Chardon, and P. Vandewalle (1994) Biomechanics of Feeding in Vertebrates. Adv. Comp. Environ. Physiol. 18. SpringerVerlag, Berlin. Bemis, W. E. (1986) Feeding systems of living Dipnoi: anatomy and function. J. Morph. (Suppl.) 1:249-275. Bloot, J., A. B. Boekelaar, G. J. Groen, and B. Baljet (1985) A mesoscopic technique for the study of the development of the peripheral nervous system in rat foetuses. Acta. Morphol. Neerl.-Scand. 23:35-46. Bock, W. J., and C. R. Shear (1972) A staining method for gross dissection of vertebrate muscles. Anat. Anz. 130:222-227. Bock, W. J., and G. von Wahlert (1965) Adaptation and the formfunction complex. Evolution 19:269-299. Bookstein, F. L. (1991) Morphometric Tools for Landmark Data. Cambridge Univ. Press, Cambridge. Bookstein, F. L. (1994) Can biometrical shape be a homologous character? Pp. 197-227. In: Homology: The Hierarchical Basis of Comparative Biology. B. Hall (ed.). Academic Press, San Diego. Bramble, D. M., and D. Wake (1985) Feeding mechanisms of lower tetrapods. Pp. 230-261. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Brown, L. (1993) The New Shorter Oxford English Dictionary, Vols. 1 and 2. Clarendon Press, Oxford.
1. T e t r a p o d F e e d i n g in Vertebrate M o r p h o l o g y Biirger, R. (1986) Constraints for the evolution of functionally coupled characters: a nonlinear analysis of a phenotypic model. Evolution 40:182-193. Cole, F. J. (1944) A History of Comparative Anatomy From Aristotle to the Eighteenth Century. MacMillan and Co., London. Condon, K. (1987) A kinematic analysis of mesokinesis in the Nile monitor {Varanus niloticus). Exp. Biol. 47:73-87. Darwin, C. (1859) On the Origin of Species by Means of Natural Selection. John Murray, London (1964 facsimile reprint of the first edition. Harvard Univ. Press, Cambridge, MA). Davis, D. D. (1960) The proper goal of comparative anatomy. Pp. 4 4 50. In: Proceedings of the Centenary and Bicentenary Congress of Biology, University of Malay, Singapore. R. D. Purchon (ed.). Univ. of Malay Press, Singapore. Davis, D. D. (1964) The giant panda: a morphological study of evolutionary mechanisms. Fieldiana: Zool. Mem. 3:1-339. de Jongh, H. J., R. Dantuma, and H. M. J. Sluijsmans (1989) The shape of the mandible in the domestic sheep: a biomechanical analysis using E.M.G. as an estimator of muscle force. Acta Morphol. Neerl.-Scand. 27:63-73. Denny, M. W. (1990) Terrestrial versus aquatic biology: the medium and its message. Am. Zool. 30:111-121. Denny, M. W. (1993) Air and Water. The Biology and Physics of Life's Media. Princeton Univ. Press, Princeton. DiDio, L. J. A. (1986) Anatomy as art and science. Anat. Anz. 161: 197-208. Driesch, H. (1908) The Science and Philosophy of the Organism. Black, London. Drost, M. R., M. Muller, and J. Osse (1998) A quantitative hydrodynamical model of suction feeding in larval fishes: the role of frictional forces. Proc. Roy Soc. Lond. B 234:263-281. Drucker, E., and G. V. Lauder (1997) Aquatic propulsion in fishes by vortex ring production. Am. Zool. 37:77A. Dullemeijer, P. (1959) A comparative functional-anatomical study of the heads of some Viperidae. Morph. Jb. 99: 881-985. Dullemeijer, P. (1974) Concepts and Approaches in Animal Morphology. Van Gorcum, Assen, The Netherlands. Dullemeijer, P. (1980) Functional morphology and evolutionary biology. Acta Biotheor. 29:151-250. Dullemeijer, P. (1989) On the concept of integration in animal morphology. Pp. 3-18. In: Trends in Vertebrate Morphology. H. Splechtna and H. Hilgers (eds.). Gustav Fischer, Stuttgart. Dullemeijer, P. (1994) Conclusion: a general theory for feeding mechanics? Pp. 347-358. In: Biomechanics of Feeding in Vertebrates. Adv. Comp. Environ. Physiol. 18. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Emerson, S. B. (1991) The ecomorphology of Bornean tree frogs (family Rhacophoridae). Zool. J. Linn. Soc. 101:337-357. Filipski, G. T., and M. V. H. Wilson (1984) Sudan black B as a nerve stain for whole cleared fishes. Copeia 1984:204-208. Filipski, G. T., and M. V. H. Wilson (1985) Staining nerves in whole cleared amphibians and reptiles using Sudan black B. Copeia 1985:500-502. Frazzetta, T. H. (1962) A functional consideration of cranial kinesis in lizards. J. Morph. Ill:287-319. Frazzetta, T. H. (1983) Adaptation and function of cranial kinesis in reptiles: a time-motion analysis of feeding in alligator lizards. Pp. 222-244. In: Advances in Herpetology, and Evolutionary Biology: Essays in Honor of Ernest E. Williams. Museum of Comparative Zoology, Cambridge, MA. Frazzetta, T. H. (1994) Feeding mechanisms in sharks and other elasmobranchs. Pp. 31-57. In: Biomechanics of Feeding in Vertebrates. Adv. Comp. Environ. Physiol. 18. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin.
17
Fung, Y. C. (1993) Biomechanics: Mechanical Properties of Living Tissues, 2nd Ed. Springer-Verlag, New York. Galis, F. (1992) A model for biting in the pharyngeal jaws of a cichlid fish: Haplochromis piceatus. J. Theor. Biol. 155:343-368. Galis, F. (1993) Interactions between the pharyngeal jaw apparatus, feeding behaviour, and ontogeny in the cichlid fish, Haplochromis piceatus: a study of morphological constraints in evolutionary ecology J. Exp. Zool. 267:137-154. Galis, F (1996) The application of functional morphology to evolutionary studies. Trends Ecol. Evol. 11:124-129. Galis, F, and E. G. Drucker (1996) Pharyngeal biting mechanics in centrarchid and cichlid fishes: insights into a key evolutionary irmovation. J. Evol. Biol. 9:641-670. Gans, C. (1969) Functional components versus mechanical units in descriptive morphology. J. Morph. 128:365-368. Gans, C. (1974) Biomechanics: An Approach to Vertebrate Biology. Lippincott, Philadelphia. Gans, C. (1986) Functional morphology of predator-prey relationships. Pp. 6-23. In: Predator-Prey Relationships: Perspectives and Approaches from the Study of Lower Vertebrates. M. E. Feder and G. V. Lauder (eds.). Univ. of Chicago Press, Chicago. Gans, C. (1988) Adaptation and the form-function relation. Am. Zool. 28:681-697. Gans, C. (1992) Electromyography. Pp. 175-204. In; Biomechanics— Structures and Systems: A Practical Approach. A. A. Biewener (ed.). IRL/Oxford Univ. Press, Oxford. Gegenbauer, C. (1878) Elements of Comparative Anatomy, 2nd Ed. MacMillan and Co., London. Gerking, S. D. (1994) Feeding Ecology of Fish. Academic Press, San Diego. Ghiselin, M. T. (1980) The failure of morphology to assimilate Darwinism. Pp. 180-193. In: The Evolutionary Synthesis: Perspectives on the Unification of Biology. E. Mayr and W. B. Provine (eds.). Harvard Univ. Press, Cambridge, MA. Goodwin, B. (1989) Evolution and the generative order. Pp. 89-100. In: Theoretical Biology: Epigenetic and Evolutionary Order from Complex Systems. Edinburgh Univ. Press, Edinburgh (1992 paperback edition by The Johns Hopkins Univ. Press, Baltimore). Gosline, W. A. (1996) Structures associated with feeding in three broad-mouthed, benthic fish groups. Environ. Biol. Fish. 47: 399-405. Gould, S. J. (1966) Allometry and size in ontogeny and phylogeny. Biol. Rev. 41:587-640. Gould, S. J., and R. C. Lewontin (1979) The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme. Proc. Roy. Soc. Lond. B 205:581-598. Greaves, W. S. (1995) Functional predictions from theoretical models of the skull and jaws in reptiles and mammals. Pp. 99-115. In: Functional Morphology in Vertebrate Paleontology. J. Thomason (ed.). Cambridge Univ. Press, Cambridge. Greene, H. W. (1982) Dietary and phenotypic diversity in lizards: why are some organisms specialized? Pp. 107-128. In: Environmental Adaptation and Evolution. D. Mossakowski and G. Roth (eds.). Gustav Fischer, Stuttgart. Greene, H. W. (1986a) Diet and arboreality in the emerald monitor, Varanus prasinus, with comments on the study of adaptation. Fieldiana Zool. New Ser. No. 31:1-12. Greene, H. W. (1986b) Natural history and evolutionary biology. Pp. 99-108. In: Predator-Prey Relationships: Perspectives and Approaches from the Study of Lower Vertebrates. M. E. Feder and G. V. Lauder (eds.). Univ. of Chicago Press, Chicago. Greene, H. W. (1994) Systematics and natural history, foundations for understanding and conserving biodiversity. Am. Zool. 34: 48-56.
18
Kurt Schwenk
Greene, H. W. (1997) Snakes: The Evolution of Mystery in Nature. Univ. of California Press, Berkeley, CA. Hall, B. K. (1994) Homology: The Hierarchical Basis of Comparative Biology. Academic Press, San Diego. Hall, B. K. (1996) Bauplane, phylotypic stages and constraint: why there are so few types of animals. Pp. 215-261. In: Evolutionary Biology. M. K. Hecht, R. J. Maclntyre, and M. T. Clegg (eds.). Plenum Press, New York. Hall, B. K. (1998) Evolutionary Developmental Biology, 2nd Ed. Chapman and Hall, London (distributed by Kluwer Academic). Hanken, J., and B. K. Hall (1993) The Skull, Vols. 1-3. Univ. of Chicago Press, Chicago. Hanken, J., and R. Wassersug (1981) The visible skeleton. Func. Photog. July/August: 22-26,44. Herrel, A., P. Aerts, and F. De Vree (1998) Ecomorphology of the lizard feeding apparatus: a modelling approach. Neth. J. Zool. 48:1-25. Hiiemae, K. M., and A. W. Crompton (1985) Mastication, food transport and swallowing. Pp. 262-290. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Hildebrand, M. (1968) Anatomical Preparations. Univ. of California Press, Berkeley, CA. His, W. (1888) On the principles of animal morphology. Proc. Roy. Soc. Edinburgh 15:287-298. Ho, M.-W., and P. T. Saunders (1979) Beyond neo-Darwinism: an epigenetic approach to evolution. J. Theor. Biol. 78:573-591. Ho, M.-W., and P. T. Saunders (1984) Beyond Neo-Darwinism: An Introduction to the New Evolutionary Paradigm. Academic Press, London. Hopson, J. A., and L. B. Radinsky (1980) Vertebrate paleontology: new approaches and new insights. Paleobiology 6:250-270. Huxley, J. S. (1932) Problems of Relative Growth. Methuen and Co., London (1972 reprint by Dover Publications, New York). Larson, A., and J. B. Losos (1996) Phylogenetic systematics of adaptation. Pp. 187-220. In: Adaptation. M. R. Rose and G. V. Lauder (eds.). Academic Press, San Diego. Lauder, G. V. (1981) Form and function: structural analysis in evolutionary morphology. Paleobiology 7:430-442. Lauder, G. V (1982a) Introduction. Pp. xi-xlv. In: Form and Function: A Contribution to the History of Animal Morphology, by E. S. Russell. Univ. of Chicago Press, Chicago. Lauder, G. V. (1982b) Patterns of evolution in the feeding mechanism of actinopterygian fishes. Am. Zool. 22:275-285. Lauder, G. V. (1985) Aquatic feeding in lower vertebrates. Pp. 210229. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Lauder, G. V. (1990) Functional morphology and systematics: studying functional patterns in an historical context. Annu. Rev. Ecol. Syst. 21:317-340. Lauder, G. V. (1991) Biomechanics and evolution: integrating physical and historical biology in the study of complex systems. Pp. 119. In: Biomechanics and Evolution. J. M. V. Rayner and R. J. Wootton (eds.). Cambridge Univ. Press, Cambridge. Lauder, G. V. (1995) On the inference of function from structure. Pp. 1-18. In: Functional Morphology in Vertebrate Paleontology. J. Thomason (ed.). Cambridge Univ. Press, Cambridge. Lauder, G. V. (1996) The argument from design. Pp. 55-91. In: Adaptation. M. R. Rose and G. V. Lauder (eds.). Academic Press, San Diego. Lauder, G. V, C. Connon, and D. Dunn-Rankin (1996) Visualization of flow behind the tail of swimming fish: new data using DPIV techniques. Am. Zool. 36:7A. Lauder, G. V, R. B. Huey, R. K. Monson, and R. J. Jensen (1995) Sys-
tematics and the study of organismal form and function. BioScience 45:696-704. Lauder, G. V, and H. B. Shaffer (1993) Design of feeding systems in aquatic vertebrates: major patterns and their evolutionary interpretations. Pp. 113-149. In: The Skull, Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Liem, K. F. (1990) Aquatic versus terrestrial feeding modes: possible impacts on the trophic ecology of vertebrates. Am. Zool. 30: 209-221. Liem, K. R (1993) Ecomorphology of the teleostean skull. Pp. 422452. In: The Skull, Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Liem, K. F., and D. B. Wake (1985) Morphology: current approaches and concepts. Pp. 366-377. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Linden, R. A. (1998) The Scientific Basis of Eating. Taste, Smell, Mastication, Salivation, and Swallowing and Their Dysfunctions. Frontiers of Oral Biology Series, Vol 9. Karger, Basel. Loeb, G. E., and C. Gans (1986) Electromyography for Experimentalists. Univ. of Chicago Press, Chicago. Lombard, R. E. (1991) Experiment and comprehending the evolution of function. Am. Zool. 31:743-756. Losos, J. B., and H. W. Greene (1988) Ecological and evolutionary implications of diet in monitor lizards. Biol. J. Linn. Soc. 35: 379-407. Losos, J. B., and D. B. Miles (1994) Adaptation, constraint, and the comparative method: phylogenetic issues and methods. Pp. 6 0 98. In: Ecological Morphology: Integrative Organismal Biology. P. C. Wainwright and S. M. Reilly (eds.). Univ. of Chicago Press, Chicago. Losos, J. B., T. R. Jackman, A. Larson, K. de Queiroz, and L. Rodriquez-Schettino (1998) Contingency and determinism in replicated adaptive radiations of island lizards. Science 279:21152118. Marcus, L. R, M. Corti, A. Loy, G. J. P Naylor, and D. E. Slice (1996) Advances in Morphometries. Plenun\ Press, New York. Maynard Smith, J., R. Burian, S. Kauffman, P. Alberch, J. Campbell, B. Goodwin, R. Lande, D. Raup, and L. Wolpert (1985) Developmental constraints and evolution. Quart. Rev. Biol. 60:265287. Mayr (1980) Morphology. P. 173. In: The Evolutionary Synthesis: Perspectives on the Unification of Biology. E. Mayr and W. B. Provine (eds.). Harvard Univ. Press, Cambridge, MA. McLellan, T., and J. A. Endler (1998) The relative success of some methods for measuring and describing the shape of complex objects. Syst. Biol. 47:264-281. Motta, P. J. (1988) Functional morphology of the feeding apparatus of ten species of Pacific butterflyfishes (Perciformes, Chaetodontidae): an ecomorphological approach. Env. Biol. Fish. 22: 39-67. Motta, P. J., and K. M. Kotrschal (1992) Correlative, experimental, and comparative evolutionary approaches in ecomorphology. Neth. J. Zool. 42:400-415. Norton, S. F., and E. L. Brainerd (1993) Convergence in the feeding mechanics of ecomorphologically similar species in the Centrarchidae and Cichlidae. J. Exp. Biol. 176:11-29. Nyhart, L. K. (1995) Biology Takes Form: Animal Morphology and the German Universities 1800-1900. Univ. of Chicago Press, Chicago. Oelrich, T. M. (1954) The anatomy of the head of Ctenosaura pectinata (Iguanidae). Misc. Publ. Mus. Zool. Univ. Michigan, No. 94. Often, E. (1989) Mathematical and philosophical reflections on motor control systems. Acta Morphol. Neerl.-Scand. 27:163-169. Owen, R. (1848) On the Archetype and Homologies of the Vertebrate Skeleton. John Van Voorst, London.
1. T e t r a p o d F e e d i n g in Vertebrate M o r p h o l o g y Owen, R. (1849) On the Nature of Limbs. John Van Voorst, London. Owen, R. (1866) On the Anatomy of Vertebrates, Vol. I. Fishes and Reptiles. Longmans, Green, and Co., London. Padian (1995) Form versus function: the evolution of a dialectic. Pp. 264-277. In: Functional Morphology in Vertebrate Paleontology. J. Thomason (ed.). Cambridge Univ. Press, Cambridge. Piaget, J. (1970) Structuralism. Basic Books, New York. Presnell, J. K., and M. P. Schreibman (1997) Humason's Animal Tissue Techniques, 5th Ed. The Johns Hopkins Univ. Press, Baltimore, MD. Przibram, H. (1931) Connecting Laws in Animal Morphology. Univ. of London Press, London. Rainger, R. (1989) What's the use: William King Gregory and the functional morphology of fossil vertebrates. J. Hist. Biol. 22: 103-139. Raff, R. A. (1996) The Shape of Life: Genes, Development, and the Evolution of Animal Form. Univ. of Chicago Press, Chicago. Rayner, J. M. V, and R. J. Wootton (1991) Biomechanics and Evolution. Cambridge Univ. Press, Cambridge. Reif, W.-E., R. D. K. Thomas, and M. S. Fischer (1985) Constructional morphology: the analysis of constraints in evolution. Acta Biotheor. 34:233-248. Riedl, R. (1978) Order in Living Organisms: A Systems Analysis of Evolution. Wiley, New York. Rieppel, O. (1979) A functional interpretation of the varanid dentition. Gegenbaurs Morphol. Jahr. 125:797-817. Rieppel, O. (1986) Atomism, epigenesis, preformation and pre-existence: a clarification of terms and consequences. Biol. J. Linn. Soc. 28:331-341. Rieppel, O. (1990) Structuralism, functionalism, and the four Aristotelian causes. J. Hist. Biol. 23:291-320. Rohlf, F. J., and F. L. Bookstein (1990) Proceedings of the Michigan Morphometries Workshop. Special Publication No. 2, Mus. of Zool., Univ. of Michigan. Romer, A. S. (1956) Osteology of the Reptiles. Univ. of Chicago Press, Chicago. Roth, B., and D. Wake (1985) Trends in the functional morphology and sensorimotor control of feeding behavior in salamanders: an example of the role of internal dynamics in evolution. Acta Biotheor. 34:175-192. Roth, V. L. (1989) Fabricational noise in elephant dentition. Paleobiology 15:165-179. Roth, V L. (1993) On three-dimensional morphometries, and on the identification of landmark points. Pp. 41-61. In: Contributions to Morphometries. L. F. Marcus, E. Bello, and A. Garcia-Valdecasas (eds.). Museo Nacional de Ciencias Naturales CSIC, Madrid. Russell, A. P., and J. J. Thomason (1993) Mechanical analysis of the mammalian head skeleton. Pp. 345-383. In: The Skull, Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Russell, A. P , R. L. Walker, and A. M. Bauer (1988) A technique for visualization of the circulatory system in small lizards. Copeia 1988:797-800. Russell, E. S. (1916) Form and Function: A Contribution to the History of Animal Morphology. John Murray, London (1917 American edition, E. P. Dutton and Co., New York). Sanderson, S. L., and R. Wassersug (1993) Convergent and alternative designs for vertebrate suspension feeding. Pp. 37-112. In: The Skull, Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Sanderson, S. L., J. J. Cech, and M. R. Patterson (1991) Fluid dynamics in suspension-feedingblackfish. Science 251:1346-1348. Schaefer, S. A., and G. V. Lauder (1986) Historical transformation of functional design: evolutionary morphology of feeding mechanisms in loricarioid catfishes. Syst. Zool. 35:489-508. Schlichting, C. D., and M. Pigliucci (1998) Phenotypic Evolution: A Reaction Norm Perspective. Sinauer Assoc, Sunderland, MA.
19
Schmidt-Kittler, N., and K. Vogel (1991) Constructional Morphology and Evolution. Springer-Verlag, Berlin. Schwenk, K. (1994) Why snakes have forked tongues. Science 263: 1573-1577. Schwenk, K. (1995) A utilitarian approach to evolutionary constraint. Zoology 98:251-262. Schwenk, K. (1996) Why snakes flick their tongues. Am. Zool. 36:84A. Schwenk, K. (2000) Functional units and their evolution. In: The Character Concept in Evolutionary Biology. G. P. Wagner (ed.). Academic Press, San Diego. Schwenk, K., and G. P. Wagner (in prep.) The reconciliation of selection and constraint. Schwenk, K., F. A. Jenkins, Jr., and J. Sheen (in prep.) The hydrostatic trigger hypothesis of chameleon tongue projection: an in vivo test. Secor, S. M., and J. Diamond (1998) A vertebrate model of extreme physiological regulation. Nature 395:659-662. Seilacher, A. (1970) Arbeitskonzept zur Konstructions-Morphologie. Lethaia 3:393-396. Seilacher, A. (1973) Fabricational noise in adaptive morphology. Syst. Zool. 22:451-465. Seilacher, A. (1979) Constructional morphology of sand dollars. Paleobiology 5:191-221. Simpson, G. G. (1953) The Major Features of Evolution. Columbia Univ. Press, New York. Singer (1957) A Short History of Anatomy from the Greeks to Harvey, 2nd Ed. Dover, New York. Smith, K. K. (1982) An electromyographic study of the function of the jaw adducting muscles in Varanus exanthematicus (Varanidae). J. Morph. 173:137-158. Smith, K. K. (1993) The form of the feeding apparatus in terrestrial vertebrates: studies of adaptation and constraint. Pp. 150-196. In: The Skull, Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Smith, K. C. (1992) Neo-rationaHsm versus neo-Darwinism: integrating development and evolution. Biol. Phil. 7:431-451. Stouder, D. J., K. L. Fresh, and R. J. Feller (1994) Theory and Application in Fish Feeding Ecology. Belle W. Baruch Institute for Marine Biology, Univ. of South Carolina Press, Columbia. Thomason, J. J. (1995) Functional Morphology in Vertebrate Paleontology. Cambridge Univ. Press, Cambridge. Thompson, D. W. (1942) On Growth and Form: A New Edition. Cambridge Univ. Press, Cambridge (1992 facsimile reprint. On Growth and Form: The Complete Revised Edition. Dover, New York). van der Klaauw, C. J. (1945) Cerebral skull and facial skull: a contribution to the knowledge of skull structure. Arch. Neerl. Zool. 7:16-37. van der Klaauw, C. J. (1948) Ecological studies and reviews. IV Ecological morphology. Biblio. Biotheor. 4:27-111. van der Weele, C. (1993) Metaphors and the privileging of causes: the place of environmental influences in explanations of development. Acta Biotheor. 41:315-321. Vandewalle, P., A. Huyssene, P. Aerts, and W. Verraes (1994) The pharyngeal apparatus in teleost feeding. Pp. 59-92. In: Biomechanics of Feeding in Vertebrates. Adv. Comp. Environ. Physiol. 18. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. van Leeuwen, J. L. (1991) Optimum power output and structural design of sarcomeres. J. Theor. Biol. 149:229-256. van Leeuwen, J. L., and C. W. Spoor (1992) Modelling mechanically stable muscle architectures. Phil. Trans. Roy. Soc. Lond. B 336: 275-292. van Leeuwen, J. L. (1997) Why the chameleon has spiral-shaped muscle fibres in its tongue. Phil. Trans. Roy. Soc. Lond. B 352: 573-589.
20
Kurt Schwenk
Vincent, J. (1990) Structural Biomaterials, 2nd Ed. Princeton Univ. Press, Princeton. Vogel, S. (1988) Life's Devices: The Physical World of Animals and Plants. Princeton Univ. Press, Princeton. Vogel, S. (1994) Life in Moving Fluids. The Physical Biology of Flow, 2nd Ed. Princeton Univ. Press, Princeton. Wagner, G. P. (1986) The systems approach: an interface between development and population genetic aspects of evolution. Pp. 149165. In: Patterns and Processes in the History of Life. D. M. Raup and D. Jablonski (eds.). Springer-Verlag, Berlin. Wagner, G. P. (1988) The influence of variation and of developmental constraints on the rate of multivariate phenotypic evolution. J. Evol. Biol. 1:45-66. Wagner, G. P. (2000) The Character Concept in Evolutionary Biology. Academic Press, San Diego. Wagner, G. P., and L. Altenberg (1996) Complex adaptations and the evolution of evolvability. Evolution 50:967-976. Wagner, G. P., and K. Schwenk (1999) Evolutionarily stable configurations: functional integration and the evolution of phenotypic stability. Pp. 155-217. In: Evolutionary Biology, Vol. 31. M. K. Hecht, R. J. Maclntyre, and M. T. Clegg (eds.). Kluwer Academic/ Plenum Press, New York. Wainwright, P. C. (1991) Ecomorphology: experimental functional anatomy for ecological problems. Am. Zool. 31:680-693. Wainwright, P. C., and A. F. Bennett (1992) The mechanism of tongue projection in chameleons. II. Role of shape change in a muscular hydrostat. J. Exp. Biol. 168:23-40. Wainwright, P C., and G. V. Lauder (1992) The evolution of feeding biology in sunfishes (Centrarchidae). Pp. 472-491. In: Systematics, Historical Ecology, and North American Freshwater Fishes. R. Mayden (ed.). Stanford Univ. Press, Stanford, CA. Wainwright, P C., and S. M. Reilly (eds.) (1994a) Introduction. Pp. 19. In: Ecological Morphology: Integrative Organismal Biology. Univ. of Chicago Press, Chicago. Wainwright, P. C , and S. M. Reilly (1994b) Ecological Morphology: Integrative Organismal Biology. Univ. of Chicago Press, Chicago. Wainwright, P. C , and B. A. Richard (1995) Predicting patterns of prey use from morphology of fishes. Environ. Biol. Fishes 44: 97-113. Wainwright, S. A. (1988) Axis and Circumference: The Cylindrical Shape of Plants and Animals. Harvard Univ. Press, Cambridge, MA. Wainwright, S. A., W. D. Biggs, J. D. Currey, and J. M. Gosline (1976) Mechanical Design in Organisms. Edward Arnold, London. Waisbren, S. J. (1988) The importance of morphology in the evolutionary synthesis as demonstrated by the contributions of the Oxford group: Goodrich, Huxley, and de Beer. J. Hist. Biol. 21:291-330. Wake, D. B. (1982) Functional and evolutionary morphology. Persp. Biol. Med. 25:603-620. Wake, D. B. (1991) Homoplasy: the result of natural selection, or evidence of design limitations? Am. Nat. 138:543-567. Wake, D. B., and A. Larson (1987) Multidimensional analysis of an evolving lineage. Science 238:42-48. Wake, D. B., and G. Roth (1989) Complex Organismal Functions: Integration and Evolution in Vertebrates. Wiley, Chichester. Wake, D. B., G. Roth, and M. Wake (1983) On the problem of stasis in organismal evolution. J. Theor. Biol. 101:211-224.
Wake, M. H. (1991) The impact of functional morphology and biomechanics on studies of evolutionary biology. Pp. 555-557. In: The Unity of Evolutionary Biology: Proceedings of the Fourth International Congress of Systematic and Evolutionary Biology. E. C. Dudley (ed.). Dioscorides Press, Portland. Wake, M. H. (1992) Morphology, the study of form and function, in modern evolutionary biology. Pp. 289-346. In: Oxford Surveys in Evolutionary Biology, Vol. 8. D. J. Futuyma and J. Antonovics (eds.). Oxford Univ. Press, Oxford. Wassersug, R. J. (1976) A procedure for differential staining of cartilage and bone in whole formalin-fixed vertebrates. Stain Tech. 51: 131-134. Webster, G., and B. Goodwin (1996) Form and Transformation. Generative and Relational Principles in Biology. Cambridge Univ. Press, Cambridge. Weibel, E. R. (2000) Symmorphosis: On Form and Function in Shaping Life. Harvard Univ. Press, Cambridge, MA. Weibel, E. R., C. R. Taylor, and L. Bolis (1998) Principles of Animal Design: The Optimization and Symmorphosis Debate. Cambridge Univ. Press, Cambridge. Weishampel, D. B. (1993) Beams and machines: modeling approaches to the analysis of skull form and function. Pp. 303-344. In: The Skull, Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Weishampel, D. B. (1995) Fossils, function and phylogeny. Pp. 3 4 54. In: Functional Morphology in Vertebrate Paleontology. J. Thomason (ed.). Cambridge Univ. Press, Cambridge. Westneat, M. W. (1990). Feeding mechanics of teleost fishes (Labridae: Perciformes): a test of four-bar linkage models. J. Morph. 205: 269-295. Westneat, M. W. (1995a) Feeding, fimction, and phylogeny: analysis of historical biomechanics in labrid fishes using comparative methods. Syst. Biol. 44:361-383. Westneat, M. W. (1995b) Phylogenetic systematics and biomechanics in ecomorphology. Environ. Biol. Fishes 44:263-283. Whyte, L. L. (1965) Internal Factors in Evolution. George Braziller, New York. Woodger, J. H. (1929) Biological Principles: A Critical Study. Kegan Paul, Trench, Trubner and Co., London. Wu, E. H. (1994) Kinematic analysis of jaw protrusion in orectolobiform sharks: a new mechanism for jaw protrusion in elasmobranchs. J. Morph. 222:175-190. Zelditch, M. L., R L. Bookstein, and B. L. Lundrigan (1992) Ontogeny of integrated skull growth in the cotton rat Sigmodon fulviventer. Evolution 46:1164-1180. Zelditch, M. L., R L. Bookstein, and B. L. Lundrigan (1993) The ontogenetic complexity of developmental constraints. J. Evol. Biol. 6:621-641. Zweers, G. (1991) Pathways and space for evolution of feeding mechanisms in birds. Pp. 530-547. In: The Unity of Evolutionary Biology: Proceedings of the Fourth International Congress of Systematic and Evolutionary Biology. E. C. Dudley (ed.). Dioscorides Press, Portland. Zweers, G., and J. C. Vanden Berge (1997) Evolutionary transitions in the trophic system of the wader-waterfowl complex. Neth. J. Zool. 47:255-287.
C H A P T E R
2 An Introduction to Tetrapod Feeding KURT SCHWENK
Department of Ecology and Evolutionary Biology University of Connecticut Storrs, Connecticut 06269
11.
III. IV. V.
A phylogeny of the Tetrapoda and its relatives is given in Fig. 2.1. All taxon names used in this chapter are in reference to this phylogeny. More detailed phylogenies for particular clades are given in subsequent chapters.
INTRODUCTION A. Feeding Form and Function: Background B. Anatomical Terminology C. Phylogenetic Terminology MORPHOLOGY OF THE FEEDING APPARATUS A. Skull and Mandible B. Teeth C. Keratinous Structures D. Hyobranchial Apparatus E. Jaw Musculature F. Hyobranchial Musculature G. Tongue H. Pharynx I. Cheeks, Lips, and Probosces KINEMATICS OF FEEDING: THE GAPE CYCLE KINEMATICS OF FEEDING: FEEDING STAGES A. Overview B. Stages of the Feeding Cycle CONCLUDING REMARKS References
A. Feeding Form and Function: Background Feeding and excreting are common to all animal life. Feeding is required by virtue of the fact that animals, unlike plants (and other autotrophic organisms), cannot harness directly the radiant energy of the sun—the ultimate source of all energy on Earth. Excretion occurs because in every mouthful of material ingested there is some portion of it that an animal cannot assimilate. Solid matter is excreted via the gut, and following intracellular energy extraction, other molecular detritus is eliminated via the urinary system. Apart from the obvious waste implied by excretion, there is a more subtle inefficiency built into every digestive process—the conversion of energy from one form to another. Food that is processed and digested is ultimately reduced to its molecular components and these are circulated throughout the body via the blood-vascular system. Different food molecules have different fates and even the same molecule can be treated differently depending on the physiological state of the individual, but most food is eventually broken down into glucose molecules. These are transported across membranes into the cytoplasm of every living cell and are then systematically dismantled and oxidized in a process known as cellular respiration.
I. INTRODUCTION This chapter serves as a primer for the study of tetrapod feeding systems. It provides a brief overview of feeding form and function, introducing general concepts and a basic vocabulary. Its goal is to prepare the reader for the more detailed, taxon-based chapters that follow, and for the primary literature in functional and evolutionary morphology of tetrapod feeding.
FEEDING (K.SchwenKed.)
21
Copyright © 2000 by Academic Press. All rights of reproductioi\ in any form reserved.
22
Kurt Schwenk
^^-
0.5
0
^ ^ ] — Upper Jaw Tip
0.9
Lower Jaw Tip —M^^ yj^-'-^^^O
"OooCr^
J-
0.4 0
28
56
83
111
139
167
194
222
250
30 20 ^ ^
10 0
. . ^
.vv^f
lead An gle
L^^^ yW^
1
^^^
-10
'^'^^^
-20 28
56
83
111
139
167
194
222
250
278
306
Time (ms) F I G U R E 4.5. Kinematic profiles for the bolitoglossine plethodontid salamander Hydromantes platycephalus. The panels are, from top to bottom, tongue reach and gape distance, gape distance, jaw deflections, and head angle, all presented on the same time axis. Gape distance is presented in both the first and the second panels, on different ordinates, to illustrate the vast differences in excursion between the tongue and jaws. This preycapture event was performed with small prey and lacks a distinct second phase of mouth opening, which often occurs during tongue retraction.
4. Terrestrial Feeding in Salamanders form a complex, pimiate muscle wrapped around the tapered epibranchial. The protractor muscles are differentiated in the plethodontids into an anterior, parallel-fibered portion and a distinct, posterior, bulb-like muscle, which is greatly enlarged in relation to that found in members of other families. The hyolingual apparatus contains two triangular units, formed by the medial basibranchial and the ceratobranchials on either side, so in order to form a compact bundle without bending, it folds in three dimensions as it is protracted, moving along a morphological track having the geometrical form of a tractrix. This form has favorable attributes. It appears to act as an accelerator of movement and contributes to the rapidity of protraction. It also acts as a brake for the returning apparatus. The three-dimensional expansion of a tractrix is known geometrically as a bugle body, and it exists, in a limited degree, in the floor of the mouth. The sides of the bugle body are formed by the medial margins of the ceratohyals. The bottom is formed by intermandibularis muscles and the top by a strap-like, unpaired suprapeduncularis muscle (the last muscle unique to the Plethodontidae). The structural element thus formed, called the cylinder by Lombard and Wake (1977), controls the direction of tongue protraction. In species that are most proficient in tongue protraction, members of the plethodontid tribe Bolitoglossini, the cylinder is well formed and incorporates a number of muscles that serve different functions in other taxa. These muscles include the anterior fibers of the subarcualis rectus I, the geniohyoideus medialis, anterior fibers of the rectus cervicis superficialis, and the hebosteoypsiloideus. According to the biomechanical model of Lombard and Wake (1976, 1977) movement of the cylinder from side to side within the mouth is possible. The cylinder appears to rest in the floor of the mouth attached posterolateral^ to the mandible by a slender and poorly defined mandibulohyoid ligament and anteriorly by the geniohyoidius lateralis. Thus there appears to be a kind of "firing platform," and a relatively large contraction of the left geniohyoideus lateralis is hypothesized to direct the tongue toward the right. The biomechanical model also hypothesizes that a mechanical linkage between parts of the hyolingual apparatus accomplishes at least a partial rotation of the tongue pad during protraction. The very action of folding is biomechanically coupled to rotate the tongue pad around the tip of the basibranchial so that the fleshy, mucous-covered pad contacts the prey. A ligament-like bundle of connective tissue extends from the anterolateral part of the first ceratobranchial into the substance of the tongue pad. This fiber bundle extends anteriorly from each side, coalescing at the ventral midline of the anterior tip of the basibranchial and then
105
variously attaching to a flexible tip of the basibranchial, to a detached anterior part of the basibranchial known as the lingual cartilage, or extending dorsally around the tip of the basibranchial and then fanning out posteriorly into the substance of the tongue pad. When the skeleton folds as it is protracted, the fibers become taut and the tongue pad is pulled forward to rotate around the tip of the basibranchial in such a fashion that the sticky dorsal surface is presented to the prey. The cylinder of plethodontids is lined along its inner surface by serous glands, which apparently lubricate the bundle within it during protraction. Protraction can result in the bundle being rapidly propelled forward and can be so great as to result in the skeletal elements, totally evacuating the cylinder as well as the bulb formed by the subarcualis rectus muscles. Momentum of the projectile carries the epibranchial cartilages fully out of the mouth in species of the genus Hydromantes and other bolitoglossine species we have observed. In these taxa the tongue is fired ballistically from the mouth as a projectile (Fig. 4.6; Deban et ah, 1997). Auxiliary protraction mechanisms function in other, nonplethodontid taxa. The subhyoideus connects the posterolateral parts of the ceratohyals to the mandible (see Fig. 3.5B). When these paired muscles contract, the ceratohyals are moved forward as a first stage in tongue protraction. The subarcualis rectus are presumably firing at the same time so the two-stage protraction involves (1) the anterior movement of the ceratohyals carrying with them the entire tongue apparatus, and (2) the independent protraction of the articulated hyobranchial apparatus and attached tongue pad relative to the first segment (Findeis and Bemis, 1990; Miller and Larsen, 1990). This has been termed the "mobile ceratohyal system" by Findeis and Bemis (1990), who contrast it with the other main evolutionary trend involving a "stable ceratohyal system" in plethodontids (which we believe is the ancestral condition). Another auxiliary system found in salamandrids (Chioglossa and Salamandrina) involves the rotation of elongated radii (Fig. 4.2) in an arc around the tip of the strengthened, mineralized basibranchial, which is "T" shaped in cross section (Ozeti and Wake, 1969). This action carries the tongue pad forward, effectively flipping the free posterior flap of the pad well out of the mouth. c. Prehension and the Tongue Pad The tongue pad varies considerably in shape among terrestrial salamanders. In general, it is attached firmly to somewhat loosely at the front, has varying degrees of freedom along the sides, and has the greatest
106
D a v i d B. W a k e a n d S t e p h e n M. D e b a n
F I G U R E 4.6. Photograph of Hydromantes genei capturing a housefly, showing projection of the hyoHngual apparatus. The projectile includes the retractor muscle and the bundle of cartilages wrapped in a cylindrical membrane of connective tissue and epithelium. The tongue pad has enveloped the fly, but is not extended so maximal reach for the tongue would be considerably longer than is apparent in this photograph. The epibranchial has left the borders of the mouth. The bends near the distal tip of the projectile indicate the positions of the end of the basibranchial and the epibranchial-ceratobranchial joints.
degree of freedom posteriorly. The posterior part of the tongue may be extended into two limbs that are loose and form relatively long strands during full protraction. This is especially true in some salamandrids and some plethodontids. In two plethodontid clades the tongue has lost its anterior attachment (the genioglossus muscles are absent) and the tongue has a mushroom-like shape. In these species the pad typically becomes relatively small and round; these groups show the greatest distance of tongue protraction. Shaping of the tongue pad may be facilitated by the contraction of muscles inside the tongue pad that arise from the tip of the basibranchial and fan out into the posterior part of the tongue (hyoglossus), extend to the tips of the radii (basiradialis), extend between the tips of the radii (interradialis), or extend from the tips of the radii into the tongue pad (radioglossus). For the plethodontid salamanders, Lombard and Wake (1977) proposed three functional classes of tongue pad muscles: rotators (genioglossus, circumglossus, basiradialis, and intraglossus), molders (interradialis and hyoglossus), and restorers (rectus cervicis profundus). Families differ substantially with respect to relative numbers, sizes, and proportions of skeletal elements and muscles pres-
ent, and the tongue pad varies significantly among taxa; however, this diversity has not been the subject of detailed comparative analysis. The surface of the tongue is rotated by a combination of skeletal and muscular movements so that the glandular dorsal surface covered with relatively sticky mucous contacts the prey. The tongue pad shape is transformed and expanded during protraction so that it is relatively large and expansive when it contacts the prey. In those taxa having an otoglossal cartilage, movement of this element during protraction is postulated to carry the glandular field of the tongue pad dorsoanteriorly so that it contacts the prey (Larsen et ah, 1996). In ambystomatids, the tongue pad is rotated forward as protraction occurs via the connection between the first radii and the hypohyals. The mechanism of tongue protraction in plethodontids is somewhat simplified in relation to the other group that has been studied in detail, the ambystomatids (Reilly and Lauder, 1989, 1990b). In the ambystomatids the ventral throat constrictors (intermandibularis and interhyoideus) and two longitudinal muscles (geniohyoideus and genioglossus) are active during protraction. The constrictors and the geniohyoids are
4. Terrestrial Feeding in Salamanders implicated in provision of a lift vector to the hyolingual apparatus. The subarcualis rectus I provides a separate forward and upward vector, and when well protracted the geniohyoids and genioglossus provide a third forward and ventral vector to the protraction. The combined effect of these three vectors is a resultant horizontal vector that advances the entire hyolingual system, and flipping of the tongue pad is attributed in part to the action of the genioglossus. d. Retraction The articulated hyolingual skeleton has one retractor muscle attached to it at the posterior border of the anterior end of the first ceratobrancial. This is a muscle with several names, but is usually termed the sternohyoideus or the rectus cervicis lateralis (in some taxa the muscle is not well differentiated from the rectus cervicis superficialis). It arises from the sternum and is an anterior continuation of the rectus abdominis series. However, this muscle appears to be a secondary retractor, and the main retractor is a muscle that extends forward under the ventral surface of the hyolingual skeleton and passes through the triangular gap between the first and the second ceratobranchials and the basibranchial. From this position the paired muscles extend forward and then bend abruptly dorsally to enter the tongue pad, where they typically attach to a mass of connective tissue long known by the German term Sehnenplatte ("tendon plate"). This muscle is also known by various names, including the abdominohyoideus and the rectus cervicis profundus. This muscle arises from the lateral margin of the sternum, from a tendinous inscription lateral to the sternum that separates it from the rectus abdominis profundus, or as a direct and continuous anterior extension of the last named muscle, which in extreme cases (Plethodontidae) represents an undivided muscle that arises from the posterior border of the ischium and proceeds uninterrupted into the tongue pad. Retraction of the tongue appears to be exclusively the result of the action of the two different segments of the rectus cervicis series. In the most proficient tongue protractors, the bolitoglossine plethodontids, the rectus cervicis lateralis muscle is absent in two of the three major lineages, and retraction is exclusively by the rectus cervicis profundus. The omohyoideus is also missing in these taxa, and the hebosteoypsiloideus, which is part of the general retractor system in most taxa, is incorporated into a muscular cylinder through which the tongue is folded as it is protracted. e. Speed and Distance of Lingual Feeding The effectiveness of lingual feeding has not been studied in most taxa. As a generalization, species dem-
107
onstrating true tongue projectility are the fastest, the most accurate, and have the greatest range. The fastest tongue recorded to date is that of Bolitoglossa occidentalis, which takes less than 10 msec from the start of electrical activity in the subarcualis rectus I muscles to the time the tongue strikes the target. Maximum tongue extension (from tip of snout) has been measured in a number of species. In six species of Ambystoma the largest average distance was 2.4 ± 0.3 mm {A. californiense) and the smallest was 0.3 ± 0.4 mm {A. niabeei){Beneski et al., 1995). Hynobiids can extend their tongues only a few millimeters beyond the tip of the snout, but as much as 7% of snout-vent length (Larsen et al, 1989, 1996). However, Hynobius kimurae was found to protract its tongue 4 - 6 mm beyond the symphisis (J. Larsen, personal communication), and in our laboratory a large (6 to 7 cm snoutvent length) Salamandrella keyserlingii has protracted its tongue 6.6 mm, thus some hynobiids are apparently far more proficient than those reported to date. Salamandrids are apparently the most variable with respect to maximal tongue extension. Pachytriton brevipes, an aquatic species that lacks a defined tongue pad (Ozeti and Wake, 1969), retracts rather than protracts the hyolingual apparatus during terrestrial feeding (Miller and Larsen, 1990), thus in essence performing an aquatic, suction-feeding behavior on land (see Chapter 3 for details of this behavior, including Fig. 3.2). In most species of salamandrids, tongue protraction is relatively short, from 1.1 to 2.7 mm (average extension beyond snout), the latter distance in Tylototriton verrucosus (Miller and Larsen, 1990). Salamandrina terdigitata has a complex tongue protraction involving an initial protraction of the hyolingual apparatus from the mouth and then flipping of the pad (Ozeti and Wake, 1969; Miller and Larsen, 1990). The average maximal tongue extension recorded by Miller and Larsen (1990) for this species is 7.4 mm, or 20% of snout-vent length (4.7 mm is reported in the paper, but a printer's error reversed the digits; J. Larsen, personal communication). Chioglossa lusitanica probably has the longest tongue extension of the salamandrids, based on its morphology (Ozeti and Wake, 1969), but it remains largely unstudied. Maximum tongue extension is variable in plethodontids; distances are about 7% of snoutvent length in species with attached tongues {Desmognathus quadramaculatus and Plethodon glutinosus), about 15% in a species with an attached protrusible tongue {Ensatina eschscholtzii), and 30-44% in Bolitoglossa occidentalis (Thexton et ah, 1977; Larsen et ah, 1989). This species is capable of firing the tongue effectively for at least 17 mm from the head, or as much as 44% of the snout-vent length of the salamander (Thexton et ah, 1977). Hydromantes italicus, a larger species, is reported
108
David B. Wake and Stephen M. Deban
to be capable of capturing prey 40 mm from the head, and the intermediate-sized Bolitoglossa subpalmata can project its tongue 30 mm (Roth, 1987). Adult Hydromantes supramontis can project the tongue accurately to a distance of about 80% of snout-vent length (over 60 mm from the head)(Fig. 4.6; Deban et al, 1997). The prey-capture success rate has been measured in only a few species, but is higher in the proficient tongue protractors than in more generalized species. Salamandra salamandra was successful in only 39% of attempts (Luthardt-Laimer, 1983) compared with above 50% in the generalized plethodontid genera Plethodon, Eurycea, and Batrachoseps. The very long tongued bolitoglossine genera Bolitoglossa and Hydromantes "rarely miss," although no exact figures are available (Roth, 1987). These last two genera may engage in "stalking" in the laboratory to move slowly into range (Roth, 1987; personal observation). The total length of the kinematic cycle is variable among the different taxa studied. The mean gape cycle (in species with tongue protrusion) ranges from 92.8 to 115.7 msec in hynobiids (Larsen et al, 1996), 78 to 214 msec in ambystomatids (Beneski et al, 1995), 100 to 238 in salamandrids (Miller and Larsen, 1990), and 87 to 110 in plethodontids (Larsen et al, 1989; Beneski and Larsen, 1988). Gape cycle time is apparently rather similar across a wide array of morphologies. Speed of the tongue strike (tongue protraction, equivalent to phase II of Beneski and Larsen, 1988) varies greatly among taxa. The fastest tongues are found in Bolitoglossa, which has a mean duration of phase II of 5.0-7.7 msec, with Ensatina at 11.6 msec being only a little slower (although we have measured one individual Ensatina at ca. 7 msec from the time the tongue left the mouth until it touched the prey). Pseudotriton ruber, a hemidactyliine with a free tongue, has a protraction time of as little as 11 msec (Deban, 1997). The plethodontids with attached tongues are slower still: Plethodon glutinosus takes 19.3 msec and Desmognathus quadramaculatus takes 37.3 msec to protract the tongue (Larsen et al, 1989). Roughly comparable times are reported for other taxa (different papers use slightly different methods of reporting). Ambystomatids range from 16 msec for Ambystoma mabeei to 87 msec for Ambystoma cingulatum (Beneski et al, 1995). Maximum tongue protraction in Ambystoma tigrinum is reported by Reilly and Lauder (1989) to take 45 msec, and by Dockx and De Vree (1986) to take 39.6 ± 1 1 . 0 msec. Hynobiids are reported to range from 25.1 msec for Hynobius kimurae to 36.0 msec for H. nebulosus by Larsen et al (1996) [27.4 msec and 36.0 msec for the same species by Larsen et al (1989)]. Among salamandrids, tongue protraction times are reported by Miller and
Larsen (1990) to range from 22.0 msec in Salamandra salamandra to 111.9 msec in Paramesotriton hongkongensis. Dockx and De Vree (1986) report 84 ± 21.7 msec for tongue protraction in Salamandra salamandra, and Findeis and Bemis (1990) report a range of durations from 80 to 140 msec for Taricha torosa. f. Physiology Electromyographic (EMG) investigations of tongue movement (Thexton et al, 1977; Reilly and Lauder, 1990b) in two very different taxa (a bolitoglossine plethodontid with a highly protractible, free tongue and an ambystomatid with a weakly protrusible, attached tongue) show that the protractor (subarcularis rectus I) and the retractor (rectus cervicis) have essentially synchronous onset and similar motor patterns. Because electrical signals are delivered to the protractors and retractors simultaneously, the activity of the system is thought to be controlled by the peripheral organization of the system. The length-tension properties of the protractors favor immediate force generation, whereas those of the much longer (and in terrestrial plethodontids, somewhat lax and even looped) retractors have delayed biomechanical activity. Differences in contractile properties of the two muscles, such as time to peak tension, remain a possibility that is yet unstudied. The similarity in EMG patterns of the two species studied may suggest that motor patterns are phylogenetically conserved. While the electromyographic patterns of larval and metamorphosed ambystomatids suction feeding in water are remarkably similar, they are radically different from those of adults feeding on land (Shaffer and Lauder, 1988; Lauder and Shaffer, 1988). In general, there is little variability detected in experiments, and the motor patterns appear to be highly stereotyped. These facts suggest that changes in the peripheral arrangement of the muscles, skeletal elements, and other connective tissues are critical in determining biomechanical function during feeding. Most of the physiological work has involved ambystomatids (Reilly and Lauder, 1990b). The epaxial muscles connecting the neck to the head and the mandibular depressors are active prior to mouth opening (which involves raising the head slightly to greatly in all species studied) and they remain active until mouth closing. There is a second peak of activity in the epaxial muscles during tongue retraction. The jaw adductors are also active prior to mouth opening, but they have a large burst of activity again as the jaws close. The muscles responsible for tongue protraction (subarcualis rectus) and pad flipping (genioglossus) have peaks of activity during the tongue protraction phase, but unexpectedly the subarcualis rectus muscles have
109
4. Terrestrial Feeding in Salamanders a second peak during the retraction phase. Buccal muscles in the floor of the mouth are active throughout the cycle, but tend to peak early and then taper off. The main retractor muscle (rectus cervicis profundus) peaks at the time retraction begins and activity continues until after mouth closing. We take the coactivity of antagonistic muscles as a sign of fine control over tongue and jaw movements and the prolonged activity of the buccal floor muscles as an indication that they are performing a stabilizing function. In Bolitoglossa strain gauges were used to measure force of tongue impact with the prey (Thexton et ah, 1977). At relatively short distances, force varies considerably, probably indicating differences in motivation and concomitant muscle activity level. However, at distances exceeding 10 mm from the snout (animals have a maximal head-body size of about 44 mm), force fell off dramatically as a function of distance from the snout, finally failing to register at the greatest distance (19 mm). This supports the idea that there is a ballistic phase (at the end of muscular protraction) to long-distance tongue projection in the species most proficient in tongue protraction. 2. Jaw Feeding Use of the jaws to capture prey is unusual in terrestrial salamanders, but it does occur, at least in members of the families Ambystomatidae and Salamandridae (Larsen and Guthrie, 1975; Miller and Larsen, 1990). We have also observed it infrequently in various terrestrial species of the Plethodontidae in circumstances when the tongue fails to apprehend the prey. In addition, we have observed the semiaquatic plethodontid Desmognathus quadramaculatus and the hynobiids Salamandrella keyserlingi and Batrachuperus persicus using jaw prehension in water and the fully aquatic plethodontid Desmognathus marmoratus using either tongue or jaw prehension in water (see also Schwenk and Wake, 1993). Kinematics of the gape profile of salamanders using jaw prehension differs from either the generalized four-part or the specialized three-part pattern of other taxa and consists of a bell-shaped, two-part gape profile. The action is relatively rapid (on the order of 60 msec) and resembles that of fully metamorphosed ambystomatids that feed underwater without using tongue protraction (Reilly and Lauder, 1989). While the standard pattern of feeding in salamanders involves participation of the tongue in apprehending the prey, species such as Pachytriton hrevipes, which effectively lacks a tongue pad (Ozeti and Wake, 1969), and other mainly aquatic salamanders feeding on land (Miller and Larsen, 1989) can use jaw prehension, al-
though not very effectively. Jaws are used in aggressive encounters in salamanders (Staub, 1993), and it is evident that biting without tongue protraction is possible. Nonetheless, we expect that salamanders feeding in terrestrial situations typically will display a pattern of feeding that involves tongue prehension (Bramble and Wake, 1985). D . Prey Processing Prey immobilization is significant when the prey are too large to be fully engulfed at the time of capture. Large-bodied salamanders are capable of eating long and slippery (e.g., earthworms in ambystomatids) and even very large prey (e.g., mice in dicamptodontids) relative to body size. In these species, strength of the jaw-closing muscles is important, as well as size and strength of the marginal tooth-bearing bones. The mouth is closed at the end of the strike and if the prey protrudes from the mouth there may be a delay of a few milliseconds to several seconds before processing and/or swallowing proceeds. This process has been most thoroughly studied in ambystomatids and is described later. Dicamptodontids have heads that are the largest of terrestrial salamanders both absolutely and relatively, and they seize (it is not recorded if this is by lingual prehension or jaw prehension) prey such as mice and hold them in the buccal cavity until the prey suffocates before further processing proceeds. Two groups of plethodontids have special prey-processing features. Aneides includes species (e.g., A. lugubris) that have greatly enlarged adductor muscles, jaws and marginal teeth (Wake, 1963). These may well be associated mainly with aggressive behavior (e.g., Staub, 1993), but they also enable these species to eat larger food than co-occurring species of similar size (Lynch, 1985). Desmognathine plethodontids engage in a unique behavior, cranial ventroflexion or head tucking (Dalrymple et al, 1985; Larsen and Beneski, 1988; Schwenk and Wake, 1993), that is enabled by the presence of a number of morphological modifications of the head and neck regions, including a ligamentized tendon extending over either side of the skull from a specialized ridge on the atlas to the lower jaw. This behavior is characteristically performed after lingual capture of the prey and involves a sharp ventroflexion of the head relative to the neck with the prey caught in the jaws. Head tucking may occur as a final component of mouth closing following capture, or without mouth opening, and it represents an extreme form of a static pressure feeding system (Olson, 1961) in salamanders. The result is the penetration of the prey by the teeth and the effective immobilization of the prey. Both Aneides and the
110
David B. Wake and Stephen M. Deban
desmognathines have enlarged quadrapectoralis and an additional muscle (often termed gularis) that contribute to their ability to deliver a strong bite. Prey transport has been studied most extensively in Ambystoma tigrinum (Reilly and Lauder, 1990a, 1991b; Gillis and Lauder, 1994), and only fragmentary information is available for other taxa (Dockx and De Vree, 1986; Thexton et al, 1977). Four phases of intraoral prey transport are recognized in metamorphosed ambystomatids feeding on worms on land (as well as in the water as larvae): preparatory, fast operung, closing, and recovery. Transport per se is accomplished during fast opening and closing. Electromyographic recordings reveal that the preparatory phase involves electrical activity in muscles of the buccal floor (both the longitudinal genioglossus and geniohyoideus, and the transverse intermandibularis and interhyoideus) and then there is a short period of electrical silence before a fast opening of the mouth that is accompanied by simultaneous activity of all muscles studied, even antagonists such as the mandibular depressors and adductors (Reilly and Lauder, 1991b). Reilly and Lauder (1990b) divided the preparatory phase into two parts. In the first part the prey item is pressed against the roof of the mouth by the elevated hyoid apparatus. Gape increases slowly but is never great during phase one. During the second phase, which is shorter in duration, the gape is held constant. Then the fast opening phase occurs, followed quickly by a closing phase, and in each cycle of opening and closing between 4 and 8 mm of prey is transported. Intraoral transport of the prey occurs at the beginning of mouth opening, as the rectus cervicis contract, thus retracting the hyobranchial apparatus and moving the tongue pad (to which the prey is sticking) posteriorly into the pharynx. The muscle that protracts the tongue during prey capture, the subarcualis rectus I, is also active during this time, and Reilly and Lauder (1991a) postulate that it may act antagonistically to the retractors so as to stabilize the interaction of the articulated hyobranchial apparatus and the ceratohyal and to enable the system to function as a whole in aiding the rapid posterior movement of the tongue. There is no inertial component to prey transport. The mouth closes as the prey moves backward. During recovery, hyobranchial protraction carries the tongue forward under the prey, for now the prey is held by the combination of palatal and marginal teeth. The intraoral prey transport system in Ambystoma is very similar during aquatic and terrestrial feeding. However, electromyographic patterns differ between prey capture and transport (reviewed by Lauder and Gillis, 1997). Variation in feeding kinematics is event specific rather than reflecting the environment in
which it occurs (Gillis and Lauder, 1994). Reilly and Lauder (1991) hypothesize that prey transport in terrestrial feeding retains an ancient (extending to fishes) motor pattern associated with suction feeding by larvae. The facts that prey transport behaviors are similar in fishes and salamanders and that they are faster and involve less excursion than prey capture have been used to hypothesize that the prey transport system of terrestrial salamanders may have been directly inherited from the aquatic transport behavior based on suction in aquatic ancestors (Gillis and Lauder, 1994; Lauder and Gillis, 1997). Many salamanders, especially plethodontids but also members of other families, eat small prey that are ingested fully on capture. In these species the marginal dentition does not contact the prey, and transport within the buccal cavity differs from the pattern seen in ambystomatids. Frequently the mouth is not opened again, but evidence shows that the tongue is repositioned and then retracted further, moving the prey into the pharynx. Thexton et al. (1977) recorded two to three bursts of activity in the rectus cervicis muscles following prey capture in the diminutive plethodontid Bolitoglossa occidentalis. This implies that the tongue is being protracted and retracted repeatedly with the mouth closed, and that swallowing follows. E. Modulation of Feeding Behavior Little attention has been given to modulation of feeding behavior under different conditions. Substantial modulation is possible under different circumstances, especially with respect to differences in prey (Deban, 1997; unpublished observations). The species that have been most intensively studied to date are Salamandra salamandra (earlier work reviewed by Roth, 1987; see also Reilly, 1995) and Ensatina eschscholtzii (Deban, 1997); other species studied less intensively include various salamandrids (e.g.. Miller and Larsen, 1990) and plethodontids {Hydromantes italicus, Roth, 1987; Plethodon cinereus, Maglia and Pyles, 1995). The greatest modulation occurs in newts that feed both on land (using tongue protraction) and in the water (using suction feeding) (Miller and Larsen, 1990). Bolitoglossa occidentalis showed little evidence of modulation in early studies (Thexton et al, 1977; Larsen et al, 1989), but our observations show a high degree of modulation of the timing and extent of tongue movements in several species of bolitoglossines. Among ambystomatids, some species (e.g., Ambystoma tigrinum) are apparently highly stereotyped in tongue protraction, whereas others {Ambystoma macrodactylum) show evidence of sonie modulation (Larsen and Guthrie, 1975).
4. Terrestrial Feeding in Salamanders When tested with two distinctly different kinds of prey (waxworms and termites), Ensatina eschscholtzii modulated both the timing and the magnitude of tongue and jaw movements with respect to different prey. When feeding on waxworms, the larger prey, feeding took less time and tongue and jaw movements attained a higher velocity than when feeding on termites (Deban, 1997). In Plethodon cinereus, maximal tongue extension was as great as 17% of snout-vent length when feeding on adult Drosophila (mean 10.4%), but as little as 1% (mean 4.5%) when feeding on larval Drosophila (Maglia and Pyles, 1995). In Ensatina, distance of protraction correlated with distance of the prey from the head, not prey type. In both species, prey capture is completed more quickly on large than on small prey.
IV. DIVERSITY A N D EVOLUTION A. Origins and Outgroups Extant out-group taxa for the order Caudata include the orders Anura and Gymnophiona, both of which differ profoundly with respect to feeding mechanisms from all salamanders. There has been a general consensus that the three orders of living amphibians (Lissamphibia) were derived from a temnospondyl labyrinthodont stock; one widely accepted hypothesis is that lissamphibians might be a sister taxon of the temnospondyl group Dissorophoidea (Bolt, 1977). Another alternative is that lissamphibians are derived from a lepospondylous ancestral stock, perhaps somewhere in the microsaur radiation (Laurin and Reisz, 1996). A further alternative is that lissamphibians do not form a monophyletic group with respect to fossils (Carroll and Holmes 1980). Regardless of phylogenetic considerations, living salamanders have a hyobranchial apparatus that is more generalized in morphology and more similar to known fossils (whether temnospondyls or lepospondyls) than either frogs or caecilians, and we (in accord with Lauder and Reilly, 1994) consider salamanders to be an appropriate model for the first terrestrial feeding system. B. Phylogenetic Diversity Six families of salamanders have species that metamorphose and feed on land: Hynobiidae, Rhyacotritonidae, Dicamptodontidae, Ambystomatidae, Salamandridae, and Plethodontidae. The last three families have been studied in greatest detail with respect to feeding mechanisms and the last two display great diversity.
1.
111
Hynobiidae
There is a general consensus that hynobiids are the most basal of the terrestrial taxa of the Caudata (Larson and Dimmick, 1993), and certain features of their feeding mechanism retain apparent ancestral states. The most evident of these is the retention of two branchial arches, with two epibranchials. All other terrestrial salamanders have a single epibranchial, supported by two ceratobranchials. There are, however, indications that hynobiids do not retain the ancestral structure, because they have a unique feature, the radial loop. The ceratohyals are connected to the basibranchial by means of elongated, attenuate hypohyal derivatives arranged in the form of a flat spring (Fig. 4.2). This arrangement is associated with a modest degree of tongue protraction, which occurs with great speed (Larsen et al, 1989,1996; unpublished data). The tongue pad itself has been described as "sac-like" (Larsen et al., 1996) and lacks differentiated musculature, and the basibranchial is extended forward in the radial loop, pulling it along as the hyobranchium is protracted. The extent of protraction is apparently limited by the structural connection of the the hyoid arch to the basibranchial, a connection that exists to a more limited extent in other terrestrial salamanders that do not protract their tongues so far. Tongue protraction is accompanied by a strong forward lunge of the entire body of the salamander, so the effective strike distance is relatively great. Apart from the radial loop, which represents the first radii of other families, the structural arrangement of the tongue in hynobiids is generalized. The second basibranchials are ossified, whereas the first remain cartilaginous, suggesting that the second basibranchials are the main force-transmitting elements from the protractile musculature to the tongue pad. There is a well-developed subhyoideus muscle in hynobiids, and the subarcualis rectus I muscle is wrapped around both first and second epibranchials. These two muscles, working together, serve to protract the hyobranchial apparatus, but functional morphology of the musculature has not been well studied (Severtsov, 1971). A pair of "cornua" are found at the anterior end of the basibranchial in many hynobiids (Cox and Tanner, 1989); these are paired anteriolaterally directed projections continuous with the anterior end of the basibranchial, and they are drawn out into long processes in Onychodactylus (Fig. 4.2). Cornua resemble the radii of plethodontids, but phylogenetic analysis suggests that the resemblance is homoplastic. The hynobiid condition has been proposed to be the most basal salamander feeding mechanism, based on the combination of its morphology and the fact that it
112
David B. Wake and Stephen M. Deban
displays a generalized four-part feeding cycle (Larsen et al, 1996; Beneski et al, 1995). However, the specialized nature of the radial loop precludes the possibility that it retains the ancestral morphology or function entirely. 2.
Rhyacotritonidae
The Rhyacotritonidae has only recently been recognized as a distinct taxon that is relatively basal phylogenetically within the Caudata. Rhyacotritonids have a generalized hyobranchial morphology (Fig. 4.2). The first radii are present but not elongate as in the hynobiids. There is an otoglossal cartilage, an unpaired, dorsal, medial element that either joins the tips of the two second radii or lies freely between them. The articulated elements of the hyolingual skeleton remain cartilaginous, and there is only a single pair of very short epibranchials. The musculature is also generalized. The family resembles the Hynobiidae and the Salamandridae in having a subhyoideus muscle, which may function in a two-phase tongue protraction system. 3.
Dicamptodontidae
The four species comprising this family are large, robust animals, three of which have a terrestrial stage. The tongue is unique and largely unstudied. It combines elements of the tongues of hynobiids, rhyacotritonids, and ambystomatids. The first radii are drawn into an attenuated loop that recalls the situation in hynobiids but is less extreme. The second radii are joined by a thin cartilaginous plate (Fig. 4.2) that appears to be the homologue of the otoglossal cartilage of rhyacotritonids and ambystomatids and supports the relatively enormous tongue pad that is used to capture prey of a wide range of sizes. There is no subhyoideus muscle. The single pair of epibranchials are short and ossified near their proximal ends. 4.
Ambystomatidae
This is a large family of North American salamanders that has been relatively well studied. Both otoglossal cartilages and two pairs of radii are present. However, in contrast to the situation in hynobiids, rhyacotritontids, and dicamptodontids, the first radii are neither attenuate nor drawn into a radial loop that is continuous with the hyoid arch, but their distal tips are loosely connected to the free anterior ends of the ceratohyals. The second radii may be forked and elongate in the subgenus Linguaelapsis (Fig. 4.2). Ambystomatids lack a subhyoideus muscle. They have a muscle termed the genioglossus lateralis that lies in the same general area as the subhyoideus but that has an entirely different origin and innervation
(Piatt, 1940). This muscle may be used to move the ceratohyals medially, thus orienting the tongue. 5.
Salantandridae
Basal salamandrids have two pairs of radii. The first radii are typically rather short and tapered, with free distal ends. The first radii are apparently lacking in Salamandrina and Chioglossa, in which the second radii have moved to the distal tip of the basibranchial and rotate around it as the tongue pad is flipped. The radii are moderately long and robust in Salamandrina, but very long and attenuate in Chioglossa (Fig. 4.2). In both genera the basibranchial is stout and is the only part of the articulated hyobranchial apparatus that is mineralized. In Salamandrina it is T shaped in cross section. In other salamandrid genera the second radii are located behind the first radii, and an interradial cartilage (possibly a homologue of the otoglossal of other taxa) extends between them. Tongue pad musculature is relatively complicated in the Salamandridae, and several muscles occur that are not found in other families (e.g., radioglossus) whereas others are larger than in other families (e.g., basiradialis). The most complicated tongue pad is found in Salamandrina, in which tongue pad flipping is thought to be accomplished by rapid rotation of the radii by the basiradialis muscles (Ozeti and Wake, 1969). Salamandrid genera were divided into two functional groups based on tongue structure and use of the tongue in feeding by Ozeti and Wake (1969). Most genera have a "water tongue" and they have been known as "Wassermolchen" in the German literature (Wolterstorff and Herre, 1935). In these genera the hyolingual apparatus is used for suction feeding in the water and for tongue prehension on land. Characteristically the skeleton of the hyolingual apparatus is relatively heavy and well ossified. At least one genus, Pachytriton, is permanently aquatic, has no tongue prehension ability, and has a very reduced tongue pad. Paradoxically, Pachytriton has a terrestrial eft stage (common in other newts) in which tongue prehension likely occurs (Thiesmeier and Hornberg, 1997), prior to the return as an adult to a permanently aquatic existence in which tongue prehension is impossible. The most terrestrial salamandrids, Salamandra, Mertensiella, Chioglossa, and Salamandrina, have "land tongues." They either do not enter water or feed in water using terrestrial behaviors and they have specialized tongues used for prehension of prey. Typically the hyolingual apparatus is either entirely cartilaginous or only the basibranchial is ossified. A significant part of tongue protraction in Chioglossa and Salamandrina is accomplished by flipping and extension of the large tongue pad (Miller and Larsen, 1990).
4. Terrestrial Feeding in Salamanders Salamandrids have a subhyoideus muscle (see Fig. 3.5 in Chapter 3) that is used to protract the ceratohyals and the tongue in general (Findeis and Bemis, 1990). Some salamandrids have some lateral muscle fibers arising near the genioglossus (e.g., Ozeti and Wake, 1969), and these may be homologues either of the genioglossus lateralis of ambystomatids or possibly the geniohyoideus lateralis of plethodontids. 6.
Plethodontidae
Otoglossals are absent in the Plethodontidae, and some taxa have a medial unpaired "lingual cartilage'' (Fig. 4.4). Rose (1996) has shown that the lingual cartilage may have some connection with the hypohyals of experimental animals treated with thyroxin. Wake (1966) argued that the ,cartilage is derived from the anterior extension of the basibranchial that lies in front of the attachment of the radii. Out-group taxa that have an anterior extension also have either a pair of first radii attached to it or bear a pair of cornua. It is possible that the anterior extension represents the fusion of the lingual cartilage to the basibranchial; perhaps the lingual cartilage should be considered a homologue of a long missing basihyal. The first radii of out-group taxa that have two pairs of radii lack muscular attachments, and because the single radii of plethodontids have muscular attachments they are best considered to be homologues of the second radii of other taxa. An alternative interpretation is that the radii of plethodontids may be homologues of the cornua of hynobiids, because in many plethodontids the radii are homocontinuous with the cartilage of the basibranchial, although in others they are articulated with the basibranchial, as are the second radii in members of other families. If, as argued earlier, the first radii represent hypohyals of out-group taxa, the lingual cartilage represents the basihyal, the distal (with respect to distance from the midline) element found in rhyacotritonids represents the epihyal, and portions of the middle ear complex are derived from the arch as well, all components of the entire hyoid arch are present in the Caudata, but not in any single taxon. There is no subhyoideus muscle in plethodontids. A well-developed geniohyoideus lateralis may play an important role in controlling lateral to medial movements of the ceratohyal and thus direct the firing of the tongue, but this has not been demonstrated behaviorally (Lombard and Wake, 1977). Feeding mechanisms in plethodontids have been studied extensively, and summary discussions of feeding in a phylogenetic context accompanied by evolutionary scenarios are presented by Roth and Wake (1985b), Lombard and Wake (1986), and Wake and Larson (1987). Phylogenetic analyses (Jackman et ah, 1997)
113
support the hypothesis (Wake, 1966; Lombard and Wake, 1986) that the freely projectile tongues of the supergenera Hydrotnantes and Bolitoglossa are derived independently. Both are derived independently of the even more phylogenetically remote members of the tribe Hemidactyliini, which also have evolved freely projectile tongues. Thus, within the Plethodontidae there are three lineages that have evolved highly specialized tongue projection mechanisms. The hemidactyliine tongue is folded according to one of two hypothetical possibilities (Lombard and Wake, 1976, 1977), called option 1. This folding pattern involves holding the first ceratobranchial and epibranchial coplanar during folding. It results in a relatively bulky projectile that is postulated to have more limited projectability than is involved in option 2 (Wake, 1982). This option has been hypothesized as a necessary consequence of the retention of an aquatic larval stage and associated larval suction feeding in hemidactyliines (Wake, 1982). In contrast, bolitoglossines, which all have direct terrestrial development and no aquatic larval stage, all use option 2 folding, in which the second ceratobranchial and the epibranchial are held coplanar. This option results in a slenderer projectile that is less limited in length than in hemidactyliines. There are two modifications of the option, that of Hydrotnantes, which has apparently optimized for distance by having evolved extremely long epibranchials, and that of the supergenus Bolitoglossa, which also has elongated epibranchials but substantially shorter than in Hydromantes, and is apparently optimized for speed (Larsen et ah, 1989). C. Feeding Biology and Evolution 1. Ecology and Selective Regime The evolution of terrestrial feeding in salamanders has proceeded in a great diversity of habitats and microhabitats, and in the absence of information about close sister taxa of Caudata it is impossible to reconstruct the habitat occupied by the first terrestrial feeders. However, it seems likely that early salamanders had a biphasic life cycle, involving feeding in water and on land, and that courtship and mating probably occurred in the water. Accordingly, feeding most likely entailed suction feeding both as larvae and as adults, and tongue protraction, apprehension, and capture on land. The most specialized terrestrial feeding mechanisms tend to be associated with species that do not spend much time in the water, or that are entirely terrestrial. The exception to this generalization is the family Hynobiidae, in which species with aquatic larvae and semiaquatic habits as adults have a hyolingual system that is fast and biomechanically specialized (e.g..
114
D a v i d B. W a k e a n d S t e p h e n M. D e b a n
Larsen et al, 1996). In the hemidactyhine plethodontids and in salamandrids such as Chioglossa and Salamandrina, tongues are also highly specialized for speed and distance of protraction, and these forms all have aquatic larvae. However, all are lungless or have reduced lungs, and this appears to be a necessary, but not sufficient, precondition for the high specialization of the tongue (Roth and Wake, 1985b). Rhyacotritonids, whose tongue and feeding are relatively unstudied but which have a generalized morphology, also have reduced lungs. The greatest specialization in tongue speed and protrusion is foimd in the bolitoglossine plethodontids, all of which lack an aquatic larval stage. These species have diverged far beyond the ancestral home of not only salamanders as a group but of plethodontids, and occur in a great diversity of terrestrial habitats and microhabitats, including waterless caves, cavities in trees, epiphytes, scrublands, and other settings in which foraging is restricted by environmental considerations. Whatever the environment in which tongue specialization evolved, it is a highly effective feeding strategy. 2.
Homoplasy
Homoplasy is common in the evolution of terrestrial feeding mechanisms of urodeles and has been discussed extensively (e.g.. Wake, 1966, 1982; Lombard and Wake, 1986). What has attracted the greatest attention is tongue protrusion. While some degree of tongue protrusion occurs in all salamanders that metamorphose and have a terrestrial feeding stage, longdistance tongue protrusion has evolved independently along very different biomechanical pathways within at least three families (Hynobiidae, Salamandridae, and Plethodontidae). In hynobiids, tongue protrusion is associated with modest increases in length of the two pairs of epibranchials and with the flat spring arrangement of the hyoid loop, which is attached to the articulated lingual skeleton in some way. However, the basibranchial remains short. Paired, relatively elongate cornua appear in Onychodactylus (Fig. 4.2), and while these may represent retained ancestral elements homologous to the second radii of other taxa, phylogenetic analysis suggests homoplasy. Tongue protrusion is never as great as in the other two families, but the tongue is fast and maneuverable. Larsen et al (1996) argue that the radial loops and the attachment of the ceratohyals to the suspensorium constitute a functional constraint that mechanically limits the extent of tongue protrusion. Wntiether there has been homoplasy within the Hynobiidae awaits a modem phylogenetic analysis of the family and further morphological studies.
V. OPPORTUNITIES FOR FUTURE RESEARCH Research effort has been unevenly distributed with respect to salamander clades. Biomechanical models of tongue protraction have been produced and tested for plethodontids and ambystomatids, but are largely lacking for other taxa. Detailed morphological studies date to the early part of this century and do not include several families, including Dicamptodontidae, Hynobiidae, and Rhyacotritonidae. However, comparative anatomical studies of some of the larger families, including Ambystomatidae, Plethodontidae, and Salamandridae, are relatively complete. What is needed is a broad and integrated comparative anatomical analysis of the musculoskeletal system of all the families, with special attention given to the establishment of homologies. Quantitative studies of kinematics are limited to only a few species and in general have been conducted in artificially controlled conditions and with limited, sometimes unnatural, prey. The modulation of behavior has been investigated in only a handful of species, even though most species feed on a diversity of prey and under a variety of conditions. Diet and foraging are in need of further study in most taxa. Two families, the Dicamptodontidae and the Rhyacotritonidae, have been especially understudied with respect to kinematics. References Beneski, J. T., Jr., and J. H. Larsen, Jr. (1989) Interspecific, ontogenetic, and life history variation in the tooth morphology of mole salamanders (Amphibia, Urodela, and Ambystomatidae). J. Morph. 199(1): 53-70. Beneski, J. T., Jr., J. H. Larsen, Jr. and B. T. Miller (1989) Ontogenetic alterations in the gross tooth morphology of Dicamptodon and Rhyacotriton (Amphibia, Urodela, and Dicamptodontidae). J. Morph. 199(2): 165-174. Beneski, J. T, Jr., and J. H. Larsen, Jr. (1995) Variation in the feeding kinematics of mole salamanders (Ambystomatidae: Ambystoma). Can. J. Zool. 73:353-366. Bishop, S. C. (1941) The salamanders of New York. New York State Mus. Bull. No. 324:1-365. Bolt, J.R. (1977) Dissorophoid relationships and ontogeny, and the origin of the Lissamphibia. J. Paleo. 51:235-239. Bramble, D. M., and D. B. Wake (1985) Feeding mechanisms in lower tetrapods. Pp. 230-261. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. E Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Bury, R. B. (1972) Small mammals and other prey in the diet of the Pacific Giant Salamander (Dicamptodon ensatus). Am. Midi. Nat. 87:524-525. Carroll, R. L. and R. Holmes (1980) The skull and jaw musculature as guides to the ancestry of salamanders. Zool. J. Linn. Soc. 68:1-40. Dalrymple, G. H., J. E. Juterbock, and A. L. La Valley (1985) Function
4. Terrestrial F e e d i n g in S a l a m a n d e r s of the atlanto-mandibular ligaments of desmognathine salamanders. Copeia 1985:254-257. Dawley, E. M., and A. H. Bass (1989) Chemical access to the vomeronasal organs of a plethodontid salamander. J. Morph. 200: 163-174. Deban, S. M. (1997) Modulation of prey-capture behavior in the plethodontid salamander Ensatina eschscholtzii. J. Exp. Biol. 200: 1951-1964. Deban, S. M., D. B. Wake, and G. Roth 1997. Salamander with a ballistic tongue. Nature 389:27-28. Dockx, P., and F. De Vree (1986) Prey capture and intra-oral food transport in terrestrial salamanders. Stud. Herpetol. 521-524. Driiner, L. (1901) Studien zur Anatomie des Zungebein-, Kiemenbogen- und Kehlkopfmuskeln der Urodelen. I. Theil. Zool. Jahrb. Abteil. Anat. 15:435-622. Findeis, E. K., and W. E. Bemis (1990) Functional morphology of tongue projection in Taricha torosa (Urodela: Salamandridae). Zool. J. Linn. Soc. 99:129-157. Flower, S. S. (1927) Loss of memory accompanying metamorphosis in amphibians. Proc. Zool. Soc. Lond. 1:155-156. Francis, E. B. T. (1934) The Anatomy of the Salamander. Clarendon Press, Oxford. Good, D. A. (1988) Hybridization and cryptic species in Dicamptodon (Caudata: Dicamptodontidae). Evolution 43:728-744. Good, D. A., and D. B. Wake (1992) Geographic variation and speciation in the torrent salamanders of the genus Rhyacotriton (Caudata, Rhyacotritonidae). Univ. Calif. Publ. Zool. 126:1-91. Hairston, N. G., Sr. (1987) Community Ecology and Salamander Guilds. Cambridge Univ. Press, Cambridge. Hiiemae, K. M. (1978) Mammalian mastication: a review of the activity of the jaw muscles and the movements they produce in chewing. Pp. 359-398. In: Develoment, Function and Evolution of Teeth. P. M. Butler and K. A. Joysey (eds.). Academic Press, New York. Himstedt, W. (1982) Prey selection in salamanders. Pp. 47-66. In: Analysis of Visual Behavior. D. Ingle, M. A. Goodale, and R. J. W. Mansfield (eds.). MIT Press, Cambridge, MA. Jaeger, R. G., and D. E. Barnard (1981) Foraging tactics of a terrestrial salamander: choice of diet in structurally simple environments. Am. Nat. 117:639-664. Jaeger, R. G., D. E. Barnard, and R. G. Joseph (1982) Foraging tactics of a terrestrial salamander: assessing prey density. Am. Nat. 119: 885-890. Jaeger, R. G., and A. M. Rubin (1982) Foraging tactics of a terrestrial salamander: judging prey profitability. J. Anim. Ecol. 51:167-176. Keen, W. H. (1979) Feeding and activity patterns in the salamander Desmognathus ochrophaeus (Amphibia, Urodela, Plethodontidae). J. Herp. 13:461-467. Kuzmin, S. L. (1991) Feeding of the salamander Ranodon sibiricus. Alytes 9:135-143. Larsen, J. H., Jr., and J. T. Beneski, Jr. (1988) Quantitative analysis of feeding kinematics in dusky salamanders (Desmognathus). Can. J. Zool. 66:1309-1317. Larsen, J. H., Jr., J. T. Beneski, Jr., and B. T Miller (1996) Structure and function of the hyolingual system in Hynobius and its bearing on the evolution of prey capture in terrestrial salamanders. J. Morph. 227:235-248. Larsen, J. H., Jr., J. T Beneski, Jr., and D. B. Wake (1989) Hyolingual feeding systems of the plethodontidae: comparative kinematics of prey capture by salamanders with free and attached tongues. J. Exp. Zool. 252:25-33. Larsen, J. H., Jr., and D. J. Guthrie (1975) The feeding system of terrestrial tiger salamanders (Amhystoma tigrinum melanostictum Baird). J. Morph. 147:137-154.
115
Larson, A. (1991) A molecular perspective on the evolutionary relationships of the salamander families. Evol. Biol. 25:211-277. Larson, A., and A. C. Wilson (1989) Patterns of ribosomal RNA evolution in salamanders. Mol. Biol. Evol. 6:131-154. Larson, A., and W. W. Dimmick (1993) Phylogenetic relationships of the salamander families: an analysis of congruence among morphological and molecular characters. Herp. Monogr. 7:77-93. Lauder, G. V., and S. M. Reilly (1994) Amphibian feeding behavior: comparative biomechanics and evolution. Adv. Comp. Environ. Physiol. 18:163-195. Lauder, G. V., and H. B. Shaffer (1985) Functional morphology of the feeding mechanism in aquatic ambystomatid salamanders. J. Morph. 185:297-326. Lauder, G. V., and H. B. Shaffer (1988) Ontogeny of functional design in tiger salamanders {Amhystoma tigrinum): are motor patterns conserved during major morphological transformations? J. Morph. 197:249-268. Laurin, M., and R. R. Reisz (1997) A new perspective on tetrapod phylogeny. Pp. 9-59. In: Amniote Origins: Completing the Transition to Land. S. S. Sumida and K. L. M. Martin (eds.). Academic Press, San Diego. Lombard, R. E., and D. B. Wake (1976) Tongue evolution in the lungless salamanders, family Plethodontidae. I. Introduction, theory and a general model of dynamics. J. Morph. 148:265-286. Lombard, R. E., and D. B. Wake (1977) Tongue evolution in the lungless salamanders, family Plethodontidae. II. Function and evolutionary diversity. J. Morph. 153:39-80. Lombard, R. E., and D. B. Wake (1986) Tongue evolution in the lungless salamanders, family Plethodontidae. IV. Phylogeny of plethodontid salamanders and the evolution of feeding dynamics. Syst.Zool.35(4):532-551. Luthardt-Laimer, G. (1983) Distance estimation in binocular and monocular salamanders. Zeit. Tierpsychol. 63:233-240. Luthardt-Laimer, G., and G. Roth (1983) Reduction of visual ir\hibition to stationary prey by early experience in Salamandra salamandra (L.). Z. Tierpsychol. 63:294-302. Lynch, J. F. (1985) The feeding ecology of Aneides flavipunctatus and sympatric plethodontid salamanders in northwestern California. J. Herp. 19:328-352. Maglia, A. M., and R. A. Pyles (1995) Modulation of prey-capture behavior in Plethodon cinereus (Green) (Amphibia: Caudata). J. Exp. Zool. 272:167-183. Maiorana, V. (1978) Behavior of an unobservable species: diet selection by salamander. Copeia 1978(4): 664-672. Miller, B. T, and J. H. Larsen, Jr. (1990) Comparative kinematics of terrestrial prey capture in salamanders and newts (Amphibia: Urodela: Salamandridae). J. Exp. Zool. 256:135-153. Olson, E. C. (1961) Jaw mechanisms: rhipidistians, amphibians, reptiles. Am. Zool. 1:205-215. Ozeti, N., and D. B. Wake (1969) The morphology and evolution of the tongue and associated structures in salamanders and newts (family Salamandridae). Copeia 1969:91-123. Reilly, S. M. (1995) The ontogeny of aquatic feeding behavior in Salamandra salamandra: stereotypy and isometry in feeding kinematics. J. Exp. Biol. 198:701-708. Reilly, S. M., and G. V. Lauder (1988) Atavisms and the homology of hyobranchial elements in lower vertebrates. J. Morph. 195:237245. Reilly, S. M., and G. V. Lauder (1989) Physiological bases of feeding behavior in salamanders: do motor patterns vary with prey type? J. Exp. Biol. 141:343-358. Reilly, S. M., and G. V. Lauder (1990a) The evolution of tetrapod feeding behavior: kinematic homologies in prey transport. Evolution 44:1542-1557.
116
D a v i d B. W a k e a n d S t e p h e n M. D e b a n
Reilly, S. M., and G. V. Lauder (1990b) The strike of the tiger salamander: quantitative electromyography and muscle function during prey capture. J. Comp. Physiol. A 167:827-839. Reilly, S. M., and G. V. Lauder (1991a) Experimental morphology of the feeding mechanism in salamanders. J. Morph. 210:33-44. Reilly, S. M., and G. V. Lauder (1991b) Prey transport in the tiger salamander: quantitative electromyography and muscle function in tetrapods. J. Exp. Zool. 260:1-17. Rose, C. S. (1996) An endocrine-based model for developmental and morphogenetic diversification in metamorphic and paedomorphic urodeles. J. Zool. Lond. 239:253-284. Roth, G. (1976) Experimental analysis of the prey catching behavior of Hydromantes italicus Dunn (Amphibia, Plethodontidae). J. Comp. Physiol. 109:47-58. Roth, G. (1987) Visual Behavior in Salamanders. Springer-Verlag, Heidelberg. Roth, G., K. Nishikawa, U. Dicke, and D. B. Wake (1988) Topography and cytoarchitecture of the motor nuclei in the brainstem of salamanders. J. Comp. Neurol. 278:181-194. Roth, G., K. C. Nishikawa, D. B. Wake, U. Dicke, and T. Matsushima (1990) Mechanics and neuromorphology of feeding in amphibians. Neth. J. Zool. 40:115-135. Roth, G., and D. B. Wake (1985a) The structure of the brainstem and cervical spinal cord in relation to feeding of lungless salamanders, family Plethodontidae. J. Comp. Neurol. 241:99-110. Roth, G., and D. B. Wake (1985b) Trends in the functional morphology and sensorimotor control of feeding behavior in salamanders: an example of the role of internal dynamics in evolution. Acta Biotheor. 34:175-192. Roth, G., and D. B. Wake (1989) Conservatism and innovation in the evolution of feeding in vertebrates. Pp. 7-21. In: Complex Organismal Functions: Integration and Evolution in Vertebrates. D. B. Wake and G. Roth (eds.). Wiley, Chichester. Schwenk, K., and D. B. Wake (1993) Prey processing in Leurognathus marmoratus and the evolution of form and function in desmognathine salamanders (Plethodontidae). Biol. J. Linn. Soc. 49:141162. Severtsov, A. S. (1971) The mechanism of food capture in tailed amphibians. Doklady Akademii Nauk SSSR 197:728-731. Sites, J. W. (1978) The foraging strategy of the dusky salamander, Desmognathus fuscus (Amphibia, Urodela, Plethodontidae): an empirical approach to predatiion theory. J. Herp. 12:373-383. Shaffer, H. B. and G. V. Lauder (1988) The ontogeny of functional design: metamorphosis of feeding behaviour in the tiger salamander (Amhystoma tigrinum). J. Zool. London 216:437-454. Shaffer, H. B., J. M. Clark, and R Kraus (1991) When molecules and morphology clash: a phylogenetic analysis of the North American
ambystomatid (Caudata: Ambystomatidae) salamanders. Syst. Zool. 40:284-303. Thexton, A. J., D. B. Wake, and M. H. Wake (1977) Tongue function in the salamander Bolitoglossa occidentalis. Arch. Oral Biol. 22: 361-366. Thiesmeier, B., and C. Romberg (1997) Paarung, Fortpflanzung und Larvalentwicklung von Pachytriton sp. {Pachytriton A) nebst Bemerkungen zur Taxonomie der Gattung. Salamandra 33:97-110. Titus, T. A., and A. Larson (1995) A molecular phylogenetic perspective on the evolutionary radiation of the salamander family Salamandridae. Syst. Biol. 44:125-151. Trueb, L., and R. Cloutier (1991) A phylogenetic investigation of the inter- and intrarelationships of the Lissamphibia (Amphibia: Temnospondyli). Pp. 223-313. In: Origins of the Higher Groups of Tetrapods: Controversy and Consensus. H.-P. Schultze and L. Trueb (eds.). Comstock Publ. Assoc, Ithaca. Wake, D. B. (1963) Comparative osteology of the plethodontid salamander genus Aneides. J. Morph. 113:77-118. Wake, D. B. (1966) Comparative osteology and evolution of the lungless salamanders, family Plethodontidae. Mem. Southern Calif. Acad. Sci. 4:1-111. Wake, D. B. (1982) Functional and developmental constraints and opportunities in the evolution of feeding systems in urodeles. Pp. 51-66. In: Environmental Adaptation and Evolution. D. Mossakowski and G. Roth (eds.). Gustav Fischer, Stuttgart. Wake, D. B. (1991) Homoplasy: the result of natural selection, or evidence of design limitations? Am. Nat. 138(3): 543-567. Wake, D.B. (1993) Brainstem organization and branchiomeric nerves. Acta Anat. 148:124-131. Wake, D. B., and A. Larson (1987) Multidimensional analysis of an evolving lineage. Science 238:42-48. Wake, D. B., and J. Hanken (1996) Direct development in the lungless salamanders: what are the consequences for developmental biology, evolution and phylogenesis? Int. J. Dev. Biol. 40:859-869. Wake, D. B., K. C. Nishikawa, U. Dicke, and G. Roth (1988) Organization of the motor nuclei in the cervical spinal cord of salamanders. J. Comp. Neurol. 278:195-208. Wiggers, W., and G. Roth (1991) Anatomy, neurophysiology and functional aspects of the nucleus isthmi in salamanders of the family Plethodontidae. J. Comp. Physiol. A 169:165-176. Wolterstorff, W., and W. Herre (1935) Die Gattungen der Wassermolche der Familie Salamandridae. Arch. Naturgesch. N.F. 4: 217-229. Zhao, E., and Q. Hu (1988) Studies on Chinese tailed amphibians. English translation. Pp. 1-44. In: Studies on Chinese Salamanders. E. Shao, Q. Hu, Y. Jiang, and Y. Yang (eds.). Society for the Study of Amphibians and Reptiles, Miami, OH.
C H A P T E R
5 Feeding in Frogs KIISA C NISHIKAWA Department of Biological Sciences Northern Arizona University Flagstaff, Arizona 86011
animals. One aspect of this goal is to compare the function of the feeding apparatus across the major clades of tetrapod vertebrates. For each clade, the function of the feeding apparatus can be addressed by asking the following questions: (1) Which muscles are involved in producing feeding movements and v^hat is the specific contribution of each? (2) What are the spatial and temporal patterns of muscle activation and how do they relate to movement? and (3) What are the neural mechanisms that are responsible for producing observed patterns of muscle activity? Since the early 1990s, my research has focused on understanding the evolutionary relationships among morphology, biomechanics, and neural control of movement using prey capture of anurans as a model system. This chapter provides a summary of this work and a historical perspective on hypotheses concerning the mechanisms of anuran prey capture. The chapter begins by briefly reviewing the phylogeny and natural history of anurans (Section I). Subsequent sections describe the morphology (Section II) and function (Section III) of the anuran feeding apparatus, the neural control of prey capture movements (Section IV), and the evolution of mechanisms of tongue protraction and neural control of prey capture (Section V). The chapter ends with a summary of conclusions and a description of current and future directions. It is hoped that by describing the pitfalls that have been encountered in attempting to understand the mechanisms of prey capture in frogs, this chapter will help future workers avoid similar problems in the future. It is also hoped that our attempts to understand the biomechanics and neural control of prey capture in frogs will stimulate functional morphologists to undertake similar studies
I. INTRODUCTION A. Phylogeny of Anurans B. Natural History of Anurans II MORPHOLOGY OF THE FEEDING APPARATUS A. Craniun\ and Jaw Muscles B. Mandible and Buccal Floor Muscles C. Hyoid and Associated Muscles D.Tongue and Associated Muscles III. FUNCTION OF THE FEEDING APPARATUS A. Methods for Studying the Function of the Feeding Apparatus B. Hypotheses for the Mechanism of Tongue Protraction C. Functional Diversification among Anuran Taxa IV. NEURAL CONTROL OF PREY CAPTURE A. Visual Analysis of Prey Features B. Role of Tongue Afferents C. Interactions between Tongue Afferents and Visual Input V. EVOLUTION OF THE FEEDING APPARATUS A. Evolutionary Transitions in Mechanisms of Tongue Protraction B. Morphological Correlates of Tongue Protraction Mechanisms C. Evolution of Tongue Afferents D. Evolutionary Transitions in Mechanisms of Neural Control VI. CONCLUSIONS VII. CURRENT AND FUTURE DIRECTIONS References
I. INTRODUCTION A major goal of functional morphology is to understand the evolution of functional differences among
FEEDING (K.SchwenKed.)
117
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
118
Kiisa C. N i s h i k a w a
in other groups of animals so that the generality of our results can be tested in other groups.
Ascaphidae Leiopelmatidae Bombinatoridae
A. Phylogeny of Anurans On superficial examination, frogs appear to be a morphologically homogeneous group. They are easily distinguished from other amphibians by their welldeveloped hind legs and the absence of a tail in adults. Frogs are, however, quite diverse morphologically, particularly in terms of their feeding apparatus, which has played an important role in anuran classification schemes for more than a century (Giinther, 1859; Griffiths, 1963). Recent workers recognize 21-27 families of frogs (Duellman and Trueb, 1986; Ford and Cannatella, 1993). A number of phylogenetic hypotheses have been published for the anuran families, based on morphological (e.g.. Ford and Cannatella, 1993) or genetic characters (e.g., Hillis et al, 1993; Hay et al, 1995). In these studies, the relationships of some anuran families are agreed upon generally, but relationships of the majority of families remain to be resolved definitively (Hillis efflZ., 1993). This chapter uses a recent phylogenetic hypothesis of anuran familial relationships (Fig. 5.1) that was developed using morphological characters by Ford and Cannatella (1993). In this hypothesis, the "archaeobatrachian" families (Ascaphidae, Leiopelmatidae, Bombinatoridae, and Discoglossidae) represent the most basal lineages of frogs. These families share many ancestral morphological characteristics, but are not closely related to each other and do not constitute a monophyletic group (Ford and Cannatella, 1993). On the basis of genetic characters, however. Hay et ah (1995) concluded that the Archaeobatrachia may, in fact, be monophyletic. In Ford and Cannatella's (1993) phylogeny, the Mesobatrachia (Fig. 5.1) is a monophyletic group that includes two lineages: the pelobatoids (families Pelobatidae, Pelodytidae, and Megophryidae) and pipoids (families Pipidae and Rhinophrynidae). The remaining families are placed in the clade Neobatrachia. Within the Neobatrachia, the families Hylidae, Centrolenidae, and Pseudidae appear to form a clade, as do the ranoids (families Ranidae, Arthroleptidae, Hyperoliidae, Rhacophoridae, Dendrobatidae, Hemisotidae, and Microhylidae), although the relationships among ranoid families are unresolved. The placement of the dendrobatids remains controversial and some authors believe that they are more closely related to bufonids than to ranoids (Hass, 1995; Hay et ah, 1995; Ruvinsky and Maxson, 1996). The relationships among the other neobatrachian families are also unresolved (Fig. 5.1). Most of the currently recognized anuran families
Pelobatidae Pelodytidae Megophryidae Rhinophrynidae Pipidae
_
Limnodynastinae Myobatrachlnae Sooglossidae Heleophrynidae "Leptodactylidae" Brachycephalidae Bufonidae Rhinodermatidae Centrolenidae Hylidae Pseudidae "Ranidae" Arthroleptidae Hyperoliidae Rhacophoridae Dendrobatidae Hemisotidae Microhylidae
F I G U R E 5.1. A phylogeny of the living anurans modified from Ford and Cannatella (1993). The archaeobatrachians are a grade group that share many ancestral characteristics but are not closely related to each other. The families Leptodactylidae and Ranidae are indicated in quotes because there is no evidence for their monophyly.
appear to be monophyletic, with the possible exception of Myobatrachidae, Leptodactylidae, and Ranidae (Ford and Cannatella, 1993). The family Myobatrachidae may not be monophyletic because its subfamily Myobatrachlnae shares several characteristics with the family Sooglossidae that are absent in the subfamily Limnodynastinae. For the families Ranidae and Leptodactylidae, monophyly has been rejected, but the relationships of the subfamilies to other taxa remain to be resolved (Ford and Cannatella, 1993). At present, the unresolved relationships among neobatrachian frogs are a serious impediment to understanding anuran evolution. B. Natural History of Anurans Nearly all frogs are carnivorous after metamorphosis, and invertebrates are the most common prey, although there are reports of frogs that eat fruits (Silva et ah, 1989) and leaves (Das and Coe, 1994; Das, 1995, 1996). Most anurans are also feeding generalists, eating any prey that will fit into their mouths in proportions that match the relative abundance of prey in
119
5. Feeding in Frogs nature (Toft, 1981). Dietary generalists tend to be widemouthed, sit-and-wait predators. A few species of anurans are known to specialize on a narrower range of prey types (Toft, 1981). Some forest floor species, including the genus Bufo (Smith and Bragg, 1949; Inger and Marx, 1961; Berry and Bullock, 1962; Clarke, 1974; Zug and Zug, 1979; Toft, 1981), and many burrowing anurans, such as Hemisus (Emerson, 1976a), specialize on small prey such as ants, coUembolans, termites, or mites (Simon and Toft, 1991; Toft, 1995; Caldwell, 1996), consuming them in greater proportions than their relative abundance in nature. Small prey specialists tend to lack teeth, have relatively narrow mouths, and search actively for prey (Toft, 1981). They also tend to have higher aerobic capacities and lower anaerobic capacities than dietary generalists (Taigen and Pough, 1983). Several anurans are known to consume rather peculiar types of prey, such as snails (Drewes and Roth, 1981), crabs (Premo and Atmowidjojo, 1987), fish (Knoepffler, 1976), or even other frogs (Emerson, 1985). However, these species also take a wide range of other prey types and so are not strict dietary specialists. There have been relatively few attempts to correlate the morphology of the feeding apparatus with natural diets in anurans. Emerson (1985) and colleagues (Emerson and Voris, 1992, Emerson and Bramble, 1993; Emerson et ah, 1994) have studied the relationships among skull morphology, feeding behavior, and diet among diverse species. She found that frogs that eat relatively large prey have relatively longer jaws and wider skulls than those that eat small prey. In addition, the feeding cycle is longer and more asymmetrical (i.e., mouth closing takes longer than mouth opening) in species that eat large prey than in species that eat small prey. Gray (1997) compared the diets of sympatric hylid frogs with short vs long tongues. She found that there was overlap in the types of prey taken by the two species, but long-tongued species consumed a greater proportion of rapidly moving prey than shorttongued species. The ecological significance of differences among frogs in the morphology of the feeding apparatus is a topic that deserves further study.
II. MORPHOLOGY OF THE FEEDING APPARATUS Feeding mechanisms of amphibians (frogs, salamanders, and caecilians) are extremely diverse. The tongues of caecilians are small and cannot be protruded from the mouth so jaw prehension is used to capture prey and some species possess a highly specialized jawclosing mechanism (Bemis et ah, 1983; Nussbaum,
1983). In contrast to caecilians, terrestrial salamanders and frogs use lingual protraction to catch prey. They depend heavily on lingual adhesion for prey capture. In salamanders, the articulated hyobranchial skeleton forms an internal support for the tongue, which leaves the mouth with the tongue during protraction (Lombard and Wake, 1976; Thexton et ah, 1977). Salamanders vary widely in their tongue morphology, especially with regard to projection distance, muscle attachment, and hyobranchial folding (Ozeti and Wake, 1969; Regal, 1966; Wake, 1982; Lombard and Wake, 1977, 1987). In contrast to salamanders, the anuran tongue consists only of muscles and connective tissue, with no intrinsic cartilaginous or bony skeleton. The hyobranchial apparatus is an unarticulated plate formed from the ontogenetic fusion of the cartilaginous larval elements. It cannot be folded and does not leave the mouth with the tongue during feeding (Gans, 1961, 1962; Gans and Gorniak, 1982a,b; Roth et ah, 1990). Frogs also vary in their tongue morphology, especially with regard to the anatomy of the protractor muscle (Regal and Gans, 1976; Horton, 1982) and projection distance (Deban and Nishikawa, 1992; Nishikawa et ah, 1992). Variation in the morphology of the feeding apparatus has been studied extensively in frogs (MagimelPelonnier, 1924; Regal and Gans, 1976; Emerson, 1976b, 1985; Horton, 1982; Cannatella, 1985). Until recently, there have been few attempts to map morphological characteristics of the feeding apparatus onto a cladogram of anuran taxa, particularly in a functional context. In general, although the size and shape of elements of the feeding apparatus may differ dramatically among species, there is a great deal of overall similarity in the bones and muscles of the feeding apparatus among frogs (Figs. 5.2A-5.2C). The functional significance of anatomical variation remains unknown in most cases. The feeding apparatus of frogs consists of the cranium, mandibles, hyoid, and tongue, each of which is associated with a series of muscles. The morphology of each of these elements is described next, along with brief descriptions of hypothesized functions of the musculoskeletal elements and a brief review of diversity among taxa, where known. A more detailed examination of the function of the feeding apparatus is given in Section III. A. Cranium and Jaw Muscles The morphology of the anuran cranium was studied extensively by Trueb (1973), and a fairly detailed summary can be found in Duellman and Trueb (1986). Cranial morphology is highly variable among species.
120
Kiisa C. N i s h i k a w a Hyla cinerea
Hemisus marmoratum
Bufo marinus M. submentalis
M. intermandibularis
geniohyoideus
M. intermandibularis
M. geniohyoideus
hyoid M. interhyoideus
M. intermandibularis
l\^. submentalis M. geniohyoideus
^ interhyoideus
M. intertiyoideus
M. omohyoideus
M. hyoglossus M. stemohyoideus
M. genioglossus
M. genioglossus
M. geniohyoideus medialis M. geniohyoideus lateralis
M. hyoglossus
M. hyoglossus M. hyoglossus
M. omohyoideusM. stemohyoideus
M. stemohyoideus
tongue tip " I ^ M. hyoglossus \ lingual sinus
,„„„«oo mucosa I
M. hyoglossus "^^X^iJ"^^
^- hyoglossus
M. genioglossus
u inte,h„«iHai,c / // M. intertiyoideus / (/ . ^ ._, M. submentalis ^ M. geniohyoideus
mentomeckelian \io\)Q%
hyoid plate
F I G U R E 5.2. Musculature of the feeding apparatus in Hyla cinerea (A), Bufo marinus (B), and Hemisus marmoratum (C). Superficial muscles of the feeding apparatus include M. intermandibularis, M. interhyoideus, and M. submentalis. Deep muscles include the extrinsic tongue muscles M. hyoglossus and M. genioglossus, and the hyoid protractor M. geniohyoideus and retractor M. stemohyoideus. A sagittal section of the tongue of Hemisus (D) shows the dorsoventral and longitudinal compartments of the M. genioglossus.
which exhibit a continuum from reduced ossification (e.g., Ascaphus) to hyperossification (e.g., pipid frogs). Reduction or loss of bones has occurred in numerous taxa, and a few taxa possess neomorphic cranial elements (Trueb, 1973). Bones of the neurocranium include the sphenethmoid and the paired prootics and exoccipitals. The dermal roofing bones usually include only the nasals and the frontoparietals. The frontoparietals and prootics provide attachment sites for the M. levator mandibulae anterior longus and M. levator mandibulae posterior longus (= M. temporalis). The bones of the palate include the paired vomers, which are often absent, and the parasphenoid, which is always present and covers the neurocranium ventrally. The prevomers, palatines, and pterygoids are absent in many taxa. Although quite variable in shape, the squamosal is always present. It acts as an important attachment site for the jaw levators. Maxillary arch bones always include the premaxilla and maxilla, and sometimes the quadratojugal. The premaxilla is highly
variable and is important in determining the shape of the snout (Trueb, 1973). The teeth of frogs are reduced compared to most other vertebrates (Duellman and Trueb, 1986). Maxillary and premaxillary teeth are usually, but not always, present. Vomerine teeth are usually present if the vomers are present. No other bones of the upper jaw bear teeth in frogs. Small prey specialists, such as toads (family Bufonidae), lack teeth entirely (Duellman and Trueb, 1986). The jaw muscles consist of one main depressor for opening the mouth and a complex of six levators (= adductors) for closing it. The M. depressor mandibulae originates from the dorsal fascia and/or otic capsule and inserts on the proximal tip of the mandible (Emerson, 1977), close to the jaw joint and therefore in a position of relatively low mechanical advantage. The muscles of the M. levator mandibulae complex extend from the otic capsules and squamosal to the mandible as follows: (1) the M. levator mandibulae anterior longus (= M. pterygoideus) originates on the frontoparietal
121
5. Feeding in Frogs and prootic and inserts via a tendon on the medial side of the angulosplenial; (2) the massive M. levator mandibulae posterior longus (= M. temporalis) originates from the median raphe on the skull roof, the lateral surface of the frontoparietal, and the dorsal surface of the prootic and it inserts via a tendon on the medial side of the angulosplenial; (3) the M. 1. m. posterior lateralis originates on the ventral arm of the squamosal and inserts on Meckel's cartilage and the lateral surface of the angulosplenial; (4) the M. 1. m. posterior articularis originates on the quadrate and inserts on the mandible; (5) the M. 1. m. externus originates on the zygomatic process of the squamosal and inserts laterally on the mandible; and (6) the M. 1. m. posterior subexternus also originates on the zygomatic process of the squamosal and inserts on the posterior end of the mandible (Duellman and Trueb, 1986). All of the Mm. levator mandibulae insert farther from the jaw joint, in a position of relatively greater mechanical advantage, than the M. depressor mandibulae. The Mm. levator
Hyla cinerea
mandibulae are innervated by the trigeminal nerve, whereas the M. depressor mandibulae is innervated by the facial nerve (Figs. 5.3D-5.3F). Although there is much variation in the morphology of the skull and jaws of frogs (Trueb, 1973) and considerable variation among species in the jaw musculature (Starrett, 1968), the functional significance of interspecific differences remains largely unknown (Duellman and Trueb, 1986). James Birch (personal communication) is using morphometric methods to analyze the functional consequences of changes in anuran skull shape during development as well as differences among species. B. Mandible and Buccal Floor Muscles The mandibles of frogs generally consist of three paired bony elements associated with Meckel's cartilage: the angulosplenial, dentary, and mentomeckelian bones (which are absent in pipoids and a few other
Bufo marinus
Hemisus marmoratum
F I G U R E 5.3. Camera lucida drawings of the peripheral nerves of adult Hyla cinerea (A,D), Bufo marinus (B,E), and Hemisus marmoratum (C,F) stained with Sudan black B. (Top row) The glossopharyngeal nerve is shown on the left and the hypoglossal nerve is shown on the right. (Bottom row) The trigeminal nerve is shown on the left and the facial nerve is shown on the right. The stippled area indicates the tongue pad. The glossopharyngeal nerve provides only sensory innervation of the tongue. The hypoglossal nerve innervates the Mm. genioglossus basalis and medialis (if present), the M. hyoglossus, and the M.m. geniohyoideus, sternohyoideus, and omohyoideus. The M. genioglossus basalis is innervated by the most proximal branches of the hypoglossal nerve at the base of the tongue, whereas the M. genioglossus medialis is innervated by the more distal branches. The trigeminal nerve crosses over the mandible and innervates the M. intermandibularis and M. submentalis. At the base of the mandible, the facial nerve innervates the M. interhyoideus.
122
Kiisa C. Nishikawa
species). The presence of a movable joint between the mentomeckelian and dentary is an unusual feature of the mandibles of most anurans (Regal and Gans, 1976; Nishikawa and Roth, 1991). In most species, depression of the mandibular tips results not only from downward movement of the mandibles relative to the cranium (mandibular depression), but also from downward movement of the mentomeckelian bones relative to the rest of the mandible (mandibular bending). Most frogs lack teeth on the mandible. Only one species (Amphignathodon guentheri) is known to possess mandibular teeth (Duellman and Trueb, 1986), although some species (e.g., Ceratophrys) possess tooth-like processes on the dentary. A series of three transversely oriented muscles form the floor of the buccal cavity: the M. submentalis, M. intermandibularis, and M. interhyoideus (Figs. 5.2A5.2C). The M. submentalis connects the anterior ends of the mandibles. During feeding, it bends the mandibles downward by depressing their tips. During breathing, the M. submentalis closes the nares by lifting the mentomeckelian bones upward. This upward movement deforms the alary cartilages, which closes the nares (Gans and Pyles, 1983; Nishikawa and Gans, 1996). The M. intermandibularis extends between the posterior ends of the mandibles (Figs. 5.2A-5.2C). In some species, the M. intermandibularis has supplementary elements of unknown function (Tyler, 1974; Emerson, 1976b). The M. interhyoideus lies posterior to the M. intermandibularis and supports the vocal sacs (Figs. 5.2A-5.2C). The Mm. submentalis and intermandibularis are innervated by the trigeminal nerve, whereas the M. interhyoideus is innervated by the facial nerve (Figs. 5.3D-5.3F). Contraction of the M. intermandibularis and M. interhyoideus raises the buccal floor (Gans and Gorniak, 1982b). C. H y o i d and Associated Muscles The hyobranchial apparatus of adult anurans is a broad plate that forms from the fusion of larval branchial elements (De Jongh, 1968). The slender hyalia extend from the anterior end of the plate and usually attach to the prootic bones, although the hyalia are disjunct in most mesobatrachians and are missing altogether in pelodytids (Cannatella, 1985). Alary processes and posterolateral processes are usually present on the lateral margins of the hyoid plate, and posteromedial processes flank the larynx. In most anurans, only the posteromedial processes are ossified. A pair of muscles, the M. geniohyoideus and M. sternohyoideus, serve to protract and retract the hyoid, respectively (Emerson, 1977; Gans and Gorniak, 1982b). The M. geniohyoideus originates on the pos-
terolateral processes of the hyoid and inserts near the mandibular symphysis (Figs. 5.2A-5.2C). In most species, it consists of separate medial and lateral compartments. The M. sternohyoideus originates on the sternum and inserts on the posterolateral edge of the hyoid plate (Figs. 5.2D-5.2F). The Mm. petrohyoidei anterior et posteriores and the M. omohyoideus elevate and depress the hyoid, respectively (de Jongh and Gans, 1969; Emerson, 1977). The glossopharyngeal nerve innervates the M. petrohyoideus anterior and the vagus nerve innervates the Mm. petrohyoidei posteriores, whereas the hypoglossal nerve innervates the Mm. omohyoideus, sternohyoideus and geniohyoideus (Figs. 5.3A-5.3C; Gaupp, 1896). A great deal of variation in hyoid morphology is found among anurans (Duellman and Trueb, 1986). Aquatic suction feeders (e.g., Hymenochirus, family Pipidae) possess large, articulated, ossified hyoids, whereas terrestrial lingual feeders possess small, fused, cartilaginous hyoids. A similar pattern is found among aquatic and terrestrial turtles, in which suction feeders possess large, ossified, articulated hyoids whereas terrestrial feeders possess small, cartilaginous ones (Bramble and Wake, 1985). An ossified and articulated hyoid appears to be associated with the ability to generate large forces, such as those necessary for moving large volumes of water during aquatic suction feeding. The role of the hyoid in lingual feeding varies widely among tetrapods. In salamanders, it forms a skeletal support for the tongue, and both tongue and hyoid are projected from the mouth as a unit (Lombard and Wake, 1977). In chameleons, the hyoid is a tapered rod around which an accelerator muscle contracts to project the tongue from the mouth (Gans, 1967; Wainwright et ah, 1991). In frogs, the hyoid forms a base on which the tongue rests. The hyoid does not leave the mouth with the tongue. The only connection between hyoid and tongue is the M. hyoglossus, which originates on the posterolateral process of the hyoid and inserts broadly in the tongue. Movements of the hyoid plate appear to play an important role in buccal expansion and contraction during breathing and calling in anurans (de Jongh and Gans, 1969; Martin and Gans, 1972; Emerson, 1977). The role of the hyoid in anuran feeding is less clear (see Section III,B). Cineradiographic recordings of hyoid movement during feeding in Bufo marinus show that the hyoid is stabilized in a retracted position during the initial phase of tongue protraction and that it moves anteriorly during tongue protraction (Emerson, 1977). Based on these observations, Emerson (1977) hypothesized that the hyoid acts as a stable platform for the tongue during the initial stages of protraction, that it stores potential energy during intermediate stages.
5. Feeding in Frogs and that the stored energy is imparted to the tongue during the final stages of protraction. However, this mechanism appears to be unlikely (see Section III,B). In contrast to the neobatrachians that have been studied, the hyoid appears to play a more important role in feeding in mesobatrachians. Based on anatomical observations and muscle stimulation experiments, Trueb and Gans (1983) suggested that the hyoid plays an important role in tongue protraction during feeding in the termite-eating frog, Rhinophrynus dorsalis. This hypothesis remains to be tested experimentally in feeding animals. In the spadefoot toad {Spea multiplicata), another mesobatrachian, tongue movements were impaired after bilateral denervation of the M. geniohyoideus, suggesting that hyoid protraction is necessary for normal tongue protraction in this species (O'Reilly and Nishikawa, 1995). The disjunct hyoid of mesobatrachians may be responsible for the greater role of hyoid protraction during feeding in this group. In most frogs, the cornua of the hyoid are fused to the prootic bones, which may limit forward excursion of the hyoid (Cannatella, 1985). In mesobatrachians, the cornua are continuous until metamorphosis, at which time a gap develops in the cornua, which frees the hyoid plate from its attachment to the skull (Ridewood, 1897), perhaps allowing the hyoid to move farther anteriorly during feeding than it can in other frogs. D . Tongue and Associated Muscles There are several problems that terrestrial frogs must overcome to capture prey successfully. These include contacting the prey with the tongue, ingesting prey, transporting it through the oral cavity, and finally swallowing it. In terrestrial anurans, the sticky tongue plays an important role in ingesting, transporting, and swallowing prey. There are no known instances of inertial feeding or transport among anurans, probably because they consume mostly small prey. Frogs do not masticate their food, and there is relatively little manipulation of the food by the tongue once it is in the oral cavity. Little is known about differences among anuran species in modes of oral transport and swallowing. All anurans possess a relatively simple tongue that consists only of two pairs of extrinsic muscles: the M. genioglossus and M. hyoglossus (Gaupp, 1901; Regal and Gans, 1976). In contrast to most other terrestrial vertebrates, intrinsic muscles are absent in most species. The M. genioglossus originates near the mandibular symphysis and inserts posteriorly into the tongue pad (Figs. 5.2A-5.2C). In many anurans, the M. genioglossus is subdivided by connective tissue
123
into a number of different compartments, which vary widely among species (Horton, 1982). The M. hyoglossus originates on the posteromedial process of the hyoid and inserts along the lateral margin of the tongue pad, often interdigitating with the fibers of the M. genioglossus (Figs. 5.2A-5.2C). The Mm. genioglossus and hyoglossus are innervated by the hypoglossal nerve (Figs. 5.3A-5.3C). In most frogs, the M. genioglossus is used to place the tongue on the prey (discussed in detail in Section III) and prey are returned to the mouth by the M. hyoglossus. In Bombina, Bufo, Phrynomerus, and Hemisus, denervation experiments demonstrate that the M. hyoglossus plays an important role in prey capture, oral transport, and swallowing (Ritter and Nishikawa, 1995; Tso et ah, 1995). Small prey are returned to the mouth and delivered to the esophagus by the tongue in a single movement, sometimes with the help of the forelimbs, whereas large prey are nearly always transported with the help of the forelimbs, presumably because lingual transport would be ineffective by itself (Gray et ah, 1997). Because most anurans retract the eyes into the orbit during swallowing, it has been suggested that the M. retractor bulbi also plays a role in anuran swallowing (Regal and Gans, 1976). However, this hypothesis has yet to be tested experimentally (Duellman and Trueb, 1986). There have been three broadly comparative morphological studies of the tongue musculature of anurans. Magimel-Pelonnier (1924) studied 45 species. Regal and Gans (1976) studied 12 species, and Horton (1982) studied 63 species, predominantly from Australia. In all, 61 genera are represented. From these studies, three different patterns of tongue morphology have been described. The first pattern consists of a round tongue that is broadly attached to the floor of the mouth, both anteriorly and posteriorly, so that there is no free flap posteriorly. In these tongues, the fibers of the M. hyoglossus radiate from the hyoid into the tongue pad and the ventralmost fascicles insert near the base of the tongue (Horton, 1982). The fibers of the M. genioglossus arise near the niandibular symphysis and radiate into the tongue, where they interdigitate with those of the M. hyoglossus (Regal and Gans, 1976; Horton, 1982). This morphology is found in all archaeobatrachians (i.e., Ascaphus, Leiopelma, Bombina, Alytes, and Discoglossus), some mesobatrachians {Pelobates, Pelodytes, and Rhinophrynus), and some neobatrachians, including Telmatobius, Litoria, Hyla, Gastrotheca, Rheobatrachus, and several genera of limnodynastines (Magimel-Pelonnier, 1924; Regal and Gans, 1976; Horton, 1982; Trueb and Gans, 1983). Based on their origins and insertions. Regal and
124
Kiisa C. Nishikawa
Gans (1976) hypothesized that, upon contraction, the M. genioglossus pulls the tongue toward the symphysis in these species, whereas the M. hyoglossus pulls the tongue toward the esophagus. These hypotheses have been confirmed by denervation of the M. genioglossus in Bombina (Nishikawa et ah, 1992), Discoglossus (Nishikawa and Roth, 1991), and Hyla (Deban and Nishikawa, 1992) and by denervation of the M. hyoglossus in Bombina (Tso et ah, 1995). The second pattern is found only among aquatic pipid frogs (e.g., Xenopus, Pipa, and Hymenochirus), which possess large articulated, ossified hyoids (Cannatella, 1985) and are secondarily tongueless (Cannatella and Trueb, 1988) although they possess vestiges of tongue musculature (Horton, 1982). These species use suction or ram feeding to ingest and transport prey (Sokol, 1969; Avila and Frye, 1977; O'Reilly et al, 1999). A third pattern consists of a muscular tongue that is attached anteriorly to the buccal floor, with a free posterior flap that varies in length among species (Regal and Gans, 1976) and varies to some extent among fixed specimens within a species (Magimel-Pelonnier, 1924). This type of tongue is found in some pelobatoids {Scaphiopus, Spea, and Megophrys) and some neobatrachians, such as Hyla and Limnodynastes (Magimel-Pelonnier, 1924; Regal and Gans, 1976; Horton, 1982). In these tongues, the fibers of the M. hyoglossus recurve to insert in the more distal parts of the tongue, rather than its anterior base. In contrast to the M. hyoglossus, which is relatively homogeneous among species, the arrangement of the M. genioglossus varies widely (Magimel-Pelonnier, 1924; Regal and Gans, 1976; Horton, 1982). All species retain the interdigitating element that is found in species with round, broadly attached tongues, but several additional elements also may be present (Regal and Gans, 1976; Horton, 1982). These elements fall into two groups: those with fibers that run parallel to the long axis of the tongue [i.e., the ventral, dorsomedial and superficial elements of Horton (1982)] and those with transverse fibers [i.e., the genioglossus basalis of Gaupp (1901)]. At present, there is no information concerning the functional significance of this variation in the arrangement of the M. genioglossus (but see Section III,C). As in species with round, broadly attached tongues, retraction of the tongue is accomplished by contraction of the M. hyoglossus (Gans and Gomiak, 1982a,b; Ritter and Nishikawa, 1995; Tso et al, 1995). On the basis of their anatomy, the tongues of species with free posterior flaps are presumed to rotate over the mandibular symphysis so that the dorsal surface of the tongue at rest becomes the ventral surface of the fully protracted tongue, as has been shown to occur in Rana (Gans, 1961, 1962) and Bufo (Gans and Gorniak,
1982a,b). It has been hypothesized that these "flipping tongues" have evolved convergently several times among frogs (Regal and Gans, 1976), and this idea is supported by phylogenetic analysis (Section V,A). Despite the variation in tongue morphology, most frogs (with the exception of pipids) share several important features of the tongue: (1) it is attached anteriorly near the mandibular symphysis; (2) most of the fibers in both the protractor and the retractor muscles are oriented nearly parallel to the long axis of the tongue so that their shortening would pull the tongue pad either toward the symphysis or toward the esophagus (Horton, 1982); (3) those fibers that are transverse (i.e., genioglossus basalis) are relatively short and are associated with large amounts of connective tissue (Horton, 1982); (4) the resting length of the tongue approximates the length of the mandibles; and (5) the mass of the tongue is approximately 0.5-1.0% of body mass, which is about as large in relative terms as the human heart.
III. FUNCTION OF THE FEEDING APPARATUS This section first gives an overview of methods that can be used to study the function of any morphological system, in this case the feeding apparatus. It next discusses previously published hypotheses for the mechanism of tongue protraction in anurans, including an analysis of which methods were informative, which were uninformative, and which were positively misleading. The section ends with a discussion of functional diversification of the mechanism of tongue protraction among anuran species. A. Methods for Studying the Function of the Feeding Apparatus Before undertaking a comparative and functional analysis of feeding, it is important to ask what types of experiments and observations can best be used to understand the function of the feeding apparatus. Obviously, the necessary data include a description of anatomy as well as a description of movement patterns, which usually involves some type of kinematic analysis. These often are supplemented with information about muscular and neural activity. Traditional methods include recordings of electrical activity from relevant muscles and nerves as well as electrical stimulation experiments. Kinematic studies are of three types: (1) description of films or still photos; (2) analysis of kinematic profiles in which values of kinematic variables, such as gape angle, are plotted over time (Fig. 5.4); and (3) trajectory
90 80
Hyla cinerea
Hemisus marmoratum
Bufo marinus
701
c
< Q. CO
o
60 50 40 30 20 10 0 -10 5.0
4.0 3.0 2.0 1.0 0.0 0> Q. CO
o
-1.0 -2.0 -3.0 200
M^
1901
c
CL CQ
I
40 20 /
/
/
/\ A
\-,
ii
//
0.5
1.0
1.5
2.0
Time (sec)
0.0
0.2
0.4
0.6
0.8
1.0
Time (sec)
F I G U R E 6.5. Selected sample gape cycles from (A) aquatic larval Epicrionops {n = 5), (B) terrestrial adult Ichthyophis {n = 3), (C) terrestrial adult Hypogeophis {n = 3), and (D) aquatic adult Typhlonectes {n = 3). Note the extreme length of some adult gape cycles and how jaw closing can be interupted and the gape cycle restarted multiple times before the mouth is finally closed on the prey item.
differences. The two aquatic foraging caecilians for which data are available {Typhlonectes and Hypogeophis) use jaw prehension to capture aquatic prey and cannot suction feed (O'Reilly, 1995). However, Typhlonectes does depress the hyobranchium in conjunction with mouth closing (Fig. 6.6, Right), which probably keeps the lunging caecilian from pushing prey items away ("compensatory suction" of Van Damme and Aerts 1997). Compared to terrestrial caecilians, Typhlonectes has relatively fast jaw movements for its body size (O'Reilly, 1995). Individuals of Hypogeophis increase the speed of their jaw movements when fed the same prey item in water than on land. However, this facultative increase in speed does not result in movements as quick as those seen in Typhlonectes, suggesting that the timing of feeding behavior has diverged significantly from that of its terrestrial ancestors. The degree to which the retroarticular process is re-
curved and whether it curves antiororly or posteriorly to the jaw articulation is likely to have significant functional consequences (Fig. 6.7). When the retroarticular process is straight, the interhyoideus will be most effective at large gape angles (Fig. 6.7A). As the retroarticular process becomes more recurved, bite force performance at low gape angles will improve, but at the expense of performance at high gape angles (Fig. 6.7B). The decrease in performance with increasing gape is more pronounced if the jaw curves upward anterior to its articulation with quadrate (Fig. 6.7C). Available data on maximum gape angles support this model, as Ichthyophis, which has a strongly recurved retroarticular process, has only been observed to open the jaws during feeding to about 50°, whereas Hypogeophis and Typhlonectes, with less recurved retroarticular processes, open their mouths to over 70° (O'Reilly, 1995).
159
6. Feeding in Caecilians TABLE 6.1 Species Gymnophiona Epicrionops sp.—larvae^ Hypogeophis rostratus' Ichthyophis kohtaoensis^ Typhlonectes nutans ^ Anura Ascaphus truei" Bombina orientalist Bufomarinus^ Discoglossus pictus^ Hyla cinerea" Hymenochirus curtipes^ Rana pipiens (waxworm)^ R. pipiens (earthworm)" Caudata Ambystoma dumerilii^ A. mexicanum^ A. mexicanum^ A. ordinarium^ A. tigrinum—larvae^ A. tigrinum^ A. tigrinum^ Amphiuma means ^ Bolitoglossa occidentalism' B. mexicana^ Cryptobranchus allegheniensis ^ Cynops pyrrhogaster^ Desmognathus fuscus ^' D. marmoratus^ D. quadramaculatus^ D. quadramaculatus' Dicamptodon tenebrosus—larvae^ Ensatina eschscholtzii^ Hynobius nebulosus^ H. kimurae^ Necturus maculosus ^ Pachytriton brevipes^ P. brevipes^ Plethodon glutinosus^ Pleurodeles waltl^ P. waltl' Paramesotriton hongkongensis" Salamandra salamandra" Salamandrina terdigitata^ Siren intermedia ^ Taricha torosa'' Tylototriton verrucosus''
Duration of the Gape Cycle in Amphibians^ Source
Range (msec)
Mean (± SE)
42-58 400-1733 367-4734 300-1033
51.8 ± 1.8 1028 ± 73 1819 ± 189 572 ± 41
O'Reilly O'Reilly O'Reilly O'Reilly
80-280 83-158
-140 126.5 ± 7.6 143 ± 22 134 152 ± 8.1 62.7 ± 4.2 100.8 ± 6.4 163.2 ± 9.8
Nishikawa and Cannatella (1991) O'Reilly (1995) Cans and Gorniak (1982) Nishikawa and Roth (1991) Deban and Nishikawa (1992) O'Reilly (1995) Anderson (1993) Anderson (1993)
88 ± 3.5 59 ± 1.8 69.7 ± 1.2 73 ± 2.6 83 75 107 72 ± 1.9 99.6 ± 1 . 4 108 ± 2 53.3 ± 1.1 162 ± 16.3 92.6 ± 19.6 128 136.1 ± 6.1 109.9 ± 1.1 54.4 ± 2.5 87.4 ± 4.9 115.7 ± 7.2 94.4 ± 5.7 51 ± 1 79 ± 10.9 44.4 ± 2.4 96.3 ± 6.3 75.9 ± 3.2 220 ± 21.2 211 ± 30.3 100 ± 7.1 240 ± 23.1 61 ± 2.8 246 160 ± 10.7
Shaffer and Lauder (1985) Reilly and Lauder (1992) Shaffer and Lauder (1985) Shaffer and Lauder (1985) Shaffer and Lauder (1988) Shaffer and Lauder (1988) Shaffer and Lauder (1988) Reilly and Lauder (1992) Larsen et al (1989) Larsen et al (1989) Reilly and Lauder (1992) Miller and Larsen (1990) Larsen and Beneski (1988) Schwenk and Wake (1993) Larsen and Beneski (1988) Larsen et al (1989) Reilly and Lauder (1992) Larsen et al (1989) Larsen et al (1989) Larsen et al (1989) Reilly and Lauder (1992) Miller and Larsen (1990) O'Reilly (1995) Larsen et al (1989) O'Reilly (1995) Miller and Larsen (1990) Miller and Larsen (1990) Miller and Larsen (1990) Miller and Larsen (1990) Reilly and Lauder (1992) Findeis and Bemis (1990) Miller and Larsen (1990)
92-325 42-108
84-288
50-112 25-58 50-125 144-312 64-468 60-136 90-326
72-248
(1995) (1995) (1995) (1995)
^Note how long the gape cycles of adult caecilians are relative to frogs and salamanders. ^Aquatic feeding. '^Terrestrial feeding.
C. Intraoral Transport After prey is seized, caecilians use different methods of prey transport, depending on the medium in which they are feeding (air vs water) and the size and type
of prey being eaten. Larvae of Epicrionops have only been observed using hydraulic transport (O'Reilly, 1995), despite the presence of a well-developed tongue (Wake, 1989). Metamorphosed caecilians employ two or three methods of transport, depending on the
160
James C. O'Reilly
F I G U R E 6.6. (Left) Selected video frames of terrestrial prey capture in an adult Hypogeophis rostratus. Note the forward lunge, lack of tongue protraction, and lack of hyobranchial depression. (Right) Selected video frames of aquatic prey capture in an adult Typhlonedes nutans. Note the forward lunge, lack of tongue protraction, and substantial hyobranchial depression during mouth closing.
161
6. F e e d i n g in Caecilians
AAAAAA
^
AAAAAA
AAAAAA
45°
^
90°
E o CQ
40
^
20 0,
Gape Angle (")
20
Gape Angle (°)
40
(
Gape Angle {°)
F I G U R E 6.7. Model of the relationship of mechanical advantage and gape angle in the three major arrangements of the retroarticular process and interhyoideus found in living caecilians. Dashed lines with arrows represent force applied by the interhyoideus. Arrangement A has its best mechanical advantage at high gape angles. Arrangements B and C reach peak performance at low gape angles. Model assumes identical force input and retroarticular process length across all systems.
situation. On land, caecilians use a combination of inertial transport (releasing prey and lunging over it, using the mass of the prey as a counter weight) and lingual transport on larger prey items, while using lingual transport alone on smaller items (Bemis et ah, 1983; O'Reilly, 1995). In water, adults add hydraulic transport sequences to the mix (O'Reilly, 1995). During transport, caecilians often use longitudinal spinning to help subdue prey (Tanner, 1971; Bemis et al, 1983), the function of which is still not well understood. Larger arthropods (such as large crickets) are often rubbed against the substrate to disable or remove their limbs.
IV. EVOLUTION Living caecilians represent an ancient lineage and began to diverge from one another over 100 million
years ago (Hedges et al, 1993). The only fossil material that has been attributed to Gymnophiona that includes cranial material is Eocaecilia micropodia from the early Jurassic. This species is thought to be the closest known outgroup to living caecilians Jenkins and Walsh, 1993). Unlike living caecilians, Eocaecilia had limbs and limb girdles. Its skull was solidly roofed, lending some support to arguments that the ancestor of living caecilians lacked fenestrae in the skull (Carroll and Currie, 1975; Jenkins and Walsh, 1993). However, read Wake and Hanken (1982) and Nussbaum (1983) for counterarguments and reviews of the available evidence. The issue of whether the solidly roofed skull has secondarily evolved in living caecilians as an adaptation for burrowing or is homologous to the condition seen in the first tetrapods will only be settled by a vast improvement in the available fossil record of the ancestors of living amphibians. For the purposes of this
162
James C. O'Reilly
chapter, I will largely follow the scenario of Nussbaum (1977,1983) while acknowledging that the fenestrated condition of adult rhinatrematids may have evolved within that lineage from a solid skulled ancestor through paedomorphosis. The common ancestor of living caecilians was most likely oviparous, with the female guarding her clutch of eggs until they hatched into free-living, aquatic larvae. These larvae localized prey by the combined use of olfaction and electroreception, after which they used suction feeding for prey capture. The skull had large temporal fenestrae and was strongly kinetic, especially in the case of movement of the quadrate relative to the braincase (streptostyly). The hyobranchial apparatus may have been cartilaginous or ossified, but almost certainly possessed many independent elements that articulated such that posterior-directed force applied to the basibranchial series was translated into ventrally directed force expanding the buccopharynx. The flow of water that carried prey into the mouth was focused by well-developed labial lobes. The teeth where pedicellate and best developed in the front of the mouth, where two rows where present on both the upper and the lower jaws. On the upper jaw, the outer (premaxillary-maxillary) series of teeth did not extend as far caudally as the inner (vomer-palatine) series. In contrast, on the lower jaw, the outer row of teeth was more extensive than the inner row. After capture, hydraulic transport was used to transport prey to the pharynx. After spending many months as a larva, this common ancestor metamorphosed into a terrestrial, semifossorial adult. During metamorphosis, the skull became more solidly constructed and less kinetic due to the fusion of some dermal elements that were independent in larvae. However, the skull still had large temporal fenestrae, components of the levator mandibulae complex originating on the midline of the skull and passing through a large gap between the squamosal and the parietal before inserting on the lower jaw. Although the quadrate was now connected more firmly to an expanded squamosal, it could still move mediolaterally relative to the neurocranium during prey transport. The lateral line system, not being useful on land, was lost at metamorphosis. The lacrimal (tear) ducts were used to transport chemicals directly to the vomeronasal organ from the eye, possibly allowing the animals to smell even when the nostrils were pressed into the soil during burrowing. This adult used jaw prehension to capture prey, with the levator mandibulae complex being aided during mouth closing by an unusually oriented interhyoideus muscle inserting on the ventral edge of the retroaricular process of the lower jaw. The dentition consisted of two rows of numerous
teeth in both the upper and the lower jaws, with all four rows extending caudally to the corner of the gape. The mouth opening was terminal, looking like that of an adult salamander. The first evolutionary changes from the just-described ancestral pattern concerned the terrestrial adult stage, with the characteristics of the larval stage remaining relatively constant. As the adult stage became more specialized for burrowing, some signficant changes took place in head anatomy. The mouth opening became sub terminal, with the lower jaw being recessed into the upper jaw. The tear duct chemosensory system became more elaborate, with extrinsic eye muscles and other parts of the eye being incorporated into the now protusible tentacle apparatus with an independent aperture located well in front of the orbit. Important transitions during the evolution of the caecilian feeding apparatus include changes in the arrangement of the jaw levators and the degree to which the adductor chamber is roofed by dermal bones. The ancestor of living caecilians apparently possessed a skull that was broadly open between the squamosal and the temporal. Prominent jaw levators originated on the top of the skull and traveled down through these fenestrae and inserted on the lower jaw. This primitive condition is still seen in larval caecilians and in adult rhinatrematids. In a common ancestor of all caecilians other than rhinatrematids, the squamosal became much larger, completely covering the temporal fenestra. Concurrently, the jaw levators were reduced in size, being restricted to the now din\inutive adductor chamber under the squamosal. Temporal fenestrae have subsequently reappeared convergently in three groups {Geotrypetes, Scolecotnorphus, and typhlonectids), but in all of these cases the jaw levators have remained restricted to the adductor chamber. Transitions in the arrangement of the retroarticular process and interhyoideus posterior were also important. A common ancestor of caecilians had a small retroarticular process and an interhyoideus muscle, which originated midventrally and inserted on the ceratohyals. The original shift of insertion to the retroarticular process and subsequent enlargement of the process and muscle probably originally occurred as an adaptation for breathing. As the ancestors of caecilians became more elongate in body form, the ratio of the volume of the buccal chamber to that of the lungs became progressively smaller. Moving the insertion of the interhyoideus to the retroarticular process would increase the stroke volume of the buccal pump and at least partially compensate for these changes (Carrier and Wake, 1995). This condition is closely approximated by the condition seen in living rhinatrematids.
163
6. Feeding in Caecilians After the initial shift in the insertion of the interhyoideus, ancestral caecilians made major rearrangements in the jaw apparatus as they became progressively more fossorial. First, there was a concurrent increase in the size of both the interhyoideus and the retroarticular process. At least three times (in ichthyophiids and uraeotyphlids, scolecomorphids, and Atretochoana), the relatively straight retroarticular process has been recurved radically upward. In ichthyophiids and uraeotyphlids, the process is curved posterior to the jaw articulation, whereas in scolecomorphids and Atretochoana the curve occurs before the jaw articulation, such that the articular surface lies in the vertical rather than the horizontal plane. The primitive arrangement of the jaw muscles and skull presents at least two problems for active burrowers that use their heads as the only means of penetration. First, large levators necessarily increase the diameter of the head. Because elongate limbless burrowers use compression to create tunnels, the cost of constructing a given length of tunnel is exponentially related to tunnel diameter (Gans, 1974) and they are most likely under strong selection to minimize body diameter. Because the interhyoideus is tucked behind the skull, relying on it as the primary jaw-closing muscle permitted a vast increase in maximum jawclosing forces with no increase in head diameter. Muscle is also more subject to injury when exposed to crushing forces than is bone. Moving the primary jawclosing muscles behind the skull and roofing the levators with bone would allow more violent burrowing movements than were possible in ancestral caecilians. While the unique arrangement of caecilian jaw-closing muscles is a spectacular adaptation for head-first burrowing, it may have also led to a severe constraint on caecilian cranial evolution. The interhyoideus, ancestrally functioning only as a breathing muscle, has at least one new function (jaw closing) and its position suggests that it plays a prominent role in head movements during burrowing as well (Lawson, 1965b). Condensing all of these functions to a single muscle potentially conserves an enormous amount of muscle volume that would otherwise increase body diameter. However, as single components take on multiple functions, their evolution tends to become severely constrained (Lauder and Liem, 1989). In most living caecilians the interhyodius is the dominant muscle in breathing, jaw closing, and side-to-side head movements during burrowing and, despite its great age, we see remarkably little variation in its arrangement. Interestingly, the only great experiment in caecilian head design is seen in Atretochoana, an animal that does not breathe and is apparently not an accomplished bur-
rower (Nussbaum and Wilkinson, 1995; Wilkinson and Nussbaum, 1997). V. THE FUTURE Caecilians are the most poorly known major group of tetrapods. Regardless of why this is the case, it makes them a gold mine of opportunity for any researcher willing to make the extra effort to acquire and work with these animals. The study of caecilian feeding is no exception. Despite a relatively good understanding of variation in caecilian cranial anatomy, there is a dearth of experimental studies on every aspect of caecilian feeding behavior. The neurophysiology and neuroethology of caecilian prey capture remain to be described. There are only two experimental studies on the sensory systems of caecilians (Himstedt and Fritzsch, 1990; Himstedt and Simon, 1995). Thus we still know very little about how any caecilian perceives its surroundings and finds prey, let alone how prey detection abilities vary among different species. The relative importance of different types of stimuli in different environments and the cellular physiology of various receptors are currently unknown. The motor control of feeding movements and how the neural connections of the interhyoideus have changed during its integration into the jaw closing system also await study. There is currently a single published paper (Bemis et ah, 1983) describing prey capture and transport in any detail and this treats only a single species {Dermophis). There are no detailed descriptions of the feeding behavior of larval or fetal caecilians. There are no quantitative studies on the functional significance of the differences among the five major arrangements of the adult caecilian feeding apparatus, including variation in feeding kinematics, bite force generation, or prey transport behavior. The variation in the contractile characteristics of the different jaw adductors and how this is related to the various biomechanical arrangements seen in living caecilians is also unknown. Although cranial kinesis is widespread among caecilians (Nussbaum, 1977; Wilkinson and Nussbaum, 1997), we know little about its potential functional significance. Ultimately, it would be ideal if the just-mentioned studies were placed in ecological, ontogenetic, and evolutionary contexts. Understanding how caecilians interact with their actual environment would be the ultimate test of the relevance of laboratory studies. Understanding how function varies during ontogeny will provide a complete picture of the demands placed on
164
James C. O'Reilly
the feeding system as well as the opportunities available through heterochrony during evolution. Finally, placing future studies in an explicit phylogenetic context will allow the synthesis of functional, ecological, and developmental data into a robust portrait of caecilian evolution. Acknowledgments First, I must extend my deep appreciation to Kurt Schwenk for his encouragement and patience during the writing of this chapter. Adam Summers and Nate Kley provided many helpful comments on earlier drafts of this manuscript. Ron Nussbaum and Mark Wilkinson introduced me to caecilian biology and have continued to provide mentoring and insight in all things caecilian for the last 12 years. Ron Nussbaum also supplied many of the animals on which the observations herein are based. Edmund Brodie, Jr., Jonathan Campbell, Louis Porras (Zooherp, Inc., Sandy, Utah), Rob Maclnnes (Glades Herp. Inc., Ft. Myers, Florida), and Ed Budziak aided in the acquisition of specimens. Carl Cans provided logistical and financial support during my initial forays into the topic of caecilian prey capture. Steve Deban and Dale Ritter helped in the videotaping of feeding sequences. Kiisa Nishikawa provided extensive advice and support throughout all aspects of this project. This work was supported by NSF Grants IBN-8909937 and IBN-9211310 to Kiisa Nishikawa, a grant-in-aid-of-research from Sigma Xi, and a Darwin Postdoctoral Fellowship from the Organismic and Evolutionary Biology Program at the University of Massachusetts Amherst.
References Anderson, C. W. (1993) The modulation of feeding behavior in response to prey type in the frog Rana pipiens. J. Exp. Biol. 179:1-12. Badenhorst, A. (1978) The development and the phylogeny of the organ of Jacobson and the tentacular apparatus of Ichthyophis glutinosus (Linne). Annale Universiteit van Stellenbosch, Serie A2 (Soologie) 1:1-26. Barbour, T., and A. Loveridge (1928) A comparative study of the herpetological faunae of the Uluguru and Usambara Mountains, Tanganyika Territory, with descriptions of new species. Mem. Museum Comp. Zool. 50:87-265. Bemis, W. E., K. Schwenk, and M. H. Wake (1983) Morphology and function of the feeding apparatus in Dermophis mexicanus (Amphibia: Gymnophiona). Zool. J. Linn. Soc. 77:75-96. Billo, R., and M. H. Wake (1987) Tentacle development in Dermophis mexicanus (Amphibia: Gymnophiona), with an hypothesis of tentacle origin. J. Morph. 192:101-111. Breckenridge, W. R., S. Nathanael, and L. Pereira (1987) Some aspects of the biology and development of Ichthyophis glutinosus (Amphibia: Gymnophiona). J. Zool. Lond. 211:437-449. Carrier, D. R., and M. H. Wake (1995) Mechanism of lung ventillation in the caecilian Dermophis mexicanus. J. Morph. 226:289-295. Carroll, R. L., and P. J. Currie (1975) Microsaurs as possible apodan ancestors. Zool. J. Linnean Soc. 57:229-247. Clairambault, R, M.-J. Cordier-Picouet, and C. Pairault (1980) Premieres donnees sur les projections visuelles d'un Amphibien Apode {Typhlonectes compressicauda), C. R. Acad. Sci. Ser. D. 291: 283-286. Deban, S. M., and K. C. Nishikawa (1992) The kinematics of prey capture and the mechanism of tongue protraction in the green tree frog, Hyla cinerea. J. Exp. Biol. 170:235-256.
Ducey, P. K., D. R. Formanowicz, L. Boyet, J. Mailloux, and R. A. Nussbaum (1993) Experimental examination of burrowing behavior in caecilians (Amphibia, Gymnophiona): effects of soil compaction on burrowing ability of four species. Herpetologica 49:450-457. Edgeworth, F. H. (1925) On the autosylism of Dipnoi and Amphibia. J. Anat. 59:225-264. Edgeworth, F. H. (1935) The Cranial muscles of Vertebrates. Cambridge University Press, Cambridge. Engelhardt, F. (1924) Tentakelapparat und auge von Ichthyophis. Jenaische Zeitschrift fuer Naturwissenschaft 60:241-304. Exbrayat, J.-M., and M. Delsol (1985) Reproduction and growth of Typhlonectes compressicaudus: a viviparous gymnophione. Copeia 1985:950-955. Findeis, E. K., and W. E. Bemis (1990) Functional morphology of tongue projection in Taricha torosa (Urodela: Salamandridae). Zool. J. Linn. Soc. 99:129-157. Fox, H. (1985) The tentacles of Ichthyophis (Amphibia: Caecilia) with special reference to the skin. J. Zool., Lond. (A) 205:223-234. Fritzsch, B. (1988) Phylogenetic and ontogenetic origin of the dorsolateral auditory nucleus of anurans. Pp. 561-585. In: The evolution of the amphibian auditory system. B. Fritzsch, M. Ryan, W. Wilczynski, W. Hetherington, and W. Walkowiak (eds.). Wiley, New York. Fritzsch, B., W. Himstedt, and M.-D. Crapon de Caprona (1985) Visual projections in larval Ichthyophis kohtaoensis (Amphibia: Gymnophiona). Dev. Brain. Res. 23:201-210. Fritzsch, B., U. Wahnschaffe, M.-D. Crapon de Caprona, and W. Himstedt (1985) Anatomical evidence for electroreception in larval Ichthyophis kohtaoensis. Naturwiss 72:102-104. Fritzsch, B., and M. H. Wake (1986) The distribution of ampuUary organs in Gymnophiona. J. Herp. 20:90-93. Fritzsch, B., and M. H. Wake (1988) The inner ear of gymnophione amphibians and its nerve supply: a comparative study of regressive events in a complex sensory system (Amphibia: Gymnophiona). Zoomorph. 108:201-217. Gans, C. (1974) Biomechanics: An Approach to Vertebrate Biology. The University of Michigan Press, Ann Arbor, MI. Gans, C , and G. C. Gorniak (1982) Functional morphology of lingual protrusion in marine toads {Bufo marinus). Am. J. Anat. 163: 195-222. Gaymer, R. (1971) New method of locomotion in limbless terrestrial vertebrates. Nature 234:150-152. Hebrard, J. J., G. M. O. Maloiy, and D. M. I. Alliangana (1992) Notes on the habitat and diet of Afrocaecilia taitana (Amphibia, Gymnophiona). J. Herp. 26:513-515. Hedges, S. B., R. A. Nussbaum, and L. Maxson (1993) Caecilian phylogeny and biogeography inferred from mitochondrial DNA sequences of the 12S and 16S rRNA genes (Amphibia: Gymnophiona). Herp. Monogr. 7:64-76. Hetherington, T. E., and M. H. Wake (1979) The lateral line system in larval Ichthyophis (Amphibia: Gymnophiona). Zoomorph. 93: 209-225. Himstedt, W. (1991) Zur biologie und larvenentwicklung der thailandischen blindwiihle Ichthyophis kohtaoensis (Amphibia: Gymnophiona: Ichthyophiidae). Veroff. Naturhist. Mus. Schleusingen 6:16-24. Himstedt, W. (1995) Structure and function of the eyes of the caecilian Ichthyophis kohtaoensis (Amphibia: Gymnophiona). Zool. (Jena) 99:81-94. Himstedt, W. (1996) Die Blindwiihlen. Die Neue Brehm-Bucherei, Vol. 360. Westarp/Spektrum, Madgeburg/Heidelberg. Himstedt, W., and B. Fritzsch (1990) Behavioural evidence for electroreception in larvae of the caecilian Ichthyophis kohtaoensis (Amphibia, Gymnophiona). Zoologische Jahrbiicher-Abteilung fiir Allgemeine Zoologie und Physiologie der Tiere 94:486-492.
6. F e e d i n g in Caecilians Himstedt, W., and G. Manteuffel (1985) Retinal projections in the caecilian Ichthyophis kohtaoensis (Amphibia: Gymnophiona). Cell Tissue Res. 239:689-692. Himstedt, W., and D. Simon (1995) Sensory basis of foraging behaviour in caecilians (Amphibia: Gymnophiona). Herp. J. 5:266-270. Jenkins, F. A., and D. M. Walsh (1993) An early Jurassic caecilian with limbs. Nature 365:246-250. Largen, M. J., P. A. Morris, and D. W. Yalden (1972) Observations on the caecilian Geotrypetes grandisonae Taylor (Amphibia: Gymnophiona) from Ethiopia. Monit. Zool. Ital. 4:185-205. Larsen, J. H., and J. T. Beneski (1988) Quantitative analysis of feeding kinematics in dusky salamanders (Desmognathus). Can. J. Zool. 66:1309-1317. Larsen, J. H., J. T. Beneski, and D. B. Wake (1989) Hyolingual feeding systems of the Plethodontidae: comparative kinematics of prey capture by salamanders with free and attached tongues. J. Exp. Zool. 252:25-33. Lauder, G. V., and K. F. Liem (1989) The role of historical factors in the evolution of complex organismal functions. Pp. 63-78. In: Complex Organismal Functions: Integration and Evolution in Vertebrates. D. B. Wake and G. Roth (eds.). Wiley, New York. Lauder, G. V., and S. M. Reilly (1993) Amphibian feeding behavior: comparative biomechanics and evolution. Pp. 163-195. In: Biomechanics of Feeding in Vertebrates. V. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Lawson, R. (1965a) The teeth of Hypogeophis rostratus (Amphibia, Apoda) and tooth structure in the Amphibia. Proc. Zool. Soc. London 145:321-326. Lawson, R. (1965b) The anatomy of Hypogeophis rostratus. II. The musculature. Proc. Univ. Newcastle Tyne Phil. Soc. 1:52-63. Lewis, E. R., E. L. Leverenz, and W. Bialek (1985). The Vertebrate Inner Ear. CRC Press, Boca Raton, FL. Lewis, E. R., and R. E. Lombard (1988) The amphibian inner ear. Pp. 93-123. In: The Evolution of the Amphibian Auditory System. B. Fritzsch, M. Ryan, W. Wilczynski, T. H. Hetherington, and W. Walkowiak (eds.). Wiley, New York. Loveridge, A. (1936) Scientific results of an expedition to rain forest regions in eastern Africa. VII. Amphibians. Bull. Mus. Comp. Zool. 79:369-430. Miller, B. T. and J. H. Larsen (1990) Comparative kinematics of terrestrial prey capture in salamanders and newts (Amphibia: Urodela: Salamandridae). J. Exp. Zool. 256:135-153. Moll, E. O., and H. M. Smith (1967) Lizards in the diet of an American caecilian. Nat. Hist. Miscell. Chicago Acad. Sci. 187:1-2. Moodie, G. E. E. (1978) Observations on the life history of the caecilian Typhlonectes compressicaudus (Dumeril and Bibron) in the Amazon Basin. Can. J. Zool. 56:1005-1008. Nishikawa, K. C , and D. C. Cannatella (1991) Kinematics of prey capture in the tailed frog, Ascaphus truei (Anura: Ascaphidae). Zool. J. Linn. Soc. 103:289-307. Nishikawa, K. C , and G. Roth (1991) The mechanism of tongue protraction during prey capture in the frog, Discoglossus pictus. J. Exp. Biol. 159:217-234. Nishikawa, K. C , C. W. Anderson, S. M. Deban, and J. C. O'Reilly (1992) The evolution of the neural circuits controlling feeding behavior in frogs. Brain Behav. Evol. 40:125-140. Nussbaum, R. A. (1977) Rhinatrematidae: a new family of caecilians (Amphibia: Gymnophiona). Occasional Papers of the Museum of Zoology, University of Michigan 682:1-30. Nussbaum, R. A. (1983) The evolution of a unique dual jaw-closing mechanism in caecilians (Amphibia: Gymnophiona) and its bearing on caecilian ancestry. J. Zool. 199:545-554. Nussbaum, R. A. (1985a) Amphibian Fauna of the Seychelles Archipelago. National Geographic Research Reports 18:53-62. Nussbaum, R. A. (1985b) Systematics of the caecilians (Amphibia:
165
Gymnophiona) of the family Scolecomorphidae. Occasional Papers of the Museum of Zoology, University of Michigan 713:1-49. Nussbaum, R. A. (1992) Caecilians. Pp. 52-59 In: Reptiles and Amphibians. H. G. Cogger and R. G. Zweifel (eds.). Smithmark, New York. Nussbaum, R. A., and P. K. Ducey (1988) Cytological evidence for the monophyly of the caecilians (Amphibia: Gymnophiona) of the Seychelles Archipelago. Herpetologica 44:290-296. Nussbaum, R. A., and H. Hinkel (1994) Revision of east African caecilians of the genera Afrocaecilia Taylor and Boulengerula Tornier (Amphibia: Gymnophiona: Caeciliidae). Copeia 1994:750-760. Nussbaum R. A., and M. E. Pfrender (1998) Revision of the African caecilian genus Schistometopum Parker (Amphibia: Gymnophiona: Caeciliidae). Miscellaneous Publications, Museum of Zoology, University of Michigan 187:1-32. Nussbaum, R. A., and M. Wilkinson (1989) On the classification and phytogeny of caecilians (Amphibia: Gymnophiona), a critical review. Herp. Monog. 3:1-42. Nussbaum, R. A., and M. Wilkinson (1995) A new genus of lungless tetrapod: a radically divergent caecilian (Amphibia: Gymnophiona). Proc. Royal Soc. Lond. Ser. B 261:331-335. O'Reilly, J. C. (1995) Comparative Studies on the Evolution of Prey Capture Behavior in Amphibians. M. S. Thesis. Northern Arizona University. O'Reilly, J. C. (1996) Keeping caecilians in captivity. Adv. Herpetocult. 1:39-45. O'Reilly, J. C , R. A. Nussbaum, and D. Boone (1996) A vertebrate with protrusible eyes. Nature 382:33. O'Reilly, J. C , D. A. Ritter, and D. R. Carrier (1997) Hydrostatic locomotion in a limbless tetrapod. Nature 386:269-272. Parker, H. W. (1956) Viviparous caecilians and amphibian phylogeny. Nature 178:250-252. Parker, H. W. (1958) Caecilians of the Seychelles Islands with description of a new subspecies. Copeia 1958:71-76. Parker, H. W., and E. R. Dunn (1964) Dentitional metamorphosis in the Amphibia. Copeia 1964:75-86. Reilly, S. M., and G. V. Lauder (1992) Morphology, behavior and evolution: comparative kinematics of aquatic feeding salamanders. Brain Behav. Evol. 40:182-196. Reiss, J. (1996) Palatal metamorphosis in basal caecilians (Amphibia: Gymnophiona) as evidence for lissamphibian monophyly. J. Herp. 30:27-39. Roth, G., K. C. Nishikawa, C. Naujoks-Manteuffel, A. Schmidt, and D. B. Wake (1993) Paedomorphosis and simplification in the nervous system of salamanders. Brain Behav. Evol. 42:137-70. Sanderson, I. (1937) Animal Treasure. Viking Press, New York. Sarasin, P., and F. Sarasin (1887-1890) Ergebnisse naturwissenschaftlichen Forschungen auf Ceylon in den Jahren 1884-1886. Zur Entwicklunggeschichte und Anatomie der Ceylonesischen Blindwuhle. Ichthyophis glutinosus. C. W. Kreidel, Weisbaden. Schmidt, A., and M. H. Wake (1990) Olfactory and vomeronasal systems of caecilians (Amphibia-Gymnophiona). J. Morph. 205:255268. Schwenk, K., and D. B. Wake (1993) Prey processing in Leurognathus marmoratus and the evolution of form and function in desmognathine salamanders (Plethodontidae). Biol. J. Linn. Soc. 49:141162. Shaffer, H. B., and G. V. Lauder (1985) Patterns of variation in aquatic ambystomatid salamanders: kinematics of the feeding mechanism. Evolution 39:83-92. Shaffer, H. B., and G. V. Lauder (1988) The ontogeny of functional design: metamorphosis of feeding behavior in the tiger salamander {Ambystoma tigrinum). J. Zool., Lond. 216:437-454. Summers, A. P., and J. C. O'Reilly (1997) A comparative study of the locomotor modes in two species of caecilian, Dermophis mexicanus
166
James C. O'Reilly
and Typhlonectes nutans (Amphibia: Gymnophiona). Zool. J. Linn. Soc. 121:65-76. Tanner K. (1971) Notizen zur pflege und zum verhalten einiger blindwiihlen (Amphibia: Gymnophiona). Salamandra 7 : 9 1 100. Taylor, E. H. (1968) The Caecilians of the World: A Taxonomic Review. Univ. of Kansas Press, Lawrence, KS. Taylor, E. H. (1969) Skulls of the Gymnophiona and their significance in the taxonomy group. Univ. Kansas Sci. Bull. 48:585-689. Taylor, E. H. (1970) The lateral-line system in the caecilian family Ichthyophiidae (Amphibia: Gymnophiona). Univ. Kansas Sci. Bull. 48:861-868. Van Damme, J., and P. Aerts (1997) Kinematics and functional morphology of aquatic feeding in Australian snake-necked turtles (Pleurodira: Chelodina). J. Morph. 233:113-125. Visser, M. H. C. (1963) The cranial morphology of Ichthyophis glutinosus (Linne) and Ichthyophis monochrous (Bleeker). Annale Universiteit van Stellenbosch, Ser. A 38:67-102. Wahnschaffe, U., B. Fritzsch, and W. Himstedt (1985) The fine structure of the lateral-line organs of larval Ichthyophis (Amphibia: Gymnophiona). J. Morph. 186:369-377. Wake, M. H. (1976) The development and replacement of teeth in viviparous caecilians. J. Morph. 148:33-63. Wake, M. H. (1977a) The reproductive biology of caecilians: an evolutionary perspective. Pp. 73-101. In: Reproductive Biology of Amphibians. E. H. Taylor and S. I. Guttman (eds.). Plenum, New York. Wake, M. H. (1977b) Fetal maintenance and its evolutionary significance in the amphibia: Gymnophiona. J. Herp. 11:379-386. Wake, M. H. (1978) Comments on the ontogeny of Typhlonectes obesus, particularly its dentition and feeding. Papeis Avulsos de Zoologia32:l-13. Wake, M. H. (1980a) Reproduction, growth and population structure of the Central American caecilian Dermophis mexicanus. Herpetologica 36:244-256. Wake, M. H. (1980b) Fetal tooth development and adult replacement in Dermophis mexicanus (Amphibia: Gymnophiona): fields versus clones. J. Morph. 166:203-216. Wake, M. H. (1985) The comparative morphology and evolution of the eyes of caecilians (Amphibia, Gymnophiona). Zoomorph. 105:277-295. Wake, M. H. (1987) A new genus of African caecilian (Amphibia: Gymnophiona). J. Herp. 21:6-15. Wake, M. H. (1989) Metamorphosis of the hyobranchial apparatus in Epicrionops (Amphibia, Gymnophiona, Rhinatrematidae): replacement of bone by cartilage. Ann. Sci. Nat. Zool. Biol. Anim. 10:171-182. Wake, M. H. (1992) Reproduction in caecilians. Pp. 112-120 In: Re-
productive Biology of South American Vertebrates. W. C. Hamlett (ed.). Springer-Verlag, New York. Wake, M. H. (1993a) Non-traditional characters in the assessment of caecilian phylogenetic relationships. Herp. Monog. 7:42-55. Wake, M. H. (1993b) The skull as a locomotor organ. Pp. 197-240 In: The Skull Vol. 3. J. Hanken and B. K. Hall (eds.). Univ. of Chicago Press, Chicago. Wake, M. H. (1993c) Evolution of oviductal gestation in amphibians. J. Exp. Zool. 266:394-413. Wake, M. H., and J. Hanken (1982) The development of the skull of Dermophis mexicanus (Amphibia: Gymnophiona), with comments on skull kinesis and amphibian relationships. J. Morph. 173:203223. Wake, M. H., and G. Z. Wurst (1979) Tooth crown morphology in caecilians (Amphibia: Gymnophiona). J. Morph. 159:331-340. Welsch, U., M. Miiller, and C. Schubert (1977) Elektronenmikroskopische und histochemische beobachtungen zur fortpflanzungs vivaprer gymnophionen (Chthonerpeton indistinctum). Zoologische Jahrbuch von Anatomie 97:532-549. Wever, E. G. (1975) The caecilian ear. J. Exp. Zool. 191:63-72. Wever, E. G., and C. Gans (1976) The caecilian ear: further observations. Proc. Natl. Acad. Sci. USA 73:3744-3746. Wiedersheim, R. (1879) Die Anatomie der Gymnophionen. Verlag Gustav Fischer, Jena. Wilkinson, M. (1989) On the status of Nectocaecilia fasciata Taylor, with a discussion of the phylogeny of the Typhlonectidae (Amphibia: Gymnophiona). Herpetologica 45:23-36. Wilkinson, M. (1991) Adult tooth crown morphology in the Typhlonectidae (Amphibia: Gymnophiona): a reinterpretation of variation and its significance. Zeitschrift fiir zoologische Systematische un,d Evolutions-forschung 29:304-311. Wilkinson, M. (1992a) The phylogenetic position of the Rhinatrematidae (Amphibia: Gymnophiona): evidence from the larval lateral line system. Amphibia-Reptilia 13:74-79. Wilkinson, M. (1992b) On the life history of the caecilian genus Uraeotyphlus (Amphibia: Gyrrmophiona). Herp. J. 2:121-124. Wilkinson, M. (1997) Characters, congruence and quality: a study of neuroanatomical and traditional data in caecilian phylogeny. Biol. Rev. 72:423-470. Wilkinson, M., and R. A. Nussbaum (1996) On the phylogenetic position of the Uraeotyphlidae (Amphibia: Gymnophiona). Copeia 1996:550-562. Wilkinson, M., and R. A. Nussbaum (1997) Comparative morphology and evolution of the lungless caecilian Atretochoana eiselti (Taylor) (Amphibia: Gymnophiona: Typhlonectidae). Biol. J. Linn. Soc. 62:39-109. Wilkinson, M., and R. A. Nussbaum (1998) Caecilian viviparity and amniote origins. J. Nat. Hist. 32:1403-1409.
S E C T I O N
III
^--\.-M^^^:
This Page Intentionally Left Blank
C H A P T E R
7 A Bibliography of Turtle Feeding KURT SCHWENK Department of Ecology and Evolutionary Biology University of Connecticut Storrs, Connecticut 06269
I. INTRODUCTION 11. BIBLIOGRAPHY
lonian Bauplan suggests some kind of uniformity. Our comparative ignorance of feeding function in turtles hampers our ability to discern general patterns in the history of tetrapod feeding. The fact that they are an ancient group that might retain ancestral amniote features in some aspects of their feeding system makes their further study all the more desirable.
I. INTRODUCTION The bibliography that follows is hardly exhaustive, but it does serve as an introduction to the literature on turtle feeding. General references on natural history and diet are not included, but a number of important papers on digestive physiology are. Also omitted are references on the systematics of turtles. The emphasis here is on the morphology of the feeding apparatus, feeding function, and behavioral observations of feeding in turtles. Historically, the functional morphology of feeding in turtles has been sadly neglected, but several recent contributions suggest that this is changing. In any case, turtle feeding is badly in need of investigation. One of the attractive features of turtles, as a group, is their diversity. Among other things, they span the range from fully terrestrial to fully aquatic, with everything in between. Thus, turtles, like salamanders, offer the opportunity to examine phenotypic changes in the feeding system associated with a change in the feeding medium (see Bramble and Wake, 1985; Lauder, 1985). Terrestrial turtles share with mammals and lepidosaurs a mobile, muscular tongue, but the intrinsic anatomy, biomechanics, and function of the turtle tongue are virtually unstudied. In general, there is a great deal of morphological diversity among turtle feeding systems that is largely unappreciated, perhaps because the che-
FEEDING (K. Schwenk, ed.)
II. BIBLIOGRAPHY Beisser, C J., J. Weisgram, and H. Splechtna (1995) Dorsal lingual epithelium of Platemys pallidipectoris (Pleurodira, Chelidae). J. Morph. 226:267-276. Beisser, C. J., J. Weisgram, H. Hilgers, and H. Splechtna (1998) Fine structure of the dorsal lingual epithelium of Trachemys scripta elegans (Chelonia: Emydidae). Anat. Rec. 250:127-135. Belkin, D. A., and C. Gans (1968) An unusual chelonian feeding niche. Ecology 49:768-769. Dels, V. L., and S. Renous (1991) Kinematics of feeding in two marine turtles {Chelonia mydas and Dermochelys coriacea). Pp. 73-78. In: Proceedings of the 6th Ordinary General Meeting of the Societas Europaea Herpetologica. Z. Korsos and I. Kiss (eds.). Hungarian Natural History Museum, Budapest. Bels, V. L., J. Davenport, and V. Delheusy (1997) Kinematic analysis of the feeding behavior in the box turtle Terrapene Carolina (L.), (Reptilia: Emydidae). J. Exp. Zool. 277:198-212. Bjorndal, K. A. (1980) Nutrition and grazing behavior of the green turtle Chelonia mydas. Mar. Biol. 56:147-154. Bjorndal, K. A. (1985) Nutritional ecology of sea turtles. Copeia 1985: 736-751. Bjorndal, K. A. (1986) Effect of solitary vs group feeding on intake in Pseudemys nelsoni. Copeia 1986:234-235. Bjorndal, K. A. (1987) Digestive efficiency in a temperate herbivorous reptile, Gopherus polyphemus. Copeia 1987:714-720. Bjorndal, K. A. (1989) Flexibility of digestive responses in two gen-
169
Copyright © 2000 by Academic Press. All rights of reproductior\ ir\ ar\y form reserved.
170
Kurt Schwenk
eralist herbivores, the tortoises Geochelone carbonaria and Geochelone denticulata. Oecologia 78:317-321. Bjomdal, K. A. (1990) Digestive processing in a herbivorous freshwater turtle: consequences of small-intestine fermentation. Physiol. Zool. 63:1232-1247. Bjorndal, K. A. (1990) Digestibility of the sponge Chondrilla nucula in the green turtle, Chelonia mydas. Bull. Mar. Sci. 47:567-570. Bjorndal, K. A. (1991) Digestive fermentation in green turtles, Chelonia mydas, feeding on algae. Bull. Mar. Sci. 48:166-171. Bjorndal, K. A. (1992) Body size and digestive efficiency in a herbivorous freshwater turtle: advantages of small bite size. Physiol. Zool. 65:1028-1039. Bjorndal, K. A. (1993) Digestive efficiencies in herbivorous and omnivorous freshwater turtles on plant diets: do herbivores have a nutritional advantage. Physiol. Zool. 66:384-395. Bjomdal, K. A. (1997) Fermentation in reptiles and amphibians. Pp. 199-230. In: Gastrointestinal Microbiology, Vol. 1. R. I. Mackie and B. A. White (eds.). Chapman and Hall, New York. Bjorndal, K. A. (1997) Foraging ecology and nutrition of sea turtles. Pp. 199-321. In: The Biology of Sea Turtles. P L. Lutz and J. A. Musick (eds.). CRC Press, Boca Raton, LA. Bramble, D. M. (1971) Functional Morphology, Evolution, and Paleoecology of Gopher Tortoises. Unpublished doctoral dissertation, Univ. of California, Berkeley. Bramble, D. M. (1973) Media dependent feeding in turtles. Am. Zool. 13:1342. [abstract] Bramble, D. M. (1974) Occurrence and significance of the Os transiliens in gopher turtles. Copeia 1974:102-109. Bramble, D. M. (1978) Functional analysis of underwater feeding in the snapping turtle. Am. Zool. 18:623. [abstract] Bramble, D. M. (1980) Feeding in tortoises and mammals: why so similar? Am. Zool. 20:931. [abstract] Bramble, D. M., and D. B. Wake (1985) Feeding mechanisms of lower tetrapods. Pp. 230-261. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Carmignani, M. P. A., and G. Zaccone (1975) Histochemical distribution of acid mucopolysaccarides in the tongue of reptiles. I. Chelonia {Pseudemys scripta Clark). Ann. Histochim. 20:77-88. Dalrymple, G. H. (1977) Intraspecific variation in the cranial feeding mechanism of turtles of the genus Trionyx (Reptilia, Testudines, Trionychidae). J. Herp. 11:255-285. Dalrymple, G. H. (1979) Packaging problems of head retraction in trionychid turtles. Copeia 1979:655-660. Davenport, J., M. Spikes, S. M. Thornton, and B. O'Kelly (1992) Crabeating in the diamondback terrapin Malaclemys terrapin: dealing with dangerous prey. J. Mar. Biol. Assoc. U. K. 72:835-848. Davenport, J., T. M. Wong, and J. East (1992) Feeding and digestion in the omnivorous estuarine turtle Batagur baska (Gray). Herp. J. 2:133-139. Drummond, H., and E. R. Gordon (1979) Luring in the neonate alligator snapping turtle {Macroclemys temminckii): description and experimental analysis. Z. Tierpsychol. 50:136-152. Edgeworth, F. H. (1935) The Cranial Muscles of Vertebrates. Cambridge Univ. Press, Cambridge. Fenchel, T. M., C. P McRoy, J. C. Ogden, P Parker, and W. E. Rainey (1979) Symbiotic cellulose degradation in green turtles, Chelonia mydas L. Appl. Environ. Microbiol. 37:348-350. Ferdinand, L. Prinz von Bayern (1884) Anatomic der Zunge. Fine vergleichend-anatomische Studie. Literarisch-Artistische Anstalt (Theodor Riedel), Munich. Fuchs, H. (1907) IJber das Hyobranchialskelett von Fmys lutaria und seine Entwicklung. Anat. Anz. 31:33-39. Fiirbringer, M. (1922) Das Zungenbein der Wirbeltiere insbesondere der Reptilien und Vogel. Abh. der Heidelberger Akad., math.naturw.Kl. 11:1-164.
Gaffney, E. S. (1972) An illustrated glossary of turtle skull nomenclature. Am. Mus. Novit. No. 2486:1-33. Gaffney, E. S. (1979) Comparative cranial morphology of Recent and fossil turtles. Bull. Am. Mus. Nat. Hist. 164(2): 67-376. George, J. C , and R. V. Shah (1954) The myology of the head and neck of the common Indian pond turtle, Lissemys punctata granosa Schoepff. J. Anim. Morphol. Physiol. 1:1-12. George, J. C , and R. V. Shah (1955) The myology of the head and neck of the Indian tortoise, Testudo elegans. J. Anim. Morphol. Physiol. 2:1-13. Gnanamuthu, C. P. (1937) Comparative study of the hyoid and tongue of some typical genera of reptiles. Proc. Zool. Soc. Lond. Ser.B 1937:1-63 Graper, L. (1932) Die das Zungenbein und die Zunge bewegenden Muskeln der Schildkroten I. Jena. Z. Naturwiss. 66:169-198. Graper, L. (1932) Die das Zungenbein und die Zunge bewegenden Muskeln der Schildkroten II. Jena. Z. Naturwiss. 66:274-280. Hailey, A. (1997) Digestive efficiency and gut morphology of omnivorous and herbivorous African tortoises. Can. J. Zool. 75:787794. Iwasaki, S.-I. (1992) Fine structure of the dorsal epithelium of the tongue of the freshwater turtle, Geoclemys reevesii (Chelonia, Emydinae). J. Morph. 211:125-135. Iwasaki, S.-I., T. Asami, Y. Asami, and K. Kobayashi (1992) Fine structure of the dorsal epithelium of the tongue of the of the Japanese terrapin, Clemmys japonica (Cheloia [sic], Emydinae). Arch. Histol.Cytol. 55:295-305. Iwasaki, S.-I., T. Asami, and C. Wanichanon (1996) Ultrastructural study of the dorsal lingual epithelium of the soft-shelled turtle, Trionyx cartilagineus (Cheloia [sic], Trionychidae). Anat. Rec. 246: 305-316. Iwasaki, S.-I., T. Asami, and C. Wanichanon (1996) Fine structure of the dorsal lingual epithelium of the juvenile hawksbill turtle, Eretmochelys imbricata bissa. Anat. Rec. 244:437-443. Iwasaki, S.-I., C. Wanichanon, and T. Asami (1996) Histological and ultrastructural study of the lingual epithelium of the juvenile Pacific ridley turtle, Lepidochelys olivacea (Chelonia, Cheloniidae). Ann. Anat. 178:243-250. Iwasaki, S.-I., C. Wanichanon, and T. Asami (1996) Ultrastructural study of the dorsal lingual epithelium of the Asian snail-eating turtle, Malayemys subtrijuga. Ann. Anat. 178:145-152. Kochva, E. (1978) Oral glands of Reptilia. Pp. 43-161. In: Biology of the Reptilia, Vol. 8. C. Gans and K. A. Cans (eds.). Academic Press, New York. Korte, G. E. (1980) Ultrastructure of the tastebuds of the red-eared turtle, Chrysemys scripta elegans. J. Morph. 163:231-252. Lakjer, T. (1926) Studien Uber die Trigeminus-versorgte Kaumuskulatur der Sauropsiden. C. A. Rietzel, Copenhagen. Lauder, G. V. (1985) Aquatic feeding in lower vertebrates. Pp. 210229. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Lauder, G. V., and T. Prendergast (1992) Kinematics of aquatic prey capture in the snapping turtle Chelydra serpentina. J. Exp. Biol. 164:55-78. Lee, M. S. Y. (1997) The evolution of beaks in reptiles: a proposed evolutionary constraint. Evol. Theor. 11:249-254. Legler, J. M. (1962) The Os transiliens in two species of tortoises, genus Gopher us. Herpetologica 18:68-69. Legler, J. M. (1976) Feeding habits of some Australian short-necked tortoises. Victorian Nat. 93:40-43. Legler, J. M. (1978) Observations on behavior and ecology in an Australian turtle, Chelodina expansa (Testudines: Chelidae). Can. J. Zool. 56:2449-2453. Legler, J. M. (1989) Diet and head size in Australian chelid turtles, genus Emydura. Ann. Soc. Roy. Zool. Belgique 119, Suppl. 1:1-10.
7. A Bibliography of Turtle F e e d i n g Legler, J. M. (1993) Family Chelidae. Pp. 142-152. In: Fauna of Australia, Vol. 2A. C. J. Glasby G. J. B. Ross, and P. L. Beesley (eds.). Australian Government Publ. Service, Canberra. Legler, J. M. (1993) Morphology and physiology of the Chelonia. Pp. 108-119. In: Fauna of Australia, Vol. 2A. C. J. Glasby G. J. B. Ross, and P. L. Beesley (eds.). Australian Government Publ. Service, Canberra. Lemell, P., and J. Weisgram (1997) Feeding patterns of Pelusios castaneus (Chelonia: Pleurodira). Neth. J. Zool. 47:429-441. Lubosch, W. (1933) Untersuchungen iiber die Visceral muskulatur der Sauropsiden. (Der Untersuchungen iiber die Kaumuskulatur der Wirbeltiere 3. Teil.). Gegenbaurs Morph. Jb. 72:584-666. Lubosch, W. (1938) Muskeln des Kopfes: Viscerale Muskulatur. Pp. 1011-1106. In: Handbuch der Vergleichenden Anatomie der Wirbeltiere, Vol. 5. L. Bolk, E. Goppert, E. Kallius, and W. Lubosch (eds.). Urban and Schwarzenberg, Berlin (1967 reprint, A. Asher and Co., Amsterdam). Marlow, R. W., and K. Tollestrup (1982) Mining and exploitation of natural mineral deposits by the desert tortoise, Gopherus agassizii. Anim. Behav. 30:475-478. Meyer, V., and L. Prutkin (1974) An ultrastructural study of the oral mucous membrane of the turtle, Pseudemys scripta elegans. Acta Anat. 89:89-99. Meylan, A. (1988) Spongivory in hawksbill turtles: a diet of glass. Science 239:393-395 Nalavalde, M. N., and A. T. Varute (1976) Histochemical studies on the mucins of the vertebrate tongue. VIII. Histochemical analysis of mucosubstances in the tongue of the turtle. Folia Histochem. Cytochem. 14:123-134. Owen, R. (1866) On the Anatomy of Vertebrates, Vol. 1. Longmans, Green, and Co., London. Parsons, T. S. (1968) Variation in the choanal structure of Recent turtles. Can. J. Zool. 46:1235-1263. Pevzner, R. A., and N. A. Tikhonova (1979) Fine structure of the taste buds of the Reptilia. I. Chelonia. Tsitologiya 21:132-138. Poglayen-Neuwall, I. (1953) Untersuchungen der Kiefermuskulatur und deren Innervation bei Schildkroten. Acta Zool. 34:241-291. Poglayen-Neuwall, I. (1953/54) Die Besonderheiten der Kiefermuskulatur von Dermochelys coriacea. Anat. Anz. 100:22-32. Pritchard, P. C. H. (1971) The leatherback or leathery turtle, Dermochelys coriacea. lUCN Monograph 1:1-39. Ray, C. E. (1959) A sesamoid bone in the jaw musculature of Gopherus polyphemus (Reptilia: Testudinidae). Anat. Anz. 107:85-91. Rhodin, A. G. J., F Medem, and R. A. Mittermeier (1981) The occurrence of neustophagia among podocnemine turtles. Br. J. Herp. 6:175-176. Romer, A. S. (1956) Osteology of the Reptiles. Univ. of Chicago Press, Chicago. Ruckes, H. (1937) The lateral arcades of certain emydids and testudinids. Herpetologica 1:97-103. Schumacher, G.-H. (1953/1954) Beitrage zur Kiefermuskulatur der Schildkroten. I. Mitteilung. Wiss. Z. Univ. Greifswald Math. Nat. 3:457-518. Schumacher, G.-H. (1954/1955) Beitrage zur Kiefermuskulatur der Schildkroten. II. Mitteilung. Wiss. Z. Univ. Greifswald Math. Nat. 4:501-518. Schumacher, G.-H. (1954/1955) Beitrage zur Kiefermuskulatur der Schildkroten. III. Mitteilung. Wiss. Z. Univ. Greifswald Math. Nat. 4:559-601. Schumacher, G.-H. (1956) Morphologische Studie zum Gleitmechanismus des. M. adductor mandibularis externus bei Schildkroten. Anat. Anz. 103:1-12. Schumacher, G.-H. (1956) Uber die Fascien des Kopfes der nebst einigen Bemerkungen zu der Arbeit von Tage Lakjer 1926. Zool. Anz. 156:35-54.
171
Schumacher, G.-H. (1973) Die Kopf- und Halsregion der Lederschildkrote. Anatomische Untersuchungen im Vergleich zu anderen rezenten Schildkroten. Abhandl. Akad. Wissensch. DDR, No. 2. Akademie-Verlag, Berlin. Schumacher, G.-H. (1973) The head muscles and hyolaryngeal skeleton of turtles and crocodilians. Pp. 101-199. In: Biology of the Reptilia, Vol. 4. C. Gans and T S. Parsons (eds.). Academic Press, New York. Secor, S. M., and J. Diamond (1999) Maintenance of digestive performance in the turtles Chelydra serpentina, Sternotherus odoratus, and Trachemys scripta. Copeia 1999:75-84. Sewertzoff, S. A. (1929) Zur Entwicklungsgeschichte der Zunge bei den Reptilien. Acta Zool. 10:231-341. Shah, R. V. (1963) The neck musculature of a cryptodire (Deirochelys) and a pleurodire (Ghelodina) compared. Bull. Mus. Comp. Zool. 129:343-368. Siebenrock, F (1898) Uber den Bau und die Entwicklung des Zungenbein-apparates der Schildkroten. Ann. Naturhist. Hofmus. Wien. 13:424-437. Smith, D. T. J. (1989) The Cranial Morphology of Fossil and Living Sea Turtles (Cheloniidae, Dermochelyidae and Desmatochelyidae). Unpublished doctoral dissertation, Kingston Polytechnic University, United Kingdom. Sondhi, K. C. (1958) The hyoid and associated structures in some Indian reptiles. Ann. Zool. Acad. Zool. (Lond.) 2:155-239. Spindel, E. L., J. L. Dobie, and D. F Buxton (1987) Functional mechanisms and histologic composition of the lingual appendage in the alligator snapping turtle, Macroclemys temmincki (Troost) (Testudines: Chelydridae). J. Morph. 194:287-301. Summers, A. P., K. F. Darouian, A. M. Richmond, and E. L. Brainerd (1998) Kinematics of aquatic and terrestrial prey capture in Terrapene Carolina, with implications for the evolution of feeding in cryptodire turtles. J. Exp. Zool. 281:280-287. Thompson, J. S. (1932) The anatomy of the tortoise. Sci. Proc. Roy. Soc. Dublin 20:359-461. Uchida, T (1989) Ultrastructural and histochemical studies on the taste buds in some reptiles. Arch Histol. Jap. 43:459-478. Van Damme, J., and P. Aerts (1997) Kinematics and functional morphology of aquatic feeding in Australian snake-necked turtles (Pleurodira; Chelodina). J. Morph. 233:113-125. Versluys, J. (1936) Kranium und Visceralskelett der Sauropsiden. 1. Reptilien. Pp. 699-808. In: Handbuch der Vergleichenden Anatomie der Wirbeltiere. Vol. 4. L. Bolk, E. Goppert, E. Kallius, and W. Lubosch (eds.). Urban and Schwarzenberg, Berlin (1967 reprint, A. Asher and Co., Amsterdam). Vogt, R. C , D. M. Sever, and G. Moreira (1998) Esophageal papillae in pelomedusid turtles. J. Herp. 32:279-282. Weisgram, J. (1985) Feeding mechanics of Claudius angustatus Cope 1865. Pp. 257-260. In: Functional Morphology in Vertebrates (Fortschr. Zool. Vol. 30). H.-R. Duncker and G. Fleischer (eds.). Gustav Fischer, Stuttgart. Weisgram, J., and H. Splechtna (1990) Intervertebral movability in the neck of two turtle species {Testudo hermanni hermanni, Pelomedusa subrufa). Zool. Jb. Anat. 120:425-431. Weisgram, J., and H. Splechtna (1992) Cervical movement during feeding in Chelodina novaeguinaeae (Chelonia, Pleurodira). Zool. Jb. Anat. 122:331-337. Weisgram, J., H. Ditrich, and H. Splechtna (1989) Comparative functional anatomical study of the oral cavity in two turtle species. Plzen. Lek. Sborn., Suppl. 59:117-122. Winokur, R. M. (1988) The buccopharyngeal mucosa of the turtles (Testudines). J. Morph. 196:33-52. Wocheslander, R., H. Hilgers, and J. Weisgram (1999) Feeding mechanism of Testudo hermanni boettgeri (Chelonia, Cryptodira). Neth. J. Zool. 49:1-13.
This Page Intentionally Left Blank
S E C T I O N
IV
(p/sBo^e SCUWEMK t ^ ^
This Page Intentionally Left Blank
C H A P T E R
8 Feeding in Lepidosaurs KURT SCHWENK
Department of Ecology and Evolutionary Biology University of Connecticut Storrs, Connecticut 06269
I. INTRODUCTION
I. INTRODUCTION 11. LEPIDOSAURIANPHYLOGENY AND CLASSIFICATION III. NATURAL HISTORY A. Diet B. Foraging Ecology C. Sensory Basis of Food Location and Identification IV. MORPHOLOGY OF THE FEEDING APPARATUS A. Skull and Mandible B. Dentition C. Hyobranchial Apparatus D. Jaw Musculature E. Tongue V. FEEDING FUNCTION A. Overview of Feeding B. Feeding Stages C. Feeding in Sphenodon D. Feeding in Iguania E. Feeding in Scleroglossa F. Biomechanics of Lingual Prey Capture G. Function of Cranial Kinesis VI. SPECIALIZED FEEDING SYSTEMS A. Chameleons B. Amphisbaenians C. Komodo Monitor D. Snakes VII. THE EVOLUTION OF FEEDING IN LEPIDOSAURS A. Evolution of Ingestion Mode B. Post-Ingestion Feeding Stages C. Evolution of the Gape Cycle D. Tongue Evolution E. Dietary Specialization F. Feeding Systems, Functional Units, and Evolutionary Constraint VIII. FUTURE DIRECTIONS References
FEEDING (K. Schwenk, ed.)
This chapter considers the structure, function, and evolution of the feeding system in nonophidian lepidosaurs—tuatara, lizards, and amphisbaenians. The latter two groups comprise, along with snakes, the squamate reptiles (Squamata). Although snakes are cladistically nested within squamates, their feeding systems have diverged sufficiently from other taxa to merit separate treatment (Chapter 9). They are, however, considered in this chapter in a general sense, as in the discussion of evolutionary patterns within Lepidosauria. Lepidosaurs offer a number of attributes that make them attractive subjects for study in the context of tetrapod feeding mechanisms. First, they are phylogenetically well positioned to be informative about evolutionary trends and patterns in the tetrapod clade. Perhaps more to the point is that many of them apparently retain a relatively primitive, or at least generalized, phenotype as compared to other living amniotes and so provide better structural analogues for reconstructing the ancestral-feeding mode. For example, most lepidosaurs retain an unspecialized, welldeveloped hyobranchial apparatus and a mobile, complexly muscled tongue. The latter trait they share with mammals and turtles and so we can infer its presence in the common anmiote ancestor. The muscular tongue is of considerable intrinsic interest. Squamate tongues haveprovided part of the empirical basis for the development of the muscular hydrostatic model of movement (e.g., Kier and Smith, 1985; Smith and Kier, 1989) and, due to their relatively more predictable
175
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
176
Kurt Schwenk
kinematics relative to mammal tongues, they are receiving increasing attention from modelers (e.g., Chiel et ah, 1992; van Leeuwen, 1997; van Leeuwen and de Groot, in preparation). Second, the feeding apparatus, particularly the tongue, is highly variable among lepidosaurs and so provides the grist for basic evolutionary studies, including phylogenetic analyses (e.g., Schwenk, 1988, and references therein), as well as studies of evolutionary pattern and process (e.g., Robinson, 1967; Schwenk, 1993b, 1995a,b; Bels et al, 1994; Kardong et al, 1997; Wagner and Schwenk, 2000). The lepidosaurian-feeding system is of particular interest in the latter case because there has been a fundamental shift in feeding mode within the group from tongue-based ingestion to jaw-based ingestion, corresponding cladistically to the origin of the Scleroglossa, one of two basal squamate clades (see later). The acquisition of jaw prehension, therefore, provides a model system for studying transitions in complex, functionally integrated systems (Schwenk, 1993b, 1995a, 2000b; Wagner and Schwenk, 2000). A related point is that the squamate tongue subserves two fundamentally different functions: feeding and vomeronasal chemoreception. A biomechanical consideration of tongue design suggests that a tongue phenotypically optimized for feeding is a poor chemoreceptor, whereas a tongue optimized for chemoreception is poorly designed for feeding function (Schwenk, 1993b, 1995a; Wagner and Schwenk, 1999). Hence, there is an evolutionary tension between these two functions that plays out in the evolution of tongue form. Clade-specific solutions to the "dilemma" posed by this evolutionary trade-off provide insight into underlying processes of phenotypic evolution (Schwenk, 1995a, 2000b, in preparation; Wagner and Schwenk, 2000).
IL LEPIDOSAURIAN PHYLOGENY A N D CLASSIFICATION Lepidosauria is a diverse clade of reptiles comprising approximately 7150 species of tuatara, lizards, snakes, and amphisbaenians (Pough et al, 1998). It is the sister group of Archosauria, which includes crocodilians, birds, and various extinct diapsid reptiles, such as the dinosaurs (Gauthier et al, 1988), or of turtles (Testudines) plus archosaurs (e.g.. Hedges and Poling, 1999; Kumazawa and Nishida, 1999) (Fig. 8.1). Lepidosauria is further divided into the Rhynchocephalia and Squamata, the former containing two species of tuatara, genus Sphenodon (Daugherty et al, 1990), and the latter, all remaining lepidosaurian species. Squamates are, themselves, divided into two basal clades.
the Iguania and Scleroglossa, and these, in turn, are subdivided into several suprafamilial groups (Fig. 8.1). A molecular study of nuclear and mitochondrial gene sequences suggested that Sphenodon is more closely related to archosaurs (+ turtles) than to squamates, thus splitting the Lepidosauria as presently construed (Hedges and Poling, 1999). Such a phylogenetic hypothesis is extremely unlikely in the face of morphological data. It would deny the 35 morphological synapomorphies identified by Gauthier et al (1988) uniting Sphenodon and Squamata relative to all other amniotes. Furthermore, one would find it very difficult to identify morphological synapomorphies uniting tuatara (+ fossil sphenodontids) with a clade including crocodilians, birds, and turtles. Morphology overwhelmingly supports a monophyletic Lepidosauria. Lepidosauria is an ancient group with fossil lizards known from the Upper Permian (approximately 250 mybp) and sphenodontids from the Triassic (Estes, 1983; Carroll, 1988b). Many Late Cretaceous (approximately 7b mybp) fossil species are assignable to modern families and some Late Jurassic (135+ mybp) taxa are recognizable as varanoids related to living monitor lizards and snakes (Estes, 1983). In traditional classifications, Lepidosauria is accorded the rank of subclass with Rhynchocephalia and Squamata as orders within it (e.g., Romer, 1956). Lizards (Lacertilia or Sauria), snakes (Serpentes or Ophidia), and amphisbaenians (Amphisbaenia) are given equal categorical ranking as suborders within the order Squamata, despite the fact that relationships among these groups remain poorly understood. Amphisbaenians were historically regarded as a family of "lizards" and called Amphisbaenidae in suit (e.g.. Camp, 1923), but later work suggested that these unusual, fossorial squamates were quite distinct from "typical" lizards and deserving of subordinal status in equality with lizards and snakes (e.g.. Cans, 1978; Crook and Parsons, 1980; Bellairs and Cans, 1982). Needless to say, such traditional classifications were crafted by workers in the context of "evolutionary taxonomy" rather than phylogenetic (cladistic) systematics, thus they were not overly concerned that the classification mirror phylogenetic relationships among the groups. Rather, recognition of Serpentes, Lacertilia, and Amphisbaenia as separate but equal ranks within Squamata calls attention to the relative morphological distinctness of each taxon. Indeed, this distinctness has been the basis for much of the ambiguity regarding the phylogenetic position of snakes and amphisbaenians relative to lizards. It has sometimes been suggested that each of these taxa is completely outside the others, i.e., monophyletic (e.g., Hoffstetter, 1968; Rieppel, 1978c, 1983; see Rieppel, 1988, for a review), but most
m IGUANIAI :::::!$:
% 1
a> C0
o a>
I
"D
P
®
5>
•«••••• • ,•* ,
•• , • • •
• ••
w»
••-
y**
^•*
• close
transport
protrude
,.—•••—T^».
••••-"^
•' >^*••-
LU
=)
CD
Z
O
; ; I
rvpv nontari —__ K^Jt iia\~ti ""^•~—~
1 fjtvj
full contact —
III
0.0
1
0.1
Ill
0.2
',
i i 1 1
> ••
—»
^ ^!^'
; ; —r^j • iiiii
0.3
i
0.4
EXTRA-ORAL
—« ««'**^_
^w-~
• *• •• «•
•
INTRA-ORAL
t
^
retract III
0.5
iiii
0.6
0.7
0.8
0.9
1
1
1.0
I.Hill
1.1
1.2
1.3
1.4
^
1.5
TIME (SEC) F I G U R E 8.17. Jaw and tongue kinematics during lingual ingestion and intraoral transport in Sphenodon punctatus. Solid vertical lines demarcate the phases of each gape cycle. A discrete SO II is only evident in the transport cycle. Note that the tongue makes contact with the prey item over an extended period of time as the prey is pushed into the papillary surface. As the prey was pinned to the substrate, it pushed the mandible down, as indicated by a momentary increase in gape angle (at second dotted line). A stationary phase intervenes between gape cycles. Based on high-speed (300 fps) 16-mm film (Schwenk, Frazzetta and Jenkins, in preparation). FC, fastclose; FO, fastopen; SC-PS, slow close-power stroke; SO, slowopen.
is initiated to control the more plastic behavior preceding a strike. However, CPG control of cyclicity does not imply immutability of the resulting kinematic patterns, which are found to vary among species, among stages, and within a stage according to food type and state of reduction. As such, there is circumstantial evidence for the modulation of feeding kinematics based on sensory feedback (see Chapter 2). During each feeding stage the jaws are repeatedly opened and closed. A single open-close sequence is known as a gape cycle and each feeding stage comprises one to many gape cycles. A covcvpXeie feeding sequence, or hout, includes several feeding stages, from ingestion to swallowing. A "model gape cycle" was proposed by Bramble and Wake (1985) based largely on kinematic patterns evident in lepidosaurs and other nonmammalian tetrapods (see Fig. 2.14 and discussion in Chapter 2). Its generality remains in dispute, but it at least provides a starting point for comparing gape cycles among species and among feeding stages. A single
gape cycle potentially comprises several phases determined by changes in the velocity of the jaws as they open and close (Fig. 8.17): slow open I (SO I), slow open II (SO II), fast open (FO), fast close (FC) and slow close (SC) or slow close power stroke (SC-PS). There is sometimes an intervening period between cycles known as a stationary phase. Gape cycles are reviewed in Chapter 2 and further discussed in Section VII,C.
B.
Feeding Stages
1. Prey Capture and Ingestion Prey capture is the apprehension and subjugation of a prey item and ingestion is its movement from the environment into the oral cavity (Chapter 2). In nearly all nonophidian lepidosaurs, prey capture (or prehension) and ingestion are accomplished by the mouth and combined into a single stage so that capture
8. Feeding in Lepidosaurs involves both the apprehension of a prey item and its immediate delivery into the mouth. Rare exceptions include the Komodo monitor {V. komodoensis) and, occasionally, amphisbaenians, whose unusual feeding behaviors are discussed in Section VI. Once captured, small prey are either transported and swallowed immediately or are killed by biting, shaking, or crushing against the substrate. In some venomous snakes, subjugation is further decoupled from prey capture. Prey are bitten, envenomated, and released to be located later after they are dead. Only then does ingestion begin. Snakes differ from other lepidosaurs in most aspects of their feeding and the traditional feeding stages used here do not clearly apply (see Chapter 9). A fundamental dichotomy is evident in the pattern
F I G U R E 8.18. Consecutive cine frames (48 fps) of a small agamid {Phrynocephalus helioscopus) during lingual ingestion. Note the conformation of the tongue during protrusion and the initial bite following retraction of the mealworm into the mouth.
223
of ingestion of small food items (see Sections IV,E and VII,A): Sphenodon and all iguanians use the tongue as a prehensile organ to apprehend small prey and draw it into the mouth, whereas all scleroglossans use jaws and teeth (Schwenk, 1988; Schwenk and Throckmorton, 1989; Wagner and Schwenk, 1999) (Fig. 2.16 in Chapter 2). As prey size increases the dichotomy fades and jaw prehension is almost universally employed, although taxonomic differences remain. A few scleroglossan species use a type of lingual prehension of small prey items in certain circumstances. These exceptions are discussed later. Lingual ingestion involves protraction and protrusion of the tongue concomitant with hyobranchial protraction, with the tongue tip ventrally curled so that the dorsal, papillary surface of the tongue is presented toward the prey item (Figs. 8.18 and 8.19). Tongueprey contact usually occurs on the anterior third of the tongue where papillae are longest in iguanians (Fig. 8.13A). One advantage of lingual prehension is that the tongue is protruded at the same time the head advances toward the prey item, thus their approach velocities are summed and prehension is more rapid than would be possible with the jaws alone. Tongue-prey contact can be relatively light, but most often involves a forceful impact that pushes the prey item against the substrate, fitting the tongue to the prey surface (see later). Adhesion is remarkably effective and involves a combination of interlocking and wet adhesion; in chameleons it may also involve suction (Schwenk, 1983; see Section VI,A). Retraction occurs almost
F I G U R E 8.19. Lingual prehension in an agamid Hzard (Pogona barbata) based on a 35-mm photograph. Note how the dorsal, papillose surface is curled around the end of the protruded, muscular tongue. The right anterior process of the basihyal is evident posteriorly, bulging as the tongue is protracted. The larynx moves forward with the tongue, presumably pulled by the laryngohyoid ligament. Sublingual plicae containing salivary glands form the sides of a well in which the tongue sits at rest.
224
Kurt Schwenk
FIGURE 8.20. Jaw prehension in a scleroglossan lizard (Varanus niloticus). Prey capture is with the tips of the jaws (more pronounced with smaller prey). Note complete retraction of the tongue. The apparatus on the lizard's head is a goniometer designed to measure mesokinetic flexion and the arrow in the upper left indicates the position of the gage (see Fig. 8.25). From Condon (1987), Exp. Biol 47, 73-87, © Springer-Verlag, with permission.
immediately following prey contact and is extremely rapid. The prey item comes to lie within the mouth on the surface of the tongue or between the tooth rows for transport or swallowing. Jaw prehension of prey involves grasping of the prey in the jaws with the anterior teeth (Fig. 8.20). Sometimes the side of the mouth is used, but grasping usually occurs with the jaw tips. Prey are moved into the mouth for further processing by the tongue or an inertial toss. Larger prey are usually pinned or dragged against the substrate or are subdued by violent shaking before subsequent feeding stages commence. 2. Processing Processing refers to the mechanical reduction or modification of food within the mouth before it is swallowed (Chapter 2). If a prey item is very small it is often transported and swallowed directly without processing. Conversely, large prey consumed by inertial feed-
ers (e.g., Varanus) are often oriented in the mouth and swallowed whole, without processing (although usually after subjugation or killing by biting or suffocating). However, in most cases some processing follows immediately upon ingestion. After ingestion a food item often comes to lie across a tooth row where it can be bitten immediately (Fig. 8.18), but sometimes it must be positioned between the teeth or shifted to a more advantageous bite point. Such manipulative cycles are usually mediated by lingual movement, including tongue twisting, but they are sometimes accomplished inertially with a lateral jerk of the head. Repositioning is also used to orient large prey items head first so that limbs, scales, or fur fold posteriorly, facilitating prey movement into the esophagus. Reduction usually involves repeated bites so that the teeth (or occasionally the palate) pierce, crush, or shear the prey item as it is coated in saliva, rendering it softened but usually whole. Most lepidosaurs are well supplied with oral salivary glands that produce a variety of secretory products (Gabe and Saint Girons, 1969; Kochva, 1978), but there is no indication thus far that these compounds initiate chemical digestion or do anything more than lubricate the bolus for swallowing (with the exception of venom glands in Heloderma, which are used primarily for defense). Such chewing behavior is called puncture-crushing in contrast to mammalian-type mastication in which a food item is reduced to tiny particles. As such, comminution of food is not a necessary outcome of most lepidosaurian chewing, although it sometimes occurs, whereas it is the raison d'etre of most mammalian mastication. This difference has a profound effect on the form and evolution of the feeding system in the two groups. Transverse movements of the mandible relative to the upper jaw and a "masticatory orbit" (e.g., Hiiemae et al, 1978; Chapter 13) are virtually never observed in nonophidian lepidosaurs and remain unique features of the mammalian masticatory mechanism (e.g., Davis, 1961; Throckmorton, 1980). Despite the absence of a masticatory orbit as in mammals, it is an unsung fact that many, if not most, lepidosaurs chew on one side at a time and often alternate between sides. Side switching is particularly evident in Sphenodon (Gorniak et al, 1982; Schwenk et al, manuscript in preparation) and iguanians, especially acrodonts (Schwenk and Throckmorton, 1989; personal observation), but it is also observed in scleroglossans (e.g.. Smith, 1984). Chewing asymmetry and the intraoral manipulation it requires have been regarded as exclusively mammalian traits associated with the presence of mobile, muscular cheeks and lips, and the absence of a lingual process within the tongue (e.g., Davis, 1961), thus its ubiquitous occurrence in
8. Feeding in Lepidosaurs lepidosaurs has significance for interpreting the evolution of mammalian mastication (see Throckmorton, 1976; Crompton, 1989). Chewing asymmetry also has implications for muscle activity patterns on active versus balancing sides, mandibular bending, the transmission of forces from one side to the other, the nature of joint reaction forces, and the position of the bite point (e.g., Druzinsky and Greaves, 1979; Gorniak et ah, 1982; Greaves, 1995). These factors have only rarely been considered in lepidosaur feeding. Chewing and intraoral transport cycles tend to be kinematically similar and are often not separated in functional analyses. The basis for such similarity is discussed in the following section. 3. Intraoral
Transport
Intraoral transport is the movement of food through the oral cavity to the pharynx for swallowing (Chapter 2). In many ways, intraoral transport is the "purest'' form of gape cycle in which the cyclical, coordinated movements of the jaws, tongue and hyobranchium are most apparent, unsullied by the mechanistic vagaries of prehension and reduction. Possibly for these reasons. Bramble and Wake (1985) based their model gape cycle on intraoral transport and suggested that it might represent an ancestral tetrapod pattern (see Chapter 2). However, Bramble and Wake (1985) also included "chewing'' or processing cycles here as well. In the vast majority of lepidosaurs, intraoral transport is accomplished by cyclical movements of the tongue and hyobranchial apparatus in coordination with the jaws (Bramble and Wake, 1985). Hyolingual transport involves anterior movement of the tongue and hyobranchial apparatus underneath the food item during SO I while it is held against the roof of the mouth. The tongue is usually deformed at the end this process (SO II) so that the bolus comes to lie within a depression in its surface, or the tongue is elevated in front of the bolus. Although the prey item is often oriented transversely across the tooth row in chewing cycles, in pure transport cycles it is oriented longitudinally on the tongue to clear the teeth and corners of the mouth (e.g., Delheusy and Bels, 1992). SO II may also be used to integrate sensory information on food position and condition to modulate the next cycle (Bramble and Wake, 1985). As the jaws are parted during FO, the bolus is freed from contact with the palate or teeth, and the hyobranchium and tongue rapidly shift posteriorly so that at jaw closure (FC and SC) the prey item has come to lie farther back in the mouth where it is once again fixed against the palate by the tongue. The tongue repeats its anterior movement beneath the prey item in preparation for the next cycle. According to
225
Bramble and Wake (1985), the presence of SC, especially with high levels of adductor activity indicating a power stroke (SC-PS), is contingent on the teeth engaging the prey item at the end of FC. They thus describe a chewing cycle as construed here, rather than a "pure" transport cycle. Indeed, an SC phase is variably present during putative transport cycles (see later). However, an SC phase is sometimes apparent even when contact of the marginal teeth with the prey item does not occur. This may represent controlled slowing of jaw closure or contact of the prey item with the palate as the jaws engage. If the latter, the presence of an SC phase during intraoral transport should be related to bolus size and condition, implying that SC should be more evident with large prey items and in earlier cycles. In some taxa, head jerking and prey inertia are substituted for hyolingual movement (Gans, 1969b). During inertial transport the prey item is momentarily released while the animal rapidly shifts its head anteriorly. The prey item thus comes to lie farther back in the mouth as the jaws close. For smaller prey, head and jaw movements are imparted to the food item itself so that it moves back relative to the ground, whereas a larger prey item remains more or less stationary while the head moves forward over it. Among lepidosaurs, only varanids are obligate inertial feeders due to extreme reduction of the tongue for chemoreception and its limited participation in feeding (e.g.. Smith, 1986; Condon, 1987; Elias et al, 2000). Other large, carnivorous lizards (e.g., Tupinambis) frequently employ inertial feeding as well (e.g., MacLean, 1974; Smith, 1984; McBrayer and White, manuscript in preparation), but these also use hyolingual transport some or most of the time (e.g., MacLean, 1974; personal observation). Inertial transport is occasionally used by other taxa, particularly when feeding on relatively large prey items (e.g., Sphenodon, see later), but small species, such as gymnopthalmids, eating relatively small prey also use it (e.g., MacLean, 1974). It is possible that the frequency of inertial feeding by lepidosaurs has been overestimated in the literature. This is because virtually all functional analyses have been undertaken in the laboratory with captive animals often fed unnatural diets (see Chapter 1). Mice are frequently used in feeding trials with larger lepidosaur species, but such vertebrate prey are rarely eaten in the wild (see Section III,A). The vast majority of prey items for such species are small invertebrates that are more likely to be manipulated with hyolingual movements. Intraoral transport and chewing cycles often overlap temporally and are difficult to distinguish kinematically (Bels et al, 1994; see later). In most taxa, they grade one into the other so that many gape cycles are a combination of the two. A pure chewing cycle would
226
Kurt Schwenk
involve no posterior movement of the prey item during the gape cycle and a pure transport cycle would involve no crushing of the prey by the teeth. However, in most cases, chewing cycles include hyolingual movement of the prey item to place it between the teeth for crushing at the end of each cycle. There is cyclical anteroposterior movement, but no net posterior translation, or some posterior movement of the prey item along the tooth row occurs between bites. Thus jaw and hyolingual movements during chewing cycles are likely to mimic transport cycles kinematically in most taxa. For these reasons, Schwenk and Throckmorton (1989) followed Bramble and Wake in suggesting that processing should be considered a type, or subset, of intraoral transport, a conclusion in which Bels et al (1994) concurred. However, functional work has shown that there may be important quantitative (e.g., Herrel 1997a; Smith, 1982; Cans et al, 1985; Cans and De Vree, 1986) and qualitative (e.g.. Smith, 1982,1984; Condon, 1987; So et al, 1992; Herrel et al, 1997b) differences between the two cycle types (see later). Given that chewing is a facultative behavior and that chewing and transport have different mechanical outcomes and can be kinematically decoupled (as in chameleons), it seems worth preserving the distinction between chewing and intraoral transport stages here, and in future functional analyses, insofar as possible. 4. Pharyngeal Emptying
(Swallowing)
Swallowing refers to the movement of food into the esophagus, where peristalsis takes over the process of transporting the bolus to the gut (see Chapter 2). In lepidosaurs, swallowing potentially comprises two separate substages: pharyngeal packing and pharyngeal compression. Although these stages serve the same function as deglutition in mammals, swallowing is so different mechanistically in the two groups that it is unlikely to be homologous in any meaningful sense (see Smith, 1992; Bramble and Wake, 1985; see Chapters 2 and 13). Pharyngeal packing occurs universally among nonophidian lepidosaurs, but not pharyngeal compression, which is most common in scleroglossans. Occurrence of the latter stage in some species may be facultative, presumably depending on bolus size and characteristics (personal observation). It is often difficult to observe, sometimes only evident as a single, brief compressive cycle without an accompanying gape cycle. Because of confusion about the mechanics and nomenclature of swallowing in lepidosaurs, it is usually not clear whether the failure to mention pharyngeal compression in literature accounts of swallowing reflects its actual absence or merely the failure to observe or document it. To complicate matters further.
terminal intraoral transport cycles sometimes grade into pharyngeal packing cycles so that kinematic distinctions between transport and swallowing stages can be blurred during the transition (see later). In part because of these ambiguities, the literature on lepidosaurian swallowing is inconsistent and sometimes muddled. Some authors do not consider pharyngeal compression at all, note it briefly in passing, or consider it "aberrant" (Herrel et al, 1997b:387). Bramble and Wake (1985), who set the tone for nearly all lepidosaurian feeding studies to date, considered pharyngeal compression to be exceptional and limited to specialized inertial feeders, such as varanids and large teiids. In some studies, the discussion of pharyngeal packing is included under intraoral transport, and a separate stage known as "cleaning" is identified, although cleaning is actually a part of pharyngeal packing (see later). The literature is also inconsistent in its application of technical terms. For example. Smith (1984) equated "swallowing" with pharyngeal compression and considered pharyngeal packing a separate, preswallowing stage, whereas Herrel et al (1999b) equated "swallowing" with pharyngeal packing and considered pharyngeal compression a postswallowing stage! Some studies (e.g., Urbani and Bels, 1995) erroneously apply the term "deglutition" to swallowing behavior in lizards, usually in the context of pharyngeal packing [see Smith (1994) and Chapter 2]. These problems make extraction of data for a comparative synthesis difficult. I have proposed a standard nomenclature here that should facilitate comparative studies. Future functional analyses should take care to watch for pharyngeal compression following pharyngeal packing and to distinguish these two components of swallowing when they occur. a. Pharyngeal Packing Pharyngeal packing is characteristic of lepidosaurs and possibly some turtles (Chapter 2). During this stage the bolus is pushed into the pharynx and the anterior part of the esophagus (Smith, 1984, 1992; Bramble and Wake, 1989). This is almost always accomplished by the tongue and may be the principal function of the tongue's posterior limbs, which serve to tamp the bolus into the pharynx. Typically (but not always) during pharyngeal packing cycles the tongue is protruded out of the mouth tip first and often appears to "lick" the labial scales and snout. A "cleaning" or "lip-licking" function is often attributed to this behavior (e.g., Throckmorton, 1980; Cans et al, 1985; Bels and Baltus, 1988; Goosse and Bels, 1992), but in most cases such tongue protrusion results from lingual positioning for asymmetrical tamping by the posterior limbs
227
8. Feeding in Lepidosaurs and probably has nothing to do with cleaning the labial scales (Smith, 1984; personal observation). The extent of tongue protrusion varies depending on the size of the bolus and its position relative to the esophagus. Packing cycles with tongue protrusion push the bolus farther back into the pharynx than cycles in which protrusion does not occur (Delheusy and Bels, 1992). Although tip-first tongue protrusion clearly distinguishes some pharyngeal packing cycles from transport cycles, early packing cycles can blend temporally and kinematically with preceding intraoral transport cycles. Both transport and packing involve hyolingual movement of the bolus posteriorly, but they differ in the position of the bolus relative to the tongue; in the former it lies on the tongue, whereas in the latter it is posterior to it. Other kinematic variables distinguish them as well, but the two phases may be largely overlapping in quantitative attributes, at least in some taxa (e.g., Herrel et al, 1995). Smith (1984), however, found the cycle types to be "distinct" in their patterns of tongue-jaw coordination and the shapes of the orbits described by tongue and hyobranchium. In packing, the hindtongue moves upward and backward during SO rather than upward and forward, and the tongue orbits become elongate and largely anteroposterior. Gape profiles become small and spiked, losing their differentiation into discrete phases. The extent to which these disagreements reflect phylogenetic differences, experimental conditions, or criteria for identifying cycle types in the first place remains to be determined. The extent to which pharyngeal packing moves food into the esophagus must be variable. Smith (1984) found that during pharyngeal packing, food collected in the pharynx only and pharyngeal compression was necessary to squeeze the bolus into the esophagus (swallowing sensu stricto). In contrast, Herrel et ah (1996a) observed that packing moved food into the esophagus as well. Because pharyngeal compression often does not occur, pharyngeal packing must be sufficient in many cases to move food fully into the esophagus to initiate peristaltic transport. Undoubtedly, bolus size and condition influence swallowing behavior, as well as taxonomic differences.
tive in taxa with posterior limbs of the tongue reduced or missing (see Section VII,B). Lepidosaurs and all other nonmammalian vertebrates lack the pharyngeal musculature characteristic of mammals (Smith, 1992; see Chapters 2 and 13) and therefore cannot compress the pharyngeal cavity internally to squeeze the bolus into the esophagus. Thus, if pharyngeal packing fails to tamp food far enough into the esophagus to initiate peristalsis, lepidosaurs resort to external compression of the pharynx by means of cervical flexure and/or contraction of the constrictor colli muscle. The constrictor colli forms a sling around the pharynx and its contraction elevates the hyobranchium and constricts the gular region. Smith (1984) noted that in order for compression to succeed, the bolus must lie behind the basihyal so that it is squeezed posteriorly and not returned into the mouth. It is conceivable that anterior and posterior intermandibularis muscles join the constrictor colli during swallowing so that their serial contraction creates a compressive wave along the pharynx from front to back, squeezing it like a tube of toothpaste, but this is purely speculative. If the bolus is posterior enough, simple constriction of the pharynx would suffice. Neck flexure also is used to compress the pharynx in some species, alone or in addition to constrictor colli constriction. Its pattern and extent vary among taxa, but it can involve ventral bending (head tucking), lateral bending, or both. In short-necked forms the bending appears to be simple flexure at the atlanto-occipital joint, but in elongate forms, lateral neck bending is sinuous and sinusoidal, i.e., it appears that a propagated wave is generated rather than simple flexure (personal observation). Sinusoidal movements continue into the trunk in some species. It is possible that internal concertina movements also aid in pharyngeal emptying in long-necked forms, especially anguimorphans, as they do in some snakes (Kley and Brainerd, 1996; N. Kley, personal communication). Such movements involve sinusoidal bending of the body axis internally, including vertebral column and esophagus, independent of the outer body wall (see suggestive Xray photo of Varanus, Fig. 89 in Greer, 1989).
h. Pharyngeal Compression Accumulating evidence indicates that pharyngeal compression is widespread among lepidosaurs (and other reptiles) and is not restricted to inertial feeders, as suggested by Bramble and Wake (1985). It occurs in both iguanian and scleroglossan squamates and although undescribed, it may be present in Sphenodon as well. It is more common or pronounced in scleroglossans, which accords well with their generally reduced hindtongues—pharyngeal packing may be less effec-
C. Feeding in
Sphenodon
1. Ingestion Lingual ingestion in tuatara was noted anecdotally by earlier authors (e.g., BuUer, 1878; Dawbin, 1962; Farlow, 1975) and has been analyzed by Gorniak et al (1982) and Schwenk et al. (manuscript in preparation). Gorniak et al. (1982) found that lingual prey prehension was used when feeding on small prey (insects).
228
Kurt Schwenk
but that large prey items (mice) were grasped directly by the jaws. Ingestion is usually initiated by prey movement (see earlier discussion). Movement elicits head cocking and visual fixation of the prey item. The head is moved toward the prey, usually forward and downward toward the substrate, but prey suspended by a thread above the head are also easily taken (Schwenk et ah, manuscript in preparation). As the head moves closer, the jaws begin to part as the tongue is protracted. Cineradiography shows that tongue protraction is caused by hyobranchial protraction (Gorniak et ah, 1982; Schwenk et al, manuscript in preparation). There is no evidence of hydrostatic elongation or sliding of the tongue along the lingual process. The tongue is slightly elevated initially and extended tip first. The initial tip-first orientation of the tongue differs from iguanians (see later). Immediately upon crossing the mandibular symphysis the tongue tip curls ventrally and is fixed to the mandible so that as the tongue continues to be pushed out of the mouth by the hyobranchium it becomes arched with its dorsal surface presented anteriorly toward the prey item. This kinematic pattern is consistent with the fact that anterior genioglossus medialis fibers run into the foretongue to the tongue tip; contraction of these fibers, or simply their inextensibility, would force the tongue tip downward and anchor it to the mandible (Schwenk, 1986; see Section V,F). As the tongue is protruded the larynx advances correspondingly (Fig. 4 in Gorniak et ah, 1982). This suggests that the laryngohyoid ligament is inelastic and when the tongue is protracted it pulls the larynx with it (see earlier discussion). Prey contact occurs on the dorsal surface of the foretongue, which is now directed anteroventrally. The zone of contact corresponds to the region of the tongue where filamentous papillae are longest (Schwenk, 1986). Differences in papillary height and orientation across the width of the tongue create a modest lingual sulcus in the midline. The sulcus may enhance the number of papillae making contact with the prey item. The prey item is usually only lightly contacted by the tongue surface, but this is sufficient to cause adhesion. In some cases, however, contact is forceful enough to push the mandible slightly downward (Fig. 8.17). Tongue-prey contact triggers tongue retraction, but the jaws continue to open as the tongue and adherent prey are rapidly withdrawn into the mouth. As tongue and prey clear the mandible, the jaws are snapped close. Tongue retraction is coupled to hyobranchial retraction. Ingestion of mice is accomplished by biting and lifting with the anterior teeth. According to Gorniak et ah (1982:337), "The tongue plays no role during capture nor is it protruded." Unfortunately, Gorniak and col-
leagues do not indicate the nature of tongue movement and conformation within the mouth during jaw prehension. It is possible either that the tongue is fully retracted to avoid prey contact, as in scleroglossans, or that it curls and contacts the prey item within the mouth concomitant with jaw prehension, as in iguanians (Schwenk and Throckmorton, 1989). Feeding studies on intermediate prey sizes are necessary. These would help determine whether the transition from tongue to jaw prehension is a graded response, as in iguanians, or a threshold response. Lingual ingestion gape cycles are highly variable in duration and kinematic pattern in a single individual (Schwenk et ah, manuscript in preparation). Most cycles roughly conform to the Bramble-Wake model and are similar to the iguanian pattern (Fig. 8.17). The tongue and hyobranchium are protracted during SO I and SO II, then retracted during FO and FC. SO II is characterized by a very high gape angle, perhaps more so than in iguanians (Schwenk and Throckmorton, 1989; see later), presumably to allow clearance for the tongue and prey as they are retracted. Thus the difference in gape between SO II and maximum gape at the end of FO is usually quite small. Herrel et ah (1995) suggested that Sphenodon exhibits no distinct SO, but this is not accurate. There is a tendency for SO I and SO II to be poorly differentiated and for SO II to blur into FO so that a distinct "plateau phase" (Schwenk and Throckmorton, 1989) during SO II is less frequently evident in Sphenodon than in iguanians. Nevertheless, an SO phase is evident in many ingestion cycles (Schwenk et ah, manuscript in preparation). A typical ingestion sequence is slow relative to most squamates (0.5 to 1.0 sec); however, tuatara are relatively large and this may be an effect of body size. Although most investigators dutifully report quantitative data on cycle duration and make comparisons among taxa, scaling effects on tetrapod-feeding kinematics are virtually unstudied. Absolute differences in cycle duration among taxa may have little biological significance. Gorniak et ah (1982) provided electromyographic (EMG) data for feeding in Sphenodon, but unfortunately did not sample hyolingual muscles. During ingestion the jaws are opened by the bilateral activity of anterior and posterior portions of the depressor mandibulae; however, the anterior part is most active earlier and the posterior part later. Both are silenced as jaw closing begins and the adductors become active, although there is some overlap. The anterior portion of the superficial external adductor is first active, followed by its posterior portion, the medialis, the pseudotemporalis superficialis, and several parts of the complex pterygoideus. Adductor activity is maximal when the teeth contact
229
8. Feeding in Lepidosaurs the prey item and crush it. They are then silenced during a stationary phase when the mouth opens slightly, probably from recoil of the prey item between the teeth. 2. Processing Tuatara use a unique form of mandibular translation to shear their food rather than the typical puncture-crushing of lizards (Gorniak et al, 1982). After two or three manipulative or killing bites, an insect is moved by the tongue to the corner of the mouth on one side where it is chewed for two to four cycles. The prey item remains fixed relative to the upper tooth row during chewing. It is then moved by the tongue and placed between upper and lower teeth of the same or opposite side and the chewing cycles are repeated. Multiple chewing clusters are repeated (approximately 15 times) until the insect is "reduced to a saliva-covered, mushy mass" (Gorniak et al, 1982:333). Each reduction cycle includes a period of mandibular translation along the articular surface of the fixed quadrate so that the lower teeth slide between the two parallel, upper tooth rows (Fig. 8.2) (see Section IV,B), shearing the prey as well as crushing it. The gape profile is typical with all phases usually evident. However, the jaws usually continue to open during SO II so that there is usually no distinct "plateau phase." SO involves simple mandibular depression. During FO, however, the mandible slides posteriorly along the quadrate-articular joint. The jaws close during FC, but there is no mandibular translation. The prey item is crushed during SC-PS as the jaws are pulled into full closure. At the end of SC-PS, a distinct "shearing phase" occurs during which the mandible is translated anteriorly. Larger prey items are slightly rolled during translation. Shearing ends when the mandible is slightly depressed at the start of a stationary or resting phase. Typically, a resting or stationary phase is present between each chewing cycle. These vary in time, but can be as long as a second. After three to five cycles of chewing, insects are shifted laterally to the opposite tooth row with twisting movements of the tongue. As chewing progresses, cycle durations decrease. Chewing cycles blend into intraoral transport. When feeding on mice, Sphenodon follows ingestion with several crushing or killing bites. Inertial transport is used to position the mouse transversely across the tooth rows of both sides at the back of the mouth for the symmetrical killing bites. Lateral jerks of the head are then used to position the mouse inertially between the teeth on one side and a series of four to seven chewing cycles begin to reduce the mouth. Initially, side
shifts and prey manipulation are inertially based, but as chewing proceeds, the tongue becomes involved in manipulatory movements and reduction cycles resemble those for insects. This probably represents a transition from processing to intraoral transport (see later). If limbs or other parts of the mouse come to lie outside the tooth rows following ingestion and manipulation, these are bitten off and not consumed. EMG recordings (Gorniak et al, 1982) showed that initial jaw closure during chewing is driven by the adductor mandibulae externus superficialis, externus profundus, and pseudotemporalis superficialis. Crushing during SC is caused by the addition of the pseudotemporalis and the pterygoideus. The shearing phase is driven by a portion of the pterygoideus whose fibers insert on the mandible with a large anterior component so that they act as a protractor, as well as an adductor, of the mandible. Chewing is highly asymmetric in Sphenodon with frequent side switching. Working and balancing side muscle activity varied from bite to bite during reduction, apparently in response to prey position and texture. 3. Intraoral
Transport
It is difficult to separate chewing from transport cycles in Sphenodon as both occur simultaneously. As reduction proceeds, the bolus is moved farther and farther back in the mouth. Chewing cycles then grade into pure intraoral transport cycles that Gorniak et al (1982) referred to as "terminal movements." However, because these cycles seem to grade quickly into pharyngeal packing cycles, the distinctions are difficult to make, at least based on the descriptions of Gorniak et al (1982). Further analysis of our cineradiographic films may help clarify differences among cycle types (Schwenk et al, manuscript in preparation). Transport is hyolingual with high amplitude anteroposterior movements of the hyoid (Schwenk et al, manuscript in preparation; see Fig. 11.23 in Chapter 11). Transport gape profiles are typical, with welldefined phases (Fig. 8.17). The bolus is released from the marginal teeth and held centrally on the tongue. During SO the tongue and hyobranchium move slowly forward, then rapidly so during FO. At FC the tongue and hyoid reverse direction and the tongue's dorsal surface is arched so that it scrapes the palate as it is rapidly retracted, pushing the bolus behind it toward the pharynx. Palate scraping does not occur with mice, presumably because they did not tend to adhere to the roof of the mouth due to their greater mass. Gape angles tend to be larger during transport cycles than during reduction, and mandibular translation is minimal (Gorniak et al, 1982). Muscle activity levels are
230
Kurt Schwenk
reduced during transport; however, EMG patterns are similar to reduction cycles with the exception of the pterygoideus muscle. The pterygoideus exhibits a great deal of regional heterogeneity, with some parts active during jaw opening. 4.
Swallowing
a. Pharyngeal Packing As noted earlier, Gorniak et al (1982) did not distinguish between late intraoral transport cycles and pharyngeal packing. However, during "terminal movements" the tongue tip is often slightly protruded between the anterior teeth prior to retraction. Gorniak ei al. (1982) interpreted this behavior as a method of "cleaning" the anterior portion of the tongue, but more likely it reflects the larger anteroposterior excursions typical of pharyngeal packing cycles in other lepidosaurs (see later). It seems to be a way of positioning the bolus as far back on the tongue as possible so that the posterior limbs of the tongue can be used to tamp the bolus into the pharynx. h. Pharyngeal Compression Unfortunately, Gorniak et al. (1982) did not describe compression cycles in Sphenodon, nor do our films continue long enough into the feeding sequence to include this stage (Schwenk et al, manuscript in preparation). Gorniak et al. (1982) implied that swallowing is accomplished by the "terminal movements" described earlier, interpreted here as intraoral transport and pharyngeal packing cycles based on comparison to other lepidosaurs (see later). Our cineradiographic films indicate that the bolus may not clear the pharynx during these cycles, thus it is possible that tuatara use pharyngeal compression and possibly cervical flexure to push the bolus fully into the esophagus to complete swallowing, but this remains undetermined. Knowledge of the presence or absence of a pharyngeal compression stage in Sphenodon would help clarify the evolution of swallowing mechanisms in lepidosaurs. D . Feeding in Iguania 1. Ingestion Lingual ingestion in chameleons is treated separately (see Section VI), but is mentioned briefly here in comparison to other iguanians. This section focuses on the more generalized iguanian families, Iguanidae and Agamidae. Lingual ingestion in these taxa was noted anecdotally in a number of studies (e.g., Abel, 1952; Cooper et al, 1970; lordansky, 1973; Smith, 1984; Frazzetta, 1986), but its universality in Iguania was pointed out by Schwenk (1988) and Schwenk and Throckmor-
ton (1989), who showed that lingual ingestion of small prey uniquely characterizes this monophyletic taxon in contrast to its jaw-feeding sister taxon, Scleroglossa. Functional treatments or complete descriptions of lingual ingestion are available for the following taxa: Iguanidae, Anolis (Bels and Baltus, 1989; Bels, 1990a; Bels and Goosse, 1990); Dipsosaurus (Schwenk and Throckmorton, 1989); Iguana (Throckmorton, 1976; Schwenk and Throckmorton, 1989); Opiums (Delheusy and Bels, 1992); Phrynosoma (Schwenk and Throckmorton, 1989); Sauromalus (Schwenk and Throckmorton, 1989); Agamidae, Agama (including Plocederma) (Kraklau, 1991; Herrel et al, 1995); Phrynocephalus (Schwenk and Bell, 1988; Schwenk and Throckmorton, 1989); Pogona (Schwenk and Throckmorton, 1989); and Uromastix (Throckmorton, 1976; Schwenk and Throckmorton, 1989; Herrel and De Vree, 1999a). As noted earlier, lingual prehension is limited to relatively small food items, whereas larger prey are grasped in the jaws (Smith, 1984; Schwenk and Throckmorton, 1989), as in Sphenodon. Nonetheless, only one study has provided any functional data for iguanian jaw prehension (Bels and Goosse, 1990). Bels and Goosse (1990) asserted that the small iguanid, Anolis carolinensis, used jaw prehension for blowfly larvae, but they did not document the behavior photographically. It is possible that 1-cm larvae were too large for lingual prehension by the small lizards, but it is also possible that poor film resolution prevented its detection. Some lingual feeding sequences involve minimal tongue protrusion and close approximation of the jaws before tongue-prey contact (particularly when prey are relatively large; see later). In small taxa, such as A. carolinensis, this is very difficult to see, even in good films (personal observation). Furthermore, studies of other Anolis species (Bels and Baltus, 1989; Bels, 1990), as well as my own observations of A. carolinensis (unpublished), confirm the use of lingual prehension. In any case, the gape profile given for this sequence was typical of lingual ingestion sequences in other taxa. The ambiguity of this case only highlights the dearth of functional data for jaw prehension in iguanians. Cropping of plants by herbivorous species may be a form of jaw prehension in some cases, but it virtually always involves initial or simultaneous lingual prehension (Schwenk and Throckmorton, 1989). This is described further later. Jaw prehension may occur in other unusual circumstances. For example, streamside basilisk lizards (Basiliscus) have been observed to capture fish under water after a lunge from their perch (Echelle and Echelle, 1972). Although underwater lingual prehension does occur in some salamanders (Schwenk and Wake, 1988, 1993; S. Deban and J. Larsen, personal communication), it is more likely that the
8. F e e d i n g in L e p i d o s a u r s
basilisks use their jaws for prehension given the mechanics of lingual prehension (see later). Montanucci (1989) observed that some individuals of Phrynosoma solare, an extreme ant specialist, eventually "learned" to use the jaws to capture unnaturally large prey, but only after repeated attempts with the tongue failed. Schwenk and Throckmorton (1989) suggested that prey size-dependent differences in iguanian prehension mode are quantitative and not qualitative. As such, tongue-prey contact always occurs, but the distance the tongue is protracted and protruded is modulated according to prey size. As prey size increases, the tongue is protruded less and less until at large prey sizes tongue-prey contact occurs within the margins of the jaws at the same time as jaw-prey contact. This scenario is supported by Throckmorton's (1976) observations of Uromastix and my own observations of Gambelia (Iguanidae) feeding on lizards (unpublished results). It also suggests the way jaw prehension may have evolved in Scleroglossa from lingual-feeding ancestors (Wagner and Schwenk, 2000; see Section VII,A). Due to the lack of functional data on jaw prehension in iguanians, the following account is limited to lingual ingestion (references given earlier). The kinematics of lingual ingestion are very similar to those described for Sphenodon, with a few exceptions. Most lizards are more likely to charge active prey than are tuatara. Prey motion typically alerts the lizard and triggers monocular fixation of the prey so that the head is often first tilted to one side. A feeding sequence begins with orientation of the snout toward the prey item (and presumably binocular fixation in most species) and approach. If the prey item is already within range, the strike usually involves forward rotation of the body over the forelimbs with the head moving down and forward toward the prey. Sometimes tongue flicking and further observation precede the start of feeding, but often it is initiated immediately. As such, the jaws may begin to part and the tongue protruded as the lizard approaches, but before it is within striking range (e.g., Kraklau, 1991). If the prey item ceases motion or moves away, the lizard may remain with its tongue partially protruded for some time before the feeding attempt is completed or aborted; sometimes the tongue moves in and out while the prey moves in and out of range (Schwenk and Throckmorton, 1989). Variation in this phase of a feeding sequence is responsible for extensive variance in ingestion cycle duration. As protraction begins, the tongue tip is immediately curled ventrally and the dorsal surface arched (Fig. 8.18). This is in contrast to Sphenodon in which the tongue is pointed during the earliest stages of protraction. As it passes the mandibular symphysis, the
231
tongue tip is more or less anchored, and as protrusion continues the arched dorsal surface of the tongue extends outside of the mouth and is presented anteroventrally (Figs. 8.18 and 8.19). Schwenk and Throckmorton (1989) found a taxonomic difference between iguanids and agamids in the orientation of the tongue tip during protrusion, but additional study has not supported a consistent dichtomy. Tongue-prey contact occurs at maximum tongue extension on the anterior third of the tongue's dorsal surface (Figs. 8.19 and 8.21). The contact zone corresponds to the region of greatest papillary length and epithelial rugosity (Figs. 8.13A and 8.14B). As contact is made the lizard usually continues to advance so that the prey item is pinned to the substrate. This behavior not only immobilizes the prey, but importantly, it forces the prey item into the deep, papillary cushion of the foretongue so that the lingual surface is literally "formed" to the prey item (Figs. 8.21A and 8.21B). This maximizes the surface area of contact and may help promote adhesion. It might further help to absorb impact energy so that the prey item is not pushed away, giving the tongue time to form an adhesive bond. The mechanism of lingual adhesion and prehension is developed in detail in Section V,F. Following upon tongue-prey contact, the jaws open rapidly, largely by means of cranial elevation at the atlanto-occipital joint, and the tongue is retracted. As soon as the tongue and adherent prey cross the tip of
FIGURE 8.21. Lingual prey capture in a horned lizard, Phrynosoma cornutum (Iguanidae). (A and B) A small cricket is first hit with the tongue and then pinned to the substrate. The cricket's body is pushed into the papillary cushion of the tongue's contact zone, maximizing the surface area of contact and ensuring interlocking of the tongue with the prey surface. (C and D) Sequential frames of retraction in another feeding bout. Note how the tongue's papillary surface is straightened and rolled around the end of the tongue like a conveyor belt. The cricket is "flipped" rapidly into the mouth on the tongue's dorsal surface. Both sequences filmed at 250 fps (4 msec between frames).
232
Kurt Schwenk
the mandible the jaws snap closed. High-speed films of Phrynosoma reveal that, during tongue withdrawal, its dorsal papillary surface is quickly retracted by "rolling" around the stiff, protracted column of lingual muscle like a conveyor-belt (Schwenk, manuscript in preparation). The prey item is "flipped" around the end of the tongue and into the mouth, its retraction accelerated by summing the independent velocities of the papillary surface and whole-tongue retraction (Fig. 8.21C and 8.21D). Bell (1989) showed that the papillary surface is moved independently around the muscular column of the accelerator muscle in chameleons, and films of other iguanians suggest a similar pattern during retraction, hence this may be a general, iguanian trait. In herbivorous species that feed by biting pieces off of whole plants (e.g.. Iguana and Uromastix), lingual ingestion succeeds initially only in getting a part of the plant into the mouth. Once there it must be cropped by the teeth (Throckmorton, 1976, 1978; Schwenk and Throckmorton, 1989). Ingestion may require several cycles to get a sufficient portion of the plant into the mouth before it is cropped (Throckmorton, 1976,1978). Multiple ingestion cycles are unique to this situation. However, one might more accurately regard cycles subsequent to the first as part of intraoral transport, and cropping as a type of reduction. In some cases, a whole plant that is fixed in place can be regarded as a "large" food item that necessitates jaw prehension for cropping. For example, the Galapagos land iguana {Conolophus subcristatus) uses both lingual and jaw prehension when feeding on cactus pads. Pads are usually scraped clean of spines and tough outer cuticle first, then bitten (see Section V,D,2). Often the tongue draws the pad in for cropping by the teeth, but sometimes it is bitten directly (H. L. Snell, personal communication). Either way, initial cropping may not completely separate a piece from the cactus. In this case the exposed cactus flesh is scraped off the tougher, outer cuticle by the teeth. The related marine iguana {Amblyrhynchus cristatus) may use jaw prehension predominantly. It feeds underwater or in the intertidal zone by grasping algae in its jaws and cropping it with a twist and jerk of the head (Carpenter, 1966). This cropping action may be enhanced by an extremely foreshortened facial skeleton and procumbent (protruding), spatulate teeth (personal observation). However, it is possible that the tongue initially contacts the algae to draw it into the mouth for cropping (see later). Because feeding frequently occurs under water, lingual prehension is probably minimally important. Nonetheless, semidomesticated individuals were said to feed terrestrially on crickets (K. Angermeyer, in Carpenter, 1966) and it seems likely that lingual protrusion would be used then, but this is unknown.
Four possible mechanisms of tongue protrusion have been proposed (Smith, 1984, 1988; Schwenk and Bell, 1988; Schwenk and Throckmorton, 1989): (1) The tongue is "pushed" out of the mouth by protraction of the hyobranchium, primarily through action of the mandibulohyoideus (geniohyoideus) muscles. (2) The tongue is "pulled" anteriorly by the genioglossus muscle. This would either slide the tongue along the lingual process or pull the hyobranchium simultaneously due to a tongue-hyobranchial linkage and inextensibility of the tongue. (3) Intrinsic verticalis contraction reduces the diameter of the central lumen around the lingual process, exerting pressure on the incompressible fluid there and causing the tongue to slide forward on the tapered process. (4) The tongue is lengthened hydrostatically by a reduction in its diameter. Clearly, these mechanisms are not mutually exclusive and they may combine to produce various tongue movements. Furthermore, they may vary in their importance along the length of the tongue. For example, the hindtongue may be more tightly coupled to hyobranchial movement than the foretongue, which might use hydrostatic lengthening to a greater extent (see Section VII,D). It is reasonably well established that lapping and tongue flicking involve primarily hydrostatic elongation of the foretongue in all squamates and that lingual movements during ingestion, intraoral transport, and swallowing in iguanians are coupled to hyobranchial movement, but the precise mechanism of tongue protrusion during lingual ingestion is problematic. The anatomy of the tongue-hyobranchial connection suggests that independent movement of the tongue should be limited in most iguanians (see earlier discussion). Schwenk and Throckmorton (1989) presented circumstantial cinegraphic data for Pogona (Agamidae) suggesting that tongue protrusion is coupled to hyoid protraction (model 1), as it is in Sphenodon and during intraoral transport in iguanians. In support of this, Herrel et al. (1995) showed that the mandibulohyoideus muscle is active during lingual protrusion in Agama and that the sternohyoideus is active during retraction, suggesting that tongue movement is coupled to hyobranchial movement. Films of lingual feeding in the iguanid Phrynosoma also indicate coupled tongue and hyobranchial movement during retraction (Schwenk, manuscript in preparation). However, Herrel et al. (1995) also found that the genioglossus and posterior verticalis ("ring" muscle) are active during lingual protraction in Agama with peak activities at maximum protrusion, thus indicating models 2 and 3. Together, data for Agama suggest that tongue protrusion occurs through a combination of models 1,2, and 3. Thus, current data suggest a combination of niechanisms for lingual protrusion during ingestion in iguanians, but so
233
8. F e e d i n g in L e p i d o s a u r s
few species have been adequately studied that a great deal of functional diversity remains possible. Tongue protrusion occurs during SO (Figs. 8.22 and 8.23). SO II typically occurs at higher gape angles than in subsequent feeding stages, probably to accommodate the protruded tongue (Schwenk and Throckmorton, 1989). The difference in gape angle between SO II and maximum gape at the end of FO is correspondingly
r
f
1
smaller than in other cycles. Herrel et al. (1996a) showed that ingestion in Agama is quantitatively distinct from transport and "swallowing" (pharyngeal packing) cycles, differing in timing, duration, and maximal excursion of jaws and tongue; however, Schwenk and Throckmorton (1989), Kraklau (1991), and Delheusy and Bels (1992) found significant overlap in the kinematics of ingestion and transport or chewing cycles
•!
i
INGESTION open ant
I
close post
•••
Gape
•••* 1
AAA^^
[.••
• .•*
•
A
Tongue
•
• •
• 1
depress Hyoid i
$
[..••
1
•
•••• '
1 1
1
•
•
• •• • •
1
1
1
1
1
1
•
*
t
1 •
•• •
• •
..
f
[.«• r
.»** A*"
11
A
A1
AAA ••
•
% A
A
A
A
•• • • ' j
elev
1
• ••
up
..••*
Head-Y down
I
TRANSPORT 2
TRANSPORT 1
• • •
*
•
,*••**
•*
1• • ' •
1 1
••
•••••••••••,^
•• >•
••
.••' . . . . . • •
•••• ••
post
••
I
•• \
•
Head-X
•
ant
•
• •
1*****1
•
•
•
•
• ••
• •
1
•
•
•
11
1
1
1
1
1
1
100
1
1
1
1L
1
•
1
•
E
E
^ D D D
o
O
QL
D D D
D D
D D
C3
•D •D •n
D
o
CO
• •D •n
D
D
• • •
•D • •D • • •C
•D • •D • • •D
0 (0
sQ. O
0) to
5
c E >> G
• • • • •m •
•D D• B • m •D • •H B• m •D
D D
D
D D
n n
"t:
3
X
n n
CO
D D
CO •D
CO
C
S
D D
• •D • B • m • H D a • mm
o O
D D
•D • • n • • u m D E • G
'3
q3
•D O
< D
c X
I
D
D
D
D H B
D
c
c
CO
3 D
•D •D D• •
• • •D • D m B • m g• • • • D• D a
c
CO
•D
•g
CO
3
CO
CD CO
a>
B c
K
(0 T3 C
1 •D •D >
CO
JO in
Q.
<E D D
n D D • D D n n ffl • •D D m m B D D • D mD • m D • D m • •D D• D D E?3
n absent • present B unknown / unstudied
disintegration of the ESC origin / integration of the lingual prehension ESC FIGURE 8.29. Distribution of phenotypic traits related to feeding across squamate phylogeny. The phylogeny is the same as Fig. 8.1 A except that Amphisbaenia has been added insertae sedis. Note that many taxa have not been studied at all. Taxa for which there are data are inadequately sampled, sometimes represented by a single species. Phenotypic ''traits'' listed on the left are not intended to represent "characters" in the strict sense. Rather, they summarize gross phenotypic characteristics, or suites of characters, discussed in the text. The half white and half stippled box (Scincidae/mesokinesis) indicates that mesokinesis has been shown to be absent in one, possibly atypical, species, but mesokinesis may be present in other scincids. White boxes with small black squares in their upper corners (tongue prehension) indicate those scleroglossan families in which one or two species have been shown to use some form of lingual prehension, although the primitive ingestion mode for the family in each case is jaw prehension. Bold lines indicate the distribution of lingual prehension on the phylogeny, inferred to represent a phenotypically stable functional unit known as an "evolutionarily stable configuration" (ESC). See text for explanation.
266
Kurt Schwenk (D CO T3
0) CO
]D
I
0) to T3
E m
D)
(U (0
E
to
x: O
CO
-o "c » £L
(D CO
5
(U
H
H. o c F >» C!)
CO
o
CO _i
T3
CO CO
•D
X
CO
.^ •§ o
c
CO
c CO
CO
CO
0)
CO
(55
(O (D
c
CD
CD
CO
loss of hyoiingual transport obligate inertial feeding loss of hyobranchial participation in feeding loss of hyoiingual transport obligate inertial feeding
acrodont teeth dental occlusion
bipartite tongue
ballistic tongue projection
inertial feeding common acrodont teeth propalineal jaw movement shearing mechanism
pleurodont teeth streptostyly
jaw - prehension foretongue-hyobranchium uncoupled foretongue modification mesokinesis pharyngeal compression
ANCESTRAL CONDITION: tongue - prehension foretongue-hyobranchium coupled hyoiingual transport pharyngeal packing akinetic skull pleurodont or acrodont teeth? F I G U R E 8.30. Major events in the evolution of feeding systems in lepidosaurian reptiles. Based on data shown in Fig. 8.29 and other information discussed in the text. Due to inadequate sampling of most taxa, some inferences remain tentative.
analysis (e.g., Lauder, 1981,1990). Outgroup comparison is widely acknowledged to be the best method for determining character state polarities (Watrous and Wheeler, 1981; Forey et al, 1992), i.e., establishing primitive (ancestral or plesiomorphic) vs derived (apomorphic) states. Schwenk (1987, 1988) and Schwenk and Throckmorton (1989) pointed out that lingual prehension of small prey is restricted to Iguania, whereas Scleroglossa universally employs jaw prehension (Fig. 8.29). Using Sphenodon as an outgroup, the criterion of parsimony argues unambiguously that lingual prehension is the ancestral ingestion mode for squamates (one rather than two evolutionary transformations in prehension mode). Lingual prey capture in Sphenodon is nearly identical to that seen in Iguania, with only minor differences (see earlier discussion), so there is little doubt about the homology of ingestion mechanisms in the two taxa. Thus, the commitment of scleroglossans to jaw prehension of small prey is a derived condition in Squamata (and the ancestral ingestion mode for all scleroglossans). As discussed previously, it has been shown that a
few scleroglossan species also employ a form of lingual prehension. Does this change our conclusion about the polarity of ingestion modes, i.e., is it possible that jaw prehension is primitive for squamates? The cases of lingual prehension among scleroglossans do not affect the conclusion that lingual prehension is primitive in squamates for the following reasons. First, each lingual-feeding scleroglossan species represents an isolated case within its genus or family and the tongue prehension behavior is not shared by closely related species (so far as we know). Therefore, it is clear in each case that tongue prehension is secondarily derived and that jaw prehension is the ancestral ingestion mode for the genus or family to which each species belongs. Second, with the exception of T. scincoides, lingual prehension in scleroglossans lacks "detailed similarity" to the iguanian condition and therefore fails the most important criterion of homology. In other words, lingual feeding in scleroglossans is probably not homologous to lingual feeding in iguanians (and Sphenodon), but was reinvented from a jaw feeding ancestor. In T. scincoides, it was shown earlier that kinematic similarity
8. Feeding in Lepidosaurs with iguanian lingual prehension is achieved by different means, also hinting at the lack of homology. Thus, both parsimony and homology arguments show that lingual prehension in scleroglossan squamates is secondarily derived and that similarity to the iguanian condition represents homoplasy, not symplesiomorphy. The evidence, therefore, strongly supports the conclusion that jaw prehension is the ancestral ingestion mode in Scleroglossa, but this pattern is consistent with either polarity hypothesis. Thus, the parsimony arguments presented earlier are unaffected by these data, and the conclusion that lingual prehension is the ancestral ingestion mode in squamates, based on outgroup analysis, is supported. To summarize, ancestral lepidosaurs captured small prey with their tongues. Early in squamate evolution, however, a cladistic bifurcation led to two major lineages, Iguania and Scleroglossa. Iguanians retained the ancestral ingestion mode. Within this lineage the lingual prehension mechanism was modified in small ways, and in chameleons the novel element of ballistic projection was introduced, but all systems have remained within a phenotypic space circumscribed by the lingual prehension mechanism. Scleroglossans, however, departed from the ancestral pattern very early in their history and evolved a novel method of capturing small prey with the jaws (Figs. 8.29 and 8.30). Inferences about the evolutionary processes underlying this pattern are presented in the final section of the chapter. B. Post-Ingestion Feeding Stages Despite differences in ingestion mode, lepidosaurs are remarkably consistent in the kinematics and muscle activity patterns of subsequent feeding stages (Smith, 1984; Bramble and Wake, 1985; Schwenk and Throckmorton, 1989; Bels et al, 1994; Herrel et ah, 1997a). This apparently stems from the common use of a homologous, fundamentally similar hyolingual-feeding mechanism (Bramble and Wake, 1985). If similarity in lepidosaurian-feeding function is imposed by the mechanics of the hyolingual system, it is not surprising to see the greatest deviation from the common pattern in cases where the system is circumvented. This is most apparent, for example, in jaw prehension and in inertial transport. It may also help explain the extensive remodeling of the feeding apparatus evident in snakes, as compared to other lepidosaurs. Extreme tongue and hyobranchial reduction early in snake ancestry (presumably related to chemoreception) deprived snakes of the ancestral hyolingual transport system, but at the same time released them from its mechanical constraints. Many snake-feeding specializations, including
267
the mobile jaws of scolecophidians and asymmetrical skull kinesis in alethinophidians, reflect novel solutions to the problem of prey transport and swallowing. Varanids, with superficially similar reduced tongues, might be expected to show comparable departures, but varanids differ fundamentally from snakes in their retention of a robust hyobranchial apparatus. Thus varanids deviate from the common pattern in emphasizing the jaws for prey capture and inertial transport, but pharyngeal packing and swallowing remain essentially "lizard-like" due to their continued reliance on the hyobranchium during these feeding stages. Despite general similarities, scleroglossans depart to some extent from the basal lepidosaurian pattern in the way they swallow. Iguanians rarely employ pharyngeal compression after pharyngeal packing, and when they do it is very brief and barely evident. In contrast, pharyngeal compression is commonly observed in many scleroglossan species. Given that the posterior limbs of the tongue are used to tamp food into the esophagus during pharyngeal packing, it follows that in taxa with reduced hindtongues the efficacy of pharyngeal packing is also reduced, for which they compensate by compression of the pharynx (Herrel et ah, 1999b). Iguanians, with well-developed posterior limbs, are therefore able to complete swallowing (movement of food into the esophagus) with little or no need for pharyngeal compression, whereas scleroglossans with reduced hindtongues must complete swallowing with a compressive stage. It seems generally true that scleroglossans with the most reduced hindtongues, such as varanids and teiids, exhibit the most dramatic pharyngeal compression. Other factors might contribute to the increased use of pharyngeal compression by scleroglossans, notably body and neck elongation. Body elongation and even limblessness are hallmarks of scleroglossan, but not iguanian evolution (e.g.. Camp, 1923). Neck elongation might render pharyngeal packing with the tongue less effective due to the greater length of the pharynx. At the same time it would make the constrictor colli more effective in pharyngeal compression due to its enlargement and freedom from the mandible posteriorly. Elongation would also enhance the ability to bend the head and, in extreme cases, to permit the formation of a propagated wave. Once again, snakes might represent an extreme manifestation of this tendency; sinusoidal body waves and internal concertina flexion are interpretable as derived forms of pharyngeal compression that assist in the final phases of swallowing (see Chapters 2 and 9). Flowever, the fact that short-necked scleroglossans, such as Tiliqua, typically use pharyngeal compression as well (Herrel et ah, 1999b) suggests that tongue reduction alone may be sufficient to
268
Kurt Schwenk
explain the predominance of pharyngeal compression among scleroglossans. C. Evolution of the Gape Cycle Based on a limited data base. Bramble and Wake (1985) postulated that tetrapods share a fundamental pattern of feeding behavior involving coordinated movements of the jaws, tongue and hyobranchial apparatus underlain by similar muscle activity patterns. They summarized this pattern of relationships in a generalized model cycle showing predicted movements of various elements relative to a standard gape cycle exhibiting discrete SO I, SO II, FO, FC, and SC-PS phases (see earlier discussion and Chapter 2). Studies of feeding in lepidosaurians since 1985 have tended to support the notion of a fundamental feeding pattern (e.g., Schwenk and Throckmorton, 1989; Herrel et al, 1997a, 1999b; Herrel and De Vree, 1999a). Nonetheless, some lepidosaurian studies have found deviations from the relationships predicted by the BrambleWake model, raising the question of its generality (e.g.. Smith, 1984; Delheusy and Bels, 1992; Bels et al, 1994; Herrel et al, 1996a; Herrel and De Vree, 1999a). Qualitative departures from the model are usually found in the shape of the gape profile or the timing of tongue movement relative to the jaws. Whether one believes that lepidosaurian data support the Bramble-Wake model or refute it depends on one's expectations for such a model. If taken as a literal, point-by-point prediction for all feeding cycles in all lepidosaurs, then clearly the model is rejected. However, if one's expectation is for the model to serve as a generally predictive, heuristic set of organizational rules, then the model is strongly supported by lepidosaurian studies. I take the latter view. The model was based on intraoral transport and chewing stages because these comprise rhythmic, hyolingually mediated cycles that are most plausibly related to control by a central pattern generator. It is therefore not surprising that most kinematic deviations from the model have been observed during ingestion and swallowing stages, and in cases where the hyolingual mechanism is not employed—jaw prehension and inertial feeding, for example (see earlier discussion). Paradoxically, these exceptions offer the strongest support of all because their very deviance is a predictable outcome of the model. For example, the model gape cycle predicts a characteristic sequence of discrete SO I, SO II, FO, FC, and SC-PS phases due to the coordination of the tongue and hyobranchium with the jaws. Ingestion in scleroglossans predictably deviates because during jaw prehension the jaw-tongue linkage postulated by the model is broken. It is therefore not surprising that
scleroglossan ingestion cycles are variable, often exhibiting spiked or bell-shaped gape profiles without discrete phases. Inertial feeding, likewise, breaks the tongue-jaw couple and shows variable gape patterns. More problematic is that some cycles in which the jaw-tongue linkage is maintained fail to exhibit a complete set of discrete phases. Most often this results from a blurring of SO I and SO II phases into a single SO phase and, occasionally, the loss of a discrete SO phase altogether. However, such "deviant" cycles usually precede, follow, or are interspersed with more typical (i.e., "model-like") cycles. Given the extreme extent of modulation possible during SO (a major prediction of the model), such variation in gape profiles is to be expected as bolus position and condition changes with each gape cycle. Absolute adherence to the model is an unreasonable expectation for a biological system. Critics might argue that too liberal an acceptance of variation renders the model unfalsifiable (see Smith, 1994). While this is true, the fact remains that the kinematics and muscle activity patterns of feeding in lepidosaurs conform to the predictions of the BrambleWake model often enough to demonstrate its merit. Even as we amass the inevitable exceptions to its predicted patterns, the model should continue to serve as a useful guide because in most cases it effectively points to the potential causes of deviation. Fifteen years after Bramble and Wake (1985), lepidosaurian gape cycles can be characterized in the following way: 1. Rhythmic intraoral transport and chewing cycles most often conform to the model gape pattern of discrete phases (SO I, SO II, FO, FC, SC-PS). When deviations occur they usually take one of the following forms: (i) SO I and SO II vary in their relative lengths; (ii) SO I and II sometimes merge into a single SO phase, i.e., there is no "plateau" in the gape profile; (iii) occasionally, SO is absent and the gape profile is spiked or bell shaped; (iv) SC is sometimes absent and may not include a PS component; (v) coordination of hyolingual movement sometimes varies between transport and chewing cycles, with chewing cycles tending to deviate from the model; and (vi) inertial transport cycles deviate from the model because the hyolingual apparatus does not participate, thereby breaking the j a w tongue linkage (a postulate of the model; see earlier discussion). 2. Lingual ingestion cycles in Sphenodon and Iguania typically conform to the model and are similar to transport cycles with the following exceptions: (i) hyolingual protraction during SO is more extreme and carries the tongue outside the mouth; and (ii) SO II often occurs at a higher percentage of the maximum
8. Feeding in Lepidosaurs gape angle than in transport, presumably to allow passage of the prey item past the jaws into the mouth. In T. scincoides, the only scleroglossan so far described to use iguanian-like lingual ingestion, the gape cycle is similar, although it may be arrived at by a different mechanism. 3. Jaw ingestion cycles in scleroglossans deviate from the model because the hyolingual apparatus does not participate, thereby breaking the jaw-tongue linkage. The gape cycle is typically spiked or bell shaped. Jaw ingestion cycles in iguanians are too little known to draw any conclusions. However, based on the model and Schwenk and Throckmorton's (1989) finding that iguanian jaw prehension actually represents truncated lingual prehension, iguanian jaw prehension should be similar to lingual ingestion cycles, i.e., conform to the model, with the expectation that deviations, if they occur, will be found in SO II. 4. Swallowing cycles are at present too poorly differentiated from transport cycles in most studies to make supportable generalizations. Pharyngeal packing cycles are expected to conform most often to the model; however, the introduction of tip-first, extraoral tongue protrusion and side-to-side asymmetry during this stage suggests the potential for significant deviation, particularly in later cycles. Pharyngeal compression is not expected to conform to the model cycle because it breaks the jaw-tongue linkage on which the model is predicated. The limited data available support this conclusion. Finally, the dependence of swallowing behavior on bolus characteristics (food type, size, and condition) makes a high degree of variation in swallowing cycles more likely. D . Tongue Evolution In general there is a clear distinction between iguanian and scleroglossan squamates in many aspects of tongue form (see Schwenk, 1988,1993,1995; Schwenk and Throckmorton, 1989; Wagner and Schwenk, 2000). The distinction is particularly pronounced in the foretongue. Sphenodon is similar to iguanians in most respects. Except for the tongue tip, iguanians are characterized by long, slender "high-profile'' papillae that are densely glandular. In contrast, scleroglossans always have "low-profile" papillae on the foretongue or sometimes no papillae at all, and the foretongue is always aglandular (Schwenk, 1984, 1988). Iguanians often show epithelial elaborations at the apex of each papilla that increase the tongue's rugosity (e.g., plumose cells), but in scleroglossans the epithelial surface of the foretongue is smooth, firm, and lightly keratinized. The iguanian foretongue is usually broad and deep (thick), but in scleroglossans it is always reduced in
269
one or both dimensions. It is typically tapered and in some cases is very slender. In iguanians the tongue tip is never more than slightly notched, but in scleroglossans (with the exception of dibamids), particularly autarchoglossans, it is more deeply cleft and in some taxa it is forked and attenuate. In iguanians the lingual process is typically robust, extends most of the tongue's length, and is often "kinked" at its apex, but in scleroglossans it is usually slender, rarely extends more than half the tongue's resting length, and, in some cases, is detached from the basihyal or fails to penetrate the tongue at all. The laryngohyoid ligament usually attaches to the lingual process near the tongue tip in iguanians, but attaches farther back in scleroglossans. In most iguanians the genioglossus muscles insert into the tongue anteriorly, but in scleroglossans they insert relatively farther back, sometimes extremely so. Altogether, the proportion of the foretongue that is free of the floor of the mouth and the lingual process (and therefore the hyobranchium) are much greater in scleroglossans. Circular muscles around the hyoglossus bundles, implicated in hydrostatic elongation of the tongue (Smith, 1984; Smith and Kier, 1989), are weakly developed and sometimes incomplete in iguanians, but in scleroglossans they are well developed. These statements are generalizations and the exceptions are certainly worth exploring, but for the most part they accurately characterize each group. This morphological dichotomy is mirrored precisely by the functional dichotomy discussed earlier: iguanians use the foretongue as a prehensile organ to capture relatively small prey items (and sometimes large prey as well), whereas scleroglossans (with a very few exceptions) use the jaws and teeth for prehension of virtually identical prey types. In iguanians, vomeronasal chemoreception is slightly or moderately well developed (Schwenk, 1993,1995b) and tongue flicks are limited to short, simple extensions (Gove, 1979; Bels et al, 1994; Herrel et al, 1998c). In scleroglossans, vomeronasal chemoreception is relatively more highly developed, sometimes extremely so (e.g., Schwenk, 1994e). Tongue-flick protrusion distances are typically large and flicks are often kinematically complex, including rapid, multiple oscillations in some taxa (Gove, 1979; Bels et al, 1994). In iguanians, most tongue flicks contact the substrate, but in scleroglossans, tongue flicks are frequently directed into the air to sample volatile chemicals. Thus, it is reasonable to interpret many features of the iguanian tongue in light of its role in prey prehension and scleroglossan departures from this primarily as specializations related to enhanced performance of tongue flicking and vomeronasal chemoreception (see later). My interpretation of these patterns is that there is
270
Kurt Schwenk
a functional-morphological distinction between hindtongue and foretongue in most lepidosaurs such that the hindtongue is coupled anatomically and functionally to the hyobranchium whereas the foretongue is relatively uncoupled. Taxa vary in the proportion of the foretongue that is free of hindtongue coupling. In iguanians, relatively little of the foretongue is free, whereas a much larger part is free in scleroglossans. Lingual movement that is independent of the hyobranchium is concentrated in the foretongue and is primarily hydrostatically generated. It may involve some translation along the lingual process, particularly in iguanians where the process extends for most of the tongue's length, although it primarily results from intrinsic, hydrostatic elongation with concomitant reduction in foretongue diameter. Although scleroglossans are more specialized for foretongue mobility, iguanians are also capable of limited hydrostatic elongation of the anteriormost portion of the foretongue, which is evident when the tongue is protruded tip first, as during tongue flicking and lapping. Cineradiography indicates that the hyobranchium is not protracted during these behaviors in iguanians (Smith, 1984; Schwenk, unpublished observations) and flattening of the tongue is observed (see Fig. 1 in Schwenk, 1995b), as predicted by the hydrostatic model of tongue elongation (Kier and Smith, 1985). In some iguanians, notably most agamine agamids and the iguanid genus Phrynosoma, the tongue is capable of remarkable extraoral protrusion distances during ingestion (e.g., Schwenk and Bell, 1988). However, most of this protrusion is coupled to hyobranchial protraction and protrusion of the lingual process, as during the "aiming" stage of chameleon feeding. Given the extensive attachment of the tongue to the hyobranchium, this kind of protrusion requires the laryngohyoid ligament to be stretched or the larynx to be protracted along with the tongue and hyobranchium. The latter is clearly the case in Sphenodon and some iguanians, but the situation may vary among species (see Section IV). The analysis just described has identified two dualities related to the lepidosaurian tongue: the morphological duality of foretongue and hindtongue and the functional duality of feeding and chemoreception. General patterns or "strategies" of tongue evolution reflect an interplay between these two dualities. In this context, four types of lepidosaurian tongue can be identified (Fig. 8.29). 1. Feeding type {Sphenodon and Iguania): This type represents a near total commitment of tongue form to feeding function. The tongue is broad, deep, and muscular and is covered with high-profile, prehensile pa-
pillae. The tip is unnotched or slightly notched. Most of the tongue is coupled anatomically and functionally to the hyobranchium. The posterior limbs are robust. In effect, the foretongue-hindtongue duality is minimal. Although iguanians possess a vomeronasal system, which they stimulate through tongue flicking, vomeronasal function and evolution are constrained by the lingual feeding system (see Section VII,F). The anteriormost portion of the foretongue is modified in iguanians to permit limited hydrostatic elongation and tip-first protrusion, but the minimal commitment of iguanian tongue form to chemosensory function is indicated by its extreme similarity to the condition in Sphenodon, which lacks the behavior of tongue flicking altogether. 2. Compromise type (Gekkota and Scincomorpha, including Amphisbaenia): This type represents a compromise between feeding and vomeronasal function. The foretongue no longer participates in lingual prehension and is highly modified for greater tongueflicking performance (and eye-wiping in gekkotans), but the tongue remains important in hyolingual transport and swallowing, functions served especially by the hindtongue. Although the foretongue-hindtongue duality is evident functionally, morphologically there is a continuum between them with the foretongue most modified and free of the hyobranchium, and the hindtongue tending to retain hyobranchial coupling and other plesiomorphic attributes, especially in gekkotans. The tongue tip ranges from notched to deeply forked. The posterior limbs are usually well developed, but in some derived forms they are reduced or lost. Both foretongue and hindtongue remain papillose, but papillae are low profile, especially on the foretongue. Because the tongue retains its ancestral function during postingestion feeding stages, the kinematics of these stages also tend to retain the primitive pattern. 3. Bipartite or diploglossan type (Anguimorpha except varanids and snakes): In this type both functional and morphological dualities are extremely developed and clearly evident. The tongue is literally divided into an anterior portion devoted to chemoreception and a posterior portion devoted to feeding function. There is a sharp transition between hindtongue and foretongue evident as a crease or "retractile" zone and marked by a sudden, dramatic change in papillary height and glandularity. The foretongue is slender, smooth, histologically specialized for hydrostatic elongation, and entirely free of the hyobranchium. The papillae are low profile (in Lanthanotus they are lost), aglandular, and lightly keratinized, and the tongue tip is deeply cleft or forked. The foretongue is functionally devoted to tongue flicking and chemoreception with minimal
8. Feeding in Lepidosaurs participation in feeding function at any stage. The vomeronasal system is highly developed in these taxa. In contrast, the hindtongue essentially retains the plesiomorphic condition with long, glandular papillae and a tight coupling to the hyobranchium. It participates fully in transport and swallowing stages of feeding. Thus, postingestion-feeding kinematics resemble the primitive condition due to retention of the hyolingual transport system despite radical modification of the foretongue. 4. Chemosensory type (Varanidae and snakes): In these taxa the tongue's feeding function is lost not only in ingestion, but in postingestion stages as well. Therefore, the tongue is almost entirely committed to tongue flicking and chemosensory function. The ancestral function of the hindtongue in feeding is lost and it has been modified into a part of the tongue sheath. The entire oral portion of the tongue is free of the floor of the mouth and the hyobranchium. What is left is essentially a greatly expanded foretongue devoid of papillae and deeply forked. The kinematics of varanid swallowing retain some similarity to the ancestral pattern due to continued participation of the hyobranchium, if not the tongue, but in snakes the extreme reduction of the hyobranchial apparatus is associated with complete remodeling of the feeding system and novel kinematic patterns.
E. Dietary Specialization 1. Diet vs Phenotype This topic is deserving of a detailed analysis beyond the scope of this chapter. Several general points need to be emphasized, however. Despite the widespread assumption to the contrary, there is no necessary relationship between dietary specialization and phenotypic specialization in the lepidosaurian-feeding apparatus (Greene, 1982; Schwenk, 1988). A narrow or specialized diet leads to phenotypic specialization in some taxa, but not in others. This is a critical observation because it implies that, with only a few exceptions (see Section IV,B), we cannot, with any confidence, infer lepidosaur diet from morphology (see Chapter 1). This is clearly illustrated by the example of the amphisbaenian-feeding system. Despite mechanical analyses and laboratory observations strongly suggesting adaptive specialization of the system for killing and reducing large, vertebrate prey, studies of natural diet have shown that, if anything, amphisbaenians consume a disproportionate number of small invertebrates. Armed with this dietary data it may be possible to devise testable, or a least plausible, explanatory hypotheses for the origin of the unique amphisbaenian
271
morphology (see earlier discussion), but the lesson is clear: there can be no facile connection of form to diet. The fundamental importance of natural history data to functional morphology is indicated (see Chapter 1). If diet cannot be inferred from morphology, neither can morphology be predicted from diet. The case of myrmecophagy (ant feeding) illustrates this point well. Most species of the North American iguanid genus, Phrynosoma, consume mostly ants. Most share a bizarre phenotype distinct from related iguanids: wide, dorsoventrally compressed body, splayed limbs, broad head, abbreviated snout, occipital spines, relatively slow movement, long activity period, variable body temperature, and a very large stomach relative to body size (Pianka and Parker, 1975). They are some of the only lizards known to show diet-based tongue modification (Schwenk and Sherbrooke, manuscript in preparation). Pianka and Parker (1975) related all of these traits to a diet of ants: (i) ants are small and chitinous, thus many must be eaten for adequate nutrition; (ii) this requires a large stomach to store the ants; (iii) a large stomach requires a broad, tank-like body; (iv) the need to eat many small prey requires a long activity period to extend foraging time; (v) a long activity period exposes the lizard to more predators and the tanklike body makes rapid escape behavior impossible, therefore cryptic behavior and body spines are necessary for defense; and (vi) long periods of foraging in the open require relaxed thermoregulation, resulting in a high variance in body temperature. Pianka and Parker (1975:156) concluded: "Thus, Phrynosoma platyrhinos, and perhaps other members of the genus Phrynosoma, seem to be characterized by a unique constellation of anatomical, behavioral, physiological and ecological adaptations that facilitate efficient exploitation of ants as a food source and set the horned lizards apart from most other species of lizards." There is no doubt that Phrynosoma represents a phenotypically specialized lizard and that many of its putative adaptations are related to its myrmecophagous diet. Pianka and Parker's (1975) arguments are so plausible that they are sometimes viewed as the inevitable consequence of ant specialization in lizards. Remarkable phenotypic convergence in Moloch horridus (Agamidae), Phrynosoma's ecological counterpart in Australia (Pianka and Pianka, 1970), has bolstered the popular view that this is what ant-eating lizards must look like. However, Pianka and Parker (1975), themselves, were at pains to point out that Moloch differs from Phrynosoma in several key features and that their "integrated view of Phrynosoma ecology clearly does not apply in general to all ant-eating lizards." A brief survey of myrmecophagy in lizards shows convincingly that Pianka and Parker's (1975) "inte-
272
Kurt Schwenk
grated view" not only does not apply to all ant-eating lizards, it may not apply to any other ant-eating lizards, apart from Moloch, to some extent. Table 6 lists lizard and amphisbaenian species in which 30% or more of the diet is ants; some species consume almost nothing but ants. None show phenotypic specialization comparable to Phrynosoma and all are more similar to their nonmyrmecophagous relatives than to other ant-eaters (some traits, such as length of activity period and variance in body temperature, are unknown for most). As such, they retain a generalized phenotype and do not show specialization in the feeding apparatus or otherwise for extreme myrmecophagy. Table 8.6 also lists termite-feeding species because termites are similar to ants as a prey type, and in mammalian biology ant and termite feeding are often grouped together under the rubric of myrmecophagy (see Chapter 15). There are no species showing obvious phenotypic specialization for feeding on termites, with the possible exception of amphisbaenians. Seasonal or ontogenetic stenophagy, or geographic variation in diet, may account for the failure of most ant and termite eaters to evolve phenotypic specializations (see Section I1I,A), however several of the species listed in Table 8.6 are well known to consume virtually nothing but ants or termites. Furthermore, the Pianka and Parker (1975) scenario might apply only to open habitat and desert species or it might depend on the social structure and foraging behavior of the particular ant or termite species preyed upon. However, these points only reinforce the conclusion that the "integrated view" is case specific and therefore has little explanatory value outside its specific realm. It cannot serve as a model of phenotypic specialization for myrmecophagy in lizards. 2. What Is a Specialized Diet? The amphisbaenian example discussed previously called attention to the importance of natural history data to functional analysis. Functional analyses should be cognizant of natural diet, the manner in which food is procured in the field, and the mechanical tasks relevant to animal performance implied by these. For example, Herrel and De Vree (1999a) thought that "reduction of particle size is of prime importance" for herbivorous lizards, but in fact particle size reduction is usually minimal in herbivorous species because dental adaptations allow them to crop mouth-sized portions of food during ingestion that are rapidly transported and swallowed with little or no processing (e.g., Throckmorton, 1976). The assumption that particle size reduction is a necessary part of herbivory stems from a mammalian bias and the failure to appreciate the significance of cropping behavior during ingestion in
many herbivorous lizards. In other words, the mechanical tasks actually required of the feeding system in a folivorous lizard are very different from the tasks that might be assumed to be important. Most of the apparent dental specialization in Uromastix, including dental occlusion and the development of wear facets, are better interpreted as adaptations for initial cropping function than for particle size reduction during chewing. Given that mechanical attributes of food affect the behavior of the feeding system (e.g., Bels and Baltus, 1988; Herrel et ah, 1999b), it is critical to duplicate or least, approximate, natural food type, form, and presentation in order to reveal the functional linkages underlying the evolutionary relationship between phenotype and diet. Extraordinarily few functional studies of lepidosaur feeding have attempted to do this. Another point related to natural diet and the mechanics of feeding is the perception of what constitutes dietary specialization in the first place. Herbivory, once again, serves to make the point. There is a long tradition in the herpetological literature of regarding herbivorous lizards as dietary specialists, apparently because most lizards eat a variety of invertebrates or are, to some extent, omnivorous. Yet, as pointed out previously in Section III,A, "herbivory" implicates a diversity of potential food types, including leaves, stems, shoots, flowers, seeds, pollen, nectar, and fruits of various kinds, each potentially requiring different abilities to handle. A lizard that consumes several or all of these plant parts may be just as much a dietary generalist as an omnivorous species that eats both animal and plant food, even though the herbivore consumes food items from a more restricted taxonomic group. In other words, the traditional notion of dietary specialization is based on taxonomic restriction of the foods taken, not the diversity of mechanical tasks required to eat them. This is a strictly ecological notion of dietary specialization that in many cases may not be relevant to the question of dietary specialization in a functional sense. It is the latter type of dietary specialization that relates directly to the evolutionary question of phenotypic specialization in the feeding apparatus. As such, it is possible that the taxonomically disparate food types of fallen fruit and earthworms present a common challenge to a lizard because they are both wet, compliant foods requiring similar mechanisms for ingestion and processing, whereas a slug and a snail represent radically different food types because the latter is shelled, despite the fact that both are gastropod molluscs. Clearly, organisms adapt to the mechanical tasks required to eat, not to food taxon per se, yet we persist in basing our identification of dietary specialists and generalists on taxon-based food categories. This is reasonable only to the extent that a given taxonomic group shares a common set of mechanical
273
8. F e e d i n g in L e p i d o s a u r s TABLE 8.6
A Partial List of Ant- and Termite-Eating Squamates^ Termites
Ants Species Iguanidae Anolis aeneus A. bonairensis^ A. oculatus Liolaemus monticola^ Sceloporus graciosus S. occidentalis S. olivaceus S. undulatus Tropidurusflaviceps^ T. hispidus T. (Plica) umbra^ Uma paraphygas U. scoparia
Stamps et ah (1981) T. Schoener (personal communication) Bullock etfl/. (1993) Jaksic et al (1979) Rose (1976) Rose (1976) Kennedy (1956) Hotton (1955) Vitt and Zani (1996) Vitt et al. (1996b) Witt etal {1997b) Gadsden and Palacios-Orona (1977) Pianka (1986)
Agamidae Agama hispida Ctenophorus fordi C. isolepis C. scutulatus Draco maximus ^ D. melanopogon D. obscurus D. quinquefasciatus^ D. volans^
Pianka (1986) Pianka (1986) Pianka (1971,1986) Pianka (1986) Inger(1983) Inger (1983) Inger(1983) Inger(1983) Auffenberg (1980)
Gekkonidae Cosymbotus platyurus
Auffenberg (1980)
Scincidae Apterygodon vittatus^ Egernia inornata Leiolopisma tricolor
Mori et al (1995) Pianka (1986) Bauer and De Vaney (1987)
Lacertidae Acanthodactylus erythrurus
S. Busack (personal communication)
Amphisbaenia Amphisbaena alba Monopeltis sphenorhynchus
Colli and Zamboni (1999) Broadleyetfl/. (1976)
Source
Species
Source
Iguanidae Enyalius leechii Tropidurus hispidus
Vitt effl/. (1996a) Vitt gffl/. (1996b)
Agamidae Agama impalearis Caimanops amphiboluroides ^ Draco obscurus D. melanopogon
Znari and Nagy (1997) Pianka (1986) Inger (1983) Inger (1983)
Gekkonidae Chondrodactylus angulifer Colopus wahlbergi Diplodactylus conspicullatus^ D. elderi D. pulcher^ Gehyra variegata Hemidactylus frenatus Pachydactylus bibroni P. capensis Ptenopus garrulus Rhynchoedura ornata^
Pianka (1986) Pianka (1986) Pianka and Pianka (1976) Pianka (1986) Pianka and Pianka (1976) Pianka (1986) Auffenberg (1980) Pianka (1986) Pianka (1986) Pianka (1986) Pianka and Pianka (1976)
Scincidae Cryptoblepharus boutonii Ctenotus ariadnae C. atlas C. calurus C. dux C. grandis C. helenae C. leonhardii C. pantherinus ^ C. quattuordecimlineatus C. schomburgkii Egernia depressa E. striata Lerista bipes L. mueleri Mabuya spilogaster M. frenata M. variegata Menetia greyii Morethia butler Tiliqua branchialis Typhlosaurus gariepensis^ T. lineatus^
Auffenberg (1980) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Twigg et al (1996) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Vrcibradic and Rocha (1998) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986) Pianka (1986); Brain (1959)
Lacertidae Eremias lineo-ocellata E. lugubris^ E. namaquensis Ichnotropis squamulosa Meroles suborbitalis
Pianka Pianka Pianka Pianka Pianka
Teiidae Cnemidophorus uniparens
Eifler and Eifler (1998)
Amphisbaenia Amphisbaena darwinii^ A. mertensii Cercolophia roberti Dalophia pistillum Monopeltis anchietae M. capensis M. leonhardi
Cabrera and Merlini (1990) Cruz Neto and Abe (1993) Cruz Neto and Abe (1993) Broadley^ffl/. (1976) Broadleyefd. (1976) Broadley^tfl/. (1976) Broadley^ffl/. (1976)
'^Only species in which 30% or more of the diet consists of each prey type are listed. ^Species in which ants or termites compose virtually the entire diet.
(1986) (1986) (1986) (1986) (1986)
274
Kurt Schwenk
attributes as food. This is clearly not the case for herbivory. Should we expect similar feeding adaptations in a folivorous iguana, a frugivorous varanid, and a nectivorous gecko? Dietary specialization cannot be assumed because a given species feeds upon food types belonging to a single taxonomic group. Rather, dietary specialists are those taxa that consume a set of food items presenting the same mechanical challenge, regardless of the food's taxonomic affinity. As such, herbivores are no more likely to be dietary specialists than insectivores or omnivores. As for herbivory, ''myrmecophagy" does not necessarily imply dietary specialization in a functional sense. Specialization is assumed because ants are viewed as uniformly small, chitinous prey lacking in nutritive value and defended by formic acid, biting, or stinging. In fact, formicids are highly variable in size, form, biochemical composition, social structure, foraging, and defensive behavior. For example, a single Moloch horridus consumes as many as 2500 tiny, innocuous ants {Iridomyrmex) at a time, each as small as 0.002 cc in volume (Pianka and Pianka, 1970), whereas a Phrynosoma platyrhinos typically eats fewer (r
#
5i^
/
,j-*
J"
Tetrapoda F I G U R E 10.1. (A) Phylogenetic relation of the Crocodylia to the other vertebrates (after Gauthier et al, 1988). (B) Relationships among the living crocodilians and the position of Tomistoma schlegelli according to morphological (dotted line) or biochemical (dashed line) studies. Modified from Densmore and Owen (1989) and Frey et al (1989).
broad and flat producing the cranial table, (7) antorbital fenestra reduced or absent, (8) quadrate strongly inclined and bordered anteriorly by a long, slender quadratojugal, (9) palate akinetic with pterygoids and quadrates fused to the braincase, (10) pterygoids with wide and deep wings, (11) interpterygoid vacuity absent, (12) many cranial bones and articular more or less pneumatic, (13) eustachian passages more or less en-
closed in bone, (14) posttemporal fenestrae reduced, (15) squamosal, quadrate, and paroccipital process combine to form an otic meatus, and (16) mandible deepened posteriorly with retroarticular process well developed. Nine of the character states (1,2, 5, 6, 8,10, 14, 15, and 16) are foreshadowed in some thecodonts, but are never found in any extensive combination in a single noncrocodilian taxon.
339
10. F e e d i n g in Crocodilians
To the present, only 8 of the 124 described genera survive, all being members of the suborder Eusuchia (Densmore and Owen, 1989). Twenty-two extant species are currently recognized, with Crocodylus the largest genus, containing 12 living species. Caiman the second largest (2 species and 3 subspecies), followed by Alligator (2), Paleosuchus (2), Melanosuchus (1), Osteolaemus (1), Tomistoma (1), and Gavialis (1). 2. Relationships
among the Extant Eusuchia
A major problem in resolving the systematics and evolution of the eusuchian crocodilians is their tendency toward general morphological conservatism and convergence/parallelism in cranial morphology (Densmore and Owen, 1989). The morphological conservatism is explicit in the postcranial region, where few reliable characters can be used for phylogenetic studies (Sill, 1968). Therefore, most traditional assessments of crocodilian phylogeny are based on analysis of the numerous differences in head morphology and skull structure among different species. This cranial variability strongly reflects variation in ontogeny (Steel, 1973) or habitat and diet (lordansky, 1973), and therefore stresses the importance of feeding in crocodilian evolution. An example of this variation is the
Gavialis gangeticus
Tomistoma scliiegelii
specialization for ichthyophagy, which is reflected in elongation of the snout and a reduction in tooth size. Even today, long-snouted species (Fig. 10.2) with reduced teeth, which live mainly or exclusively on fish, are found among the gharials {Tomistoma schlegelii and Gavialis gangeticus) and the crocodiles {Crocodylus johnsoni, C. novaeguinea, C. cataphractus). Similar but opposite morphological adaptations toward a broadening of the snout are associated with a more general diet (large tetrapods such as reptiles, birds, and mammals as well as fish). Broad-snouted (brevirostrine) forms can be found among alligators, caimans, and true crocodiles (Fig. 10.2). These examples clearly show that similar adaptive strategies have led to convergent skull morphology and head shape in various groups of recent and fossil crocodilians (Densmore and Owen, 1989). Such convergence in character states has long been considered important phylogenetically, but only complicates the interpretation of systematic relationships and evolution in crocodilians. For example, the ecological adaptation toward a piscivorous diet and its associated morphological consequences on head shape reoccurred many times throughout crocodilian history, presumably in widely divergent lineages (Romer, 1956). Fiowever, this does not necessarily mean that taxa showing this
Crocodylus intermedius
Alligator mississippiensis
FIGURE 10.2. Dorsal (A), lateral (B), and ventral (E) views of the skull and lateral (C) and dorsal (D) views of the mandible of the gharial, Gavialis gangeticus; the false gharial, Tomistoma schlegelii; a narrow-snouted crocodile, Crocodylus intermedius; and the extreme broad-snouted American alligator. Alligator mississippiensis. Note the gradual increase of the snout width from the longirostrine Gavialis to the brevirostrine Alligator and a decrement in the size of the supratemporal fenestrae, an increasing heterodontic appearance of the dentition, an decreasing length of the mandibular symphysis, a more pronounded undulation of the jaw margins, a less distinct verticalization process of the basisphenoid, basioccipital, and the posterior end of the pterygoid, and an increase in the cranial osteodermic relief. Modified from Mook (1921b).
340
Johan Cleuren and Frits De Vree
condition share recent ancestry. The cranial similarity between Gavialidae and Teleosauridae may be interpreted to result from convergent adaptations to ichthyophagy, but is no proof for close relationships (lordansky, 1973). Studies of protein divergence (Densmore and Owen, 1989) and molecules (Poe, 1996), while in agreement with traditional interpretations of affinities between alligators and caimans, suggest that the true (Gavialis) and false {Tomistoma) gharials are more closely related to each other than to other crocodilians, and that the true crocodiles, Crocodylus, are all very close relatives that may have diverged recently (Fig. 10.1). However, contradictory to the results of the biochemical techniques, new morphological studies have shown that the braincase structure, neural pocket, air sinus systems, jaw adductor mechanism, pelvic and hindlimb morphology, and epaxial musculature of the caudal region of Gavialis gangeticus do not correspond to the rest of the living Eusuchia and therefore may be its most primitive living member (Tarsitano et ah, 1989). B. Inertial Feeding Many reptiles transport food in the oral cavity by cyclic movements of the tongue (De Vree and Gans, 1989; Schwenk and Throckmorton, 1989; Bels et al, 1994). During these transport cycles the tongue moves forward beneath the food, lifts it up, and then retracts, drawing the food toward the esophagus. This lingual transport overcomes the mass-dependent inertial resistance of the food mediated by surface-dependent bonding mechanisms, such as adhesion and interlocking (Bramble and Wake, 1985). As the weight and size of the food item increase, the surface bonding must increase disproportionately. This requires increased surface contact between the tongue and the food, which can only be achieved by enlarging the surface relief (papulation) and/or total tongue size. However, an enlargement of the tongue produces several disadvantages. Support of large, active prey by a large tongue requires an increase in gape, which in turn decreases the grip of the jaws on the prey and thus increases the chance of prey escape. A large tongue also decreases the gular opening, providing an obstacle during swallowing. Predatory reptiles, such as crocodilians, varanid lizards (Smith, 1982,1986), and snakes (Gans, 1961; Cundall and Gans, 1979; Cundall, 1983) feed on very large and heavy prey items and use an alternative method for prey transport. They do not transport the prey with the tongue but employ, instead, inertial feeding in which the inertia of the food item is utilized in shifting the prey toward the back of the oral cavity (Gans, 1969).
Bramble and Wake (1985) note that "inertial feeding is a facultative behavior for most tetrapods that use it" and state that snakes and varanid lizards are perhaps the only tetrapods in which inertial feeding has become obligatory. Obviously, crocodilians also belong to this latter group (Cleuren and De Vree, 1992) as the wide and flat crocodilian tongue lacks the ability to transport prey. Thus, the group of obligatory users of inertial feeding consists of species with a tongue that is too simple/primitive to be suitable for true lingual transport (crocodilians) or is highly specialized for chemosensory behavior (varanids and snakes).
IL MORPHOLOGY A. Morphology of the Cranium and Mandible 1. Cranial
Osteology
a. Skull The skull of many crocodilian species has been described in detail by many authors (Briihl, 1862; Miall, 1878; Mook, 1921a, c; Wermuth, 1953; lordansky, 1964). A wonderful review of the general crocodilian craniology is presented by lordansky (1973). The crocodilian skull (Fig. 10.2) conforms to the archosaurian diapsid type and is akinetic. The most notable modifications of the crocodilian skull are the formation of the cranial table, the elongation of the jaws, and the development of a secondary palate. The cranial table is formed by the flattened dorsal, postorbital part of the cranial roof. The secondary palate is formed by palatal processes of the premaxillae, maxillae, palatines, and pterygoids. This formation results in a posterior extension of the nasal passages, which terminate in secondary choanae ventral to the base of the brain case. This specialization permits crocodilians to breathe via the dorsally placed nostrils even when the mouth is holding prey under water. Pterygoids and quadrates are attached firmly to the braincase. The anterodorsal inclination of the immobile quadrates results in a posterior displacement of the retroarticular processes. The jaw margins are undulating, forming three convex and two intermediate concave arches. This pattern is more developed in brevirostrine crocodilians than in longirostrine species (Fig. 10.2); it is practically absent in T. schlegelii and G. gangeticus (lordansky, 1964). The length of the mandibular symphysis is also related positively to the length of the snout (Fig. 10.2). In G. gangeticus it reaches the level of the 23rd or 24th dentary tooth, whereas in Crocodylus niloticus it only reaches the 4th tooth.
10. Feeding in Crocodilians
341
b. Dentition
d. Cartilago Transiliens
Thecodont teeth occur on the premaxilla, maxilla, and dentary. Crocodilian teeth have conical, pointed, thick-walled crowns that are often separated from the cylindrical root by a slight constriction. The teeth are deeply embedded in the alveolar ridges. All extant crocodilian species lack palatal teeth. Most authors (Edmund, 1969; Ferguson, 1981,1984) refer to the dentition as being pseudoheterodont, although the differentiation in tooth size differs among species. Crocodilians with moderate to broad snouts show more variation than narrow-snouted species such as Gavialis and Tomistoma, which tend to have all their mature teeth more or less the same size (Fig. 10.2). Kieser et al. (1993) identified three morphogenetic zones in each of the age classes of C. niloticus: an incisor, a canine, and a molar region. They suggest that the Nile crocodile has five premaxillary incisors, followed by five canines and six or more postcanines. In the lower jaw they identified three incisors, five canines, and up to seven postcanines, and therefore concluded that the dentition of the Nile crocodile is heterodont rather than homodont. Fieterodonty is increased by the undulation of the jaws. The largest teeth of both the upper and the lower jaws are located in the central portion of the convex arches of the undulating jaw margins, whereas the smallest teeth occupy the concave arches (lordansky, 1973). The fine structure and chemical analysis of the teeth of Alligator mississippiensis are described by Sato et al (1990) and Shimada et al (1992).
The pyramidally shaped cartilago transiliens (Fig. 10.3) consists of two triangular cartilaginous disks, which are detached in the median plane and covered by a thick tendinous sheet (fibrous pillow in lordansky, 1964). It is positioned between the torus transiliens of pterygoidal flanges and the coronoid by the presence of many tendons that insert on it. On the dorsal side the cartilago is attached to the mandibular adductor tendon (stem tendon in lordansky, 1964; tendon B in Van Drongelen and DuUemeijer, 1982), the tendon of the pseudotemporal muscle, and some fibers of the m. adductor mandibulae externus profundus. The intramandibular tendon attaches to its ventral surface. The cartilago is also connected to the angular bone and to the m. pterygoideus anterior. It thus forms a connection between the adductor tendon and the lower jaw, and a junction of the vertical tendon system of the intramandibular muscle. A certain degree of freedom is permitted. The cartilago transiliens is not uniquely found in crocodilians. It forms a part of the gliding joint in the turtles CheIonia and Caretta (Schumacher, 1973).
c. Jaw Joint The mandibular joint shows a simple hinge mechanism in which only movements in the sagittal plane are allowed. Several morphological arrangements guarantee the rotational motion in the crocodilian jaw joint by preventing lateral movements of the mandible. During closing, the medial sides of the angular are guided by the pterygoid wings to ensure sagittal movements. This guiding is necessary as the medial traction component of the muscles implies that both halves of the lower jaw are drawn toward the pterygoid, which is made possible by the flexible connection of the two halves of the lower jaw at the mandibular symphysis. The fibrous pillow of the mandibular adductor tendon (stem tendon in lordansky, 1964) serves as a special buffer between the lateral edge of the pterygoidal flange and the mandible during opening and closing. This tendon, together with the cartilago transiliens (a cartilaginous disk situated between the tendon and the surangular), thus plays a special role in the guidance of the lower jaw past the pterygoid wings. The medial components of the adductor muscles also provide a firm guide in the jaw joints (Schumacher, 1973).
2. Jaw Muscles The jaw musculature of the several crocodilian species has been described extensively by many authors: A. mississippiensis (Poglayen-Neuwall, 1953; lordansky, 1964; Schumacher, 1973; Busbey, 1989), Crocodylus niloticus, C. rhomhifer, and C. porosus (Lubosch, 1914; Lakjer, 1926; Poglayen-Neuwall, 1953; lordansky, 1964; Schumacher, 1973), and Caiman crocodilus (Schumacher, 1973; Van Drongelen and DuUemeijer, 1982). The terminology of lordansky (1964) and Schumacher (1973) will be followed. The classification of the adductors corresponds to that of the crossopterygians (Luther, 1914) and depends on their relation to the N. trigeminus. Generally, three adductors are recognized: the m. adductor mandibulae externus (subdivided in a pars superficialis, a pars medialis, and a pars profundus, although not clearly separable), the m. adductor mandibulae posterior, and the m. adductor mandibulae internus (including the m. pterygoideus anterior and posterior, the m. pseudotemporalis, and the m. intramandibularis). The jaw adductors are quite uniform in the crocodilians; only minor differences are observed. For a full description of the jaw muscles, refer to the mentioned authors. The origins and insertions are summarized in Table 10.1 and Fig. 10.3. The tendinous structure and histochemical characteristics of the jaw muscles of A. mississippiensis are described by Sato et al (1992) and Shimada et al (1993). They distinguished the fiber types as red, intermediate.
342
Johan Cleuren and Frits De Vree
mandibijlae°'^ MAMP MAMEP MAMES
M. pseudotemporalis
M. pterygoideus anterior
M. intramandibularis
M. pterygoideus posterior cartilago transiliens
F I G U R E 10.3. Dorsal and lateral views of the lines of action of the jaw muscles and the position of the cartilago transiliens in Caiman crocodilus (fr, frontal; j , jugal; 1, lacrimal; MAMP, m. adductor mandibulae posterior; MAMEP, m. adductor mandibulae extemus profundus; MAMES, m. adductor mandibulae extemus superficialis; m, maxillary; n, nasal; p, parietal; pm, premaxillary; po, postorbital; prf, prefrontal; pr.r, retroarticular process; sq, squamosal; q, quadrate; qj, quadratojugal).
white, and tonic and found the highest percentage of red and intermediate fibers in the m. intramandibularis. In the superficial and medial portions of the m. depressor mandibulae and of the m. pterygoideus, all three fiber types are present in approximately equal amounts. The highest white fiber composition is found in the m. pseudotemporalis, the m. adductor mandibulae posterior, and the m. adductor mandibulae externus. B. Morphology of the Hyolingual Apparatus and Its Associated Musculature 1. Hyobranchial Apparatus and Tongue Crocodilians have a rather simple hyobranchial apparatus that consists of a hyoid body (basihyoid or corpus hyoidei) and a pair of anterior cornua (cornu bran-
chiale I). The posterior cornua (cornu branchiale II) have been fused or lost and there is no processus lingualis (processus entoglossum). The cartilaginous hyoid body (corpus hyoidei) is the most prominent part of the hyobranchium (Figs. 10.2 and 10.4). It is a broadly rectangular, ventrally convex plate that has rounded corners and resembles a widebladed shovel. The part of the hyobranchial apparatus posterior to the articulation of the cornu branchiale I is narrower than the anterior part. The trachea and the larynx are embedded in the posterior dorsal concavity of this posterior part. In older animals, ossifications are found in the posterior part of the basihyal (Fiirbringer, 1922). The slightly dorsally curved anterior edge of the basihyal is thinner than its lateral and posterior edges. The anterior portion of the basihyal bears small incisures or fenestrations, which are most obvious in older animals and are closed by a thin membrane
343
10. Feeding in Crocodilians TABLE 10.1 Origin and Insertion of the Jaw Muscles of the Crocodylia Muscle
Origin
Insertion
MDM
The posterior edge of the parietal and supraoccipital, the posterior surface of the squamosal, and the posterodorsal part of the quadrate anterior to the paroccipital process From a groove in the lateral surface of the squamosal
The dorsal surface of the retroarticular process of the mandible The anterior part of the retroarticular process
MAMES
The ventral surface of the descending processus (formed by the quadratojugal and quadrate) The lateral part of the ventral surface of the quadrate bone
The dorsolateral edge of the surangular
MAMEM
The anterior and lateral surface of the medial lamina of the CAT The lateral part of the ventral surface of the quadrate
By means of a small aponeurosis (a part of the MAT) on the cartilago transiliens The dorsomedial surface of the surangular and the angular, medial to the insertion of the MAMES
MAMEP
The ventral side of the quadrate and from between the lateral and medial lamina of the CAT
On the dorsomedial surface of the cartilago transiliens by means of the lamina lateralis of the MAT By means of an aponeurosis to the dorsal crest of the angular
MAMP
The ventral side of the descending process of the quadrate The medial part of the ventral surface of the quadrate The posterior surface of the lamina medialis of the CAT
The medial surface of the lateral crest of the angular bone, the medial side of the surangular, and on the posterior wall of the Meckelian fossa Posterior lamina (runs backward from the angular and the articular bone) of the MAT
MPST
From the laterosphenoid
On the angular bone, by means of the dorsal surface of the posterior lamina of the MAT On the dorsal surface of the cartilago transiliens
MPTA
The dorsal surface of the cartilaginous septum area between the orbit and the nasal cavity and the ventral part of the interorbital septum From the medial surface of the maxilla, the dorsolateral surface of the palatine, the anteromedial surface of the pterygoid, the descending prefrontal pillar, and the caudoventral part of the lateral surface of the basisphenoidal rostrum
On the dorsomedial surface of the angular, by means of the lamina anterior of the MAT The dorsal surface of the cartilago transiliens
MPTP
Dorsal and ventral surface of the pterygoid and three aponeuroses attached to the posterior part of the pterygoid flange
On the surangular and articular bone, by means of three lamina of the pterygoid tendon
MIM
The ventral surface of the cartilago transiliens
The lateral surface of the angular, coronoid and splenial, the dorsal surface of Meckel's cartilage, and the medial surface of the dentary
(Schumacher, 1973). They are poorly developed or sometimes absent in Caiman and Crocodylus, but rather large in Alligator (Flirbringer, 1922). In Caiman (Cleuren and De Vree, 1992) and Gavialis gangeticus (Sondhi, 1958), the rod-shaped cornu branchiale I articulates medially with the lateral margin of the basihyoid and then extends posteromedially (Fig. 10.2). In Alligator, this articulation lies more posteriorly and in Crocodylus, more anteriorly (Flirbringer, 1922). In Caiman crocodilus (Cleuren and De Vree, 1992) the posterior part of the cornu branchiale I gradually widens, flattens, and twists toward the trachea, ending in a thin, leaf-like cartilaginous epibranchial. In G. gangeticus, a ligament connects the base of the cornua branchiales I with the basihyal (Sondhi, 1958), whereas this sheath of ligament is absent in other crocodilians (Fig. 10.2). The ossification of the cornu branchiale I progresses from proximal to distal (Schu-
macher, 1973). In crocodilians, the cornua branchiales II are not separated and are represented by the posterior corners of the basihyal (Flirbringer, 1922; Gnanamuthu, 1937). As described for A. mississippiensis (Busbey, 1989) and C. crocodilus (Cleuren and De Vree, 1992), the ventral surface of the basihyal is connected to the posterior part of the tongue by a fibrous pad and thus does not support the tongue. However, the fibrous connection between the hyobranchial apparatus and the tongue will transmit forces passively between them when the hyobranchium moves. Also, the curved anterior border of the basihyal forms a buccal fold, which can contact the gular fold on the palate. The buccal fold lies in front of the gular fold with the mouth closed. Both folds form a seal between the posterior edge of the tongue and the palate and tend to prevent flooding of the esophagus and larynx.
344
Johan C l e u r e n a n d Frits D e Vree
The crocodilian tongue lacks any specific subdivision into base, body, or apex. This condition is observed in Alligator (Busbey, 1989), Crocodylus (Tanner and Avery, 1982), Gavialis (Sondhi, 1958), and C. crocodilus (Cleuren and De Vree, 1992). The wide, roughly triangular tongue forms a fibrous pad that thickens posteriorly from the tip. It is connected to the lining of the buccal floor over its total length; only its anterior tip is free. The tongue covers nearly the whole ventral surface of the oral floor, filling the space between the mandibular symphysis and the glottis. An intrinsic tongue musculature is completely absent in the crocodilian tongue (Ferguson, 1981b). The surface of the tongue in Alligator is covered by heavily keratinized, conical papillae (Shimada et ah, 1990). 2. Hyobranchial
Muscles
The hyobranchial apparatus and its associated musculature are described for Gavialis gangeticus (Fiirbringer, 1922; Sondhi, 1958), several alligatorines {A. mississippiensis, C. crocodilus, Paleosuchus palpebrosus: Fiirbringer, 1922; Schumacher, 1973; personal observations), and crocodylines {Crocodylus acutus, C. niloticus, C. palustris, and C. rhombifer: Fiirbringer, 1922; Gnanamuthu, 1937; Schumacher, 1973; personal observations). Muscles of the hyolingual apparatus (Fig. 10.4) are subdivided into four different muscle groups: the hypoglossal muscles, the hypobranchial longitudinal muscles, the glossopharyngeal muscles, and the m. intermandibularis (Schumacher, 1973). The terminology of Lubosh (1933) is followed. Hypoglossal muscles consist of hypobranchial spinal muscles (m. branchiomandibularis spinalis) and
sternohyoideus
the muscles of the tongue (m. hyoglossus and m. genioglossus). They all form a connection between parts of the hyobranchial apparatus and the lower jaw (Fig. 10.4). The m. branchiomandibularis has its origin on the basal end of the cornu branchiale I and inserts on the medial surface of the lower jaw. Together with the fibrous pad, the m. genioglossus and the m. hyoglossus constitute the main mass of the tongue; the m. hyoglossus lying ventromedially and the m. genioglossus laterally. The m. hyoglossus arises from the posterodorsal edge of the corpus hyoidei and inserts on the ventral surface of the fibrous pad of the tongue. The m. genioglossus originates from an aponeurosis from the medial surface of the mandibular symphysis of the dentary bone. The fibers of the pars medialis insert on the medioventral surface of the corpus hyoidei, and the thicker pars lateralis inserts on the lateral area of the tongue. The hypobranchial longitudinal muscles (m. coracohyoideus, m. episternobranchiotendineus, and the m. episternobranchialis) consist of long, parallel running muscle fibers that arise from the coracoid bone or the sternum and insert on the hyoid body or the cornu branchiale I (Fig. 10.4). These muscles are extrinsic tongue muscles that act as retractors of the hyoid apparatus. The m. coracohyoideus (syn.: m. omohyoideus) originates from the lateral margin of the coracoid bone, runs anteriorly paralleling the trachea, and attaches with a tendinous aponeurosis to the caudal edge of the first ceratobranchials. The m. episternobranchiotendineus (Schumacher, 1973) originates from the anteroventral surface of the sternum and runs to the medial surface and the posterodorsal margin of the splenial bone. Because the episternobranchiotendineus does not insert on the hyoid apparatus (personal ob-
branchiomandibularis visceralis hyoglossus intermandibularis genioglossus medialis
genioglossus lateralis branchiomandibularis spinalis coracohyoideus
sternomandibularis
F I G U R E 10.4. Ventral view of the lines of action of the hyolingual muscles and the position of the hyobranchial body and the first ceratabranchials (CBI) and the tongue (dotted line) in Caiman crocodilus.
10. Feeding in Crocodilians servations), its status as a hypobranchial longitudinal muscle (Schumacher, 1973) can be questioned. Lubosch (1933) refers to the anterior part of this muscle as m. tendineomandibularis. Because none of these names reflects its true topography, a new name is suggested: m. sternomandibularis (presented in Fig. 26 of Gnanamuthu, 1937). Fibers of the m. episternobranchialis (syn.: m. sternohyoideus) originate from the anteroventral and ventrolateral surface of the episternum and run anteriorly medial to the m. episternobranchiotendineus, also paralleling the trachea. They insert on the lateral surface of the hyoid body and on the medial surface of the first branchials. Glossopharyngeal muscles (m. branchiomandibularis visceralis and m. thyrohyoideus) are small muscles connecting the hyobranchial apparatus to the pharynx. The m. branchiomandibularis visceralis (Fig. 10.4; m. mandibulohyoideus in Sondhi, 1958) originates from the lateral sides of the cornu branchiale I. The fibers run anterolaterally to the ventral edge of the mandibula to insert on the ventromedial surface of the angular bone. Some fibers arise from the flap-like, cartilaginous extension of the cornu branchiale I and insert on the fascia surrounding the pharynx. The m. intermandibularis (Fig. 10.4) consists of a thin layer of fibers that arise from the dorsomedial surface of the splenial bone and from the medial surface of the anterior part of the dental bone. The fibers extend transversely medianly and insert on a medial raphe (gular septum in Sondhi, 1958). C. Morphology of the Neck and Cervical Muscles 1. Osteology of the Cervical Region The osteology of the cervical region is described extensively by Van Bemmelen (1887), Virchow (1914), Boschma (1920), Mook (1921b), Troxell (1925), Kalin (1933, 1955), Hofstetter and Case (1969), Seidel (1978), and Frey (1988). All three crocodilian subfamilies, Crocodylinae, Alligatorinae, and Gavialinae, show a homogeneous vertebral osteology (Hofstetter and Case, 1969). As in most other crocodilian vertebrae, the cervical vertebrae are procoelous, meaning that they are concave-convex with the hollow end in front (Troxell, 1925). Only the axis, the second sacral, and the first caudal vertebrae form an exception to this rule. Confusion exists whether the neck includes the first seven or nine vertebrae. Proof for the existence of seven cervical vertebrae is given by the coelom extending as far forward as the eighth vertebra. However, according to Hofstetter and Case (1969) and Frey (1988), the first nine verte-
345
brae may be called cervical for having no connection with the sternum. This latter theory will be followed here, as the first nine cervical vertebrae seem to form a functional unity. They include the pro-atlas and atlas (referred to as C-1), the axis (C-2), and the uniform third to ninth vertebrae (C-3 to C-9). The chain of cervical centra forms a curved cylinder that is concave dorsally. This curvature, known as the cervical or nuchal curvature (Seidel, 1978), is accentuated in the neck by the neural spine tips (Fig. 10.5). 2. Cervical Muscles The crocodilian epaxial muscles (Fig. 10.5) of the cervical and occipital region have been studied in two members of the Alligatorinae—A. mississippiensis (Seidel, 1978; Frey, 1988) and C. crocodilus (personal observations)—and in one crocodyline—Crocodylus niloticus (personal observations). The three epaxial subdivisions found in the thoracolumbar region of the Crocodylia are extended into the cervical and occipital region; the transversospinalis system, the longissimus system, and the iliocostalis system. The presence of these three muscle systems represents a truly primitive condition in the Crocodylia, as it is found in all living reptiles, mammals, and birds, but not in fishes and amphibians. The nomenclature of the cervical epaxial musculature is determined by the assignment of a muscle to any of these three systems and is followed by an appendix, which is related to its topography. The appendix "dorsi" is used for muscles of the thoracal region (trunk muscles). Muscles originating and inserting on the vertebrae of the neck (cervical muscles) are assigned with the appendix "cervicis," and muscles that arise from the cervical vertebrae and insert on the cranium (cervical-occipital muscles) are assigned with the appendix "capitis." The cervical musculature is covered superficially by the complex fascia of the neck and shoulder. All three epaxial systems are divided through the formation of fascial compartments. The dorsal intermuscular septum is situated between the transversospinalis and the longissimus system. It extends superficially to enclose the transversospinalis system dorsally and the longissimus system laterally. The dorsal intermuscular septum also forms a strong connection with the fascia of the skin and osteoderms of the neck. The longissimus system is fully separated from the iliocostalis system, which is enclosed laterally and ventrally by its own fascia. Because some of the iliocostal myosepta merge with the longissimus system, the distinction between these two systems is only distinct from the occiput back to the fifth cervical vertebra (Seidel, 1978). The cervical transversospinalis system is the most
346
Johan Cleuren and Frits D e Vree MTSCa
MSCaP
MLCe
MECa C3
MICCe
C2
C1
MLCaP
MTSCa T .MECaS MECaM
L
.MSCaP MLCaS MICCa
FIGURE 10.5. Lateral view of the lines of action of the cervical muscles and the position of their insertion of the occipital region of the skull in Caiman crocodilus (C1-C9, first to ninth cervical vertebra; bsphen, basisphenoid; bocc, casioccipital; co, occipital condyle; ectopt, ectopterygoid; exocc,exoccipital; MECa, m. epistroheo-capitis; MICCa, m. ilio-costalis capitis; MICCe, m. ilio-costalis cervicis; MLCaP, m. longissimus capitis profundus; MLCaS, m. longissimus capitis superficialis; MLCe, m. longissimus cervicis; MScaP, m. spinocapitis posticus; MTSCa, m. transversospinalis capitis; p, parietal; pteryg, pterygoid; q, quadrate; qj, quadratojugal; socc, supraoccipital; sq, squamosal).
differentiated and complex system and is associated with the neural spines and pre- and postzygapophyses. It is the most dorsally positioned system, bordered medially by the neural spines and ventrally by the longissimus system. The transversospinalis system is subdivided into a "cervicis'' part, inserting on the atlas (m. transversospinalis cervicis), a "capitis" part inserting on the occiput (m. transversospinalis capitis, m. spinocapitis posticus, and m. epistropheo-capitis) and several small intervertebral muscles (m. interneuralis cervicis and m. interarticularis cervicis).
The cervical longissimus system is concerned with the cervical transverse processes and forms an extension of the longissimus system of the trunk. The system consists of three muscles: one "cervicis" muscle (m. longissimus cervicis) and two "capitis" muscles (m. longissimus capitis superficialis and m. longissimus capitis profundus). The cervical iliocostalis system is associated with cervical ribs. It consists of two muscles continuing from the iliocostalis dorsi (m. iliocostalis capitis and m. iliocostalis cervicis).
347
10. Feeding in Crocodilians
In addition to the detailed description of A. tnississippiensis, Seidel (1978) investigated the neck muscles of other crocodilians, including Caiman, Melanosuchus, Osteolaemus, Crocodylus, Tomistoma, and Gavialis, and he observed a relative uniformity or regularity of the musculature. All muscles are present in all genera. The differences are of proportion and size, without any qualitative difference. The only notable features that Seidel (1978) mentioned are that long snouts seem to be correlated with elongated musculature and that the crocodiles have a reduced spinocapitis posticus and longissimus musculature. In contrast to this, the large m. spinocapitis of the flat-snouted alligator indicates the relative importance of roll and yaw muscles. The cervical musculature of C. crocodilus and C. niloticus (personal observations) strongly resembles the descriptions of A. mississippiensis (Seidel, 1978), but differs slightly from Frey (1988). Therefore, the subdivision given by Seidel (1978) is used to summarize the origins and insertions of the most important crocodilian cervical muscles in Table 10.2 and Fig. 10.5.
III. FUNCTION A. General Feeding Behavior Crocodiles widely exceed the size of all other recent reptiles. In addition to the giants Crocodylus porosus and Crocodylus niloticus, which can reach lengths up to 10 m, remarkably small forms are found in genera Paleosuchus and Osteolaemus. The adults of these species slightly exceed a body length of 1.5 m. Because the hatchlings of the larger species generally measure only 25 cm, they possess the ability to enlarge their birth length by 40 times, by far the largest increase in length of all vertebrates (Wermuth, 1953). This massive growth is accompanied by a change in diet and feeding behavior. Hatchling crocodilians subsist predominantly on aquatic and shoreline insects of many species; during their first years they progress through a phase of frog and fish eating. Only animals exceeding 2 m rely heavily on eating mammals and birds, but do not lose the ability to feed on smaller prey. This change
TABLE 10.2 Origin and Insertion of the Cervical Musculature of Caiman crocodilus" and Alligator mississippiensis^ Muscle
Origin
Insertion
m. transversospinalis capitis
Medial part: tips of the neural spines of C-9 to the axis Lateral part: fascia of the shoulder region
Tendinous to the dorsal surface of the processus postoccipitalis (suture of the supraoccipital and squamosal)
m. spino-capitis posticus
Tendinous aponeurosis from the posterolateral surface of the neural spine of the axis Dorsolateral surfaces of the neural spines of C-3 to C-7 Tips of neural spines of C-8 and C-9
By lateral tendon to the lateral edge of the exoccipital Fleshy to the tip of the processus paraoccipitalis
m. transversospinalis cervicis
Complex system of tendinous aponeurosis, which is attached to the prezygapophyses and neural spines of C-3 to C-9 through the dorsal intermuscular septum
Fleshy to the posterior surface of postzygapophyses of C-4, C-3, and atlas
m. epistropheocapitis
Anterolateral surface of the neural spine of the axis
Fleshy to the occipital surface of the supraoccipital bone and exoccipital-squamosal suture
m. longissimus cervicis
Tendinous aponeuroses that connect the dorsal intermuscular septum to the prezygapophyses The undersurfaces of prezygapophyses C-4, C-5, C-6, and C-7
By tendon on the postzygapophyse of the atlas (same aponeurosis as the first two bundles of the m. transversospinalis cervicis)
m. longissimus capitis superficialis
Lateral surface of the neural arches of C-5 to C-8
m. longissimus capitis profundus
Dorsolateral surface of the neural arches from C-6 to the atlas Transverse processes of C-7 to C-3
By a narrow, strong tendon on the lateral surface of the processus paraoccipitalis Flat aponeurosis to the edge of the basioccipital bone Fleshy to the medial surface of the basioccipital bone
m. iliocostalis cervicis
Anterolateral surface of the cervical ribs Anterior surface of the myoseptum arising from the posterior edge of the ribs
Posterior surface of the myoseptum of the next anterior rib Tendinous aponeurosis to the posterior surface of the atlas-rib
m. iliocostalis capitis
Distal half part of atlas-rib First septum of the m. iliocostalis cervicis
By transversal tendon to the ventral edge of paraoccipital process
^Author's research. ^From Seidel (1978) and Frey (1988).
348
Johan Cleuren and Frits De Vree
in diet during development is known for all crocodilian genera (Cott, 1961; Dodson, 1975; Pooley and Cans, 1976; Taylor, 1979; Schaller and Crawshaw, 1982; Davenport et al, 1990). Examination of hundreds of stomach samples from all sizes and from many species documents that the crocodilian diet not only undergoes considerable changes with size and age, but also with habitat (Pooley, 1989). In brackish water, estuaries, and lagoons, young crocodilians feed principally on insects, as well as on mud and fiddler crabs, mud prawns, shrimps, molluscs, mudskippers, and a variety of small marine fishes. This is supported by the study of Davenport et al. (1990), which showed that C. porosus feeds on a wide range of invertebrates and vertebrates, including beetles, crabs, prawns, and small mammals. Freshwater species subsist largely on tadpoles, frogs, freshwater snails, fishes, small mammals, and possibly a greater variety of insect life (Pooley, 1989). Numerous observations suggest that crocodilian species, especially the short-snouted ones, reduce very large prey items to a size convenient for further transport by jerking and twisting motions (Pooley and Gans, 1976). Commonly, large crocodilians seize some part of the prey and then rotate about their longitudinal axis. This tactic is not used by crocodilian species that feed exclusively on small prey animals, such as small mammals, fish, crabs, and prawns, as no reduction takes place with these prey items. Davenport et al (1990) observed that C. porosus, when feeding on prey caught in the water, manipulates its prey wholly in air, with the mouth being kept clear of the water. This aerial manipulation is also present in Caiman crocodilus, but only during transport and deglutition (Cleuren and De Vree, 1992). This feeding behavior is necessary because swallowing large items of food under water would involve breaching the seal between the tongue and the palate, thereby flooding the esophagus. The feeding habits of A. mississippiensis (Busbey, 1989) and C. niloticus (Pooley and Gans, 1976) are very similar to those described for C. crocodilus (Cleuren and De Vree, 1992). B. Feeding Stages The feeding behavior in the different crocodilian species is very similar. As in other lower tetrapods (De Vree and Gans, 1989, 1994), the feeding sequence is generally subdivided into three phases: ingestion, intraoral transport, and swallowing (Cleuren and De Vree, 1992). Ingestion involves capturing of the prey with sideways bites in which the head is rotated around the ver-
tical axis so that the food item is grasped by the teeth on one side of the snout. After capture of the prey, the prey is repositioned within the mouth by a series of inertial bites. A forceful bite then follows with a welldefined crushing phase in which the prey is killed and crushed. This subset of several repositioning bites, followed by a killing/crushing bite, is repeated until the prey has been killed and reduced to a size suitable for further transport. Repositioning and crushing subsequences may be interrupted for several seconds, while the prey is held between the median teeth (Cleuren and De Vree, 1992). Instead of killing by crushing and biting, crocodilians frequently carry the struggling prey to the water to hold it submerged until struggling ceases. After drowning, repositioning and crushing follow. During intraoral transport, the prey is first oriented lengthwise and is then shifted headfirst between upper and lower jaws with rapid inertial repositioning bites. The prey is then moved toward the back of the oral cavity. Once the food is well within the gular region, swallowing cycles shift the prey into the esophagus. Swallowing is not an inertial process; the jaw apparatus plays only a minor role. It involves cyclic movements of the hyobranchial apparatus that push the food item into the esophagus. Once the prey reaches the entrance of the esophagus, it is squeezed more posteriorly by compressive movements of the gular region (Cleuren and De Vree, 1992). C. Kinematics 1.
Overview
Several researchers have examined the kinematics of the inertial feeding process in crocodilians: C. crocodilus (Van Drongelen and DuUemeijer, 1982; Cleuren and De Vree, 1992) and A. mississippiensis (Busbey, 1989). The following kinematic description is based mainly on cineradiographic records of C. crocodilus (Cleuren and De Vree, 1992). They characterize four types of bite: inertial repositioning bites, inertial killing/ crushing bites, inertial transport bites, and swallowing cycles. Each bite type is characterized by a specific displacement pattern of the neck, cranium, and hyolingual apparatus and has a unique gape profile. As in other lower tetrapods (Bramble and Wake, 1985), characteristic changes in the gape profiles are used to subdivide each open-close cycle into several kinematic phases: slow opening (SO), fast opening (FO), fast closing (FC), slow closing (SC), and crushing or power close (Cr). The change in the features of these kinematic phases or their absence/presence is used for the identification of the bite types throughout the feeding
349
10. F e e d i n g in Crocodilians
sequence. Gape profiles of repositioning bites involve a FO, FC, and an unclear SC phase. Killing/crushing bites have a profile identical to that of repositioning bites, but the terminal SC and Cr phase is well defined. Transport bites begin with a SO phase, followed by FO and FC, but lack SC. Swallowing cycles show only an opening and a closing phase. Busbey (1989) uses another terminology in the subdivision of the gape profile into kinematic phases: closing, closed, open 1, static open, and open 2. All phases can probably be correlated to the phases observed in C. crocodilus (Cleuren and De Vree, 1992): open 1 and static open corresponding with the slow-open phase, open 2 with fast opening, closing with fast closing, and closed with the crushing phase. Crocodilian gape profiles lack the stereotypy observed in mammalian chewing cycles and are affected strongly by the position of the prey between the teeth. This is represented in the gradual modification of the gape profile throughout the feeding sequence and results in the presence of intermediate bite types in the transition phases from manipulating/crushing to intraoral transport and from intraoral transport to swallowing. Repositioning and killing/crushing bites occur in the first part of the feeding sequence whenever the prey is caught between the opposed tooth rows. Once the prey has been killed and oriented lengthwise, only transport cycles occur, followed by swallowing. The number of bites depends on food type and size; feeding on large prey involves a larger number of bites. In C. crocodilus, the influence of prey size is very clear for repositioning bites (10 bites for newborn mice to 100 for large juvenile mice), whereas it is less obvious for the other bite types (between 2 and 5 killing/crushing, 3 and 8 transport, and 5 and 12 swallowing bites; Cleuren and De Vree, 1992). 2. Inertial Bites Inertial feeding in C. crocodilus (Cleuren and De Vree, 1992) proceeds at a rate of approximately three to four bites per second, which means that the average bite lasts less then 300 msec. Only killing/crushing bites exceed this duration. The onset of inertial bites involves a slow elevation of the neck and cranium, with the cervical elevation being accompanied slightly later by cranial elevation. During this time the hyolingual apparatus is slowly lifting the prey dorsad (Fig. 10.6). Fast opening results from the rapid elevation of the cranium and neck and the depression of the mandible (Fig. 10.6) and is nearly always associated with a rapid sideways head movement. Further elevation of the cranium and neck dur-
Killing/crushing
Transport
Swallowing
FIGURE 10.6. Major events occurring during killing/crushing, transport, and swallowing in Caiman crocodilus. Positions of the head, hyolingual apparatus, and prey are drawn from a cineradiographic sequence taken at 50 frames per second. The first six frames (left) of the killing/crushing sequence (0-200 msec) are identical to those observed during a repositioning bite. Frames 1 to 4 for each biting mode represent positions during the slow-opening or interbite phase. Frame 5 shows maximum gape for each mode; at this point the head is maximally pulled backward. Frame 6 shows the ensuing closed jaw position. The forward thrust of the head at this time is accompanied by a depression of the cranium in killing/crushing and repositioning bites and by a further cranial elevation in transport bites. Frame 7 represents the crushing phase in killing/crushing bites and the beginning of the interbite or slow-opening phase (cf. frames 1-4) in transport and swallowing bites. The position of the prey (juvenile mouse) is represented by the stippled oval. From Cleuren and De Vree (1992), with permission.
ing fast opening retracts the head and accelerates the prey backward. Depression of the lower jaw allows the prey to be pushed rapidly upward by the hyolingual apparatus. During the following fast-closing phase, the neck and cranium are depressed abruptly and return to their starting position with a rapid reversal of the lateral movement. Depression of the neck and cranium
350
Johan Cleuren and Frits De Vree
thrusts the head forward toward the backward moving prey. The prey is grabbed by the jaws at a more posterior position when the mandible is elevated. Fast closing retracts the hyolingual apparatus rapidly ventrally. As the jaws touch the prey, its resistance decreases the closing velocity, which marks the beginning of the slow-closing phase. Fiead and neck are depressed further during slow closing. a. Repositioning and Killing/Crushing Bites Repositioning and killing/crushing bites begin by an elevation of the neck and head, during which time the hyolingual apparatus is protracted anterodorsally until it supports the prey (frames 1-4 in Fig. 10.6). This dorsal movement of the hyobranchium and of the posterior part of the tongue has a greater excursion (approximately twice as much) than that of the anterior part. This slightly elevates the tongue posterior to the prey, forming a kind of bowl surrounding it. During the following fast opening, the prey is pushed rapidly upward (frame 5 in Fig. 10.6). The elevation of the neck and head and the upward movement of the hyolingual apparatus impart a backward and upward acceleration to the prey. This disengages the prey from the teeth so that it is "floating'' backward between the jaws. The hyolingual apparatus retracts ventrally immediately after reaching its maximal dorsal position. Fast closing then starts as the neck and cranium are depressed simultaneously, thrusting the head forward in the opposite direction of the backward moving prey. During this phase the tongue and hyobranchium are retracted further posteroventrally (frames 6 and 7 in Fig. 10.6). The buccal cavity thus enlarges so that the prey can be caught in a more advantageous position. h. Transport Bites In transport bites, the slow-opening phase is obviously subdivisible into a SO I phase in which the gape increases rather fast and a SO II phase in which there is only a slight increase in gape. The tongue and hyobranchium initially move to their maximal anterior position and slightly dorsad, pushing the prey against the palate and resulting in a depression of the lower jaw, thus increasing the gape (slow open) (shown in frames 1-3 of Fig. 10.6). They then move posteriorly, shifting the prey slightly backward into the pharynx (frame 4 of Fig. 10.6). Fast opening then follows. The upward displacement of the basihyal and the posterior part of the tongue during FO is 30-50% less than that seen during repositioning and killing/crushing bites. This reflects the more posterior position of the prey as the tongue is pushing the prey against the palate. In the fast-closing phase of transport bites the cranium is lifted further while the neck is already depressed (Fig. 10.6). Half-
way through this phase, the cranium will have reached its maximal elevation before it is depressed. Further elevation of the cranium places the jaws into a maximum vertical position, thus increasing gravitational effects. These effects associated with the ventral displacement of the hyolingual apparatus facilitate transport of prey into the pharynx and toward the esophagus. In contrast to repositioning and killing/crushing bites, transport bites show greater ventral displacement of the hyolingual apparatus during fast closing (frame 6 in Fig. 10.6). This can be explained by a partially passive movement and depends on the size and form of the prey. As the lengthwise oriented prey is compressed between the palatine and the buccal floor, it deflects the hyolingual apparatus downward during jaw closure. This movement is absent in the former bite types because the prey is oriented perpendicular to the tooth rows and thus does not exert a push against the buccal floor. The gape at the beginning of the transport bite decreases gradually in subsequent transport bites. This decreasing gape reflects the further shift of the prey into the pharynx, which no longer obstructs the closing of the jaws. 3.
Swallowing
Swallowing starts as soon as the prey has reached a position in which the hyobranchial apparatus lies anterior to it; cyclic movements then push the prey into the esophagus. Swallowing cycles consist of active protraction and retraction of the hyoid apparatus. The tongue passively follows the hyobranchium movement but does not participate in swallowing. Swallowing cycles differ from inertial bites in having a longer interbite interval between two subsequent cycles. Swallowing cycles start with a forward displacement of the hyobranchial apparatus until it reaches a position anterior to the prey. The hyobranchium then moves slightly posterodorsad to reach the anterior end of the prey (shown in frames 1 and 2 of Fig. 10.6). Its rapid posteroventrad retraction forces the prey into the esophagus during the opening and closing phase (frames 3-6 of Fig. 10.6). Halfway through retraction of the hyobranchial apparatus, the jaws open slightly to facilitate the passage of the prey. The opening phase mainly reflects depression of the lower jaw accompanied by a slight depression of the head. At the end of hyobranchial retraction, the mouth is closed by elevating the mandible relative to the simultaneously elevating cranium, which pushes the prey further into the esophagus. After mouth closure, the hyobranchial apparatus restarts its forward displacement, completing its cyclic movement.
351
10. Feeding in Crocodilians D . Role of the Hyolingual Apparatus in the Inertial Feeding Process A lot of confusion has existed on the use of the hyolingual apparatus during inertial feeding and swallowing. Sewertzoff (1929) described the tongue in crocodiles as being unable to move independently because it is fixed so firmly to its base, and he believed it to be the most primitive of reptilian tongues. The morphological relation of the tongue, hyobranchial apparatus, and buccal floor led to the assumption that the crocodilian tongue can be elevated and depressed, following passively the raising and lowering of the buccal floor. Based on the form and position of the basihyal and the anterior cornua in Gavialis gangeticus, Sondhi (1958) suggested that they are mainly responsible for the dorsoventral movements of the buccal floor, whereas the function of the posterior cornua lies in their support for the hyoglossus muscles. Busbey (1989) assigned the raising of the buccal floor in A. mississippiensis to a constriction of the m. intermandibularis, consequently enabling the tongue to immobilize food items against the palate. Generally, it is assumed that the tongue cannot be protruded and thus does not aid in manipulation and anteroposterior prey transport. Busbey (1989) ascribed a rather passive role to the tongue during the transport of food through the pharynx. The hyobranchial apparatus was supposed to be specialized for sealing the pharynx, but is not supposed to support the tongue (Busbey, 1989). However, Busbey (1989) observed the hyoid cornua pressing against the skin of the throat after the prey was partially transported into the pharynx. He also noticed that the hyobranchium may move in several small orbits, or move backward during this phase, although prey transport may not be obvious. Similar observations were done during swallowing in submerged Crocodylus porosus (Davenport et al, 1990). The presence of this stage was revealed only by throat movements, as the teeth were held tightly together. Thus, although both researchers report movements of the tongue and hyobranchial apparatus during feeding in A. mississippiensis (Busbey, 1989) and C. porosus (Davenport et al., 1990), they both assumed that the dorsoventral movements of the tongue are a passive result of the movements of the buccal floor and do not play an active role in the inertial process. In contrast to all this, cineradiographic recordings of Caiman crocodilus (Cleuren and De Vree, 1992) revealed an active role for both the tongue and the hyobranchial apparatus. The tongue aids in the inertial feeding process by pushing the prey item dorsally during the upward acceleration of the craniocervical complex, just prior to its release (Fig. 10.6). This upward
acceleration and velocity of the cranium and neck must be sufficiently rapid to overcome the downwardly directed gravitational acceleration on the food object (Gans, 1969). In other inertial feeders, the upward acceleration is imparted exclusively to the food item by an upward and backward thrust of the cranium and neck. However, in crocodilians the upward motion of the hyolingual apparatus during the FO-phase assists the posterodorsal thrust of the craniocervical region. It imparts an upwardly directed acceleration to the prey and thus increases its "upward" kinetic energy. As a result, the food item is pushed higher and thus takes longer to travel up and back downward to its starting point. Crocodilians can use this additional time in shifting their head into the most advantageous position to catch the prey, facilitating food transport. Movements of hyobranchium and tongue change gradually with position of the prey relative to the hyobranchial apparatus. A major change in hyobranchial movement occurs whenever the posterior end of the prey is right above it. Thereafter, the hyobranchium shows reduced dorsal movement during the fastopening phase and increased ventral movement during the FC phase. All repositioning, killing/crushing, and the first (1-4) transport cycles occur prior to this transition, and the other transport cycles and swallowing cycles take place thereafter. During transport of the food through the pharynx, the tongue is depressed, giving the throat and floor of the mouth the appearance of a large sack, which opens up the pharynx, despite the protruding gular fold. In the meantime, lifting of the head facilitates gravitational transport (Busbey, 1989; Cleuren and De Vree, 1992). Whenever a large proportion of the prey is in the pharynx, the jaws close and the hyolingual apparatus is retracted, after it is placed in front of the prey, effectively squeezing the prey posteriorly into the esophagus. The hyolingual apparatus thus also plays a vital role during swallowing.
E. Motor Patterns 1. Jaw Musculature Activity patterns of the jaw musculature are examined in C. crocodilus (Van Drongelen and DuUemeijer, 1982; personal observation), in A. mississippiensis (Busbey, 1989), and in C. niloticus (personal observations). The following description mostly contains data from an extensive study of C. crocodilus, which covered the complete feeding process from ingestion to deglutition (Cleuren, 1996), and supported by data on C. niloticus (personal observations). The fragmentary results of other authors will also be discussed.
352
J o h a n C l e u r e n a n d Frits D e Vree
In all bites (Fig. 10.7), jaw opening is achieved by a depression of the lower jaw by contraction of the m. depressor mandibulae. Simultaneously, the upper jaw is elevated by the contraction of several dorsal cervical muscles (see later). According to Van Drongelen and DuUemeijer (1982), jaw opening in C. crocodilus is
Mean electromyograms MDM MAMP MAMES MAMEP MPTA MIM MPST
Capture
MDM MAMP MAMES MAMEP MPTA MIM MPST
Killing/crushing
MDM MAMP MAMES MAMEP MPTA MIM MPST
Repositioning
MDM MAMP MAMES
n=l
n=12
n=31 1
—
:
Transport
MAMEP MPTA MIM MPST
n=12
MDM MAMP MAMES
Swallowing
MAMEP MPTA MIM MPST
n=16 800
1000
Time (ms) FIGURE 10.7. Mean electromyograms of the jaw muscles for capture, killing/crushing, repositioning, transport, and swallowing in Caiman crocodilus. The activity of each muscle is subdivided into three bursts; however, not every bite type contains all three bursts. In the adductors, burst 1 is the activity that is usually present during fast closing. Burst 2 is high-level activity specific for the pulsatile activity during the crushing phase (capture and killing/crushing). The postburst is low-level activity, primarily seen after full closure of the jaws. The height of each block is related to the intensity of the activity; full height equals maximal activity. Time zero is determined by the maximal gape at the end of the fast-opening phase (MDM, m. depressor mandibulae; MAMP, m. adductor mandibulae posterior; MAMES, m. adductor mandibulae externus superficialis; MAMEP, m. adductor mandibulae externus profundus; MPTA, m. pterygoideus anterior; MIM, m. intramandibularis; MPST, m. pseudotemporalis).
mainly accomplished by the contraction of the cervical rauscles, as it is rarely accompanied by activity of the m. depressor mandibulae in their experiments. This finding was supported by the fact that their animals constantly kept the lower jaw in a horizontal position. However, in our experiments the depressor muscle always shows major activity, which results in fast opening of the jaws (Fig. 10.7). This rapid increase in gape is always accomplished by both a lower jaw depression (by the m. depressor mandibulae) and a cranial elevation (by the dorsal cervical muscles) as shown by the profiles of both kinematical characteristics (Fig. 4 in Cleuren and De Vree, 1992). In all examined species, the m. depressor mandibulae shows low-level activity, simultaneous with the activity in the jaw adductors during jaw closure (Fig. 10.7). This activity reaches a peak at the start of fast closing, falls off during further closing, and might peak again at jaw closure (Busbey, 1989; personal observations). Van Drongelen and DuUemeijer (1982) assigned a strain-regulating function to this activity peak during the crushing phase. However, Cleuren et ah (1995) showed activity levels of the depressor muscle going from 0 to 19%, whereas jaw adductors showed recruitment levels from 27 to 100% (measured relative to the maximal observed value per muscle). This, together with the fact that the m. depressor mandibulae only forms a small component (7.3%, see Cleuren et al, 1995) of the total physiological cross section and the proportional role of the bite force (a negative component of only 0.3% on the total bite force), makes a strain-regulating hypothesis questionable. Simulations with a static bite model (unpublished data) support this argument. Making the depressor muscle maximally active during crushing, simultaneous with all jaw adductors, results in a significant increase in joint force (10-34% for gape 0°, 10-23% for gape 10°, depending on the angle of the food reaction force), accompanied by a decrease in bite force (4% for gape 0° and 10°), and this in a phase where bite force seems to be crucial. The main differences in muscle activity can be found in the activity patterns of the jaw adductors (Fig. 10.7). In the first part of the feeding sequence, fast jaw closure is achieved by the simultaneous contraction of most jaw adductors, in which they show 10 to 30% of their maximal activity. In acquisition bites and killing/crushing bites (Fig. 10.7), this is followed by a crushing phase, which is characterized by the presence of pulsatile high-level activity of all closers (70 to 100%). Similar tetanic potentials were first described in lizard jaw muscles by Cans and De Vree (1986) in Trachydosaurus rugosus during crushing of snails. This mechanism of synchronized tetanus proves to be widely used by vertebrates in crushing hard prey.
10. Feeding in Crocodilians Toward the end of the feeding sequence, during intraoral transport, and swallowing, fewer adductors remain active during the closing of the jaws and their activity decreases gradually (Fig. 10.7). In swallowing cycles, only four jaw adductors remain active (Fig. 10.7). The duration of low-level activity (less than 10% of the maximal activity) after full jaw closure increases toward swallowing. Thus, based on activity pattern, jaw closers can be divided into two groups: group one, containing muscles that show major activity throughout the complete feeding process—the m. adductor mandibulae posterior, the deep part of the m. adductor mandibulae externus, the m. intramandibularis, and the m. pseudotemporalis. Group 2 contains the superficial part of the m. adductor mandibulae externus, and the m. pterygoideus anterior and posterior, which are only active when group one muscles show high levels of activity (Fig. 10.7). Histochemical data for the American alligator (Sato, 1992) revealed that the muscles of the first group consist of a large amount of red muscle fibers, whereas those of the second group consist of a high percentage of white fibers. As the distributions of fiber types is not homogeneous in crocodilian jaw muscles, precise knowledge of the placement of the probing electrodes is of crucial importance to reveal the relationships between activity pattern and histochemical characteristics. Van Drongelen and DuUemeijer (1982) described unusual activity patterns, which are unique in vertebrates, i.e., during prey drowning, the jaw adductors become active before jaw opening and remain active during opening and closing. During some other unspecified feeding activity, all adductors activate prior to jaw opening and remain active during opening, and the PTA, MAME, and MAMP become silent during jaw closing. None of these patterns resemble those reported by Busbey (1989) or were observed during our experiments.
2. Cervical
Musculature
Seidel (1978) included theoretical predictions on the function of the cervical musculature in his study of the axial musculature of A. mississippiensis, based on the morphological-topographical characteristics. He assumed that lateral movements of the skull and neck are caused by unilateral contractions of the ipsilateral side of certain muscles. However, electromyographical examination of the major neck muscles in C. crocodilus revealed that sideways movements are always produced by simultaneous activation of more than one muscle and by an interaction of the ipsilateral and contralateral side (personal observations).
353
All cervical muscles show extensive activity during the inertial feeding process. During swallowing, most muscles become silent. Straight lifting of the cranium during fast opening of the jaws is mainly caused by a bilateral contraction of all three muscles of the transversospinalis system: the m. transversospinalis capitis, m. spinocapitis posticus, and m. epistropheo-capitis. When head elevation is accompanied by a sideways shift, the m. transversospinalis capitis and m. iliocostalis capitis show a bilateral activity, but with the ipsilateral muscle at a higher intensity. The ipsilateral sides of the m. spinocapitis posticus and m. longissimus capitis superficialis then also show high-level activity, whereas the contralateral sides show low-level activity or are completely inactive. This bilateral difference is only obvious during large lateral head moveraents in the m. epistropheo-capitis. Simultaneous with cranial elevation, the neck is lifted by bilateral contraction of the m. transversospinalis cervicis, m. longissimus cervicis, and m. iliocostalis cervicis. Unilateral contractions of the m. iliocostalis cervicis cause lateral flexion of the neck. During intraoral transport, "cervicis" muscles seize their activity at the end of the fast-open phase, whereas "capitis" muscles remain active. This results in a continued elevation of the head and a static position of the neck in the fast-closing phase. Because the m. longissimus profundus only shows light activity during jaw closure, the downward displacement of the skull during this phase can probably be ascribed mainly to gravitational forces. The m. longissimus profundus, the only muscle positioned to function as a depressor of the neck, only shows major activity during the lifting of the neck, probably revealing a stabilizing function for the occipital joint. The fast elevation of the heavy crocodilian cranium causes immense inertial forces at the level of this joint. As manipulation of the neck of fixated specimens revealed no mechanical restriction of dorsal movement (150° backward elevation in Virchow, 1914), these forces cannot be counteracted by the presence of bony structures or ligaments. Simultaneous activation of an antagonistic muscle allows dosing of the dorsal movement, and thus minimizes inertial forces occurring at the occipital joint. The same principle is observed in contralateral muscles during sideways movements of the cranio-cervical complex. The occipital joint is thus stabilized during all head movements. In dorsad movements, inertial forces are counteracted by the most ventral neck muscle, in lateral movements by a contraction of the contralateral muscle. Most cervical muscles have multiple functions; muscles with cranial elevation as a major function also assist in the lateral head flexion or in the elevation of the neck and vice versa. Table 10.3 summarizes the role
354
Johan Cleuren and Frits De Vree
TABLE 10.3 Importance of Cervical Muscles in Elevation and Lateral Movements of the Cranium and Neck and in the Depression and Rotation of the Head Cranium Muscle
Elevation
TSC
***
SCP EC
Lateral
* ***
** *** *
*
***
Neck
DepresElevasion Rotation tion
*
***
LCe LCS
***
LCP ICCe ICC
Lateral
*
*
*
*sf*
*
cervical muscles play in the elevation and lateral movements of the neck and cranium, as well as in axial rotation and depression of the cranium. IV. EVOLUTION A. Diet in Relation to Skull Morphology: Long Snouted versus Short Snouted Among crocodilian species, variation in diet is strongly reflected in skull morphology and head shape. Many adaptations to diet have both an ecological and a biomechanical explanation. The development of an elongated snout as in ichthyophagous species, such as the gharials {Tomistotna schlegelii and Gavialis gangeticus) and the crocodiles {Crocodylus johnstoni, C. novaeguinea, and C. cataphractus), proves to be advantageous biomechanically. Long and narrow snouts offer less resistance to the water when sweeping sideways to catch fish and are also effective in probing for crabs in subterranean burrows (Pooley, 1989). However, a slender snout is too fragile to take large prey, which explains the evolution toward broad snouts in crocodilian species feeding on a more general diet. Besides a change in head shape, many other morphological characteristics of the crocodilian skull can be related to the feeding behavior, many of them incorporating engineering principles to increase the mechanical strength. In his work on inertial feeding, Gans (1969) stated that the forces required to accelerate either food or the head will tend to induce equal reaction forces on the body and tend to shift it. For floating or swimming crocodilians, these reaction forces are critical, as they must keep their body from shifting while their head manipulates the prey. According to Gans (1969), sus-
pended animals show two ways to counteract these reaction forces. They can induce equivalent but opposed forces with the appendages or they have the evolutionary option of decreasing the ratio of head mass to body mass. Terrestrial forms are able to transmit reaction forces to the substratum, utilizing the friction of the contact zones (Gans, 1969). However, in order to minimize emerging reaction forces, terrestrial species specializing in cranioinertial feeding might also be expected to show modification for the reduction of the craniocervical mass (Bramble and Wake, 1985). This tendency to very lightly built crania is demonstrated in carnivorous lizards and birds that regularly use cranioinertial feeding, such as Varanus, Tupinambis (Smith, 1982), and pigeon (Zweers, 1982; Zweers et al, 1994). At first sight, this tendency toward cranial slenderization seems to be absent in crocodilian skull. A mechanical explanation for its firmly built appearance can be found in the substantial forces that occur during the jerking and twisting manoeuvers in feeding behavior, especially seen in short-snouted crocodilian species. As this tactic is not used by the crocodilian species that feed exclusively on small prey animals, one might expect to find a lighter built cranium in exclusively ichthyophagous "long-snouted" crocodilians (Cleuren and De Vree, 1992). A first confirmation of this assumption can be found in the presence of the cranial osteodermic relief. This relief increases the mechanical strength of the flattened skull, and consequently also its resistance to fracture. Because the longitudinal crests of the osteodermic relief of short-snouted crocodilians coincide with the loads that occur in twisting of prey, this principle may apply to crocodilians (lordansky, 1973). It is further supported by the fact that crests are absent in all longsnouted crocodilians. Apparently, osteodermic relief, indeed, increases the mechanical strength of the crocodilian skull. According to Lanyon and Rubin (1985), local increases in mass can avoid points of potentially high stress. Many crests, lines, tuberosities, or local cortical thickenings in the crocodilian skull may thus be interpreted as local reactions that reduce stress concentrations. Their absence in longirostrine crocodilians supports the hypothesis that their cranium is not subjected to stresses equivalent to those in brevirostrine species (Cleuren and De Vree, 1992). Analogous to this, one would also expect that younger animals possess more lightly built skulls, as they commonly feed on relatively small animals, such as insects, fish, crabs, and small rodents. Mook (1922a) confirmed this hypothesis by stating that the skulls of young crocodilians show a relatively smiooth surface. In medium-sized skulls, the pitting is deeper and the surface rough. In old animals, the pitting and rugose
10. Feeding in Crocodilians condition of the surface of many of the bones is emphasized greatly. Other specific characters, such as oblique ridges in front of the orbits, median elevations of the snout, and facial ridges, are usually also emphasized in older animals. This ontogenetic variation in age also applies to the thickness of the bone (Mook, 1922a). Dodson (1975) registered the belief that isometry in the length of the skull with respect to body length and positive allometry of the jaws is an adaptation to everincreasing size of prey. The shape and proportions of the upper temporal fenestrae change dramatically during ontogeny and differ in long-snouted and shortsnouted species. Gavialis, for example, shows enlarged temporal fenestrae in contrast to the short-snouted Alligator. Apparently, the demand for thicker and more solid bone, and an increase in the osteodermic relief, is also related to age and may be associated with a change in diet. This hypothesis is further supported by observations of Webb and Messel (1978). They observed that Crocodylus porosus over 120 cm in total length eat more vertebrates; the change in diet is associated with broadening of the head. The secondary bony palate has considerable importance for the respiratory function, enabling the animal to breathe from the surface even when the mouth is open underwater. In addition to this function, it has an important second advantage in that it braces the long snout against heavy stresses engendered by the capture of large prey (Buffetaut, 1989). Ferguson (1981a,b) also recognized the engineering principle of tubular reinforcements in the form of the palate of the American alligator. Consistent with maintaining a light anterior snout, maxillary sinuses may serve as extra strengthening for the flat skull. B. Skull Morphology in Relation to the Bauplan of Jaw Adductors and the Cervical Musculature The evolutionary potential of feeding behavior is limited by the mechanical restrictions (physical arrangement of the muscles, tendon, bone, joint, etc.) on the capabilities of the musculature due to the morphology of the cranial and vertebral bones. With a given morphology, the crocodilian head and neck must adequately perform such varied and mechanically complex functions, such as capture of prey, manipulation and swallowing of prey, nest building, and care of the hatchlings. Solutions to all these problems require a high degree of refined adaptation. Considering the large sizes attained by some crocodilians, it will become apparent that the crocodilian neck is a highly specialized structure that meets its functional demands (Seidel, 1978).
355
The posthatchling skull undergoes a verticalization process caused by a downward growth of the basisphenoid and basioccipital (Romer, 1956) and the transformation of the diffuse sinus into a basicranial tube system (Tarsitano, 1985,1988). Among all living crocodilians, this verticalization process only differs in Gavialis gangeticus in the presence of a large, midsagittal, anterior pocket above the braincase. The verticalization of the basisphenoid also requires the verticalization of the posterior end of the pterygoid, which is likely to change the angle of force application of the m. pterygoideus and allows a larger volunie of this muscle (Tarsitano, 1985, 1989). Together with the enlarged volume in G. gangeticus, the origin of the pterygoid muscle is shifted posteriorly by a posterior elongation of the processus retroarticularis (Fig. 10.2). This elongation also lengthens the moment arm of the depressor muscle and therefore improves its force efficiency. The morphology of the m. pseudotemporalis is related to the head shape, as it has been enlarged at the expense of the m. pterygoideus anterior in longirostrine species, whereas the reverse trend is observed in brevirostrine crocodilians. This suggests quicker and stronger muscle contractions in short-snouted species (lordansky, 1964). This is demonstrated further in the G. gangeticus (Tarsitano, 1989) and Tomistoma schlegelii (Kalin, 1933), which differ from other living crocodilians in having larger, fairly vertical supratemporal fenestrae (Fig. 10.2). In Gavialis, the expanded volume of the supratemporal fenestra allows thickening of the pseudotemporalis muscle, a strategy for muscle enlargement that differs from the one observed in tomistomines (despite the presence of enlarged supratemporal fenestrae) and other crocodylines and alligatorines. In these forms the width of the skull table is reduced and the supratemporal fenestrae have moderate dimensions or may be closed entirely by an expansion of the parietal, postorbital, and squamosal (sometimes in Osteolaemus, usually in Paleosuchus; Kalin, 1933), which results in a different method of housing the m. pseudotemporalis. In most eusuchian genera, except in Crocodylus porosus (Tarsitano, 1989), the pseudotemporal muscle is elongated posteriorly as it extends posteriorly within the temporal fenestra along a pulley or trochlear surface (Lakjer, 1926; Schumacher, 1973). lordansky (1964) promoted the hypothesis that the cartilago transiliens can be used as a locking mechanism to keep the jaws open without muscle activity of the depressor muscle. He supposed that this behavior functions during thermoregulation, as it is frequently observed in crocodilians when lying on a riverbank, sunbathing with fully opened jaws. With a wide gape, the pterygoidal flanges are displaced from under the
356
Johan Cleuren and Frits De Vree
mandibular adductor tendon and are placed immediately above and in front of this tendon. By activation of the m. intramandibularis, the tendon is drawn forward, downward, and outward, together with the attached cartilago transiliens. This fixes the position of the widely opened jaws because the passage of the pterygoidal flanges under the adductor tendon will be blocked. Activation of dorsal adductor muscles will normalize the position of the tendon and thus close the jaws (lordansky, 1964). Van Drongelen and Dullemeijer (1982) extended this locking function to the closed jaw position as they observed no adductor activity during drowning of prey. This suggests that the cartilago can be manipulated in dorsal and ventral direction by the attached muscles. To test their hypothesis, they injected amalgam in the cartilago to determine its position on lateral X-ray photographs. With fully closed jaws, the cartilago is positioned caudodorsal to the edge of the torus transiliens (ventral extension of the pterygoids), whereas with fully opened jaws, it lies ventral to it. These observations, together with manipulation experiments, confirmed their hypothesis. During electromyographical experiments on C. crocodilus (personal observations), it was observed that jaws can be kept open without any muscle activity and that a change in jaw position is often preceded by activity of the m. intramandibularis. These observations further support the hypothesis. Schumacher (1973) and Ferguson (1981) suggested that the m. intramandibularis, which runs exclusively in the inferior dental canal alongside the persistent Meckel cartilage, may act as an antagonist to the mandibular adductors during jaw opening, thus preventing distortion of the cartilago transiliens and the mandibular adductor tendon. Contraction of the m. intramandibularis during adduction would stretch the fibers of the mandibular adductor muscles (m. pterygoideus anterior and m. adductor mandibulae externus profundus), thus broadening the length-tension curve of these muscles, giving them a larger range of isometric contraction. This hypothesis is supported by irregular activity in the m. intramandibularis prior to a changing gape (Busbey, 1989; personal observations). The position of the lower jaw in crocodilians is controlled by eight jaw adductors and one opener, all pulling in different directions. To allow determination of the role each muscle plays during the five observed bite types and to estimate the force that each muscle can exert, a static bite model was developed by Sinclair and Alexander (1987). Their simulation was based on the assumption that muscle forces are proportional to the physiological cross section and that all muscles are fully active simultaneously. As these conditions conflict with reality, Cleuren et ah (1995) improved the
model by using the actual recruitment levels of the jaw muscles, which were determined by a quantitative electromyographical analysis. Given a range of orientations of the food reaction force, the magnitude of the bite force and the orientation and magnitude of the joint forces were calculated. Their model showed that bite forces are largely deterniined by changes in the orientation of the muscle forces, a finding with two important biological implications. By using different compartments of complex muscles, crocodilians are able to modulate bite force extensively, and slight morphometric differences may determine a shift in the feeding ecology of closely related species. The model also showed that the different direction of pull and the modification of the force level of each individual muscle not only affected bite force but also determined the magnitude and angle of the forces occurring at the level of the jaw joint. The orientation of the joint forces always fits within the heavily ossified triangle at the level of the jaw suspension. The anteriorly pointing leg of this triangle is formed by the massive quadrate, which inclines medially. The quadratojugal and jugal form the other leg, i.e., the lower temporal bar, a strong bony strut pointing rostrally in a sagittal plane. This means that joint forces in C. crocodilus result in compressive loading of both bony legs of the triangle, irrespective of the orientation of the food reaction forces. The more they point forward, the higher the lower temporal bar will be loaded, as forward pointing food reaction forces coincide with increasing joint force magnitudes and decreasing joint force angles, which tend to come in line with the lower temporal bar. The sagittal position of the lower temporal bar ensures pure axial loading during symmetrical muscle activity (observed during holding and crushing; Cleuren et ah, 1995). In the case of the quadrate, the joint forces participate in a bending moment too. This might explain why, despite the much smaller axial loading, the quadrate appears to be stronger built than the lower temporal bar. As the orientation in which the caiman can expect and thus must also absorb joint forces is highly determined by its jaw muscle morphology, reinforcements of the skull can be limited to the essential structures and therefore minimized. This also fits into the hypothesis of Bramble and Wake (1985) that terrestrial species specializing in cranioinertial feeding are expected to show modification for the craniocervical mass in order to minimize inertial forces on the body. A study of the form of the lower jaw of C. crocodilus (Van Drongelen and Dullemeijer, 1982) provided further evidence for this hypothesis. For each food-intake action, the amount of bony material necessary to resist muscle force and the required specific shape of the mandible was calculated
10. Feeding in Crocodilians
in this study. Except for the difference in the level of the mandibular fenestrae, the integrated shape highly resembled the form of the actual mandible. As further distribution of material is impossible due to the support for dentition, the articulation, and the muscle attachments, a minimization of the required material for the same mechanical demand is only possible in the area between the joint and the dentition. This finding is evident in the position and shape of the mandibular fenestra. As the cervical muscles play a very important role in the rapid head movements during inertial feeding, they may also affect the morphology of the crocodilian skull. The presence of the many powerful neck muscles could result in the evolutionary tendency toward enlargement and reinforcement of the insertion area, and thus an increase in the size and mass of the occipital region. However, this is not observed as it would influence the inertial feeding process negatively. As the insertions of the neck muscles are nearly always tendinous, an increase in musculature is permitted without the need for an expansion of the attachment areas. The placing of these occipital insertion points as far as possible away from the occipital condyle also achieves the maximum length of lever arm. Given the length of the fibers of the cervical-occipital muscles, an increase of the moment arm results in a mechanical advantage in terms of reduced force requirements.
References Bels, V. L., M. Chardon, and K. V. Kardong (1994) Biomechanics of the hyolingual system in Squamata. Pp. 197-240. In: Biomechanics of Feeding in Vertebrates. Vol. 18. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer Verlag, Berlin. Boschma, H. (1920) Das halsskelet der Krokodile. Leiden: Tijdschrift der Nederlandsche Dierkundige Vereniging, serie 2, 18:85-123. Bramble, D. M., and D. B. Wake (1985) Feeding mechanisms of lower tetrapods. Pp. 230-261. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge. Briihl, C. B. (1862) Das Skelett der Krokodilinen, dargestelt in 20 Tafeln (Icones ad Zootomiam illustrandam). Wilhelm Braumiiller, Wien. Buffetaut, E. (1989) Evolution. Pp. 26-41. In: Crocodiles and Alligators. C. A. Ross (ed.). Merehurst Press, London. Busbey, A. B. (1989) Form and function of the feeding apparatus of Alligator mississippiensis. J. Morphol. 202:99-127. Cleuren, J. (1996) Functionele morfologie van het craniocervicaal en hyolinguaal apparaat van Caiman crocodilus tijdens de inertiele voedselopname. Unpublished Ph.D. Dissertation, University of Antwerp (UIA), Belgium. Cleuren, J., and F. De Vree (1992) Kinematics of the jaw and hyolingual apparatus during feeding in Caiman crocodilus. J. Morphol. 212:141-154. Cleuren, J., P. Aerts, and F. De Vree (1995) Bite and joint force analysis in Caiman crocodilus. Belg. J. Zool. 125:79-94. Cott, H. B. (1961) Scientific results of an inquiry into the ecology
357
and economic status of the Nile crocodile {Crocodylus niloticus) in Uganda, northern Rhodesia. Trans. Zool. Soc. Lond. 29:211-356. Cundall, D. (1983) Activity of head muscles during feeding by snakes: a comparative study. Am. Zool. 23:383-396. Cundall, D., and C. Cans (1979) Feeding in water snakes: an electromyographic study. J. Exp. Zool. 209:189-208. Davenport, J., D. J. Grove, J. Cannon, T. R. Ellis, and R. Stables (1990) Food capture, appetite, digestion rate and efficiency in hatchling and juvenile Crocodylus porosus. J. Zool. Lond. 220:569-592. Densmore, L. D., Ill, and R. D. Owen (1989) Molecular systematics of the order Crocodilia. Am. Zool. 29:831-841. De Vree, R, and C. Cans (1989) Functional morphology of the feeding mechanisms in lower tetrapods. Pp. 115-127. In: Fortschritte der Zoologie, Vol. 35. (H. Splechtna and H. Hilgers (eds.). Gustav Fisher Verlag, New York. De Vree, R, and C. Cans (1994) Feeding in tetrapods. Pp. 93-118. In: Biomechanics of Feeding in Vertebrates, Vol. 18. (V L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Dodson, P. (1975) Functional and ecological significance of relative growth, in Alligator. J. Zool. (London) 175:315-355. Edmund, A. G. (1969) Dentition. Pp. 117-200. In: Biology of the Reptilia, Vol. 1. (C. Gans, A. d'A. Bellairs and T. S. Parsons (eds.). Academic Press, London. Ferguson, M. W. J. (1981a) The value of the American alligator. Alligator mississippiensis as a model for research in craniofacial development. J. Craniofac. Genet. Dev. Biol. 1:123-144. Ferguson, M. W. J. (1981b) The structure and development of the palate in Alligator mississippiensis. Arch. Oral. Biol 26:427-443. Ferguson, M. W. J. (1984) Craniofacial development in Alligator mississippiensis. Symp. Zool. Soc. Lond. 52:223-273. Frey, E. (1984) Aspects of the biomechanics of crocodilian terrestrial locomotion. Pp. 93-98. In: Third Symposium on Mesozoic Terrestrial Ecosystems W. E. Reif and F. Westphal (eds.). Attempto Verlag, Tubingen. Frey, E. (1988) Anatomie des Korperstammes von Alligator mississippiensis Daudin (Anatomy of the body stem of Alligator mississippiensis Daudin). Stuttgarter Beitr. Naturk. Ser. A. Nr. 424:1-106. Furbringer, M. (1922) Das Zungenbein der Wirbeltiere, insbesondere der Reptilien und Vogel. Abh. Heidelb. Akad. Wis. 11:1-164. Gans, C. (1961) The feeding mechanism of snakes and its possible evolution. Am. Zool. 1:217-227. Gans, C. (1969) Comments on inertial feeding. Copeia 4:855-857. Gans, C. (1992) Electromyography. Pp. 175-204. In: Biomechanics Structures and systems. A. A. Biewener (ed.). Oxford Univ. Press, Oxford. Gans, C , and F. De Vree (1986) Shingle-back lizards crush snails shells using temporal summation (tetanus) to increase the force of the adductor muscles. Experientia 42:387-389. Gauthier, J. A., A. G. Kluge, and T. Rowe (1988) Amniote phylogeny and the importance of fossils. Cladistics 4:105-209. Gnanamuthu, C. P. (1937) Comparative study of the hyoid and tongue of some typical genera of reptiles. Proc. Zool. Soc. Lond. 107:1-63. Hecht, M. K., and S. F. Tarsitano (1983) On the cranial morphology of the Protosuchia, Notosuchia and Eusuchia. N. Jb. Geol. Paleont.Mh. 11:657-668. Hoffstetter, R., and J. P. Gasc (1969) Vertebrae and ribs of modern Reptiles. Pp. 201-310. In: Biology of Reptilia, Vol. 1. C. Gans, A.d'A. Bellairs, and T. S. Parsons (eds.). Academic Press, London. lordansky, N. N. (1964) The jaw muscles of the crocodiles and some relating structures of the crocodilian skull. Anat. Anz. 115:256280. lordansky, N. N. (1973) The skull of the crocodiles. Pp. 201-262. In Biology of the Reptilia, Vol. 4. C. Gans (ed.). Academic Press, London.
358
J o h a n C l e u r e n a n d Frits D e Vree
Kalin, J. A. (1933) Beitrage zur vergleichenden Osteologie des Crocodilidenschadels. Zool. Jb. Abt. Anat. 57 (4): 535-714. Kalin, J. A. (1955) Crocodilia. Pp. 695-783. In: Traite de Paleontologie, Vol. 5. J. Piveteau (ed.). Masson et Cie., Paris. Kieser, J. A., C. Klapsidis, L. Law, and M. Marion (1993) Heterodonty and patterns of tooth replacement in Crocodylus niloticus. J. Morphol. 218:195-201. Lakjer, T. (1926) Studien iiher die Trigeminus-versorgte Kaumuskulatur der Sauropsiden. C. A. Reitsel, Copenhagen. Langston, W., Jr. (1973) The Crocodilian skull in historical perspective. Pp. 263-284. In: Biology of the Reptilia, Vol. 4. C. Cans (ed.). Academic Press, London. Lanyon, L. E., and C. T. Rubin (1985) Functional adaptation in skeletal structures. Pp. 1-25. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge. Lubosch, W. (1914) Zwei vorfaufige Mitteilungen iiber die Anatomie der Kaumuskeln der Krokodile. Jena. Zschr. Naturw. 51:697-706. Lubosch, W. (1933) Untersuchungen liber die Visceralmuskulatur de Sauropsiden. Morph. Jb. 72:584-666. Luther, A. (1914) tjber die vom N. trigeminus versorgte Muskulatur der Amphibien, mit einem vergleichenden Ausblick iiber den Adductor mandibulae der Gnathostomen, und einem Beitrag zum Verstandnis der Organisation der Anurenlarven. Acta Soc. sc.Fenn. 44(7): 1-115. Miall, L. C. (1878) The skull of the crocodile, a manual for students. Stud. Comp. Anat. 1:1-50. Mook, C. C. (1921a) Individual and age variations in the skulls of recent Crocodilia. Bull. Am. Mus. Nat. His. 44:51-66. Mook, C. C. (1921b) Notes on the postcranial skeleton in the Crocodilia. Bull. Am. Mus. Nat. His. 44:67-100. Mook, C. C. (1921c) Skull characters of recent Crocodilia, with notes on the affinities of the recent genera. Bull. Am. Mus. Nat. His. 44: 123-268. Poe, S. (1996) Data set incongruence and the phylogeny of crocodilians. Syst. Biol. 45:393-414. Poglayen-Neuwall, I. (1953) Untersuchungen der Kiefermuskulatuur und deren Innervation an Krokodilen. Anat. Anz. (Jena) 99:257277. Pooley, A. C , and C. Cans (1976) The nile crocodile. Sci. Am. 234: 114-124. Pooley, A. C. (1989) Food and feeding habits. Pp. 76-91. In: Crocodiles and Alligators. C. A. Ross (ed.). Merehurst Press, London. Romer, A. S. (1956) Osteology of the Reptiles. University of Chicago Press, Chicago. Rowe, T. (1986) Homology and evolution of the deep dorsal thigh musculature in birds and other Reptilia. J. Morphol. 189: 327-346. Sato, I., K. Shimada, A. Yokoi, J. C. Handal, N. Asuwa, and T. Ishii (1990) Morphology of the teeth of the American Alligator {Alligator mississippiensis): fine structure and chemistry of the enamel. J. Morphol. 205:165-172. Sato, I., K. Shimada, T. Sato, and T. Kitagawa (1992) Histochemical study of jaw muscle fibers in the American Alligator {Alligator mississippiensis). J. Morphol. 211:187-199. Schaller, G. B., and P G. Crawshaw (1982) Fishing behavior of Paraguayan Caiman {Caiman crocodilus). Copeia 1:66-72. Schumacher, G. H. (1973) The head muscles and hyolaryngeal skeleton of turtles and crocodilians. Pp. 101-199. In: Biology of the Reptilia, Vol. 4. C. Gans (ed.). Academic Press, London. Schwenk, K., and G. S. Throckmorton (1989) Functional and evolutionary morphology of lingual feeding in squamate reptiles: phylogenetics and kinematics. J. Zool. (Lond.) 219:153-175.
Seidel, R. (1978) The Somatic Musculature of the Cervical and Occipital Regions of Alligator mississippiensis. Ph.D. Dissertation, City University of New York, NY. Sewertzoff, S. A., Jr. (1929) Zur Entwicklungsgeschichte der Zunge bei den Reptilien. Acta Zool. 10:231-341. Shimada, K., I. Sato, A. Yokio, T. Kitagawa, M. Tezuka, and T. Ishii (1990) The fine structure and elemental analysis of keratinized epithelium of the filiform papillae analysis [sic] on the dorsal tongue in the American alligator {Alligator mississippiensis). Okajimas Folia Anat. Japan 66:375-392. Shimada, K., I. Sato, and H. Moriyama (1992) Morphology of the tooth of the American Alligator {Alligator mississippiensis): the fine structure and elemental analysis of the cementum. J. Morphol. 211:319-329. Shimada, K., I. Sato, and H. Ezure (1993) Morphological analysis of tendinous structure in the American alligator jaw muscles. J. Morphol. 217:171-181. Sill, W. D. (1968) The zoogeography of the Crocodilia. Copeia 1968: 76-88. Sinclair, A. G., and R. McN. Alexander (1987) Estimates of forces exerted by the jaw muscles of some reptiles. J. Zool. (London) 213: 107-115. Smith, K. K. (1982) An electromyographic study of the function of the jaw adducting muscles in Varanus exanthematicus (Varanidae). J. Morphol. 173:137-158. Smith, K. K. (1986) Morphology and function of the tongue and hyoid apparatus in Varanus (Varanidae, Lacertilia). J. Morphol. 187: 261-287. Sondhi, K. C. (1958) The hyoid and associated structures in some Indian reptiles. Ann. Zool. Agra. 2:155-240. Steel, R. (1973) Crocodilia. Handbuch der Paleontologie 16:1-116. Tanner, W. W., and D. F Avery (1982) Buccal floor of reptiles, a summary. Great Basin Nat. 42 (3): 273-349. Tarsitano, S. F (1985) Cranial metamorphosis and the origin of the Eusuchia. N. J. Geol. Palaont. 170(1): 27-41. Tarsitano, S. F, E. Frey, and J. Reiss (1989) The evolution of the Crocodilia: a conflict between morphological and biomechanical data. Am. Zool. 29:843-856. Taylor, J. A. (1979) The foods and feeding habits of subadult Crocodylus porosus Schneider in Northern Australia. Aust. Wildl. Res. 6:347-359. Troxell, E. L. (1925) Mechanics of Crocodile vertebrae. Bull. Geol. Soc. Am. 36:605-614. Van Bemmelen, J. F (1887) Beitrage zur kenntniss der Halsgegend bei Reptilien. I. Anatomischer theil. P. W. M. Trap, Amsterdam. Van Drongelen, W., and P. Dullemeijer (1982) The feeding apparatus of Caiman crocodilus, a functional-morphological study. Anat. Anz. 151:337-366. Virchow, H. (1914) Uber die AUigatorwirbelsaule. Arch. Anat. 1914: 103-142. Webb, G. J. W., and H. Messel (1978) Morphometric analysis of Crocodylus porosus from the north coast of Arnhem Land, northern Australia. Aust. J. Zool. 26:1-27. Wermuth, H. (1953) Systematik der rezenten Krokodile. Mitt. Zool. Mus. Berlin 29:375-514. Zweers, G. A. (1992) Pecking of the pigeon {Columba livia L.). Behavior 81:173-230. Zweers, G. A., H. Berkhoudt, and J. C. Vanden Berge (1994) Behavioral mechanisms of avian feeding. Pp. 241-279. In: Biomechanics of Feeding in Vertebrates, Vol. 18. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin.
C H A P T E R
11 Feeding in Paleognathous Birds CAROLE A. BONGA TOMLINSON Department of Organismic and Evolutionary Biology Museum of Comparative Zoology Harvard University Cambridge, Massachusetts 02138 I. INTRODUCTION II. MATERIALS AND METHODS III. MORPHOLOGY OF THE HYOLINGUAL APPARATUS A. Neognathous Birds B. Paleognathous Birds IV. FUNCTION OF THE HYOLINGUAL APPARATUS A. Lingual Feeding in a Generalized Neognath (Meleagris gallopavo) B. Cranioinertial Feeding in Paleognaths C. Comparison of Ratite Cranioinertial and Neognathous Lingual Feeding V. EVOLUTION OF THE FEEDING SYSTEM A. Avian Phylogeny and Outgroup Choice B. Primitive Condition of the Neornithine Hyolingual Apparatus C. Changes in Feeding Function during the Theropod-Bird Transition D. Proposed Functional Evolution of Early Avian Transport Mechanisms E. Evolutionary Morphology: An Overview F. Conservation of Pattern Generation G. Phylogenetic Relationships References
1988; Padian and Chiappe, 1998; Sibley and Ahlquist, 1990). The large, sometimes giant, flightless ratites include 10 species in six extant genera {Struthio, the African ostrich; Rhea and Pterocnemia, South American rheas; Dromaius, the Australian emu; Casuarius, Nev^ Guinea and Australian cassowraries; and Apteryx, the kiw^is of New Zealand) and two extinct groups (the elephantbirds of Madagascar and Africa and the moas of New Zealand). Ratites are believed to have reached their modern pattern of distribution on southern land masses by means of vicariance and dispersal via land routes across Antarctica during the late Cretaceous and/or early Tertiary, approximately 80 to 50 million years ago (Cracraft, 1973,1974,1986,1988; van Tuinen, 1998; see also Sampson et ah, 1998). Tinamous are moderate to small-sized, volant birds (nine genera, 47 species) restricted to the Neotropics and savannas of South America. Superficially they resemble the neognathous galliforms (pheasants, fowl). Parkes and Clark (1966) proposed that a "proto-tinamou" was ancestral to all ratites and tinamous. Kurochkin (1995) listed fossil birds that he considered paleognaths that occurred worldwide in the Cretaceous and early Tertiary (see also Alvarenga, 1983; Alvarenga and Bonaparte, 1988; Houde, 1988; Houde and Haubold, 1987; Houde and Olson, 1981; Peters, 1988; Tambussi, 1995), but he included no tinamou. Feduccia (1996) stated that the earliest tinamou fossils date only from the Miocene of South America. Paleognathous birds possess a small tongue, a mostly cartilaginous hyobranchial skeleton and feed cranioinertially (Bramble and Wake, 1985; Ftirbringer, 1922; Lang, 1956; McLelland, 1979; Mtiller, 1963; Parker, 1866; Parker, 1891; Pycraft, 1900; Webb, 1957). Neognathous birds are primarily hyolingual (tonguebased) feeders. Nonetheless, cranioinertial feeding is
L INTRODUCTION Within Neornithes (modern birds), monophyly of the nominal taxa Paleognathae (ratites and tinamous; Pycraft, 1900) and Neognathae (all other modern birds, >8600 species; Sibley and Monroe, 1990) is disputed. Divergence of the two putative lineages may have occurred as long ago as 120 million years ago during the Cretaceous period (Cooper and Penny, 1997; Rambaut and Bromham, 1998; van Tuinen et ah, 1998), but there is no consensus on which group is more phenotypically primitive (see Cracraft and Mindell, 1989; Houde, FEEDING (K. Schwenk,ed.)
359
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
360
Carole A. Bonga Tomlinson
increasingly evident in neognaths as food size increases (e.g., Columba; Zweers, 1982a,b), thus all birds may be capable of using cranioinertial feeding to some extent (Bramble and Wake, 1985). Few neognathous birds are known to be obligate cranioinertial feeders (e.g., Egretta; Homberger, 1989), but all paleognathous taxa are presunied to share this behavior because the small tongue would seem to preclude lingual feeding (see Gussekloo and Zweers, 1997; Tomlinson, 1997a). Bock and Blihler (1988) proposed that tongue reduction occurred independently in at least two paleognathous groups (ostrich vs all other paleognaths) as a result of similar feeding patterns in which large food items are swallowed whole (thus reducing the functional role of the tongue). The ancestral condition for birds is assumed to be an unreduced tongue and an extensively ossified hyobranchium (Bock and Biihler, 1988). Ecological studies of paleognathous species, however, show that extant paleognaths are neither uniform in the habitats they occupy (arid grasslands and tropical forest) nor in their omnivorous diets. They choose foods of various sizes, and in many taxa, large food items are broken up by pecking before ingestion (del Hoyo et al, 1992). Furthermore, primitive character states of the avian hyolingual apparatus (tongue, hyobranchium, and musculature) remain undetermined, a problem exacerbated by the fact that the closest relevant outgroups (toothed Mesozoic birds, theropod dinosaurs, or thecodont reptiles) are extinct (see Chiappe, 1995; Cooper and Penny, 1997; Feduccia, 1994,1996; Hecht, 1985; Martin, 1983,1985,1987; Molnar, 1985; Ostrom, 1969, 1973, 1985, 1991; Padian and Chiappe, 1998; Welman, 1985). The modern avian hyobranchial apparatus differs significantly from those of other extant reptiles, yet its evolution has not been addressed. For example, the ceratohyals characteristic of nonavian reptiles (henceforth referred to simply as "reptiles" for convenience) have been lost in birds, and a novel element, the paraglossal, occurs within the tongue (Crompton, 1953). The muscular tongue of most reptiles is absent in birds and intrinsic hyolingual muscles that connect hyobranchial elements have taken the place of intrinsic lingual muscles. The hyobranchium is located in the neck region of most reptiles and retractor muscles are attached to the sternopectoral region (Busbey, 1989; Cleuren and De Vree, 1992; Delheusy et al, 1994; Kesteven, 1944; Oelrich, 1956; Schumacher, 1973; Smith, 1984, 1986; Sondhi, 1958), whereas in birds the hyobranchium is located immediately behind and beneath the mandible, and the major retractors originate on the laterocaudal surfaces of the mandible (see Baumel et ah, 1993; Bhattacharyya, 1980; Burton, 1984; Homberger, 1986; Homberger and Meyers, 1989; Zweers, 1982b). Changes in hyolingual function that accompanied
this structural transformation are also largely unknown, although the basic patterns of cranioinertial transport and swallowing are believed to be similar in birds and some extant reptiles (Smith, 1992; see Suzuki and Nomura, 1975). Nevertheless, mechanisms of intraoral transport clearly differ between birds and reptiles: reptiles possess teeth and (with the exception of crocodilians) a fleshy, muscular tongue capable of movement independent of the hyobranchium (e.g.. Smith, 1984,1988; Schwenk, 1986,1988; see Chapters 2 and 8), whereas in modern birds, teeth are absent and tongue movements depend entirely on movements of the hyobranchial skeleton (Zweers, 1974, 1982a,b). Nearly all reptiles employ some form of hyolingual feeding, with cranioinertial feeding exceptional (e.g., in crocodilians. Chapter 10; and Varanus, Chapter 8; Cans, 1969). The muscular, manipulative tongue of parrots, with highly differentiated, intrinsic hyolingual musculature, is uniquely derived and presumably associated with their ability to position seeds within the beak for husking (Homberger, 1986; see Chapter 2). Thus, the neornithine feeding apparatus is significantly different from that of other modern reptiles, and within birds there is a deep phenotypic and phylogenetic divergence between paleognathous and neognathous forms. This divergence offers an opportunity to examine evolutionary transformations in the avian feeding apparatus in the light of outgroup comparison. Although higher-level phylogenetic relationships among birds are highly contentious and largely unresolved (see Chapter 12), Fig. 11.1 presents one generally accepted phylogeny for the relationships of the paleognath taxa relative to the Neognathae. Although not all phylogenetic analyses agree with this hypothesis, it is used here because it is most consistent with form and function of the hyolingual apparatus (see later). The purpose of this chapter is twofold. First, the morphology and function of the paleognathous hyolingual apparatus are described and compared to the generalized neognathous condition. Second, these data are compared to comparable data for fossil and extant reptilian outgroups in order to determine whether the paleognathous or the neognathous condition is representative of the primitive condition for neornithine birds. The evolutionary origins of the modern avian hyolingual apparatus and of two basic types of avian feeding—ratite (obligate) cranioinertial feeding and avian lingual feeding—are discussed. IL MATERIALS A N D M E T H O D S Morphological and functional comparisons are based on dissections and cineradiographic films of
361
11. F e e d i n g in P a l e o g n a t h o u s Birds
Paleognathae
I Toothed Mesozoic Early Theropods Birds Paleognaths Emu
Other ratites
Ostrich
Rhea
1
Lithornithlds Tinamous
Neognathae
F I G U R E 11.1. Phylogenetic relationship of paleognathous birds that is most consistent with form and function of the feeding apparatus.
feeding in three ratite species (South American greater rhea, Rhea americana; Australian emu, Dromaius novaehollandiae; and African ostrich, Struthio camelus) in comparison to a neognathous species (North American wild turkey, Meleagris gallopavo). Additional anatomical data were obtained for a Chilean tinamou {Nothoprocta perdicaria). These observations were supplemented by reference to the literature wherever possible. The wild turkey is assumed to be representative of generalized, ground-pecking, lingual-feeding birds. It belongs to the order Galliformes, widely regarded as basal within the neognath clade (see Sibley and Ahlquist, 1990; Zweers, 1985, 1991a,b; Zweers et al, 1994). Morphological character state polarities were determined by comparison of data to descriptions of hyolingual structure in modern reptiles, as well as fossils of theropod dinosaurs and toothed Mesozoic birds. In addition, hyobranchial function in a lepidosaurian reptile (tuatara, Sphenodon punctatus; also see Chapter 8) is compared to bird data in an effort to discern polarities in avian functional patterns. High-speed cineradiographic films (100 fps: Struthio, Sphenodon; 200 fps: Rhea, Dromaius, Meleagris) were recorded using a Siemens cine X-ray machine with a Sirecon image intensifier attached to an Eclair GV-16 camera {Struthio and Sphenodon films were made by
earlier workers) and digitized using a Vanguard motion analyzer and custom software. Functional data for other taxa are taken from the literature on neognathous birds, reptiles, and fossil taxa, as noted. III. MORPHOLOGY OF THE HYOLINGUAL APPARATUS A general description of the neognathous hyolingual apparatus (tongue, hyobranchial skeleton, and associated musculature) given later provides a basis for comparison with the paleognathous condition that follows. Muscle terminology follows Nomina Anatomica Avium (Baumel et al, 1993) and/or Homberger and Meyers (1989) wherever possible, but modifications and descriptive clarifications are added to supplement the incomplete, inadequate, and often inaccurate literature on paleognathous taxa (Bock and Biihler, 1988; Kesteven, 1945; Webb, 1957; see Table 11.1). The terms for muscle groups (e.g., intrinsic and extrinsic hyolingual, hyolaryngeal) in neognathous taxa differ in Baumel et al (1993), Homberger and Meyers (1989), and Zweers (1982b), and here follow Baumel et al (1993), with some exceptions (see Table 11.2). Avian terminology for muscle groups, however, differs from that
362
Carole A. Bonga Tomlinson
used in reptiles and mammals, and attempts are made to clarify these differences in the text. Muscle innervation in neognathous taxa follows Bubien-Waluszewska (1981) and Kontges and Lumsden (1996) and in paleognathous taxa follows Kesteven (1945) and Webb (1957). A. Neognathous Birds The hyolingual apparatus is well documented for several lingual- and filter-feeding neognathous taxa (e.g., Bhattacharrya, 1980; Burton, 1984; Homberger and Meyers, 1989; Kallius, 1905; Zweers, 1974,1982b). Hyolingual morphology in the turkey, M. gallopavo, is used to characterize the general neognathous pattern for comparison to paleognathous species (Figs. 11.2 and 11.3). The dorsal surface of the tongue is cornified and covered by filamentous papillae (Fig. 11.2B).
The hyobranchium consists of seven elements. There are three articulated median elements: a paraglossal within the tongue, a basihyal within the tongue base, and a urohyal anteroventral to the larynx. Paired, lateral elements consist of the hyoid horns (cornua), which articulate with the lateral surface of the basihyal. Each cornu is formed by a ceratobranchial that curves around the larynx anteroventrally and laterally, articulating distally with an elongate epibranchial. The epibranchial curves upward from below the mandible to the occipital region, where it is attached to the skull by connective tissue (fascia vaginalis hyoideus; Homberger, 1986; Homberger and Meyers, 1989). In neognathous embryos, the basihyal and urohyal form a single cartilaginous anlage that later separate; the basihyal always ossifies, but the urohyal may not. The paraglossal originates as two cartilaginous "paraglossalia" that fuse and ossify last (Kallius, 1905; Fiirbringer, 1922).
TABLE 11.1 Synonymous Terms for Muscles Acting on the Hyolingual Apparatus in Paleognathous Species Described in This Study and Previous Works ^ Present study (Rhea, Struthio, and Dromaius)
Kesteven (1945) {Struthio and Dromaius)
Webb (1957) (Struthio)
Bock and Buhler (1988) (all paleognathous species)
M. constrictor colli cervicalis
Second dorsal superficial constrictor
Constrictor colli (cucullaris)
M. constrictor colli intermandibularis
Second ventral superficial constrictor, pars posterior
Constrictor colli (cucullaris)
M. intermandibularis
First ventral superficial constrictor
Mylohyoideus
M. branchiomandibularis
Hyomandibularis
Ceratomandibularis
Branchiomandibularis Genioglossus, medial slip
M. genioglossus M. genioceratohyoideus*
Geniohyoideus
M. serpihyoideus
Second ventral superficial constrictor, pars anterior
M. hyomandibularis* (== M. H. lateralis* in Dromaius) (absent in Rhea)
Genioglossus
Serpihyoideus Hyomandibularis medialis
M. hyomandibularis medialis* (absent in Rhea and Struthio)
Interhyoideus
M. cricohyoideus
? Thyro-hyoideus + ? Ceratohyoideus
M. ceratocricoideus* (absent in Rhea)
Ceratothyroideus
Genioglossus
? Stylohyoideus
' Stylohyoideus
M. basiarytaenoideus* M. ceratohyoideus
Ceratohyoideus
M. ceratoglossus
Ceratoglossus
M. hyoglossus* (sling absent in Struthio)
Hypoglossus obliquus (absent in Struthio)
''An attempt was made to standardize nomenclature in accordance with Nomina Anatomica Avium (Baumel et ah, 1993; see Table 11.2). An asterisk (*) indicates a muscle that is unique to paleognathous species. Uncertain homologies are indicated with a question mark. A blank indicates that the muscle was not described.
363
11. Feeding in Paleognathous Birds TABLE 11.2 S y n o n y m s for Avian H y o l i n g u a l M u s c l e s and M u s c l e Groups^ in Paleognathous Species, W i l d Turkey, and Other N o n s p e c i a l i z e d N e o g n a t h o u s Species^ Paleognathous species This study {Rhea, Struthio, and/or Dromaius) External hyolingual mm M. constrictor colli cervicalis M. constrictor colli intermandibularis M. Intermandibularis
Neognathous species This study (Meleagris)
Constrictor colli cervicalis Constrictor colli intermandibularis Intermandibularis
Zweers (1982b) {Columba) External hyoid mm^ Cutaneous colli
Intermandibularis ventralis caudalis Intermandibularis ventralis
Homberger and Meyers (1989) (Gallus), except as noted Superficial neck mm Constrictor colli cervicalis Gular mm Constrictor colli intermandibularis Mylohyoideus
Baumel et al. (1993) {Nomina Anatomica Avium) External hyolingual mm Constrictor colli
Constrictor colli, pars intermandibularis Intermandibularis ventralis
Extrinsic hyolingual mm Protractive extrinsic hyolingual mm Extrinsic lingual mm Extrinsic hyoid mm Branchiomandibularis Branchiomandibularis M. branchiomandibularis Branchiomandibularis Geniohyoideus (rostralis and caudalis) (anterior and posterior) (rostralis and caudalis) Genioglossus M. genioglossus Genioglossus Geniophar3mgealis (external hyoid mm) M. genioceratohyoideus* Retractive extrinsic hyolingual mm M. serpihyoideus Serpihyoideus
Serpihyoideus (external hyoid mm)
Serpihyoideus
Serpihyoideus
Stylohyoideus
Stylohyoideus
Stylohyoideus
Cricohyoideus
Cricohyoideus
M. hyomandibularis (including M. H. lat. and med.) Stylohyoideus Hyolaryngeal mm M. cricohyoideus M. ceratocricoideus* M. basiarytaenoideus* Intrinsic hyolingual mm M. ceratohyoideus M. ceratoglossus M. hyoglossus*
Cricohyoideus
Ceratohyoideus Ceratoglossus Hyoglossus obliquus Hyoglossus anterior
Extrinsic laryngeal mm Cricohyoideus
Ceratohyoideus Intrinsic hyoid mm Ceratoglossus Hyoglossus obliquus Hyoglossus anterior
Intrinsic lingual mm Ceratohyoideus
Intrinsic hyolingual mm Ceratohyoideus
Ceratoglossus
Ceratoglossus
Hypoglossus obliquus Hypoglossus anterior
Hyoglossus obliquus Hyoglossus rostralis
^Showninbold. ^An asterisk (*) indicates a muscle that is unique to paleognathous species. '^Some external hyoid muscles are listed under "extrinsic hyoid m m / '
The paraglossal remains cartilaginous throughout life in many neognaths, but ossifies in others (Homberger, 1986, 1989, 1999). In most cases the epibranchials remain cartilaginous throughout life. The paraglossal is arrow shaped in dorsal view, flattened dorsoventrally and may include cartilaginous processes (anterior and/or posterolateral), as in the turkey. It occupies most of the dorsal region of the avian tongue and is overlain by a tough, cornified, papillose epithelium (Fig. 11.2B). The remainder of the tongue consists of salivary glands and "intrinsic hyolingual muscles." Intrinsic hyolingual muscles are complex in some
neognathous taxa (e.g., parrots; Homberger, 1986), but in generalized taxa such as the wild turkey and chicken, only four muscle pairs are present (Fig. 11.3A): (1) the hyoglossus (innervated by c.n. XII) is divided into anterior and oblique segments connecting the articulated paraglossal and basihyal; the hyoglossus rostralis inserts ventrally on the anterior paraglossal process and originates on the ventral surface of the basihyal; (2) the hyoglossus obliquus originates on the ventral surface of the basihyal and inserts on the ventral surface of the paraglossal; (3) the ceratoglossus (c.n. XII) originates on the ceratobranchial rostrally and inserts by tendon on the ventral surface of
364
Carole A. Bonga Tomlinson
A.Rhea
Cb
mg
B. M e l e a g r i s
F I G U R E 11.2. Sagittal sections through the head of (A) a paleognathous species, Rhea americana, and (B) a neognathous species, Meleagris gallopavo, showing relative positions of the hyolingual apparatus and palatal bones. Superimposed on each section are the left ceratobranchial and epibranchial of the hyobranchium and the pterygoid and basipterygoid process of the skull. The position of the basipterygoid process and a mesial segment of the palatine bone are shown in relation to the pterygoid and vomer. Cartilaginous hyoid elements are stippled (note that this is the opposite of subsequent figures). Bh, basihyal; BP, basipterygoid process; Cb, ceratobranchial; Eb, epibranchial; mg, mucus glands; Pa, palatine bone; Pg, paraglossal; PR, parasphenoid rostrum; Pt, pterygoid bone; RE, rostral esophagus; sg, salivary glands; V, vomer.
the paraglossal; and (4) the ceratohyoideus (c.n. XII, VII?) originates on the medial surface of the ceratobranchial; the muscles of both sides join in a median raphe ventral to the urohyal and larynx. The floor of the mouth ventral to the free portion of the tongue contains salivary glands that lie between the intermandibularis (ventral) and the Mm. genioglossus (dorsal) (see extrinsic hyolingual muscles, later). Fibers of the intermandibularis (c.n. V) originate along the mandibular rami and join at a midventral raphe. It does not extend as far rostrocaudally as in ratites [ratites shown in Fig. 11.11; Bhattacharyya (1980) and Homberger and Meyers (1989) illustrate the intermandibularis in neognaths]. A cricohyoideus muscle (c.n. IX, X, XII), connecting the cricoid cartilage to the dorsal surface of the basihyal, is always present (Fig. 11.3B). Kinematic data suggest that muscles originating on
the mandible and inserting on the hyobranchial apparatus either protract or retract the hyobranchium relative to the mandible (see Table 11.1: "extrinsic hyolingual muscles") (Fig. 11.3C). Protractor muscles are innervated by cranial nerve XII and/or IX, whereas retractors are innervated by cranial nerve VII. The main protractors (Mm. branchiomandibularis rostralis and caudalis) originate on the mandible and run caudally to encircle the distal end of the ipsilateral hyoid horn. The genioglossus (c.n. XII, IX?) originates on the mandibular symphysis and inserts on the epithelium at the root of the tongue. This muscle is small in Meleagris and in other neognathous taxa it is either small or absent (Burton, 1984; Homberger and Meyers, 1989). Three extrinsic retractor muscles originate on the laterocaudal mandible: (1) the serpihyoideus, (2) the constrictor colli intermandibularis, and (3) the stylohyoideus. All are innervated by cranial nerve VII and have a common developmental origin with the constrictor colli cervicalis, a superficial dermal muscle that extends rostrally from the side of the upper neck and fans out in the throat region where the two sides meet in the midline (Fig. 11.3C; superficial constrictor not shown) (Kesteven, 1945; Noden, 1983a,b). Paired serpihyoideus muscles arise from the posterolateral margins of the mandible and run anteromedially to join at a midventral raphe ventral to the urohyal. The constrictor colli intermandibularis has a similar origin, with some of its fibers taking origin from the fascia overlying the serpihyoideus and the depressor mandibulae muscle (not shown) and from the tough connective tissue surrounding the external ear opening. It also inserts on a midventral raphe. Some fibers of the serpihyoideus and the constrictor colli intermandibularis overlap, with the latter more ventral. The common midventral raphe is connected to a fascial sheet that attaches to the rostroventral surfaces of the ceratobranchials and the ventral surface of the basihyal between the ceratobranchial articulations (see Homberger and Meyers, 1989). The latter two retractor muscles and their conimon raphe lie ventral to the ceratohyoideus (an intrinsic hyolingual muscle; see earlier discussion). The constrictor colli intermandibularis occurs in all reptiles and birds (Kesteven, 1944). The serpihyoideus is known only in birds and is characteristic of all known species (Kesteven, 1945). In contrast, the stylohyoideus (see later) occurs only in neognathous birds. The stylohyoideus originates on the mandible just rostral to the serpihyoideus and runs anteromedially to insert on the dorsal surface of the basihyal (Figs. 11.3A and 11.3C). It passes ventral to the branchiomandibularis caudalis and rostralis muscles and dorsal to the ceratoglossus. "Stylohyoideus" is a misno-
11. F e e d i n g in P a l e o g n a t h o u s Birds
365
F I G U R E 11.3. Hyolingual apparatus in a generalized neognathous bird, based on the wild turkey, Meleagris gallopavo (Galliformes). Stippling indicates ossification. (A) Ventral view of intrinsic hyolingual muscles and hyobranchium. (B) Dorsal view of main hyolaryngeal muscle, hyobranchium, larynx, and anterior end of the trachea. (C) Ventral view of extrinsic hyolingual muscles, hyobranchium, and mandible. Protractors are shown on the left, retractors on the right, apgp, anterior paraglossal process; Bh, basihyal; Cb, ceratobranchial; crl, cricoid (larynx); Eb, epibranchial; epr, epithelium at root of tongue; gl, glottis (larynx); MBmr, M. branchiomandibularis rostralis; MBmc, M. branciomandibularis caudahs; MCg, M. ceratoglossus; MCh, M. ceratohyoideus; MCrh, M. cricohyoideus; MGg, M. genioglossus; MHgo, M. hyoglossus obHquus; MHgr, M. hyoglossus rostralis; Mn, mandible; MSph, M. serpihyoideus; MSth, M. stylohyoideus; Pg, paraglossal; ppgp, posterior paraglossal process; tr, trachea; Uh, urohyal.
mer because birds lack a styloid process. However, it is maintained because the name has been in common use since the 19th century (refer to Table 11.2; see also Homberger, 1986; Homberger and Meyers, 1989). Other workers have noted that the stylohyoideus in neognathous birds can insert on the basihyal or the ceratobranchial (Burton, 1984). This distinction is important because muscle forces acting directly on the basihyal will have different mechanical consequences than forces acting on the ceratobranchials (see Section IV). The avian stylohyoideus as described here is known in no other vertebrate taxa and can be considered a synapomorphy of neognathous birds. 1. Summary of the Neognathous
Condition
The tongue of neognathous birds is extremely variable (McLelland, 1979), but its surface is often cornified and its length usually closely matches the length of the beak and oral cavity. In the wild turkey, filamentous papillae occur on the dorsal surface, and salivary glands occur within the base, but not the body, of the tongue.
Features of the neognath hyobranchial apparatus can be summarized as follows: (1) the paraglossal is ossified in the turkey, but it remains cartilaginous in other species (Homberger, 1986,1989,1999); (2) the basihyal is always ossified; (3) the paraglossal and basihyal meet to form a movable bony articulation; (4) the basihyal and urohyal separate during development; (5) the epibranchials are elongate and curve strongly upward; and (6) if the condition described for the chicken (Homberger and Meyers, 1989) and wild turkey (both galliforms) is representative, epibranchials connect to the occipital region by means of a complex and extensive fascia vaginalis hyoideus. Intrinsic hyolingual musculature connecting the paraglossal and basihyal is divided into hyoglossus rostralis and obliquus muscles. Extrinsic hyolingual protractor musculature (branchiomandibularis rostralis and caudalis) originates on the middle and anterior portion of the mandible and runs posteriorly to insert on the epibranchials. A third, small protractor muscle, the genioglossus, is variably present. It runs from the tip of the mandible to the base of the tongue. Extrinsic hyolingual retractor musculature consists of
366
Carole A. Bonga Tomlinson
muscles innervated by cranial nerve VII that are present in all reptiles and birds (constrictors), in all birds (serpihyoideus), or in neognathous birds alone (stylohyoideus). The stylohyoideus inserts directly onto the basihyal. B. Paleognathous Birds The small paleognathous tongue is supported by a hyobranchium that is unique among living birds—it is entirely cartilaginous except for the ceratobranchials (Figs. 11.2A, 11.4-11.6, 11.12) (Lang, 1956; MiiUer, 1963; Parker, 1866, 1891; Pycraft, 1900; Webb, 1957). Certain features of the neognathous condition described earlier are absent in all paleognathous taxa examined, including ossification of the paraglossal, basihyal, and urohyal, upwardly curved epibranchials attached to the occipital region, and a basibranchial insertion for the stylohyoideus muscle. A common paleognathous pattern occurs in extrinsic hyolingual protractor musculature that is distinct from neognathous taxa. However, variations among paleognathous taxa
Cb
F I G U R E 11.5. Dorsal view of the hyobranchium and tongue in the emu, Dromaius novaehollandiae, relative to the mandible. Note the fringed margins of the tongue. Stippling indicates ossification. Buh, basiurohyal; Cb, ceratobranchial; Eb, epibranchial; Mn, mandible; Pg, paraglossal; Tg, tongue.
occur in the form of the hyobranchium and extrinsic retractor musculature (see later). Among the four paleognathous species examined, the rhea appears to possess the most generalized condition of the paleognathous hyobranchial skeleton and displays the least complex extrinsic hyolingual retractor muscles among all birds. Thus, the condition of the hyolingual apparatus in the rhea is described first in most detail, and those features that distinguish the paleognathous condition from the basic neognathous condition are noted. Distinctive aspects of the hyolingual apparatus in the emu, ostrich, and a tinamou are then enumerated. 1. The Greater Rhea
F I G U R E 11.4. Dorsal view of the hyobranchium and tongue in the rhea, Rhea americana, relative to the larynx and mandible. Ossification occurs in the ceratobranchials only. Bh, basihyal; Cb, ceratobranchial; Eb, epibranchial; Lx, laryngeal glottis; Mn, mandible; Pg, paraglossal; Tg, tongue.
The tongue consists of a thick, rough epithelium containing mucus-secreting cells closely applied to the paraglossal; there are no salivary glands in the body of the tongue (Fig. 11.2A). The shape of the tongue mirrors the dorsoventrally flattened, arrow-like shape of the paraglossal. Three globose papillae occur at the posterolateral corners of the tongue (Fig. 11.4). In embryos, the paraglossal is the last element to form (Miiller, 1963). The basihyal is cylindrical, with its anterior end rounded. No urohyal portion projects caudal to the ceratobranchial articulations (Figs. 11.2A, 11.4, 11.7B, and 11.8B; see later). The rod-like ceratobranchials are
11. Feeding in Paleognathous Birds
367
FIGURE 11.6. Dorsal (left) and lateral (right) views of the hyobranchium and tongue in the ostrich, Struthio camelus, shown in relation to the larynx and trachea. Stippling indicates ossification; shading denotes the position and shape of the lingual pocket. Buh, basiurohyal; Cb, ceratobranchial; Eb, epibranchial; gl, glottis; Lp, lingual pocket; Ix, larynx; Pgp, paraglossalia.
the only ossified hyobranchial elements and, in resting position, lie medioventral to and parallel with the mandible. The epibranchials are short and do not extend beyond the caudal end of the mandible (Figs. 11.2A, 11.4, and 11.19C). Each cornu is attached to the vicinity of the external ear opening by means of tough connective tissue. This tissue is analogous to the fascia vaginalis hyoideus of neognaths, but it does not extend as far as the occipital region. The ventral surface of the paraglossal is covered by a tough, elastic connective tissue sheet that forms a tube-like space beneath the basihyal, completely investing the hyoglossus muscle and the basihyal. The sheet attaches to the posteroventral surface of the basihyal and the ceratobranchial-basihyal articulations
(Figs. 11.5 and 11.7B). The hyoglossus forms a muscular sling looping around the basihyal from the ventrolateral surface of the paraglossal. The hyoglossus and its connective tissue investment form the only connection between the basihyal and the paraglossal. A ceratoglossus muscle runs anteriorly from the ventrolateral surface of the ceratobranchial, over (ventral to) the connective tissue sheath covering the hyoglossus, to insert on the ventral surface of the paraglossal at two points anterior and lateral to the hyoglossus (Fig. 11.9B). According to Bock and Biihler (1988), the hyoglossus sling occurs in all tinamous and ratites, with the exception of the ostrich. An unusual connective tissue "collar" loosely encircles the anterior end of the basihyal (Figs. 11.8B and
FIGURE 11.7. Ventral view of intrinsic hyolingual muscles in ratites. (A) The emu, Dromaius; (B) the rhea, Rhea; and (C) the ostrich, Struthio. The genioglossus muscle is also shown. Stippling indicates ossification. Bh, basihyal; Buh, basiurohyal; Cb, ceratobranchial; cts, connective tissue sheath; Meg, M. ceratoglossus; Mgg, M. genioglossus; Mhg, M. hyoglossus; Pg, paraglossal; Pgp, paraglossalia.
368
Carole A. Bonga Tomlinson
MCrh MCrh
F I G U R E 11.8. Ventral view of hyolaryngeal muscles in ratites. (A) The emu, Dromaius; (B) the rhea, Rhea (caudal laryngeal papillae not shown); and (C) the ostrich, Struthio (only right paraglossalium shown). Not to scale; stippling indicates ossification. Bh, basihyal; bhs, basihyal sheath (connective tissue); Buh, basiurohyal; Cb, ceratobranchial; crl, cricoid cartilage (larynx); Eb, epibranchial; MCrh, M. cricohyoideus (right muscle only shown in A and B); Pgp, paraglossalium.
11.9B). Dorsal to the basihyal, paired, slender, straplike muscles, the basiarytaenoideus (new muscle; Table 11.1) extend caudally from the collar and attach to the epithelium of the ipsilateral arytaenoid cartilage of the larynx (Fig. 11.9B). No comparable muscles have been described in neognathous taxa. Cricohyoideus muscles insert on the rostral ceratobranchials ventrally and the mediocaudal surface of the basihyal (Figs. 11.8B and 11.9B). The ceratohyoideus is narrow and strap-like in contrast to its significant breadth in the wild turkey (Fig. 11.10; shown as if cut close to its origin). There are no salivary glands beneath the smooth
epithelium covering the floor of the mouth. The floor of the oral cavity between the mandibular rami is formed primarily by two thin muscles: the intermandibularis muscle (originating on the mandible and running medially to a midventral raphe) and a prominent, paired extrinsic protractor muscle, the genioceratohyoideus (new muscle; see Table 11.1 and later) (Figs. 11.2A, ll.lOB, and l l . l l B ) . Extrinsic protractor musculature is distinct as compared to all known neognathous taxa (Fig. 11.1 OB). The branchiomandibularis is undivided and analogous only to the "caudalis'' portion in neognathous
Buh
MCrh
F I G U R E 11.9. Dorsal view of hyolaryngeal muscles and tendons in ratites. (A) The emu, Dromaius; (B) the rhea, Rhea; and (C) the ostrich, Struthio. Not to scale; stippling indicates ossification. Bh, basihyal; bhs, basihyal (connective tissue) sheath; Buh, basiurohyal; Cb, ceratobranchial; gl, glottis; crl, cricoid cartilage (larynx); MBa, M. basiarytaenoideus; MCrh, M. cricohyoideus; Pgp, paraglossalium; TBa, basiarytaenoideus tendon; Tu, unified tendon of TBa.
369
11. F e e d i n g in P a l e o g n a t h o u s Birds
MGch MGch
MBm
MBm MSph
MSph
A
B
FIGURE 11.10. Ventral view of extrinsic hyolingual muscles in ratites. (A) The emu, Dromaius; (B) the rhea, Rhea; and (C) the ostrich, Struthio. Protractors (genioceratohyoideus and branchiomandibularis) are shown on the left of each figure, retractors are shown on the right. The genioglossus, also a protractor, arises dorsally from the genioceratohyoideus and is shown on the right of each figure. Also shown are the ceratocricohyoideus and ceratohyoideus muscles. MBm, M. branchiomandibularis; MCcr, M. ceratocricohyoideus; MCh, M. ceratohyoideus; MGch, M. genioceratohyoideus; MHm, M. hyomandibularis; MHml, M. hyomandibularis lateralis; MHmm, M. hyomandibularis medialis; MSph, M. serpihyoideus.
taxa. The genioglossus is small and arises from a much larger genioceratohyoideus muscle (new name; Table 11.1) (Fig. II.IOB), also innervated by cranial nerve XII (Webb, 1957). Both muscles were called the genioglossus by Bock and Biihler (1988) for all modern paleognaths. The strap-like genioceratohyoideus, however, originates on the mandibular symphysis and inserts on the caudoventral surface of the ceratobranchial; no analogous muscle is known in neognathous birds (Fig. 11.1 OB). The genioglossus originates from the middorsal surface of the genioceratohyoideus, runs caudally to the root of the tongue near the midline, turns laterally, and enters the tongue to insert on the posteroventral margin of the paraglossal. The genioglossus is the most ventral muscle within the tongue. Extrinsic hyolingual retractor muscles consist of the constrictor colli intermandibularis and the serpihyoideus (Figs. ll.lOB and l l . l l B ) . The constrictor colli intermandibularis originates on fascia attaching to the
caudal surface of the mandible and the external ear region, and the serpihyoideus originates on the caudalmost surface of the mandible. As in the neognathous wild turkey, both muscles insert on a midventral raphe that connects indirectly to the posteroventral surface of the basihyal and anteroventral surfaces of the ceratobranchials by means of a fascial sheet. No retractors insert on the basihyal because it is covered by the hyoglossus, the connective tissue sheath investing it and the additional "collar" of connective tissue anchoring the basiarytaenoideus muscles (see earlier discussion). Thus, the neognathous stylohyoideus muscle is absent in paleognaths. Absence of a urohyal segment in the rhea is unique among paleognaths. Miiller (1963) reported a separate globular body ("copula 11") posterior to the basihyal in the rhea embryo that is presumably homologous to the urohyal. It is apparently lost during later development, i.e., it is absent in juveniles and adults.
370
Carole A. Bonga Tomlinson
F I G U R E 11.11. Constrictor musculature in ratites. (A) The emu, Dromaius; (B) the rhea, Rhea; and (C) the ostrich, Struthio. MCci, M. constrictor colli intermandibularis; MIm, intermandibularis; Mime, M. intermandibularis, caudal portion; Mimr, M. intermandibularis, rostral portion.
2. The Emu In all paleognathous taxa other than the rhea (see earlier discussion), the basihyal and urohyal form a continuous, cartilaginous structure termed the basiurohyal (Figs. 11.4-11.8). The emu differs from the rhea additionally in lingual embellishment, hyolaryngeal connections, and aspects of the extrinsic hyolingual retractor musculature. The paraglossal, epibranchials, the fascia vaginalis hyoideus, and intrinsic and extrinsic protractor musculature are similar in the rhea and emu. Differences are described later. Protruding posterolaterally from the lateral margins of the emu tongue are five dorsoventrally flattened, curved papillae that taper to rounded points (Fig. 11.5). Basihyal-paraglossal and basihyal-laryngeal articulations are as in the rhea with some exceptions. The ceratoglossus originates more anteriorly on the ceratobranchial and inserts on the paraglossal lateral to the hyoglossus (Fig. 11.7A). In place of the basiarytaenoideus muscles, a paired tendon (T. basiarytaenoideus) originates directly on the dorsal surface of the basihyal
and a connective tissue collar is absent from the anterior end of the basihyal (Figs. 11.8A and 11.9A). An unusual hyolaryngeal muscle, the ceratocricoideus (new name; Table 11.1), originates on the ceratobranchial (medial to the posterior end of the genioceratohyoideus insertion and posterior to the ceratohyoideus insertion) and inserts on the cricoid cartilage of the larynx (Fig. 11.10A). Extrinsic protractor musculature is essentially the same as in the rhea, although the genioceratohyoideus is broader and its insertion on the ceratobranchial extends further rostrally (Fig. 11.10A). The emu possesses an extrinsic retractor in addition to the serpihyoideus and constrictor colli intermandibularis muscles (Figs. II.IOA and l l . l l A ) . The hyomandibularis (Webb, 1957) originates on the posterior end of the mandible anterior to the serpihyoideus, runs anteromedially, and divides into two portions (medialis and lateralis) that insert on the urohyal and midceratobranchial, respectively (Fig. 11.10A). Both branches pass ventral to the branchiomandibularis, but the anterior portion of the lateral branch alone passes
11. Feeding in Paleognathous Birds dorsal to the genioceratohyoideus. This configuration of the hyomandibularis has not been reported in other avian taxa. 3. The Ostrich The ostrich hyolingual apparatus differs from that of the rhea in several ways. The epibranchials are long and curve conspicuously downward, lying alongside muscles of the neck (Fig. 11.6). This condition is unique to the ostrich. The fascia vaginalis hyoideus connects the ear region to the hyoid horns at the level of the ceratobranchial-epibranchial junction. Other differences are described later. The unique tongue of the ostrich is exceedingly short and virtually fixed in position immediately anterior to the laryngeal glottis. There are no lingual papillae (Fig. 11.6). Paraglossal form is unique among birds; it comprises two narrow, wing-like cartilages called the paraglossalia (Fiirbringer, 1922; inaccurately referred to as paraglossal "processes" by Bock and Btihler, 1988). The anterodorsal surface of each paraglossalium is attached by connective tissue to the anteroventral surface of a very broad basiurohyal. The paraglossalia do not contact one another in the midline. The paraglossalia project posterolateral^ and somewhat ventrally, but the tips curve upward so that they lie in a plane dorsal to the basiurohyal. Thus, as compared to other birds, the paraglossal is displaced ventrally and laterally relative to the basihyal. Only in parrots is the paraglossal also formed by two separate elements, but these are ossified and located rostral to the basihyal (Homberger, 1986). According to Fiirbringer (1922), the ostrich paraglossalia retain the form and position of the cartilaginous anlagen (paraglossalia) of the adult paraglossal present during embryonic development in all birds (see Kallius, 1905). This suggests the possibility that the condition of the ostrich paraglossal arose through paedomorphosis (evolutionary juvenilization) (Elzanowski, 1986). The position of the paraglossalia relative to the basiurohyal in the ostrich leaves the anterior tip of the basiurohyal to form the tip of the tongue. The lingual epithelium encloses a "lingual pocket" on the dorsum of the tongue with an opening facing posteroventrally (Fig. 11.6). The pocket appears grossly to be lined with the same type of epithelium found on the tongues outer surfaces and is similar to the rhea. The structure and conformation of the lingual pocket suggest that ancestrally there was a portion of the tongue anterior to the basiurohyal that eventually folded over the base of the tongue. This scenario suggests that the lining of the pocket represents the ancestral dorsal surface of the
371
tongue and that the epithelial surface exposed dorsally in the ostrich was ancestrally the ventral surface of the tongue. The paraglossalia are located within the ventralmost portion of the tongue and extend to the ventrolateral margins of the pocket. The lingual pocket changes shape during intraoral transport (see later), apparently in response to muscular action, suggesting that the ostrich tongue is specialized for an as yet unexplained biological role (Bock and Biihler, 1988). The ostrich tongue appears to have been secondarily derived from a more general paleognathous condition as evident in the rhea. The hyoglossus muscle is absent in the ostrich. The ceratoglossus inserts on the ipsilateral paraglossalium anterodorsally, intervening between the paraglossalium and the anterior basihyal (Fig. 11.7C). The genioglossus inserts on the posterior surface of the paraglossalium (Figs. 11.7C and ll.lOC). The cricohyoideus has a broad zone of insertion on the dorso- and ventrolateral surfaces of the basihyal (Bock and Biihler, 1988). Thus, this "hyolaryngeal" muscle is located partially within the tongue in the ostrich, a condition unknown in other taxa (Figs. 11.8C and 11.9C). Extrinsic protractor musculature is essentially the same as in the other paleognaths. Insertion of the genioceratohyoideus on the ceratobranchialis occurs more anteriorly than in the rhea (Fig. 11.IOC). Regarding extrinsic retractor muscles, the serpihyoideus (Fig. 11.8D) inserts directly onto the ventral surface of the cricoid cartilage, a configuration unique among known avian taxa. The identity of the muscle was determined by its origin on the mandible—anterior to the origin of the constrictor colli intermandibularis and caudal to the origin of the hyomandibularis— and by its position ventral to the branchiomandibularis (compare Figs. 11.IOC and l l . l l C ) . An additional retractor, the hyomandibularis, originates on the posterolateral surface of the mandible anterior to the serpihyoideus and inserts on the ceratobranchial anterodorsally (Figs. II.IOC and l l . l l C ) . Along the way it passes ventral to the branchiomandibularis and dorsal to the genioceratohyoideus. Webb (1957) called this muscle the hyomandibularis medialis (see Table 11.1). Simplification of the name was justified by the fact that there is no lateral branch. Recall, however, that in the emu the hyomandibularis is divided into lateral and medial moieties (Fig. II.IOA; see earlier discussion). Bock and Biihler (1988) may have referred to the ostrich hyomandibularis as the stylohyoideus (Table 11.1). The latter name, however, is reserved for the muscle in neognathous taxa that inserts on the basihyal (Table 11.2; see earlier discussion).
372
Carole A. Bonga Tomlinson
4. The Chilean Tinamou The tongue and lingual epithelium in this small bird resemble that of the rhea, with the exception that the tongue tip ends in a rounded point and there are no lingual or laryngeal papillae present (Fig. 11.12A). The paraglossal is paper thin and its margins are scalloped. The lingual epithelium is thick relative to the paraglossal and is not scalloped, but it follows the general outline of the paraglossal (compare in Figs. 11.12A and 11.12B). The paraglossal of Nothoprocta contrasts strongly with that of another tinamou, Tinamus (Parker, 1866), which appears to resemble that of the rhea. The floor of the oropharyngeal cavity is deeply folded and forms a "shelf" that curves from the rostral end of the mandible to the posterior end of the oropharynx, around the tongue and larynx (Fig. 11.12A). In addition, a thickened epithelial fold forms another shelf between the tongue and the larynx. This shelf encloses a pocket that is lined by the oral epithelium. The pocket extends rostrally beneath the base of the tongue
but dorsal to the basiurohyal, thus it lies within the lingual base (the lingual-base pocket; Fig. 11.12A; compare Figs. 11.12A and 11.12B; see also Figs. 11.2A and 11.2B). The 'Tingual-base pocket" differs from the "lingual pocket" in the ostrich by virtue of its position, which is entirely ventral, or caudoventral, to the paraglossal and body of the tongue. A lingual-base pocket has not been previously reported in birds. Connection of the paraglossal and basiurohyal by means of connective tissue and the hyoglossus is the same as in the rhea and emu. In the region between its articulations with the ceratobranchials, the basiurohyal is partially ossified. The distal ends of the short epibranchials extend slightly posterior to the mandible, where they curve only slightly upward and attach by means of the fascia vaginalis hyoideus to the ear region (Fig. 11.12B). The short epibranchials oi Nothoprocta are in marked contrast to the elongate epibranchials depicted for Tinamus (Parker, 1866). Extrinsic hyolingual musculature is similar to that of the rhea (Fig. 11.12B). Presence or absence of the basiarytaenoideus muscles, a basihyal collar and the condition of the cricohyoideus (presumed to be present), could not be determined. 5. Summary of the Paleognathous
FIGURE 11.12. Tongue, hyolingual apparatus, and mandible in a tinamou, Nothoprocta perdicaria. (A) Dorsal view of the tongue and lingual-base pocket (see text) relative to mandible, larynx, and anterior end of the trachea; position of the pocket ventral to the tongue is denoted by a dashed line. (B) Ventral view of extrinsic hyolingual muscles. Except for the genioglossus, protractors are shown on the left of figure (with the exception of the genioglossus, which is shown on the right), retractors are shown on the right. A dashed circle denotes the attachment site of the sling-like hyoglossus muscle on the ventral surface of the paraglossal (see description of the rhea in text). Stippling indicates ossification. Buh, basiurohyal; Cb, ceratobranchial; Eb, epibranchial; crl, cricoid cartilage (larynx); sas, attachment site for muscular sling; epf, epithelial fold on floor of oral cavity; gl, glottis; Ibp, lingual-base pocket; MBm, M. branchiomandibularis; MGch, M. genioceratohyoideus; MGg, M. genioglossus; MSph, M. serpihyoideus; Pg, paraglossal; Tg, tongue.
Condition
The paleognath taxa described earlier are assumed to be representative of the group as a whole, but this is not demonstrated. The ratite cassowaries (Casuarius) and kiwis (Apteryx) and 46 additional species of tinamou remain undescribed, but based on Pycraft's (1900) description of the hyobranchium and tongue, these are unlikely to deviate significantly from the conditions described here. Because Australasian ratites are likely to resemble one another, the description of the emu condition may approximate that of Casuarius and Apteryx (Cooper et al, 1992). The lesser rhea {Pterocnemia pennata) is assumed to be similar to the greater rhea (JR. americana) described previously, and the close relationship of the tinamous (Prager and Wilson, 1976) suggests that the description for N. perdicaria is generally applicable, except where noted. The tongue in all paleognathous taxa examined here is composed of rough epithelium (in the rhea, it contains mucous cells) with few papillae, which appears to represent the ancestral paleognathous condition. The fringed lateral margins on the tongue in the emu may be autapomorphic, but conditions in other Australasian ratites are unknown. In the ostrich, the tongue appears to be secondarily reduced and a pocket (lingual pocket) forms within the lingual epithelium. The lingual pocket is uniquely derived in Struthio. A different epithelial "pocket" forms within the lingual base of
11. Feeding in Paleognathous Birds the tinamou, but these structures do not appear to be homologous. It is unknown whether the tongue-base pocket is present in other tinamou species. Globose papillae posterolateral to the tongue are found in the rhea only and may represent an autapomorphy for rheas. However, the condition in Pterocnemia is unknown. The hyobranchial skeleton in the paleognathous ratites demonstrates a consistent pattern in which the paraglossal is cartilaginous, the basihyal or basiurohyal is cartilaginous, the basiurohyal forms a single structure, the paraglossal does not join the basi(uro)hyal by means of an articulated joint, epibranchial curvature is usually downward and never upward, and the fascia vaginalis hyoideus is less extensive than in neognaths and connects the hyoid horns (ceratoepibranchials) to the ear region. Tinamous show a condition intermediate between ratites and neognaths in having more ossification in the hyobranchium and a slight upward curvature at the tips of the epibranchials. The overall similarity between ratites and tinamous, however, suggests that they descended from a common ancestral condition. The hyobranchial skeleton in the emu may represent the ancestral paleognathous condition because it contains the full complement of hyobranchial elements known to occur in the group, including the species described earlier, as well as a kiwi (Parker, 1891), a cassowary, and another tinamou (Parker, 1866). The hyobranchial skeleton in the rhea {Rhea) is virtually identical to that in the emu except that a urohyal portion is absent due to its disappearance posthatching (Mliller, 1963) (it is unknown if this is also true for the lesser rhea). The paraglossal in the tinamou, N. perdicaria, differs from the putative ancestral type in its possession of scalloped margins, but this trait may not be characteristic of all tinamous—the paraglossal in Tinamus is shown to be similar to the rhea by Parker (1966). In the ostrich, the paraglossal comprises two separate elements, the paraglossalia, which occur ventral and lateral to the basihyal, seemingly displaced from the position of the paraglossal in all other birds. It is thus autapomorphic for Struthio. Also uniquely derived in the ostrich are elongate epibranchials, which curve strongly downward. In all paleognathous taxa except the ostrich, the intrinsic hyolingual musculature is similar—an undivided hyoglossus sling ensheathed by connective tissue forms the sole connection between the paraglossal and the basihyal. This form of connection between the paraglossal and the basihyal seems to represent the ancestral paleognathous condition, with loss of the hyoglossus derived in the ostrich. Extrinsic hyolingual protractor musculature consists of the same three muscle pairs in all paleognaths
373
(branchiomandibularis, genioceratohyoideus, genioglossus), one of which (the genioceratohyoideus) is unknown in neognathous taxa. The genioglossus inserts on the paraglossal (or paraglossalium) in all paleognaths. These shared features are likely to represent the ancestral paleognathous condition. Extrinsic hyolingual retractor musculature varies, but in no instance does a retractor insert on the basihyal. In the rhea and tinamou there are no retractor muscles other than the serpihyoideus. This may represent the ancestral paleognath condition (see later). The presence of the hyomandibularis in the ostrich and its division into medial and lateral portions in the emu are derived relative to the proposed ancestral condition. Muscular and/or tendinous connections between the basihyal and the arytaenoid cartilage are unknown in other tetrapods. Thus the presence of a basiarytaenoideus muscle (or tendon) may be a synapomorphy of paleognaths, but its presence in a tinamou could not be confirn\ed in the present study. Thus a basihyalarytaenoid connection may be unique to ratites or may have been present in the common ancestor of ratites and tinamous. A ceratocricoideus muscle connecting the ceratobranchial to the cricoid cartilage is found uniquely in the emu and ostrich. The ancestral condition is clearly absence of such connections, as evident in the rhea and tinamou. IV. FUNCTION OF THE HYOLINGUAL APPARATUS The following descriptions focus primarily on movement patterns of the basi(uro)hyal (protraction, retraction, orbit) and the hyoid horns (depression, elevation) as measured relative to the mandible. As such, movements of the hyobranchium are shown independent of mandibular movement (as if the mandible was stationary). Tongue position is dependent on the position of the basihyal. Extrinsic hyolingual muscles act on the hyoid horns (cerato-epibranchials) to affect movement, but in neognathous species (such as the turkey) they can act on the basihyal directly. Synovial joints between the ceratobranchials and the basi(uro)hyal permit considerable movement, but the hyoid horns move symmetrically, each in concert with the other, suggesting bilaterally symmetric motor patterns in the hyobranchial muscles. The ceratoepibranchials act as third-class levers as they move the basi(uro)hyal. The caudal suspensorium of each epibranchial by fascia serves as a movable fulcrum (= primary fulcrum) for each lever (Fig. 11.13). Curvature of the epibranchials, location of their fulcra, and
374
Carole A. Bonga Tomlinson
Primary hyoid fulcra —
BasihyaH»ratobranchial articulation viewed from lateral —
Fascial attachment site of caudal end of epibranchial
Hyoid horn (comu; lateral view)
External ear region
Occipital region
Rostral
Rostral Loose fascia
Basihyal
Protraction
#
Late protraction
/
/ B a s i h y a l path
Basihyal orbit
Paleognath
Retraction
Neognath
F I G U R E 11.13. Effect of primary hyoid fulcra on motion of the comua (curved bars) and basihyal (black circles) during an intraoral transport cycle. Ratite cranioinertial feeding is shown on the left (with a 'loose fulcrum") and neognathous lingual transport shown on the right (with a ''firm fulcrum"). Solid bar in paleognath shows position during early protraction, in neognath during retraction. Heavy dashed bars indicate fully protracted positions. Light dashed bar shows fully retracted position in a paleognath. Arrows indicate direction of basihyal orbit.
insertion sites of protractor and retractor muscles thus circumscribe the range of possible movements potentially exhibited by the hyoid horns in response to forces exerted by extrinsic muscles originating on the mandible. One consequence of this restricted movement is that the hyoid horns are seen to depress and elevate relative to the mandible (see Figs. 11.17 and 11.21; discussed later). Two other types of hyolingual movement pattern are observed. Intrinsic hyolingual muscles flex the tongue, but flexion is more important in lingual feeding by neognathous birds than in ratite cranioinertial feeding, as noted earlier. Second, the distance between the larynx and the basi(uro)hyal fluctuates during the hyolingual cycle, i.e., during protraction this distance increases whereas during retraction it decreases. Retraction is pronounced in the wild turkey and limited in the ratites, probably reflecting the importance of this movement during hyolingual transport in neognathous feeding. Because the mandible depresses and elevates (= jaw
cycle) as the hyolingual apparatus moves, the hyolingual cycle {= protraction and retraction) is superimposed on the jaw cycle (Figs. 11.15,11.16,11.19,11.20, and 11.22). The jaw cycle is indicated by gape distance (distance between rostral tips of the upper and lower beaks) in Figures 11.15 and 11.22. Because gape distance in birds is a function of movements by both mandible and upper beak, movement of each beak tip relative to the cranium is also shown (Figs. 11.16, 11.19, and 11.20). Hyolingual position is measured from the point of articulation between the ceratobranchials and the basi(uro)hyal; protraction and retraction of the hyobranchium are measured relative to the mandible, as noted previously. Tongue movement patterns mirror those of the basi(uro)hyal. As the mandible is depressed and elevated, so too are the hyoid horns relative to the mandible. The summation of hyobranchial protraction-retraction movements and hyoid horn depression-elevation movements on the basihyal results in a regular, cyclic orbit relative to the mandible (Fig. 11.14). Distinct kinematic
11. Feeding in Paleognathous Birds
375
F I G U R E 11.14. Basihyal orbit relative to a fixed mandible during intraoral transport in neognathous (A) and paleognathous (B-D) birds. (A) Wild turkey, Meleagris; (B) emu, Dromaius; (C) rhea, Rhea; and (D) ostrich, Struthio. Position of the basihyal is indicated when the beak tips release the food (R), at maximum jaw gape (Mx), and when the mandible completes elevation (C). In the three ratites, early protraction (epr) precedes food release, and late protraction (Ipr) occurs during a jaw open-close cycle; in the turkey, all protraction takes place while the jaws are closed.
patterns involving these three movement parameters occur in lingual and cranioinertial feeding. Descriptions are based on lateral views of the head as recorded on high-speed cineradiographic film. Kinematic patterns are based on digitized points (Figs. 11.15, 11.16, 11.19, 11.20, 11.22, and 11.23), traced frame sequences (Figs. 11.13,11.17,11.18, and 11.21) or a combination of the two (Fig. 11.14). "FiyolinguaP' refers to both tongue and hyoid and therefore includes the paraglossal, whereas "hyoid" is used here to refer to the parts of the hyobranchial skeleton [the basi(uro)hyal and/or the cerato-epibranchials] acted on by extrinsic hyolingual protractor and retractor muscles originating on the mandible. A. Lingual Feeding in a Generalized Neognath (Meleagris gallopavo) Five movement parameters of lingual feeding are shown in Figs. 11.15 (small food transport) and 11.16
(large food transport)—head, food (relative to ground and passage through the oral cavity), hyobranchium, and gape. After ingestion with the tips of the beak, a small food item (pellet; Purina Turkey Chow) is transported intraorally by the tongue in stepwise fashion from beak tips to the posterior palate, or pharynx. Usually three steps are required to complete transport to the pharynx. Each step is accomplished with one complete hyolingual cycle (retraction-protraction) and one gape cycle (open-close) (see later). In the initial step, food is transported by the retracting tongue tip as the mandible is depressed and the tongue is lifted from the floor of the oral cavity. In subsequent steps, food is pushed caudally by the upwardly bulging midregion of the retracting tongue. Between each retraction phase, the mandible is held closed as the tongue is protracted beneath the food item, pushing it against the palate and completing the hyolingual cycle. The food is held in place by posteriorly protruding palatal papillae. The food becomes more rounded as
376
Carole A. Bonga Tomlinson
14H 13 O
1?
o> E 10 o
Head
k_
CD O
c (fl (fi •O
"(5 o tQ
>
8 6 4
Beginning of lingual tranport
2 H
Food dropped
11 m 11111
Gape
-0.5 H
$
1 M1111111111M1111111111111111111111111111111111111111111111111111111111111111111 n 111111111111 n n 11111 n I'
-1-5-
.-.
•
Lower beak tip
• ••
1
> 6
11 • 1 1 1 1 1 1
• •
II Ml l i l l l
1111111111111 MII1111111111111111111111
Beak tips 0)
.9-
^CO (D
^
E o
^
•4—
CD O
c
0) "D "cO
*-* o
N c o X k»
J
2
'•.. Food
4 Protraction
6
T
Retraction
8 10 12 lllllllllllllllll
0
10
I i 111111111111111111111111111111111111111111111111111111111
0
10
20 30 1 frame = 5 msec
40
50
FIGURE 11.18. Kinematic profile of cranioinertial transport in a paleognathous ratite, Rhea americana. (Top) Vertical displacement of the head and food item. (Center) Vertical displacement of upper and lower beak tips relative to a fixed point on the cranium showing gape distance. (Bottom) Horizontal displacement of hyobranchium and food item relative to a fixed mandible. The initial stages of the transport cycle shown on the left were obtained from a separate film sequence.
11. Feeding in Paleognathous Birds 16-
i. i
the basihyal in a dorsally convex arc that enables the tongue to scrape the palate (see Figs. 11.20C and 11.20D) to affect swallowing during each hyolingual cycle. Tongue movement in ratites is variable. Extension of
Head y / ' ^ N .
1412-
/
10-
Food ^""^N^
X
..••'
381
"••..^^
864 -
Reha
Food released by beak tips
2-
..' 11111111II1 1 i 111111II111II If 11 Ml 11111 I I I i l l 1111 • 1111II1IIIIII11II111
20
10
I
1
30
40
50
"•••^
7-
•..
Upper beak •
m
^m'
1
•::::i
..-.•:: m
'•
8-
60
u
•
-••
•
Lower beak
9 •
10•••••
I I I l l | M i l l 1111i111111••11111111H11 M111111111111i 111 i1111111111111
-10
0
10 20 30 Frame (1 fr = 5 msec)
Beak tips
1:161. Harlid, A., A. Janke, and U. Arnason (1997) The mtDNA sequence of the ostrich and the divergence between paleognathous and neognathous birds. Mol. Biol. Evol. 14:754-761. Hecht, K. (1985) The biological significance of Archaeopteryx. Pp. 149-160. In: The Beginnings of Birds. Proc. Int. Archaeopteryx Conf., Eichstatt, (1984). M. K. Hecht, J. H. Ostrom, G. Viohl, and P. Wellnhofer (eds.). Freunde des Jura-Museums Eichstatt, Willibaldsburg, Eichstatt. Hedges, S. B. (1994) Molecular evidence for the origin of birds. Proc. Natl. Acad. Sci. USA 91:2621-2624. Hedges, S. B., P H. Parker, C. G. Sibley, and S. Kumar (1996) Continental breakup and the ordinal diversification of birds and mammals. Nature 381:226-229. Hedges, S. B., D. Simmons, A. van Dijk, G.-J. Caspers, W. W. de Jong, and C. G. Sibley (1995) Phylogenetic relationships of the hoatzin.
392
Carole A. Bonga Tomlinson
an enigmatic South American bird. Proc. Natl. Acad. Sci. USA 92: 11662-11665. Hiiemae, K., and A. W. Crompton (1985) Mastication, food transport and swallowing. Pp. 262-290. In: Functional Vertebrate Morphology. M. Hildebrand, D. M. Bramble, K. F. Liem, and D. B. Wake (eds.). Harvard Univ. Press, Cambridge, MA. Hiiemae, K. M., A J. Thexton, and A. W. Crompton (1978) Intra-oral transport: a fundamental mechanism of feeding? Pp. 181-208. In: Muscle Adaptation in the Craniofacial Region. Monograph No. 8, University of Michigan. Homberger, D. G. (1986) The lingual apparatus of the African grey parrot, Psittacus erithacus Linne (Aves: Psittacidae): description and theoretical mechanical analysis. Omith. Monogr. No. 39: 1-233. Homberger, D. G. (1988) Comparative morphology of the avian tongue. Pp. 2427-2435. In: Acta XIX Congressus Internationalis Ornithologici, Vol. II. National Museum of Natural Sciences, Univ. of Ottawa Press. Homberger, D. G. (1989) Correlations between morphology of the lingual apparatus and feeding mechanics in birds. Forsch. Zool. 35:14-150. Homberger, D. G. (1999) The avian tongue and larynx: multiple functions in nutrition and vocalization. In: Proc. 22nd Int. Ornithol. Congr. N. Adams and R. Slotow (eds.). Univ. of Natal, Durban. Homberger, D. G., and R. A. Meyers (1989) Morphology of the lingual apparatus of the domestic chicken. Callus gallus, with special attention to the structure of the fasciae. Am. J. Anat. 186:217-257. Houde, P. (1986) Ostrich ancestors found in the Northern Hemisphere suggest new hypothesis of ratite origins. Nature 324: 563-565. Houde, P. W. (1988) Paleognathous Birds from the Early Tertiary of the Northern Hemisphere. Publications of the Nuttall Ornithological Club, No. 22. Cambridge, MA. Houde, P., and H. Haubold (1987) Paleotis weigelti restudied: a small Middle Eocene ostrich (Aves: Struthioniformes). Palaeovertebrata 17:27-42. Houde, P., and S. L. Olson (1981) Paleognathous carinate birds from the Early Tertiary of North America. Science 214:1236-1237. Jean, A. (1984) Brainstem organization of the swallowing network. Brain Behav. Evol. 25:109-116. Jollie, T. (1957) The head skeleton of the chicken and remarks on the anatomy of this region in other birds. J. Morph. 100:389-436. Kallius, E. (1905) Beitrage zur Entwickelung der Zunge. II. Teil. Vogel. {Anas bochas L., Passer domesticus L.). Anat. Hefte. 85/86(2/3): 21-586. Kesteven, H. L. (1944) The evolution of the skull and the cephalic muscles: a comparative study of their development and adult morphology. III. The Sauria (Reptilia). Austral. Mus. Mem. 8: 237-269. Kesteven, H. L. (1945) The evolution of the skull and the cephalic muscles: a comparative study of their development and adult morphology. III. The Sauria (Aves). Austral. Mus. Mem. 8:270293. Kontges, G., and A. Lumsden (1996) Rhombencephalic neural crest segmentation is preserved throughout craniofacial ontogeny. Development 122:3229-3242. Kurochkin, E. N. (1995) Morphological differentiation of palaeognathous and neognathous birds. Courier Forschungsinstitut Senckenberg 181:79-88. Lang, C. (1956) Das Cranium der Ratiten mit besonderer Beriicksichtigung von Struthio camelus. Zeit. Wissensch. Zool. 159:165224. Lee, K., J. Feinstein, and J. Cracraft (1997) The phylogeny of ratite birds: resolving conflicts between molecular and morphological
data sets. Pp. 173-208. In: Avian Molecular Evolution and Systematics. D. P. Mindell (ed.). Academic Press, San Diego. Lowe, P. R. (1928) Studies and observations bearing on the phylogeny of the ostrich and its allies. Proc. Zool. Soc. Lond. 1928: 185-247. Martin, L. D. (1983) The origin and early radiation of birds. Pp. 291353. In: Perspectives in Ornithology. Martin, L. D. (1985) The relationship of Archaeopteryx to other birds. Pp. 177-183. In: The Beginnings of Birds. Proc. Int. Archaeopteryx Conf., Eichstatt, (1984). M. K. Hecht, J. H. Ostrom, G. Viohl, and P. Wellnhofer (eds.). Freunde des Jura-Museums Eichstatt, Willibaldsburg, Eichstatt. Martin, L. D. (1987) The beginning of the modern avian radiation. Docum. Lab. Geol. Lyon 99:9-19. Martin, L. D., and J. D. Stewart (1985) Homologies in the avian tarsus. Nature 315:159-160. McDowell, S. (1948) The bony palate of birds. I. The Palaeognathae. Auk 65:520-549. McGowan, C (1984) Evolutionary relationships of ratites and carinates: evidence from ontogeny of the tarsus. Nature 307:733-735. McGowan, C. (1985) Tarsal development in birds: evidence for homology with the theropod condition. J. Zool. Lond. 206:53-67. McLelland, J. (1979) Digestive system. Pp. 69-181. In: Form and Function in Birds, Vol. 1. A. S. King and J. McLelland (eds.). Academic Press, New York. Molnar, R. E. (1985) Alternatives to Archaeopteryx: a survey of proposed early or ancestral birds. Pp. 209-217. In: The Beginnings of Birds. Proc. Int. Archaeopteryx Conf., Eichstatt, (1984). M. K. Hecht, J. H. Ostrom, G. Viohl, and P. Wellnhofer (eds.). Freunde des Jura-Museums Eichstatt, Willibaldsburg, Eichstatt. Miiller, H. J. (1963) Die Morphologie und Entwicklung des Craniums von Rhea americana Linne. II Viszeralskelett, Mittelohr, und Osteocranium. Zeit. Wissensch. Zool. 168:35-118. Nei, M. (1996) Phylogenetic analysis in molecular evolutionary genetics. Annu. Rev. Genet. 30:371-403. Noden, D. M. (1983a) The embryonic origins of avian cephalic and cervical muscles and associated connective tissues. Am. J. Anat. 168:257-276. Noden, D. M. (1983b) The role of the neural crest in patterning of avian cranial skeletal, connective, and muscle tissues. Dev. Biol. 96:144-165. Oelrich, T. M. (1956) The anatomy of the head of Ctenosaura pectinata (Iguanidae). Misc. Publ. Mus. Zool. Univ. Michigan, No. 9 4 : 1 122. Olson, S. L. (1985) The fossil record of birds. Pp. 79-238. In: Avian Biology, Vol. VIII. D. S. Farner, J. R. King, and K. C. Parkes (eds.). Academic Press, London. Olson, S. L. (1986) Aspects of global avifaunal dynamics during the Cenozoic. Pp. 2023-2029. In: Acta XIX Congressus Internationalis Ornithologici, Vol. II. Ottawa. Ostrom, J. H. (1969) Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of Montana. Peabody Mus. Nat. Hist. Bull. 30:31-165. Ostrom, J. H. (1973) The ancestry of birds. Nature 242:136. Ostrom, J. H. (1985) The meaning of Archaeopteryx. Pp. 161-176. In: The Beginnings of Birds. Proc. Int. Archaeopteryx Conf., Eichstatt, 1984. M. K. Hecht, J. H. Ostrom, G. Viohl, and P Wellnhofer (eds.). Freunde des Jura-Museums Eichstatt, Willibaldsburg, Eichstatt. Ostrom, J. H. (1991) The question of the origin of birds. Pp. 467-484. In: Origins of the Higher Groups of Tetrapods, Controversy and Consensus. H.-P Schultze and L. Trueb (eds.). Comstock, Ithaca, NY. Padian, K., and L. Chiappe (1998) The origin and early evolution of birds. Biol. Rev. 73:1-42.
11. Feeding in Paleognathous Birds Parker, T. J. (1891) Observations on the anatomy and development of Apteryx. Phil. Trans. Roy. Soc. Lond. 182(B): 25-134. Parker, W. K. (1866) On the structure and development of the skull in the Ostrich Tribe. Phil. Trans. 1866:113-183. Parkes, K. C , and G. A. Clark (1966) An additional character linking ratites and tinamous and an interpretation of their monophyly. Condor 68:459-471. Peters, D. S. (1987) Mechanische Unterschiede palaognather und neognather Vogelschadel. Natur und Museum 117:173-182. Peters, D. S. (1988) Ein vollstandiges Exemplar von Paleotis weigelti (Aves, Palaeognathae). Cour. Forsch.-Inst. Senckenberg 107: 223-233. Prager, E. M., A. C. Wilson, D. T. Osuga, and R. E. Feeney (1976) Evolution of flightless land birds on southern continents: transferrin comparison shows monophyletic origin of ratites. J. Mol. Evol. 3:283-294. Pycraft, W. P. (1900) On the morphology and phylogeny of the Palaeognathae (Ratitae and Crypturi) and Neognathae (Carinatae). Trans. Zool. Soc. Lond. 15:149-290. Rambaut, A., and L. Bromham (1998) Estimating divergence dates from molecular sequences. Mol. Biol. Evol. 15:442-448. Sampson, S. D., L. Witmer, C. A. Forster, D. W. Krause, P. O'Connor, P. Dodson, and F. Ravoavy (1998) Predatory dinosaur remains from Madagascar: implications for the Cretaceous biogeography of Gondwana. Science 280:1048-1051. Schumacher, G.-H. (1973) The head muscles and hyolaryngeal skeleton of turtles and crocodilians. Pp. 101-199. In: Biology of the Reptilia, Vol. 4. C. Gans (ed.). Academic Press, New York. Schwenk, K. (1986) Morphology of the tongue in the tuatara, Sphenodon punctatus (Reptilia: Lepidosauria), with comments on function and phylogeny. J. Morph. 188:129-156. Schwenk, K. (1988) Comparative morphology of the lepidosaur tongue and its relevance to squamate phylogeny. Pp. 569-598. In: Phylogenetic Relationships of the Lizard Families. R. Estes and G. Pregill (eds.). Stanford Univ. Press, Stanford, CA. Schwenk, K. (2000a) Functional units and their evolution. In: The Character Concept in Evolutionary Biology. G. P. Wagner (ed.). Academic Press, San Diego. Schwenk, K. (2000b) Intrinsic versus extrinsic lingual muscles: a false dichotomy? Bull. Mus. Comp. Zool. Schwenk, K., and G. S. Throckmorton (1989) Functional and evolutionary morphology of lingual feeding in squamate reptiles: phylogenetics and kinematics. J. Zool. Lond. 219:153-175. Sibley, C. G., and J. E. Ahlquist (1990) Phylogeny and Classification of Birds: A Study in Molecular Evolution. Yale Univ. Press, New Haven, CT. Sibley, C. G., and B. L. Monroe, Jr. (1990) Distribution and Taxonomy of Birds of the World. Yale Univ. Press, New Haven, CT. Simonetta, A. M. (1960) On the mechanical implications of the avian skull and their bearing on the evolution and classification of birds. Quart. Rev. Biol. 35:206-220. Smith, K. K. (1984) The use of the tongue and hyoid apparatus during feeding in lizards (Ctenosaura similis and Tupinambis nigropunctatus). J. Zool. Lond. 202:115-143. Smith, K. K. (1986) Morphology and function of the tongue and hyoid apparatus in Varanus (Varanidae, Lacertilia). J. Morph. 187: 261-287. Smith, K. K. (1988) Form and function of the tongue in agamid lizards with comments on its phylogenetic significance. J. Morph. 196:157-171. Smith, K. K. (1992) The evolution of the mammalian pharynx. Zool. J. Linn. Soc. 104:313-349. Sondhi, K. C. (1958) The hyoid and associated structures in some Indian reptiles. Ann. Zool. 2:155-240.
393
Stapel, S. O., J. A. Leunissen, Versteeg, J. Wattel, and W. W. de Jong (1984) Ratites as oldest offshoot of avian stem-evidence from alp/ifl-crystallin A sequences. Nature 311:257-259. Starck, J. M. (1995) Comparative anatomy of the external and middle ear of palaeognathous birds. Adv. Anat. Embryol. Cell Biol. 131: 1-137. Suzuki, and S. Nomura (1975) Electromyographic studies on the deglutition movement in the fowl. Jap. J. Vet. Sci. 37:289-293. Takagi, N., and Sasaki (1974) A phylogenetic study of bird karyotypes. Chromosoma (Berl.) 46:91-120. Takagi, N., Itoh, and Sasaki (1972) Chromosome studies in four species of Ratitae (Aves). Chromosoma (Berl.) 36:281-291. Tambussi, C. P. (1995) The fossil Rheiformes from Argentina. Courier Forschungsinstitut Senckenberg 181:121-129. Thexton, A. J., A. W. Crompton, and R. Z. German (1998) Transition from suckling to drinking at weaning: a kinematic and electromyographic study in miniature pigs. J. Exp. Zool. 280:327343. Throckmorton, G. S. (1976) Oral food processing in two herbivorous lizards. Iguana iguana (Iguanidae) and Uromastix aegyptius (Agamidae). J. Morph. 148:363-390. Throckmorton, G. S. (1980) The chewing cycle in the herbivorous lizard Uromastix aegyptius (Agamidae). Arch. Oral Biol. 25:225-233. Tomlinson, C. A. (1997a) Anomalous pattern of jaw-hyolingual synchrony during true cranioinertial feeding by paleognathous birds. J. Morph. 232:330. Tomlinson, C. A. (1997b) Evolutionary transformation of prokinetic neognathous birds from ''akinetic" paleognathous grade based on feeding apparatus. J. Morph. 232:331. Van den Heuvel, W. R (1992) Kinetics of the skull in the chicken {Callus gallus domesticus). Neth. J. Zool. 42:561-582. van Tuinen, M., C. G. Sibley, and S. B. Hedges (1998) Phylogeny and biogeography of ratite birds inferred from DNA sequences of the mitochondrial ribosomal genes. Mol. Biol. Evol. 15:370-376. Vrba, E. S. (1989) What are the biotic hierarchies of integration and linkage? Pp. 379-401. In: Complex Organismal Functions: Integration and Evolution in Vertebrates. D. B. Wake and G. Roth (eds.). Wiley, Chichester. Wagner, G. P., and K. Schwenk (1999) Evolutionarily stable configurations: functional integration and the evolution of phenotypic stability. In: Evolutionary Biology, Vol. 32. M. K. Hecht, R. J. MacIntyre, and M. T. Clegg (eds.). Plenum Press, New York. Wake, D. B., and G. Roth (1989) The linkage between ontogeny and phylogeny in the evolution of complex systemanuscript in preparation Pp. 361-377. In: Complex Organismal Functions: Integration and Evolution in Vertebrates. D. B. Wake and G. Roth (eds.). Wiley, Chichester. Webb, M. (1957) The ontogeny of the cranial bones, cranial peripheral and cranial parasympathetic nerves, together with a study of the visceral muscles of Struthio. Acta Zool. 38:81-203. Weber, V. E. (1992) Zur Evolution basicranialer Gelenke bei Vogeln, insbesondere bei Hiihner- und Entenvogeln (Galloanseres). Z. Zool. Syst. Evolut.-forsch. 31:300-317. Weishampel, D. B., P. Dodson, and H. Osmolska (1990) The Dinosauria. Univ. of California Press, Berkeley, CA. Welman, J. (1995) Euparkeria and the origin of birds. S. Afr. J. Sci. 91: 533-537. Witmer, L., and L. D. Martin (1987) The primitive features of the avian palate, with special reference to Mesozoic birds. Docum. Lab. Geol. Lyon 99:21-39. Zeigler, H. P., R. Bermejo, and R. Bout (1994) Ingestive behaviour and the sensorimotor control of the jaw. Pp. 182-200. In: Perception and Motor Control in Birds. M. N. O. Davies and P. R. Green (eds.). Springer-Verlag, Berlin.
394
Carole A. Bonga Tomlinson
Zeigler, H. P., P. Welch Levitt, and R. R. Levine (1980) Eating in the pigeon {Columba livia): movement patterns, stereotypy, and stimulus control. J. Comp. Physiol. Psych. 94:783-794. Zeigler, H. P., Miller, and R. R. Levine (1975) Trigeminal nerve and eating in the pigeon {Columba livia): neurosensory control of the consummatory responses. J. Comp. Physiol. Psych. 89:845858. Zhou, Z. (1995) The discovery of Early Cretaceous birds in China. Courier Forschungsinstitut Senckenberg 181:9-22. Zusi, R. L. (1984) A functional and evolutionary analysis of rhynchokinesis in birds. Smiths. Contr. Zool. 395:1-40. Zweers, G. A. (1974) Structure, movement, and myography of the feeding apparatus of the mallard {Anas platyrhynchos L.): a study in functional anatomy. Neth. J. Zool. 24:323-467. Zweers, G. A. (1982a) Pecking in the pigeon {Columba livia). Behaviour 81:173-230.
Zweers, G. A. (1982b) The Feeding System of the Pigeon (Columba livia L). Springer-Verlag. Zweers, G. A. (1985) Generalism and specialism in avian mouth and pharynx. Pp. 189-201. In: Functional Morphology in Vertebrates (Fortschritte der Zoologie, Vol. 30). H.-R. Duncker and G. Fleischer (eds.). Gustav Fischer, Stuttgart. Zweers, G. A. (1991a) Pathways and space for evolution of feeding mechanisms in birds. Pp. 530-547. In: The Unity of Evolutionary Biology. Dioscorides Press. Zweers, G. A. (1991b) Transformation of avian feeding mechanisms: a deductive method. Acta Biotheor. 39:15-36. Zweers, G. A., H. Berkhoudt, and J. C. Vanden Berge (1994) Behavioral mechanisms of avian feeding. Pp. 241-279. In: Biomechanics of Feeding in Vertebrates. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Adv. Comp. Environ. Physiol., Vol. 18. Springer-Verlag, Berlin.
C H A P T E R
12 Feeding in Birds: Approaches and Opportunities MARGARET RUBEGA Department of Ecology and Evolutionary Biology University of Connecticut Storrs, Connecticut 06268 covered a previously unknown feeding mechanism, surface tension transport (STT) (Fig. 12.2), in phalaropes (Phalaropus), despite decades of previous study on their foraging (e.g.. Bent, 1927; Tinbergen, 1935; Mercier and Gaskin, 1985; Jehl, 1986). Similarly, Piersma et ah (1998) demonstrated the existence of a previously unknown prey detection mechanism in red knots {Calidris canutus). Quite aside from our interest in the evolution of feeding mechanisms per se, this lack of understanding has important consequences for the evolutionary and ecological study of birds. First, without an informed understanding of feeding mechanisms, we may seriously err in our ideas about the dietary and energetic strategies available to birds, and hence about one of the most fundamental aspects of the selective regimes they operate, and evolve, under. Note that I am distinguishing here (and hereafter) between our (frequently extensive) information about what birds eat, and our relatively poor understanding of how they eat it ("food acquisition'' sensu Zweers et ah 1994) , and of how bill morphology influences the latter. Second, our knowledge of avian feeding mechanics circumscribes our ability to understand how foraging relates to other behaviors. Prey intake rates are an important component of many behavioral and energetic models of the process of habitat choice (Krebs and Davies, 1991; Sutherland, 1996). Indeed, much of optimal foraging theory was built upon studies of avian subjects (Stephens and Krebs, 1986). Yet because we rarely understand the functional relationship of feeding movements in birds to actual ingestion rate, our data frequently constitute estimates of intake rates with unknown error terms. Often we are unable even to distinguish the specific prey being taken.
I. INTRODUCTION 11. PATTERNS OF ANALYSIS A. Systematics and Choice of Taxa B. Inferring Function from Structure versus Tests of Hypotheses C. Statistical Analysis, Sample Sizes, and the Importance of Variation III. CONCLUSION References
L INTRODUCTION The feeding structures of birds are probably more diverse than those in any animals except insects (Fig. 12.1). This dramatic modification of the feeding structures in birds has attracted a good deal of attention, historically, on the basis of the idea that where there is a crossbill, there must be an interesting feeding mechanism. In addition, bird beaks and their workings have long been attractive subjects because students of evolution reasonably presume that extreme modification of structures as fundamental to survival as mouthparts is likely the result of strong selection. Indeed, Darwin's (1859) ideas about evolution by natural selection were influenced by variation in beak size and shape in Galapagos finches (Geospizinae). Studies of the influence of this variation on survival via the ability to crack hard seeds in hard times remain a classic demonstration of evolution in the wild (Boag and Grant, 1981; Grant, 1985). Nonetheless, to a great degree, avian feeding mechanics and functional morphology remain poorly understood. For example, Rubega and Obst (1993) disFEEDING (K. Schwenk, ed.)
395
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
396
Margaret Rubega
no review that examines the entire published literature with a view specifically toward identifying areas where new efforts are liable to produce new insights into the evolution of avian feeding mechanisms. Rather than merely repeat an overview of the existing literature, this chapter aims to (1) identify patterns in the nature of past and present work on avian feeding mechanics and (2) suggest areas where new investigations might be particularly informative, and approaches which I believe will be especially productive.
IL PATTERNS OF ANALYSIS A. Systematics and Choice of Taxa A survey of the literature (see earlier list of reviews) reveals that feeding structure and/or mechanisms have been investigated in a wide variety of avian taxa, but opportunities for significant new contributions are abundant. The majority of the published analyses I could locate are single species ("idiographic"; see Chapter 1) studies, and I have used the term "analysis" very broadly. In a large number of cases, the analysis consisted of an examination of beak morphology and the subsequent generation of (frequently untested) hypotheses about the functional significance of features of the beak (see Section II,B), rather than direct examinations of feeding patterns and mechanics.
FIGURE 12.1. Diversity of feeding structures in birds. (A) Hyacinth macaw (Anodorhynchus hyacinthinus), (B) southern giant petrel {Macronectes giganteus), (C) parakeet auklet {Aethia psittacula), (D) wrybill (Anarhynchus frontalis), (E) Andean avocet (Recurvirostra andina), (F) whippoorwill {Caprimulgus vociferus), and (G) African spoonbill {Platalea alba). Feeding structures and feeding mechanics of these species are unstudied. Drawings by M. J. Spring.
Because of the long history of interest in feeding in birds, there is a large literature, with multiple reviews published since the mid-1980s. Gottschaldt (1985) reviewed sensory receptors in the bill; Berkhoudt (1985) summarized information on taste receptors; Zusi (1984) presented a detailed review and analysis of avian rhynchokinesis; Vanden Berge and Zweers (1993) reviewed the myology of the avian feeding apparatus; and Zweers et al. (1994) reviewed behavioral aspects of feeding mechanisms. To date, however, there has been
FIGURE 12.2. Surface-tension feeding in phalaropes. Small invertebrate prey are seized in the tips of the jaws and are transported in the water that adheres to the bill. Water is adhesive to the surface of the bill; by spreading its jaws the bird stretches the drop. The increase in potential energy resulting from the increase in the surface area of the drop drives the drop and prey along the bird's bill into the buccal cavity. Reproduced from Rubega (1997), with permission.
12. Feeding in Birds Throughout the remainder of this chapter I refer only to feeding analyses of the following types: (1) detailed anatomical descriptions of the feeding apparatus [e.g., Homberger's (1986) now-classic treatment of the tongue in the African grey parrot, Psittacus erithacus], (2) experimental (or at least controlled) examinations of motor patterns and feeding mechanisms in live animals (e.g., Tomlinson's analysis of cranioinertial feeding in paleognaths. Chapter 11), or (3) studies that combine the two [e.g., Zweers et al.'s (1977) brilliant and comprehensive analysis of feeding and the feeding apparatus in mallards. Anas platyrhynchos]. I specifically exclude uni- or bidiraensional comparisons of bill size [e.g., tables of bill lengths, commonly found in, but not restricted to, identification guides, such as Prater et al.'s (1984) guide to Holarctic waders], casual observations of free-living birds, and untested speculation about feeding mechanisms based on either. Species in approximately 19 of 25 orders have been the subject of some form of feeding analysis. Although this may seem like rather extensive coverage of the class, it should be noted that the majority of bird families remain completely unexamined. Published work to date covers only about 49 of 158 families, i.e., details of feeding structure and mechanics are unknown for almost 70% of all families of birds. Table 12.1 identifies the orders and families of birds for which no published work on the details of either bill morphology or feeding mechanism could be located. It can be assumed that I have failed to locate every published feeding study on birds. Also, my assessment of the degree to which we are uninformed about avian feeding depends on the classification of birds used. I have used a traditional classification (Morony et ah, 1975; del Hoyo et ah, 1992), rather than a newer, still controversial classification (Sibley et ah, 1988, Sibley and Ahlquist, 1990) with fewer orders and families (see Section II,C). Nonetheless, even if my estimate of the number of published studies was doubled, my overall conclusion would not change: there are about 9000 extant species of birds, and we know little or nothing about feeding structure and mechanics for the majority of them. This survey reveals that the field lacks a phylogenetic strategy with respect to the taxa investigated. In a few cases, systematic and purposeful within-clade comparative work has been done [e.g., passerines, Passeriformes (Bock, 1960); waterfowl, Anatidae (Goodman and Fisher, 1962); woodpeckers, Piciformes (Spring, 1965); kingfishers and allies, Coraciiformes (Burton, 1984)]; however, most of the literature on avian feeding seems to have been largely driven by (a) convenience [e.g., the investigator works with a common, or commonly available, species such as
397
chickens, Gallus domesticus (e.g., Calhoun, 1933; White, 1968; Lucas and Stettenheim, 1972; McClelland, 1979; Bhattacharyya, 1980; Berkhoudt, 1985; Homberger and Meyers, 1989; Van den Fleuvel, 1992), or domestic pigeons, Columba livia (e.g., Lucas and Stettenheim, 1972; Zeigler et al, 1980; Zweers, 1982a,b, 1985; Bermejo et al, 1989), thus each of these are disproportionally well known relative to their phylogenetic importance]; (b) serendipity, in which the investigator is studying something else and makes a chance observation [e.g., observations leading to the discovery of surface tension feeding in phalaropes (Rubega and Obst, 1993) were made during the making of an educational film (University of California 1985)]; or (c) the allure of the extreme [e.g., flamingo, Phoenicopteriformes, feeding mechanisms have been much more intensively studied (e.g., Jenkin, 1957; Kear and Duplaix-Hall, 1975; Zweers et al, 1995) than those of tyrant flycatchers, Tyrannidae, for example, which are far more speciose (~ 374 species: del Fioyo et ah 1992) and widely distributed)]. To be sure, these criteria have produced a wealth of information about the diverse ways in which birds capture and process their food. Yet available data are so thinly scattered across taxa that it would be impossible to confidently assert anything about the higher-level evolution of avian feeding systems (Table 12.1). In fact, to date the field not only lacks a widely accepted general theory explaining the evolution and diversity of avian feeding mechanisms (Lauder, 1989), but lacks a core group of plausible hypotheses, which are being systematically evaluated. For example, in the context of attempting to construct a general theory, Zweers (1991a,b) has stated that pecking mechanisms occur in all modern birds, and thus pecking is the ancestral condition (Zweers et ah, 1994; Zweers and Gerritsen, 1997; Zweers and Vanden Berge, 1997). This assertion is intuitively attractive, but made in the absence of information about the feeding mechanics in two-thirds of the families of birds, its accuracy remains to be demonstrated. The hypothesis is certainly true if pecking is defined sufficiently broadly. This is not mere hairsplitting; if defined sufficiently broadly, pecking is present in all reptiles as well. What, if anything, makes avian pecking characteristically avian, as opposed to reptilian in nature? Is there only one kind of avian pecking, arising from one conserved motor pattern underlying this approach to food grasping? This would be an interesting and impressive finding. If not, how many kinds of pecking are there, how are they distributed among taxa, and what is their relationship to the diversity of feeding structures in birds? An even more compelling question is what, if anything, makes avian feeding, as a whole, characteristically avian?
398
Margaret Rubega TABLE 12.1
Unstudied Orders and Families in the Class Aves''
Common name
Approx. no. of species in family
Order
Family
Struthioniformes
Struthionidae Rheae Casuariidae Dromaiidae Apterygidae
Ostrich Rheas Cassowaries Emu Kiwis
Tinamiformes
Tinamidae
Tinamous
47
Sphenisciformes Gaviiformes
Spheniscidae Gaviidae
Penguins Loons (divers)
17 4
Podicipediformes
Podicipedidae
Grebes
22
Diomedeidae Procellariidae Hydrobatidae Pelecanoididae Phaethontidae Pelicanidae Sulidae Phalacrocoracidae Anhingidae Fregatidae
Albatrosses Petrels, shearwaters Storm petrels Diving petrels
14 70 20 4
Tropic birds Pelicans Gannets, boobies Cormorants Darters Frigate birds
3 7 9 39 2 5
Ardeidae Scopidae Ciconiidae Balaencipitidae Threskiornithidae
Herons Hamerkop Storks ShoebiU Ibises, spoonbills
60 1 19 1 32
Procellariformes
Pelecaniformes
Ciconiiformes
Phoenicopteriformes
1 2 3 1 3
Order
Family
Common name
Approx. no. of species in family 8 2 1 7 1 13 9
Charadriidae Scolopacidae Thinocoridae Chionididae Stercorariidae Laridae Rynchopidae Alcidae
Jacanas Painted snipe Crab plover Oystercatchers Ibisbill Avocets, stilts Thick knees Coursers, pratincoles Plovers Sandpipers, snipe Seedsnipe Sheathbills Skuas Gulls, terns Skimmers Auks
Columbiformes
Pteroclididae Columbidae
Sandgrouse Pigeons, doves
16 283
Psittaciformes
Loriidae Cacatuidae Psittacidae
Lories Cockatoos Parrots
55 18 271
Cuculiformes
Musophagidae Cuculidae
Turacos Cuckoos
19 130
Strigiformes
Tytonidae Strigidae
Barn owls "Typical" owls
12 134
Caprimulgiformes
Steatornithidae Podargidae Nyctibiidae Aegothelidae Caprimulgidae
Oilbird Frogmouths Potoos Owlet-nightj ars Nightjars
1 13 5 8 76
Apodidae Hemiprocnidae Trochilidae
Swifts Tree swifts Hummingbirds
82 4 338
Charadriiformes
5
Jacanidae Rostratulidae Dromadidae Haematopodidae Ibidorhynchidae Recurvirostridae Burhinidae Glareolidae
16 64 86 4 2 5 90 3 23
Phoenicopteridae
Flamingoes
Anseriformes
Anhimidae Anatidae
Screamers Ducks, geese, swans
Falconiformes
Cathartidae Pandionidae Accipitridae Sagittariidae
7 New World vultures, Osprey 1 217 Hawks, eagles Secretary bird 1
Apodiformes
Coliiformes
Coliidae
Mousebirds
Galliformes
Megapodiidae Cracidae
Trogoniformes
Trogonidae
Trogons
37
44 213 1
Coraciiformes
Phasianidae Opisthicomidae
Megapodes Guans, chachalacas. currasows Pheasants, grouse Hoatzin
Mesitornithidae Turnicidae Pedionomidae Gruidae Aramidae Psophiidae Rallidae Heliomithes Rhynochetidae Eurypygidae Cariamidae Otididae
Mesites Button quails Plains wanderer Cranes Limpkin Trumpeters Rails, coots Finfoots Kagu Sunbittern Seriemas Bustards
3 14 1 15 1 3 133 3 1 1 2 24
Alcedinidae Todidae Motmotidae Meropidae Coraciidae Brachypteraciidae Leptosomatidae Upupidae Phoeniculidae Bucerotidae
Kingfishers Todies Motmots Bee eaters Rollers Ground rollers Cuckoo roller Hoopoe Woodhoopoes Hornbills
90 5 9 21 11 5 1 1 8 44
Piciformes
Galbulidae Bucconidae Capitonidae Indicatoridae Ramphastidae Picidae
Jacamars Puffbirds Barbets Honeyguides Toucans Woodpeckers
17 34 81 14 33 204
Gruiformes
3 147
12
6
(continues)
399
12. Feeding in Birds TABLE 12.1 (continued)
Order
Family
Passeriformes
Eurylaimidae Dendrocolaptidae Furnariidae Formicariidae Conopophagidae Rhinocryptidae Cotingidae Pipridae Tyrannidae Oxyruncidae Phytotomidae Pittidae Xenicidae Philepittidae Menuridae Atrichornithidae Alaudidae Hirundinidae Motacillidae Campephagidae Pycnonotidae Irenidae Laniidae Vangidae Bombycillidae Dulidae Cinclidae Troglodytidae Mimidae Prunellidae Muscicapidae
Common name
Approx. no. of species in family
14 Broadbills 52 Woodcreepers Ovenbirds 218 Antbirds 228 Gnateaters 11 Tapaculos 30 Cotingas 79 57 Manakins Tyrant flycatchers 374 Sharpbill 1 Plantcutters 3 Pittas 24 New Zealand wrens 4 Asities 4 Lyrebirds 2 Scrub birds 2 Larks 77 Swallows, martins 80 Wagtails, pipits 54 Cuckooshrikes 70 Bulbuls 123 Leafbirds, ioras. fairy bluebirds 14 Shrikes 74 Vanga shrikes 13 Waxwings 8 Palmchat 1 Dippers 5 Wrens 59 Mockingbirds, thrashers 31 Accentors 12 Thrushes and allies 1423
Order Passeriformes (continued)
Family Aegithalidae Remizidae Paridae Sittidae Certhiidae Rhabdornithidae Climacteridae Dicaiedae Nectariniidae Zosteropidae Meliphagidae Emberizidae Parulidae Drepanididae Vireonidae Icteridae Fringillidae Estrilididae Ploceidae Sturnidae Oriolidae Dicruridae Callaeidae Grallinidae Artamidae Cracticidae Ptilonorhynchidae Paradisaeidae Corvidae
Common name
Approx. no. of species in family
Long-tailed tits Penduline tits Tits, chickadees Nuthatches Treecreepers Philippine creepers Australian creepers Flowerpeckers Sunbirds White eyes Honeyeaters Buntings, cardinals. tanagers New World warblers Hawaiian honeycreepers Vireos New World blackbirds Finches Waxbills Weavers, sparrows Starlings Orioles Drongos Wattlebirds Magpie-larks Woodswallows Butcherbirds Bowerbirds Birds of paradise Crows, jays
8 10 27 25 6 2 6 58 116 83 171 558 126 23 43 95 122 127 143 111 28 20 3 4 10 8 18 42 105
^Taxa for which I could identify no published studies of the functional morphology of the feeding structures or feeding mechanics are bold faced. Classification is traditional and follows del Hoyo et al (1992) and Morony et ah (1975).
Progress currently is hampered by our lack of focus, phylogenetically speaking. A long list of authors have persuasively stated the case for a phylogenetic approach to understanding the evolution of complex behaviors and their ecological relevance (see citations summarized in Brooks and McClennan, 1991; Losos and Miles, 1994). It is therefore striking to note that the modern study of feeding mechanics in all vertebrates, except birds (and possibly mammals), is proceeding within an explicitly evolutionary framework, using phylogenetic tools and approaches (see other chapters in this book). What factors have prevented investigators of avian
feeding mechanisms from following suit? As with other groups of vertebrates, the bulk of all work to date was done prior to the rise of phylogenetic methods. More recently, ornithologists have been hampered by the lack of a rigorous phylogeny for the class. The ordinal level relationships of birds are still poorly understood (Raikow, 1985; Cracraft, 1988). The development of a phylogeny for the class is impeded by an insufficient inventory of cladistic characters (Cracraft, 1988). Such a phylogeny is essential to understanding the evolution of avian feeding mechanisms, as a basis for the generation of sampling schemesn, and for mapping character states.
400
Margaret Rubega
This is not to say that no higher-level phylogenetic analyses exist; Sibley and associates have provided a phylogenetic hypothesis for the whole class (Sibley et ah, 1988; Sibley and Ahlquist, 1990). Unfortunately, their characters, methods, and assumptions have serious weaknesses and have been widely criticized (e.g., Krajewski, 1991; O'Hara, 1991; Raikow, 1991; Lanyon, 1992). It has been argued, however, that this phylogeny at least presents us with a working hypothesis (O'Hara, 1991), and thus a starting place, and an opportunity for tests of hypotheses about the evolution of feeding mechanisms. To date, there are no alternative higherlevel phylogenies for the whole class. Nonetheless, ornithologists have been among the most active vertebrate systematists, and phylogenies below the ordinal level now are available for many groups of birds. A shortage of phylogenies, of course, is not the only factor preventing a shift toward phylogenetic (cladistic) methods. Some of the most active investigators of avian feeding methods and evolution have deliberately eschewed a phylogenetic approach in order to pursue alternate strategies of investigation. Zweers and colleagues, the most prolific and productive group currently working on avian feeding mechanics, have been developing an approach that deduces likely pathways of phenotypic transformation in avian feeding systems (Zweers, 1991a,b; Zweers and Vanden Berge, 1997; Zweers and Gerrittsen, 1997). They use functional optimality as the criterion for morphological and mechanistic change in order to generate testable ideas about the domain of all possible feeding mechanisms and evolutionary pathways to those mechanisms (see Chapter 1). In other words, they are using mechanical principles to generate ideas about how feeding mechanisms and bill morphology might have evolved, rather than looking at the distribution of feeding characters on a phylogeny, and then inferring the direction and nature of evolution. This approach shows some promise as a tool for generating hypotheses, but suffers from potential circularity. (The choice of optimal morphologies and mechanisms is unavoidably drawn from extant examples, which are then mapped onto the resulting transformation scheme, and are found to match.) Such a method will ultimately require grounding in a phylogenetic framework if it is to serve as a tool for understanding the actual evolution of existing feeding mechanisms. For instance, Zweers and associates (Zweers, 1991a,b; Zweers et al, 1994; Zweers and Gerritsen, 1997; Zweers and Vanden Berge, 1997) have proposed that avian filter-feeding mechanisms arose via modification of a charadriiform-like bill structure, and of the motor patterns associated with pecking. This idea posits the following sequence of events: (1) Shorebird-like bill structures are a modification of a more generalized
(pigeon-like) ancestral beak. In this first modification, elongation and slenderization contributed to improved probing performance. (2) Surface-tension transport, a mechanism by which birds transport prey along the bill using the physical properties of water droplets (Rubega and Obst, 1993; Rubega, 1997), arose as an epiphenomenon of this change in bill morphology. (3) Further modification of the bill structure and motor patterns arose as a consequence of specialization. (4) Modifications that increased the volume of prey and water processed in one feeding cycle led to flamingo- and duck-like bill morphologies, and hence to filter-feeding mechanisms. The initial generation of these ideas took place in a strictly biomechanical rather than phylogenetic context, and thus was not accompanied by the directed sampling of feeding structure and mechanism necessary to test most of the resultant hypotheses. An additional problem is that trophic mechanisms or morphologies are not coded in a manner that would lend itself to cladistic analysis. Zweers and Vanden Berge (1997) do provide a phenogram in which key trophic mechanisms and transitions are overlaid with the names of taxa that putatively have the mechanisms (Fig. 12.3), and then compare their scheme of "trophic radiation" to available phylogenetic analyses. Unfortunately, their phenogram includes many taxa for which detailed analysis of feeding mechanisms have not been conducted (e.g., stone curlew, Burhinus oedicnemus; spoonbill sandpiper, Eurynorhynchus pygmeus; crab plover. Dramas ardeola; ruff, Philomachus pugnax; Eurasian curlew, Numenius arquata; the screamers Anhimidae), or can never be done (e.g., Preshyornis, a fossil bird with a duck-like head and a shorebird-like axial skeleton). Also, they fail to map their character states (mechanisms or morphologies) directly onto existing phylogenies. Thus, it is difficult to evaluate their conclusion that their scheme of phenotypic transformation is largely congruent with cladistically produced phylogenies. Nonetheless, this detailed set of hypotheses can provide a basis for designing a sampling scheme that would contribute to our understanding of the evolution of feeding mechanisms in the shorebirds (Charadriiformes). There are a number of explicit predictions resulting from the Zweers model that can be tested. First, the model postulates that the ability to use surface-tension prey transport is simply a consequence of the basic shorebird bill structure. If true, then the capacity to employ this feeding mechanism should be found not only in the species in which it was discovered, the highly aquatic red-necked phalarope {Phalaropus lohatus), but in every shorebird with a straight, needlelike bill. Initial steps in a survey of the whole shorebird clade indicate that surface-tension transport of prey is indeed available to other phalaropes (Wilson's phala-
401
12. Feeding in Birds
organic ooze scraping mechanism
t
catching fish I shoveling shells 1 ^ grazing m's \ 1 A dabbling m's
\lt/
scaling of filter-feeding mechanisms
scaling of filter-feeding mechanisms
\t/
\t/
M
size & hardness recording touch mechanism: along jaw rami L- , « , . back & forth pump mechanism
at jaw tips
scaling of remote touch mechanisms stretching curved beak curving beak ^ combined vertical wedge mechanism \ remote 1 touch & \ 1 penetration filter-feeding m.
stab & crunch/ split & cut mechanisms
M e r g u s (merganser)
Polysticta Somateria (eider) Aythya (scaup)
Limnodromus (dowitcher) Scolopax (woodcock) Calidris (stint, sandpiper) Gallinago (snipe) T r i n g a (shank)
A n a s (wigeon)
L i m o s a (godwit) Numenius (curlew) Philomachus (ruff)
\u
\
grubbing '^'''®^^ ^^^^^^ hunting m's horizontal wedge mechanism curving/scaling/widening' mechanism
t suction pressure pump mechanism
\\t
grazing & filter-feeding (water) hole inspection mechanism mechanisms A T / scaling curved beak / probing mechanisms i ^
substrate penetration mechanisms
deep probe-hunting mechanism
grasp-pump graze-filter
\f
fetch and carry mechanism
^
»-
inspection/ turn-over/ chase m's
combined sight-peck & touch-probe m.
sit-watch & run-peck m. & superficial-probe m. stretch-catch m. & scaling , ^ _ ® T walk & peck mechanism top-soil breaking m. , ^ ^ fastened ingestion m. I sight-peck mechanism
f
grasp mechanism shores/ wetlands
Ciconidae (storks) Ardeidae (herons)
Burhinidae (stone curlews) shorebird-like ancestor
G a l l i f o r m e s (fowl) C o l u m b i f o r m e s (doves) pecking ancestor
FIGURE 12.3. A hypothesis of phenotypic transformations of avian feeding systems resulting in probing and filter feeding. (A) The branching pattern and the mechanisms along it were deduced by modifications of a pecking mechanism, optimized for probing and filter-feeding functions. (B) A phenogram of hypothetical evolutionary change in avian feeding systems, produced by overlaying taxa with appropriate feeding systems on (A). Reproduced from Zweers and Vanden Berge (1997), with permission.
rope, P. tricolor), as well as other species of shorebirds, including western {Calidris mauri) and least (C. minutilla) sandpipers (Rubega, 1997). The latter two species generally feed by probing in sandy or muddy substrates, hence it is unlikely that STT is a specialization for an aquatic lifestyle. Second, it follows from the optimization criteria used by Zweers that character states within the shorebirds that deviate from this basic needle-like bill morphology are derived and thus will be accompanied by improved performance of some other (new) feeding mechanism. The physical model for STT requires that deviations from a straight needle-like bill will result in a reduction in performance of surface-tension feeding (Rubega and
Obst, 1993). Some evidence for intra- and interspecific STT performance variation exists (Rubega, 1996,1997), but sampling of a broader array of shorebird bill morphologies would be informative. One interesting observation points to the importance of detailed and quantitative performance testing: American avocets {Recurvirostra americana) hatch with a needle-like bill that subsequently develops into a structure that is markedly dorsoventrally flattened and recurved. Hatchlings employ surface-tension transport throughout the transition from one morphology to another (Harker, Rubega, and Oring, unpublished observation). Field observations indicate that mean feeding performance increases as chicks grow (i.e., during
402
Margaret Rubega
the transition from a needle-shaped bill to one that is dorsoventrally flattened and recurved) (Harker and Rubega, unpublished observation). This appears to contradict the prediction that STT performance should decrease with deviations from a needle-shaped bill. These field observations, however, cannot distinguish improved performance at the individual level from improvements in the mean performance of age classes, which could result from the elimination of poorly performing individuals from the population by selection. Only quantitative measures of clearly defined performance characters under controlled conditions (i.e., in the laboratory) allow us to test directly the relationships between variation in morphology and feeding mechanics (e.g., Rubega, 1996). A third prediction of Zweers' model for the evolution of filter feeding is that character transformations, including increased bill volume, a tongue-based water pump, and straining structures, lead to a filter-feeding mechanism (Zweers et al., 1994). In its general outline of progression from a simple bill with low internal volume to a higher-volume, complex bill with filtering structures, this model is plausible, even compelling. As a phylogenetic hypothesis for the evolution of filter feeding in extant birds, the model is at odds with phylogenetic information, as it appears to suggest that the Anseriformes (the avian lineage in which filter feeding is most widespread and developed) arose from a shorebird (charadriiform) ancestor (Zweers and Vanden Berge, 1997; see Fig. 12.3B). All available evidence points to a sister-group relationship between the Anseriformes and Galliformes, which together form a clade that is the sister group of all other neognaths, with no special relationship to the charadriiform clade (Ho et al, 1976; Sibley et al, 1988; Cracraft, 1988). Nonetheless, it is conceivable that, despite their galliform relationship, anseriform ancestors had a simple, plover-like bill, which might have been subsequently modified for filter feeding. There is no paleontological evidence for or against this idea. With the proper sampling opportunities, however, we could test Zweers' ideas about the evolution of filter feeding within the Anseriformes. Although there are no anseriform extant taxa with Zweers' hypothesized ancestral plover-like (simple) bill morphology, the basic ideas of the model may be testable within the charadriiform lineage instead. For example, red phalaropes {Phalaropus fulicaria), which are so closely related to red-necked phalaropes as to be virtually indistinguishable genetically (Dittman et al, 1989; Dittman and Zink, 1991), have a bill that is wider and deeper (i.e., has a larger internal volume). Red phalaropes also have small internal bill structures that may be simple filtering systems (personal observation). Evidence shows that they select prey within
a rather narrow size range (Dodson and Egger, 1980; Mercier and Gaskin, 1985), as would be expected if filter size limits prey-capture performance. Red phalaropes thus provide a putative intermediary in which to examine the mechanistic predictions of the Zweers model for the evolution of filter feeding. If these predictions appear to be supported, it would be of interest to consider what factors may have prevented further evolution of filter feeding, which is otherwise not known to be present among the shorebirds. B. Inferring Function from Structure versus Tests of Hypotheses Possibly no group of vertebrates has a richer history of anatomical description than birds. Beginning with Aristotle, a long line of investigators has been carefully dissecting and describing birds in an attempt to understand their anatomy. From the 16th century onward, the detailed description oLanatomy was particularly important, second only to plumage descriptions (which were frequently all an investigator had, prior to the discovery of methods to preserve tissues) as a means of classification (Stresemann, 1975). When binoculars became widely available early in the 20th century, the focus of mainstream ornithology shifted to avian behavior, but beautiful and useful anatomical descriptions continued to be produced, particularly among German and Dutch investigators (e.g., Fiirbringer, 1922). While the anatomical descriptions in these studies were masterful and comprehensive, they constituted only the first step in understanding the role morphological structures play in avian feeding. Attempts to understand the anatomy they were describing naturally led investigators to formulate hypotheses about the function and evolutionary significance of various structures and structural complexes. Unfortunately, there has been a tendency in the ornithological literature to elevate such hypotheses to the status of fact. Nowhere has this practice been more evident than in the description of the avian feeding apparatus. Because of the obvious and dramatic modifications of the bill, there has been much speculation on the functional relationship between bill structures and feeding mechanics. In the most common approach, feeding mechanics are inferred from morphological features revealed by anatomical dissection, rather than directly observed or experimentally verified (see discussion in Chapter 1). This approach, dubbed "adaptive storytelling" in Gould and Lewontin's (1979) now-famous "Spandrels" paper and extensively criticized since, has been slow to fade in the avian literature and is still surprisingly common. An examination of the beaks of different birds that feed on the same prey provides
403
12. F e e d i n g in Birds
ample demonstration that many tools can do one job (Fig. 12.4) and implies that they are unlikely to do the job in simple, easily predicted ways. It seems likely that adaptive storytelling about feeding in birds is driven more by what is already known about the diets of birds than by informed understanding about the relationship between feeding structures and function. Occasionally, the storytelling approach is given a veneer of experimentalism via manipulation of specimens in ways meant to reveal the functional relationship among structures (e.g., pulling on a muscle to see if the mouth opens). The weakness of this approach is nicely illustrated by Dial's (1992) study of the avian flight apparatus, which showed clearly that flapping flight in birds is powered by different muscles, firing at different times, than anatomical description and manipulation of dead specimens had led investigators to believe. The frequency of the earliest adaptive storytelling is not surprising, not least because many of the hypotheses generated by anatomists would have been difficult or impossible to test without modern technology. Nonetheless, adaptive storytelling appears to have persisted for a series of obvious and not-so-obvious reasons. First, it is, quite simply, easier to formulate a hypothesis than it is to test it. Even in the post-Spandrel era, when most investigators have learned to label clearly untested hypotheses as speculation, such hypotheses are often untested. For example, Zusi (1984) pointed out that hypotheses concerning cranial kinesis arising from his anatomical analysis of bony hinges in bird skulls should be tested against observations of bill movement in living birds. It appears they never have been. Sometimes this failure to follow-up tends to occur because our ideas about the relationship between form and function in avian feeding have not been translated into formalized, falsifiable null and alternative hypotheses, and thus are weak generators of testable predictions. For instance, ornithologists have long guessed that the unique spinning behavior of phalaropes serves to "stir u p ' ' prey from the bottoms of ponds and lakes where they were feeding, but this idea failed to explain why birds spin while at sea over water many fathoms deep. It was only when this idea was formalized into testable hypotheses about the specific patterns of water flow generated by spinning that it became possible to show that spinning by phalaropes does draw prey to the surface, but by creating an upwelling rather than by stirring {Ohstetal,1996). As in other vertebrates, it often is difficult even to guess how intricate structural complexes might function (e.g., there are many unanswered questions about the functioning of the avian tongue in feeding and drinking; Homberger, 1988), hence it is difficult to generate clear, testable predictions about the relation-
K ^ ^ . l } . . . . , "I rry
F I G U R E 12.4. An example of the diversity of feeding structures associated with feeding on a single prey type. All these species eat fish. (A) Brown pelican {Pelecanus occidentalis), (B) horned puffin {Fratercula comiculata), (C) common loon (Gavia immer), (D) shoebill {Balaeniceps rex), and (E) red-breasted merganser {Mergus senator). Drawings by M. J. Spring.
ship between parts and feeding mechanics. There appears to be no cure for this problem but creativity and empiricism.
404
Margaret Rubega
To some extent, untested hypotheses based only on descriptions of structure have also accumulated because birds are uniquely vagile organisms, thus the range of feeding circumstances is huge, and the opportunities for observation (much less manipulation) of live, feeding birds are limited. Taxa that feed in midair (e.g., flycatchers, see discussion later) or underwater (e.g., penguins, Spheniscidae, and auks, Alcidae) are particularly poorly known. Even when observation is possible and good hypotheses exist, in many cases there are significant technical barriers to testing hypotheses directly, e.g., the feeding event happens too rapidly to be discernable with the naked eye. For example, in rednecked phalaropes, surface tension transport can be completed in as little as 0.002 sec (Rubega and Obst, 1993) and would have been undetectable without the aid of high-speed videography. Occasionally, we fail even to discard hypotheses that have been falsified. For example, ornithologists have long guessed that the brush-like array of rictal bristles (feathers modified into fine, rather stiff, whiskers; Fig. 12.IF) present around the mouth margins of many species of fly-catching birds function to funnel prey into the gape. Lederer (1972) presented strong evidence that this is unlikely, and Conover and Miller's (1980) study of willow flycatchers {Empidonax traillii) clearly demonstrated that this is not the case. Birds caught prey equally well both before and after rictal bristles had been removed. Yet textbooks continue to assert that rictal bristles function as insect nets (e.g.. Gill, 1995). This may simply be a demonstration of the difficulty inherent to the dissemination of results at a time when investigators are more overwhelmed with new literature than ever before. Alternatively, this example may merely demonstrate that we are fonder of a good story than of the facts. This is unfortunate, as the facts generally are more interesting than any story we could make up. What are rictal bristles for? They are present to a greater or lesser extent in many birds (Lederer, 1972) (including birds such as kiwis, Dinornithiformes, flightless birds, which forage in leaf litter), but dense, basketlike arrays of them around the margins of the mouth have apparently arisen independently more than once in birds. A partial list of birds with prominent rictal bristles includes New World flycatchers (Tyrannidae), Old World flycatchers (Muscicapidae), shrikes (Lannidae), and frogmouths (Caprimulgidae). In all cases, insect capture on the wing is a significant part of the feeding biology. Conover and Miller (1980) presented evidence that rictal bristles may function to protect the eyes from strikes by missed prey or from parts of prey that may break up when seized. Increased emphasis on experimentalism and availability of new tools (e.g.. X-ray cineradiography and high-speed film and video) have contributed to a wel-
come and growing tendency to approach avian feeding with formalized hypothesis testing. Some recent examples include Benkman's elegant analysis of crossbill {Loxia sp.) feeding (Benkman, 1987,1988; Benkman and Lindholm, 1991); the impressive body of work amassed by Zweers and associates on greater flamingoes Phoenicopterus ruber (Zweers et al, 1995), pigeons (Zweers, 1982), ducks {Anas platyrhynchos, A. clypeata, and Ay thya fuligula) (Zweers et ah, 1977; Kooloos et ah, 1989), and sandpipers (Calidris sp.) (Gerritsen et ah, 1983; Gerritsen and van Heezik, 1985; Gerritsen and Meiboom, 1986); my own work on feeding mechanics in phalaropes (Rubega and Obst, 1993; Rubega, 1996,1997); Hulscher and Ens's (1991) analysis of the functional significance of bill shape in Eurasian oystercatchers {Haematopus ostralegus), and the clever experimental work of Piersma et al. (1998) on red knot prey detection mechanisms. In all these cases, real progress in our understanding of feeding in birds was achieved by the observation of live animals at close range under controlled (i.e., laboratory) conditions, inventive experimental approaches, the application of appropriate technology to reveal details of feeding mechanics, or all three. Most importantly, all these tools were employed in the deliberate testing of formalized, falsifiable hypotheses about the relationship of feeding structures to mechanisms of food capture and processing. C. Statistical Analysis, Sample Sizes, and the Importance of Variation Historically, investigators of feeding in birds have tended to base their studies on observations of few individuals. In some situations this is acceptable, but most of our understanding of avian feeding mechanisms is hampered by reliance on small sample sizes. Studies of avian feeding can be broken down into (1) those that primarily describe phenomena and (2) those that compare groups of organisms. Descriptions of phenomena do not require large sample sizes. A sample of one is sufficient to demonstrate that a structure or mode of feeding exists. Even in these studies, however, assessing whether the phenomenon occurs in more than one or two individuals is important to ensure that the observations are not aberrant. When one wants to compare groups (e.g., comparing structure among species or trying to relate performance to variation in morphology), a rigorous statistical analysis becomes important (Shaffer and Lauder, 1985a,b). Results of statistical analyses are only meaningful when applied to appropriate sample sizes. The vast majority of published studies of the feeding apparatus and feeding function, however, are based on fewer than five in-
12. Feeding in Birds dividuals (I am guilty of this myself: Rubega and Obst, 1993,1997); in many cases, the sample size is one. Why do avian feeding specialists persist in presenting results from such small samples? One of the most obvious reasons for this problem is the difficulty in obtaining, and keeping, sufficient numbers of live, healthy specimens. This problem is not unique to birds, but perhaps uniquely complicated by their volant nature. Birds can be much more difficult to catch than fish, lizards, or small mammals. Once caught, all but the smallest species of birds also require significantly more space and attention for captive maintenance. Experimental feeding setups for some species of birds (e.g., pursuit diving birds) can be too demanding of space and resources anywhere outside of a zoological park. Birds held in zoos are only rarely available for manipulative experiments. These problems are real, but by no means sufficient to explain the widespread lack of statistical rigor in the field. For example, warblers are completely unstudied with respect to feeding mechanics. Yet many species are widespread, abundant, easily caught in nets (as evidenced by the thousands banded yearly for studies of movement patterns), and require no more space for captive maintenance than a typical lizard or snake. The same is true for many other families of passeriform birds. Our failure to direct our attention to the opportunities present in these taxa is probably due to patterns identified earlier (see Section II,A). An important contributor to the lack of statistical rigor in the field is that journal editors and reviewers have continued to allow investigators of avian feeding mechanics to publish with small samples. This appears to be due, at least in part, to a tradition of belief that feeding patterns are "hardwired" (genetically inherited, rather than learned, and therefore largely invariant), thus a sample of one is as representative, and as informative, as a larger sample. As I have repeatedly pointed out, we actually have very little detailed information on the feeding process in birds, but we have enough to know that the notion of feeding patterns as invariant within a species must be at least partly false. First, although there is certainly a genetic component to control of the feeding process and development of the feeding apparatus in birds, it would be surprising in the extreme to find complete genetic fixation for most traits in the feeding complex. The huge range of variance in bill shape and feeding patterns across the class Aves attests, at a minimum, to the historical availability of population-level variation in feeding structures and selection for their modification. Further, evidence shows that extrinsic factors may influence adult bill morphology (and presumably feeding performance, if not pattern) via developmental plasticity James (1983) and NeSmith (1984; cited in Travis,
405
1994), for example, showed that variation in temperature and humidity induces variation in bill shape of nestling red-winged blackbirds {Agelaius phoeniceus). The existence and importance of interspecific feeding variation have mostly been assumed on the basis of observed variation in bill morphology among related species or inferred from observations of differential habitat use and diets. Few direct comparisons of multiple species employing a common feeding mechanism on a standard food type have been conducted. Exploratory analyses (of data from a small sample of individuals) indicate that there is significant interspecific variation among four species of shorebird (red-necked phalarope, Wilson's phalarope, western sandpiper, and least sandpiper) in the performance of surface-tension feeding. Motor patterns are similar, but vary quantitatively among species (Rubega, 1997). Variance in feeding structures, process, and performance clearly exists within species as well. For example, it has been demonstrated repeatedly that juvenile birds exhibit poor feeding performance (usually expressed as feeding efficiency, or the catch-to-attack ratio; poor performance is also inferred from diet restricted to prey that is presumed to be less preferred) relative to their adult conspecifics (for reviews, see Marchetti and Price, 1989; Wunderle, 1991). Increasing age is associated with improved feeding performance. Explanations offered for this pattern of an ontogenetic feeding shift include learning, physical maturation of the feeding apparatus, variance in the nutritional status of juveniles relative to adults, and competitive suppression of juvenile feeding by adults. To date, none of these have been accompanied by formalized, testable hypotheses linking them to the feeding mechanism itself. Additional explanations that are well worth pursuing include the effects of neurological maturation, the possibility that juveniles may exhibit superior performance (relative to adults) of "juvenile" feeding mechanisms, and the likelihood that the perceived improvement in mean feeding performance with age is due to the elimination of poorly performing individuals from the population due to selection. Finally, significant variation in feeding within groups (among individuals) has also been demonstrated. Red-necked phalaropes exhibit significant among-individual variation in the performance of surface-tension feeding as a function of morphological variation of the inside of the upper jaw (Rubega, 1996). It should be apparent by this point in this chapter that, aside from the importance of accounting for variability when assessing the generality of our conclusions about avian feeding mechanisms, variance in avian feeding structures and mechanics is (or should be), in itself, a statistic of interest to us. This is especially true given the extreme degree of variation in feeding in birds overall relative to other vertebrates. In any group of
406
Margaret Rubega
vertebrates, variation (especially at the individual level) is the basis for selection and thus reflects the opportunity for, and outcome of, the evolution of feeding mechanics. In particular, attention to individual-level variation may be revealing with respect to the direction of selection, which in turn may provide us with clues as to which aspect of function is selectively most important.
III. C O N C L U S I O N The exceptional degree of variation in the avianfeeding apparatus has stimulated a large and interesting literature on beak and tongue morphology, feeding mechanics, and behavioral aspects of foraging. Much, however, remains to be done. I have tried to show that opportunities for significant new contributions to our understanding of feeding in birds are abundant. There is no widely accepted general theory explaining evolution of the observed diversity in avian feeding mechanisms. This is at least partly due to the complete lack of information about many species of birds: detailed analyses of feeding structures and function are lacking for more than half of all families of birds. To date our choice of taxa has been largely opportunistic. Significant advances will be made only when a phylogenetic strategy is applied to the problem of choosing study taxa. Further advances will also require the controlled testing of formal hypotheses about the relationship of feeding structures to some aspect of function or performance, and coding of feeding mechanisms in ways that allow cross-taxa comparisons. Although the difficulties inherent in maintaining birds in captivity are not trivial, our confidence in the outcome of comparative studies will depend on statistically appropriate sample sizes, however difficult they may be to attain. Some of the tools required to achieve these goals (such as high-speed video cameras and family-level phylogenetic hypotheses) are increasingly available. Widely and properly applied, they could produce a renaissance in the study of avian feeding. Even in the absence of a renaissance, we stand to learn a great deal more about feeding in birds. References Benkman, C. W. (1987) Crossbill foraging behavior, bill structure, and patterns of food profitability. Wilson Bull. 99:351-368. Benkman, C. W. (1988) On the advantages of crossed mandibles: an experimental approach. Ibis 130:288-293. Benkman, C. W., and A. K. Lindholm (1991) The advantages and evolution of a morphological novelty. Nature 349:519-20.
Bent, A. C. (1927) Life histories of North American shore birds. Order Limicolae (Part I). U. S. National Museum Bulletin 142. Berkhoudt, H. (1985) Structure and function of avian taste receptors. In: Form and Function in Birds, Vol. 3. A. S. King and J. McClelland (eds). Academic Press. Bermejo, R., R. W. Allan, D. Houben, J. D. Deich, and H. P Zeigler. (1989) Prehension in the pigeon. I. Descriptive analysis. Exp. Brain Res. 75:569-576. Bhattacharyya, B. N. (1980) The morphology of the jaw and tongue musculature of the common pigeon, Columba livia, in relation to its feeding habit. Proc. Zool. Soc. Calcutta 31:95-127. Boag, P. T, and P. R. Grant (1981) Intense natural selection in a population of Darwin's finches (Geospizinae) in the Galapagos. Science 214:82-85. Bock, W. J. (1960) The .palatine process of the premaxilla in the Passere. Bull. Mus. Comp. Zool. 122(8):361-488. Burton, P. J. K. (1984) Anatomy and evolution of the feeding apparatus in the avian orders Coraciiformes and Piciformes. Bull. Brit. Mus. (Nat. Hist.) 47(6): 331-443. Brooks, D. R., and D. A. McLennan. (1991) Phylogeny, Ecology and Behavior; a Research Program in Comparative Biology. University of Chicago, Chicago. Calhoim, M. L. (1933) The microscopic anatomy of the digestive tract of Gallus domesticus. Iowa State Coll. J. Sci. 7:261-382. Conover, M. R., and D. E. Miller (1980) Rictal bristle function in willow flycatcher. Condor 82:469-471. Cracraft, J. (1988) The major clades of birds. In: The Phylogeny and Classification ofTetrapods, Vol. 1. M. J. Benton (ed). Systematics Association Special Volume No. 35A. Clarendon Press, Oxford. Darwin, C. 1859. On the Origin of Species hy Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. John Murray, London. Del Hoyo, J., A. Elliott, and J. Sargatal (eds.) (1992) Handbook of the Birds of the World Vol. 1. Lynx Edicions, Barcelona. Dial, K. P. (1992) Avian forelimb muscles and nonsteady flight: can birds fly without using the muscles in their wings? Auk 109: 874-885. Dittman, D. L., and R. M. Zink (1991) Mitochondrial DNA variation among phalaropes and allies. Auk 108:771-779. Dittman, D. L., R. M. Zink, and J. A. Gerwin. (1989) Evolutionary genetics of phalaropes. Auk 106:326-331. Dodson, S. I., and D. L. Egger. (1980) Selective feeding of red phalaropes on zooplankton of Arctic ponds. Ecology 61:755-763. Fiirbringer, M. (1922) Das Zungenbein der Wirbeltiere insbesondere der Reptilien und Vogel. Abhandlungen der Heidelberger Akademie, math.-nathurw. Kl. Abt.B 11:1-164. Gerritsen, A. F. C , and A. Meijboom (1986) The role of touch in prey density estimation by Calidris alba. Neth. J. Zool. 36:530-562. Gerritsen, A. F. C , and Y. M. van Heezik (1985) Substrate preference and substrate related foraging behavior in three Calidris species. Neth. J. Zool. 35:671-692. Gerritsen, A. R C , Y. M. van Heezik, and C. Sweenen (1983) Chemoreception in two further Calidris species: Calidris maritima and C. canutus; a comparison of the relative importance of chemoreception during foraging in Calidris species. Neth. J. Zool. 33: 485-496. Gill, R B. (1995) Ornithology, 2nd Ed. Rreeman, New York. Goodman, D. C , and H. I. Fisher (1962) Functional Anatomy of the Feeding Apparatus in Waterfowl (Aves: Anatidae). Southern Illinois University Press, Carbondale, IL. Gottschaldt, K. M. (1985) Structure and function of avian somatosensory receptors. In: Form and Function in Birds, Vol. 3. A. S. King and J. McClelland (eds). Academic Press, New York. Gould, S. J., and R. C. Lewontin (1979) The spandrels of San Marco
12. Feeding in Birds and the Panglossian paradigm: a critique of the adaptationist programme. Proc. R. Soc. Lond. B 205:581-598. Grant, P. R. (1985) Selection on bill characters in a population of Darwin's finches: Geospiza conirostris on Isla Genovesa, Galapagos. Evolution 39:523-532. Ho, C. Y.-K., E. M. Prager, A. C. Wilson, D. T. Osuga, and R. E. Feeney (1976) Penguin evolution: protein comparisons demonstrate phylogenetic relationships to flying aquatic birds. J. Mol. Evol. 8: 271-82. Homberger, D. G. (1986) The Lingual Apparatus of the African Grey Parrot Psittacus erithacus Linne (Aves: Psittacidae): Description and Theoretical Mechanical Analysis. Ornithological Monograph No. 39. American Ornithologists' Union, Washington. Homberger, D. G. (1988) Comparative morphology of the avian tongue. In: Acta XIX Congressus Internationalis Ornithologici, Vol. II. H. Ouellet (ed). University of Ottawa Press, Ottawa. Homberger, D. G., and R. A. Meyers (1989) Morphology of the lingual apparatus of the domestic chicken, Gallus gallus, with special attention to the structure of the fasciae. Am. J. Anat. 186:217-257. Hulscher, J. B., and B. J. Ens (1991) Somatic modifications of feeding system structures due to feeding on different foods with emphasis on changes in bill shape in Oystercatchers. Acta XX Congr. Inter. Ornith. Symposium 13:889-896. James, F. C. (1983) Environmental component of morphological differentiation in birds. Science 221:184-186. Jehl, J. R. (1986) Biology of the red-necked phalarope (Phalaropus lobatus) at the western edge of the Great Basin in fall migration. Great Basin Nat. 46:185-197. Jenkin, P. M. (1957) The filter feeding and food of flamingoes (Phoenicopteri). Phil. Trans. Roy. Soc. Lond. B 240:401-493. Kear, J. and N. Duplaix-Hall (1975) Flamingos. Poyser, Berkhamsted, UK. Kooloos, J. G. M., A. R. Kraaijeveld, G. E. J. Langenbach, and G. A. Zweers (1989) Comparative mechanics of filter feeding in Anas platyrhynchos, Anas clypeata, and Aythya fuligula (Aves, Anseriformes). Zoomorphology 108:269-290. Krajewski, C. (1989) Phylogeny and classification of birds: a study in molecular evolution. Auk 108:987-990. Krebs, J. R., and N. B. Davies (1991) Behavioural Ecology: An Evolutionary Approach. Blackwell, Oxford. Lanyon, S. M. (1992) Phylogeny and classification of birds: a study in molecular evolution. Condor 94:304-307. Lauder, G. V. (1989) How are feeding systems integrated, and how have evolutionary innovations been introduced? In: Complex Organismal Functions: Integration and Evolution in Vertebrates. D. B. Wake and G. Roth (eds). Wiley, Chichester. Lederer, R. J. (1972) The role of avian rictal bristles. Wilson Bull. 84: 193-197. Losos, J. B., and D. B. Miles (1994) Adaptation, constraint, and the comparative method: phylogenetic issues and methods. In: Ecological Morphology, P. C. Wainwright and S. M. Reilly (eds). University of Chicago, Chicago. Lucas, A. M., and P. R. Stettenheim (1972) Avian Anatomy, Integuement. Agricultural Handbook No. 362. U. S. Government Printing Office, Washington, DC. Marchetti, K., and T. Price (1989) Difference in the foraging of juvenile and adult birds: the importance of developmental constraints. Biol. Rev. 64:51-70. McClelland, J. (1979) Digestive system. In: Form and Function in Birds, Vol. 1. A. S. King and J. McClelland (eds). Academic Press, New York. Mercier, F., and D. E. Gaskin (1985) Feeding ecology of migrating red-necked phalaropes {Phalaropus lobatus) in the Quoddy region. New Brunswick, Canada. Can. J. Zool. 63:1062-1067
407
Morony, J. J., W. J. Bock, and J. Farand. (1975) Reference List of the Birds of the World. American Museum of Natural History, New York. NeSmith, C. C. (1984) The Effect of the Physical Environment on the Development of Red-Winged Blackbird Nestlings: A Laboratory Experiment. M.S. thesis, Florida State University, Tallahassee, FL. Obst, B. S., W. M. Hamner, E. Wolanski, P. P. Hamner, M. A. Rubega, and B. Littlehales (1996) Kinematics and fluid mechanics of spinning in phalaropes. Nature 384:121. O'Hara, R. J. (1991) Phylogeny and classification of birds: a study in molecular evolution. Auk 108:990-993. Piersma, T., R. van Aelst, K. Kurk, H. Berkhoudt, and L. R. M. Maas (1998) A new pressure sensory mechanism for prey detection in birds: the use of principles of seabed dynamics? Proc. Roy. Soc. Lond. B 265:1377-1383. Prater, A. J., J. H. Marchant, and J. Vuorinen (1984) Guide to the Identification and Aging of Holarctic Waders. British Trust for Ornithology Field Guide 17. Raikow, R. (1985) Problems in Avian Classification. Current Ornith. 2:187-212. R. J. Johnson (ed). Plenum, New York. Raikow, R. (1991) Phylogeny and classification of birds: a study in molecular evolution. Auk 108:985-987. Rubega, M. A. (1996) Sexual size dimorphism in red-necked phalaropes and functional significance of the nonsexual bill structure variation for feeding performance. J. Morph. 228:45-60. Rubega, M. A. (1997) Surface tension prey transport in shorebirds: how widespread is it? Ibis 139:488-493. Rubega, M. A., and B. S. Obst (1993) Surface tension feeding in phalaropes: discovery of a novel feeding mechanism. Auk 110:169178 + frontispiece. Shaffer, H. B., and G. V. Lauder (1985a) Aquatic prey capture in ambystomatid salamanders: patterns of variation in muscle activity. J. Morphol. 183:273-326. Shaffer, H. B., and G. V Lauder (1985b) Patterns of variation in aquatic ambystomatid salamanders: kinematics of the feeding mechanism. Evolution 39:83-92. Sibley, C. G. J. E. Ahlquist, and B. L. Monroe, Jr. (1988) A classification of the living birds of the world based on DNA-DNA hybridization studies. Auk 105:409-423. Sibley, C. G., and J. E. Ahlquist (1990) Phylogeny and Classification of Birds: A Study in Molecular Evolution. Yale Univ. Press, New Haven, CT. Spring, L. W. (1965) Climbing and pecking adaptations in some North American woodpeckers. Condor 67:457-488. Stephens, D. W., and J. R. Krebs (1986) Foraging Theory. Princeton Univ. Press, Princeton, NJ. Stresemann, E. (1975) Ornithology: From Aristotle to the Present. Harvard Univ. Press, Cambridge. Sutherland, W. J. (1996) From Individual Behaviour to Population Ecology. Oxford Univ. Press, Oxford. Tinbergen, N. (1935) Field observations of east Greenland birds. I. The behavior of the red-necked phalarope (Phalaropus lobatus, L.) in spring. Ardea 26:1-42. Travis, J. (1994) Evaluating the adaptive role of morphological plasticity. In: Ecological Morphology. P. C. Wainwright and S. M. Reilly (eds). University of Chicago Press, Chicago. University of California. (1985) Phalarope Feeding Behavior (film). From the film series Aspects of Animal Behavior. Office of Instructional Development, University of California, Los Angeles. Van den Heuvel, W. R (1992) Kinetics of the skull in the chicken (Callus gallus domesticus). Neth. J. Zool. 42:561-582. Vanden Berge, J. C , and G. A. Zweers (1993) Myology. In: Handbook of Avian Anatomy. J. J. Baumel (ed). Nuttall Ornithological Club, Cambridge.
408
Margaret Rubega
White, S. S. (1968) Mechanisms involved in deglutition in Gallus domesticus. J. Anat. 104:177. Wunderle, J. M., Jr. (1991) Age-specific foraging proficiency in birds. In: Current Ornithology, Vol. 8. D. M. Power (ed). Plenum Press, New York. Ziegler, H. P., P W. Levitt, and R. Levine (1980) Eating in the pigeon {Columba livia): movement patterns, stereotypy and stimulus control. J. Comp. Physiol. Psychol. 94:783-794. Zusi, R. L. (1984) A functional and evolutionary analysis of rhynchokinesis in birds. Smith. Contrib. Zool. 395:1-40. Zweers, G. A. (1982a) Pecking of the pigeon {Columba livia L.). Behaviour 81:173-230. Zweers, G. A. (1982b) The feeding system of the pigeon (Columba livia L.) Adv. Anat. Embryol. Cell Biol. 73:VII+108. Zweers, G. A. (1985) Generalism and specialism in the avian mouth and pharynx. Fortschr. Zool. 30:189-201. Zweers, G. A. (1991a) Transformation of avian feeding mechanisms: a deductive approach. Acta Biotheor. 39:15-36.
Zweers, G. A. (1991b) Pathways and space for evolution of feeding mechanisms in birds. In: The Unity of Evolutionary Biology, E. C. Dudley (ed). Dioscorides Press, Portland. Zweers, G. A., H. Berkhoudt, and J. C. Vanden Berge (1994) Behavioral mechanisms of avian feeding. Pp. 241-279, In: Advances in Comparative and Environmental Physiology, Vol. 18. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Zweers, G. A., F. de Jong, H. Berkhoudt, and J. C. Vanden Berge (1995) Filter feeding in flamingos (Phoenicopterus ruber). Condor 97:297-324. Zweers, G. A., and A. F. C. Gerritsen (1997) Transitions from pecking to probing mechanisms in waders. Neth. J. Zool. 47:161-208. Zweers, G. A., A. F. C. Gerritsen, and P. J. van Kranenburg-Vood (1977) Mechanics of feeding of the mallard (Anas platyrhynchos L.; Aves, Anseriformes). Contrib. Vert. Evol, Vol. 3. Karger, Basel. Zweers, G. A., and J. C. Vanden Berge (1997) Evolutionary transitions in the trophic system of the wader-waterfowl complex. Neth. J. Zool. 47:255-287.
S E C T I O N
VI
This Page Intentionally Left Blank
C H A P T E R
13 Feeding in Mammals KAREN M. HIIEMAE Department of Bioengineering and Neuroscience Institute for Sensory Research Syracuse University Syracuse, New York 13244
tion can be associated with the availability of habitats previously occupied by the highly successful dinosaur radiation whose species then occupied most, if not all, the ecological niches now occupied by mammals. Survival to reproductive age, followed by successful reproduction, requires an adequate food intake: in this context "adequate food intake" is defined as that which yields more biochemical energy than required for its collection, ingestion, and digestion. While that statement holds true for all vertebrates, warm-blooded mammals and birds require a higher and a sustained level of energy to maintain homeostatic mechanisms. This positive energy balance is needed to support tissue turnover and the maintenance of all body systems, as well as intrauterine fetal development followed by lactation (a uniquely mammalian reproductive mechanism). For most mammals in nontropical climates, there is also a need for the acquisition of metabolic reserves to be drawn upon in seasons where the available food supply diminishes and drives the metabolic equation into negative (e.g., winter in high latitudes or drought in all arid regions). It follows that not only must food be available to meet physiological demand, but it must be accessible and processable. Some mammals make food caches, others brown fat, and still others migrate in search of food and water. For survival:
I. INTRODUCTION 11. MAMMALIAN FEEDING SYSTEM A. Overview B. Approaches to the Study of Feeding in Mammals III. THE "PROCESS MODEL" FOR MAMMALIAN FEEDING IV. MECHANICAL PROPERTIES AND TEXTURES OF FOODS V. THE FEEDING APPARATUS A. Jaw Complex B. Oropharyngeal Complex VI. FEEDING FUNCTION A. Tongue-Jaw Linkages B. Food Manipulation and Movement VII. CONTROL OF FEEDING BEHAVIORS References
I. INTRODUCTION About 60 million years ago, an explosive evolutionary radiation of mammals took place (Romer, 1974). Although dental and cranial evidence shows that the earliest mammals had appeared some 120 million years ago, the extraordinary proliferation of mammalian genera occurring during and after dinosaur extinc-
metabolic metabolic cost metabolic cost reserves reserves yield of = of food + of body + for a n d / o r for seasonal food intake acquisition maintenance reproduction food shortage
FEEDING (K.Schwenk,ed.)
411
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
412
K a r e n M. H i i e m a e
During the evolution of mammals, the reptilemammal transition was characterized by the unique development of several fundamental and linked structural innovations. First, the dentary expanded to become the single lower jaw bone, articulating with the squamosal to form a new jaw joint. The quadrate and articular bones of the reptilian jaw joint were coopted to form the ossicular chain of the mammalian middle ear (see Crompton, 1995). Second, the palatal processes of the premaxilla and maxilla fused in the midline, forming a true hard palate completely separating the oral and nasal cavities. A new structure, the muscular soft palate, attached to the posterior margin of the hard palate, appeared as a mobile flap separating the airway from the oral cavity (see Smith, 1992). Third, the appearance of a highly differentiated tooth row restricted to, at most, a deciduous and a permanent dentition developed for both the acquisition and the processing of food (mechanical digestion). There were also associated fundamental changes in oropharyngeal soft tissues. The fiber directions of the adductor muscles of the jaw provided for mandibular movement in the anteroposterior and mediolateral directions, as well as the vertical. A functionally integrated intrinsic and extrinsic tongue musculature allowed for complex tongue movements, including differential expansion and contraction. A completely new neuromuscular complex of longitudinal and circular muscles forming the pharynx provided a new method of swallowing (see Smith, 1992). This system allowed for the more or less continuous movement of tidal air (respiration) as well as the intermittent transmission of swallowable material into the gastrointestinal tract. Although the taxonomy and putative monophyly of most mammalian orders is well established and reasonably stable, higher-level phylogenetic relationships of mammals (i.e., among orders) remain contentious (for reviews, see Novacek, 1992; Honeycutt and Adkins, 1993; also Meng et al., 1994; Springer ei al., 1997). Difficulties in resolving higher-level relationships might relate to the explosive nature of mammalian adaptive radiation during the Cretaceous (but see Hedges ei al., 1996). Such extreme rapidity of cladogenesis and phenotypic evolution may have led to extensive homoplasy in both molecular and morphological characters, thus confounding cladistic character analysis. Figure 13.1 illustrates a generally accepted phylogeny of mammals based on Novacek (1992), but it must be acknowledged that this cladogram is neither wholly agreed upon now nor likely to persist unchanged for long. Possibly as a consequence of uncertainty in the higher-level relationships of mammals, virtually no study has attempted an overarching, phylogenetic
MONOTREMATA MARSUPIALIA PHOLIDOTA XENARTHRA CARNIVORA INSECTIVORA MACROSCELIDEA LAGOMORPHA RODENTIA PRIMATES SCANDENTIA DERMOPTERA CHIROPTERA TUBULIDENTATA ARTIODACTYLA CETACEA PERISSODACTYLA HYRACOIDEA SIRENIA PROBOSCIDEA
F I G U R E 13.1. Phylogenetic relationships among mammalian orders based on Novacek (1992). See text for discussion.
analysis of feeding system evolution in mammals. Rather, there is a general dogma that most eutherian orders arose from a generalized, insectivorous ancestor with subsequent divergence and specialization. Thus, comparative approaches to mammal feeding typically are typological in the sense that feeding systems are characterized order by order (e.g., TurnbuU, 1970) with little attention paid to evolutionary transformations among systems. With a phylogenetic approach it should now be possible to reconstruct aspects of the feeding system at ancestral nodes and to examine patterns of character evolution. Unfortunately, such an analysis is beyond the scope of this chapter. Rather, this chapter establishes the fundamentals of mammalian feeding and reviews much of the known diversity in feeding systems, with emphasis on those relatively few taxa for which significant functional data are
13. Feeding in Mammals
available. Chapters 15 and 16 provide detailed coverage of specialized myrmecophagous and marine feeding systems, respectively. It is a telling fact that some mammal orders are named for dietary habit, e.g., Carnivora and Insectivora, indicating both the importance of feeding system characters in mammal taxonomy and the presumed stability of feeding system phenotype within (in contrast to among) orders. However, it is important to note that the actual diets of species in a given order range across the available food source spectrum (Table 13.1 and Fig. 13.2). Despite various modifications to the basic ordinal Bauplan in response to the mechanical demands of food collection or processing, members of each order retain the fundamental characters diagnostic of its group. For example, Ailuropoda (the giant panda) retains features identifying its carnivoran (ursid) origins, despite its highly specialized diet of bamboo. In short, phylogeny dictates the overall musculoskeletal anatomy of the orofacial complex, but the details can be very specific to genera, even species, sub-
TABLE 13.1 An Overview of the Range of Food Sources Utilized by Members of the Major Orders of Terrestrial Mammals'' Order
Diet
Marsupialia
Insectivores, omnivores, carnivores, herbivores
Insectivora
Insects, small vertebrates, blood, pollen, nectar, fruit (extreme specialization: anteaters^)
Chiroptera
Insects, blood, fruit, honey
Primates
Insects, fruit, leaves, nuts, small mammals (termites) (extreme generalization: H. sapiens)
Carnivora
Insects, Crustacea, fish, small/large animals, including carrion, fruit, honey/nectar, leaves [extreme variant: baleen whales (Spermaceti)]
Perissodactyla and Artiodactyla
Some omnivores; insects, roots, bulbs, fruit, buds, shoots, young and old leaves, grasses, aquatic vegetation (extreme variants: elephants,^ dugong, manatee'')
Hystricomorpha and Lagomorpha
Grasses, leaves, nuts, fruits
Rodentia
Insects, small animals, fish, nuts, leaves, grasses, seeds, bark, fruit
Edentata
Sloths: leaves Anteaters: termites (see Chapter 15)
^This list is not intended to be exhaustive, but rather indicative of the dietetic opportunism within groups. ^Indicates members of separate "specialized" orders derived from the primary order cited.
413
sequently adapted for the acquisition and digestion of specific foods. The American opossum, Didelphis marsupialis, has been used as the extant exemplar of the ancestral mammalian condition (Turnbull, 1970; Hiiemae and Crompton, 1985; earlier references cited therein). Whether or not modern Didelphis has itself evolved (other than in its opportunistic ability to take advantage of all available food sources, i.e., its northward spread into suburbia raiding garbage cans), it remains true that the oropharyngeal behavioral mechanisms of Didelphis provide a valuable baseline for other mammalian forms. In his "heroic" (Herring, 1993) monograph. Mammalian Masticatory Apparatus, Turnbull (1970) attributed the anatomy of the jaws, teeth, and the major adductor jaw muscles to four dietetic groups: (1) generalized group, exemplified by Didelphis and Echinosorex, including the primates; (2) specialized group I (carnivores), typified by the domestic cat {Telis domesticus); (3) specialized group II (ungulates), typified by the horse {Equus caballus), a deer {Odocoileus virginianus), and a sheep {Ovis aries); (4) specialized group III ("rodent/gnawing mammals"), exemplified by a squirrel {Sciurus niger), a rat (Rattus norvegicus), and a porcupine (Hystrix). Turnbull was also forced to recognize a miscellaneous group of "oddball" mammals that did not fit into his major categories. His primary focus was on the jaw musculature and its bony attachments, not on the dentition, still less on the actual mechanics of the feeding process. Nor was Turnbull much concerned about the evolutionary history of the mammals he used as exemplars, as noted earlier. However, he made an invaluable contribution, in part because he included primary source references for mammals other than those he used as exemplars for his groups. Since the 1950s, there has been an important shift in the focus of studies of the mammalian feeding apparatus from morphology (shape and structure) to mechanism (behavior and biomechanics), especially of the teeth and jaws (see reviews by Hiiemae, 1978; Hiiemae and Crompton, 1985; Herring, 1993; Weijs, 1994). Nevertheless, Turnbull's classification of dietary types is used here as the scaffold on which to base this review. Mammals are so diverse and their feeding mechanisms so varied that this chapter is designed to discuss (1) the processes involved in food acquisition and mechanical digestion; (2) the biomechanics of that process in the context of an "archetypal" primitive mammal; the American opossum (Didelphis virginiana); (3) the hroad brush variations on the mammalian Bauplan associated with broad dietetic categories; and (4) what is now known about the linkages between jaw and tongue functions in feeding. Primary source references
414
K a r e n M. H i i e m a e
OMNIVORES
ANIMAL = high protein
VEGETABLE = variable protein and lignin/celiulose
i
Mature leaves,^ Grasses Buds; S h o o t s /
Marine mammals/,^ large fish V ^
^^
Seeds, Nuts
Fish ^
Aquatic herbivores
Small mammals, ,,SbiKlS.
Moliusca Krill
^^ASW N9
Insects
Nectar
Eggs
BODY SIZE
Pollen
Generalized Mammal e.g. Opossum
BODY SIZE
F I G U R E 13.2. Schematic diagram showing broad categories of food sources available to mammals. Although there is some correlation betw^een food type and body size, this is in no w^ay absolute. The largest w^hales (Spermaceti) consume the smallest of Crustacea (krill). There is also no precise correlation between phylogenetic history and dietary habit (see Table 13.1).
are cited, but emphasis has been given to recent review articles to facilitate access to the relevant literature for those wishing to pursue a given area in depth. Substantial emphasis is placed on "process/' as recent experimental work has shown that jaw and tongue movements are linked and that this linkage has a far greater effect on jaw movement patterns than generally recognized to date (Hiiemae and Palmer, 2000). Given societal imperatives. Homo sapiens has become essentially a "species apart," a focus for anthropologists (human evolution), social scientists (behavior), and clinicians (disease). Although the only true facultative biped, the human is, otherwise, only "peculiar" in having evolved adaptations in the orofacial complex to accommodate a fundamental change in dominant oropharyngeal function from feeding to speech. Those changes are worthy of some attention. IL M A M M A L I A N FEEDING SYSTEM A. Overview The mammalian feeding system, like that of other vertebrates, can best be characterized as a tube or luminal space of variable dimensions, lined by epithelia with different properties, into which the products of exocrine glands are poured in the presence of food. In mammals, this tube, whose lumen is, sensu strictu, extracorporeal, subserves several functions: its anterior or rostral end is the oral cavity, where food ingestion and, almost uniquely among tetrapods, food break-
down occurs as a function of patterned jaw movements and tooth-food-tooth interaction (physiologically, "mechanical digestion"); immediately distal is the oropharynx, a lumen used both for food transport to the gastrointestinal tract and for movement of tidal air in respiration; the esophagus is a simple transthoracic/ transdiaphragmatic tube leading to the stomach, in turn to the intestines, all of which are involved in chemical digestion, in some cases with the assistance of symbiotic bacteria, before fecal formation in the colon for emission through the rectum and anus. Importantly, this tube is interrupted by a series of sphincters, dividing it into separable sections. These sphincters control the rate of passage of material along its length. This tube is supported directly by skeletal elements (oral cavity and oropharynx) or indirectly by suspension (peritoneum) from the posterior abdominal wall. Its included volumes are altered by the actions of the jaw and tongue muscles (oral cavity); by the striated longitudinal and circular striated pharyngeal muscles (pharynx); and then, in the lower esophagus and gastrointestinal (GI) tract, by the circular and longitudinal muscles of the gastric and intestinal walls. It follows that the upper part of the system is under "voluntary control," whereas motility in the remainder is governed by the electrotonic coupling of smooth muscle cells and is controlled by their intrinsic neural networks associated with the autonomic nervous system. This feeding system subserves four vital physiological functions: (1) the ingestion of nutrients (liquids
415
13. Feeding in Mammals and solids), (2) the preparation of solids for transmission across the oropharynx to the esophagus, (3) the breakdown of materials such that their included nutrients can be absorbed across the intestinal epithelium, and (4) the elimination of waste products, including indigestible materials. While for most mammals, the system serves as a "one-way channel," this is not always the case. For example, in the ruminant herbivores, material that has been transported to the stomach is regurgitated for further mechanical digestion. Conversely, some mammals ingest foods that are not processed in the oral cavity and are made nutritionally useful only when they reach the GI tract (e.g., ants, krill, see Chapters 15 and 16). The biochemistry of digestion and the not so subtle variations in gastrointestinal anatomy and physiology associated with particular diets cannot be addressed here. Similarly, adaptations in metabolism, indicated by food intake patterns to form reserves of brown fat, e.g., dictated by overwintering strategies such as hibernation, are far beyond the scope of this chapter. Important elements in the investigation of the interactions among the ecology, anatomy, and physiology of mammals are body size, metabolic rate, and the "metabolic yield" of preferred foods. These have to be factored into the interpretation of behavior, and the underlying physiological imperatives, especially when modeling causation for evolutionary change. Conversely, sympatric species may be able to coexist simply because each has developed subtle specializations in dietary preference within the same, even very specialized, food source habitat. As Birt et ah (1997) pointed out for Australian megachiropteran bats, this degree of niche specialization may guarantee the extinction of species as habitats are isolated and lost as a result of environmental degradation coupled with the deraands of expanding human populations, especially when the animals are competitors with farmers for a fruit crop (e.g., a BBC World News Service 12/19/98 report about the cherry crop and a cull of fruit bats in northern Australia). Ironically, the success of H. sapiens, now a competitor for resources with many tetrapod species, is itself a product of an extreme specialization of the oropharyngeal complex coupled with expansion of the central nervous system (CNS) for a new function, speech. Nevertheless, when considering the history of mammalian taxa and assessing the environmental factors that may have produced gradual or stochastic change in craniofacial anatomy, interpretations of climatic change may be important. For example, evidence shows that preferred diets changed between wet and dry seasons in some fossil east African mammals. The sophisticated techniques now available for the deter-
mination of the botanical environment in fossil sites are helping in such evaluations. That said, traditional methods, such as comparative anatomy and, importantly, comparative physiology, cannot be relegated to the realm of "old hat." DNA analysis, however seductive a means of attempting to establish not only relationships among taxa, but the temporal pattern of their divergence, is still a very inexact science. We do not have a "genetic blueprint" for feeding mechanisms. We almost certainly never will. Given the complexity of "the system, reliance will have to continue to be placed on careful experimentation and observation (see Chapter 1 for further discussion). B. Approaches to the Study of Feeding in Mammals Differences in the shape of mammalian skulls, including jaws and their associated teeth, have intrigued naturalists since before written history (cf. Lascaux and comparable cave art). Given that such observations were rarely formally codified until the creation of the scientific journal and the scientific meeting (initiated by the Royal Society and the Academie Frangaise in the 17th century), it is difficult to establish the genesis of preceding observations that are now taken as "given." However (and importantly), major studies of mammalian comparative anatomy were published beginning in the late 1700s, continuing through the 1800s and to the present. We continue to rely on them, e.g., Magendie (1825) on swallowing, Dobson (1882) on the digastric muscle, and Edgeworth (1911, 1914) on cranial and hyoid muscles. Comparative studies of the tongue are more recent (e.g., Doran and Baggett, 1971; Doran, 1975). TurnbuU (1970) includes an invaluable bibliography of the source literature for jaw muscle anatomy and associated feeding mechanisms and Smith (1992) for the oropharyngeal soft tissues. If one wants to know how a system functions, then the obvious approach is to study it in action. This approach was used in the last century, but without the technology to create an analyzable permanent record of events. Mammalian cyclical feeding behaviors occur too fast for accurate visual recording, so no quantifiable measures of events could be obtained. The first calibrated records of jaw movements were made in either 1899 or 1908 with a human subject (Lord, 1913). Recently (post-1960), two parallel trains of investigation into feeding mechanisms in mammals have been ongoing (see Herring, 1993): (1) experimental and theoretical studies of selected species or groups and (2) studies seeking to explore morphological changes during evolution. Both have adopted, in the last two to three decades, very sophisticated approaches, but are
416
Karen M. Hiiemae
always dictated by technological developments applicable to the problem. It is proper to say that the "silicon revolution" has allowed major changes in the methods available for (1 and 2), but that comparative and evolutionary anatomists (2) are still constrained by the specimens available to them. Legislative efforts, such as the U.S. Endangered Species Act and the international conventions on trade in endangered species, as well as the vocal concerns of those objecting to experiments on animals, have restricted what is now possible in the laboratory. Offsetting these limitations, the extraordinary expansion of public interest in "natural history" has led organizations such as the National Geographic Society to sponsor major expeditions to a wide range of environments, which have yielded a treasure trove of film and tape showing feeding behavior in natural habitats for many mammals, which can greatly illuminate the evidence to be gleaned from bones, teeth, and soft tissues. Specific examples of the experimental methods changing functional analysis are (1) the development of the fine wire in-dwelling electrode (Basmajian and Stecko, 1957); (2) the use of strain gauges to measure forces on the lower jaw and skull (e.g., Hylander, 1977, 1984); (3) the development of the scanning electron microscope (SEM), allowing high magnification studies of tooth surfaces (e.g., Rensberger, 1978; von Koenigswald, 1982; Teaford and Runestad, 1992); (4) the advent of small powerful computers capable of manipulating complex data sets and able to support modeling software for techniques such as motion analysis and finite element analysis (e.g., De Jongh et ah, 1989; Hart et al, 1992; Spears and Macho, 1998); and (5) refinements in scintillation technology, coupled with video techniques (particularly S-VHS), which have improved not only the image intensifier (the fineresolution screen required for recording movement events using radiography) but which have also allowed a drastic reduction in radiation exposure making videofluorography (VFG) applicable to research on normal humans within the constraints dictated by federal regulations for the involvement of human subjects. It remains true that the use of X-ray techniques pioneered by Ardran and Kemp (1958) provide the primary data base for the interpretation of intraoral behaviors and mechanisms in mammals. Similar methodology has been used to establish the mechanisms of food transit through the GI tract. Mammal species studied to date using X-ray include rabbit (Ardran and Kemp, 1958; Weijs and Dantuma, 1981, Anapol, 1988; Cortopassi and Muhl, 1990), rats (Hiiemae and Ardran, 1968; Weijs and Dantuma, 1975), opossum (Hiiemae and Crompton, 1971; Hiiemae et al, 1978), cat (Hiiemae et al, 1981; Thexton et al, 1982; Thexton and Mc-
Garrick, 1988, 1989), pig (Herring and Scapino, 1973), hyrax (Janis, 1979; Franks et al, 1985; German and Franks, 1991), tenrec (Oron and Crompton, 1985), goat (de Vree and Gans, 1976), and bats (Kallen and Gans, 1972; de Guelde and de Vree, 1984). In the period between World Wars I and II, A. V. Hill, as well as Sherrington and his student Mountcastle, made seminal discoveries about the behavior of muscle and the central nervous system, respectively. Bremer (1923) was the first to demonstrate that rhythmic behaviors in feeding, such as chewing, were controlled by a "centre de correlation," now called a central pattern generator (CFG). The existence of such a CFG was confirmed in an elegant study by Dellow and Lund (1971). Since then, Jean (1984) and others have developed the concept of a "swallowing center." An as yet important unresolved question is the issue of how these two centers, both located in the pontinemedullary region of the hindbrain, are connected to produce the smooth integration of food processing and swallowing. Given the comparatively recent discovery (Hiiemae et al, 1978; Hiiemae and Falmer, 2000) that rhythmic tongue movements are linked to those of the jaw, but can occur independently of jaw movement as in suckling (Chapter 14) and at some stages in feeding (Hiiemae et al, 1996), substantial questions as to the neural control of feeding remain. Given the "process model" (see Section III), the stages in food processing, i.e.. Ingestion, Stage I Transport, Frocessing (Reduction), Stage II Transport, and Swallowing, have to be addressed sequentially. Our current knowledge base is uneven. For example, far more is known about the biomechanics of food processing than of ingestion or bolus formation and deglutition. Because the processes of ingestion, reduction, and swallowing largely involve the same musculoskeletal elements, a "generalized mammalian model" is proposed; significant variants associated with dietetic specialization (insectivory, frugivory, carnivory, herbivory) are addressed in each of the following sections.
III. THE "PROCESS MODEL" FOR M A M M A L I A N FEEDING The process model is shown in Figure 13.3. Originating in cinefluorographic studies of the rat (Hiiemae, 1967), but fully developed from data on opossum and cat (see Hiiemae and Crompton, 1985; Hiiemae et al, 1978), it has been tested in other mammals, including the macaques (Hiiemae et al, 1995) and humans (Hiiemae and Palmer, 1999), although H. sapiens displays one important difference (see later). Descriptions of
417
13. F e e d i n g in M a m m a l s
INGESTION Food moved into front of mouth
TRANSPORT ?
EJECTION
•NO-
YES
n
-i2
ii-
STAGE I TRANSPORT I to postcanine area ^^^^A^XM
PROCESSING Chew or tonguepalate compression
'by mastication'
TRANSPORT ?
¥
NO
YXAX/'X/yA/yyj
YES
1
STAGE II TRANSPORT through fauces for bolus formation
^
'
r^
I
THRESHOLD?
H. Sapiens Liquids and semi-solids only
•f
SWALLOW
Bolus formation continues
»<x/x;ovxxx^ NO
J vxxt^
Liquids, soft semi-solids
5?5^
Solids
F I G U R E 13.3. The process modeL First developed to explain feeding in the rat (Hiiemae, 1967), this model has been revised over the years as more data have become available. It now includes what is known of dual bolus formation and the swallowing mechanism in humans (see text). The distinction between mechanical elements of the model and postulated sensorimotor ''gates'' is indicated by shaded boxes.
the feeding process in other mammals indicate that the process model applies to a wide variety of species. The model is predicated on the fact that (a) most mammals ingest both liquids and solids, but (b) solids are typically not swallowable as ingested, i.e., are reduced by some form of processing. The model embodies two major hypotheses, one mechanical, the other sensorimotor. The mechanical hypothesis argues that ingested material is transported through the oral cavity to the
pharynx and then the GI tract, using two transport mechanisms (which may not be identical): stage I transport, in which food is moved from the incisal area to the postcanine region, followed by stage II transport, in which swallowable food is passed through the fauces for bolus formation and deglutition. The sensorimotor hypothesis postulates a series of sensorimotor "gates" that regulate the process. It argues that gate 1 provides for an evaluation of the nature of the ingested
418
Karen M. Hiiemae
material (i.e., "palatability")- Noxious or otherwise unacceptable material can be ejected ("spat out" in common parlance). Gate 2 postulates a sensorimotor mechanism for the determination of the "swallowability" of food in the postcanine area. If the material is not suitable for bolus formation, then the mechanisms for processing are triggered. Gate 3 postulates a "threshold" below which a swallow does not occur. This is the most problematical feature of the model. While based on data for Didelphis and cat, there is no corroborating evidence available for other nonhuman mammals. The substantial literature on swallowing behavior in humans does not appropriately address the problem as the paradigm on which the human experiments were designed is based on liquid swallows rather than swallows of processed natural bites (for an exhaustive review of swallowing in mammals, see Thexton and Crompton, 1998). The model accommodates the natural variations in initial food consistency and feeding behavior among mammals. Clearly, liquids are swallowable "as is": giraffes at a watering hole are lapping up water and swallowing in a definite rhythm. Water and maternal milk (see Chapter 14) clearly meet the "swallowable standard" (gate 2). In fact, the processes of stage I and stage II transport become a continuum. In the more artificial laboratory environment, semisolid foods such as tinned cat food mixed to varying consistencies may be treated in much the same way (Thexton and McGarrick, 1988). In these circumstances, the ratio between "lap" (ingestion of an aliquot of fluid) and swallow may change (Hiiemae et al, 1978; see also Thexton and Crompton, 1998). However, at some consistency threshold, this transport mechanism is interrupted because the ingested material cannot be formed into a swallowable bolus. With the exception of those mammals that swallow their prey whole (e.g., insects, eggs, or whole fish by dolphins), this triggers processing, defined as the reduction of the food material to a consistency "acceptable" for swallowing. The model makes no attempt to define "acceptable consistency." Clearly, this must differ among taxa and types of food. For example, it is alleged that felid carnivores can swallow fairly large lumps of fresh prey meat, whereas it is well known that herbivorous grazers spend considerable time chewing the foliage they have ingested. However, at some point a proportion, if not all, of the ingested material can be swallowed. In rendering it swallowable, the effect of the mixing of the food particles with saliva cannot be ignored. Saliva has a "wetting function." Indeed Prinz and Lucas (1997) argue that for a bolus (in humans) to be swallowed, certain rheological criteria must be met. Those criteria involve a combination of food particle sizes and salivary wetting with
their combined surface tension and packing effects. Their model is compelling. In this context, variations in the pattern can be noted. De Gueldre and De Vree (1984) reported that the frugivorous, megachiropteran bat, Pteropus, swallows only the juice extracted from the ingested fruit, spitting out pellets of cellular residue. Many fruits have hard shells or peel. Incisors are used to break open the fruit, accessing the pulp. During this process, pieces of peel or shell may enter the anterior mouth only to be ejected (primates, rodents). When fed 1-cm^ lumps of hard liver, cats cut them into two lumps, ejected one and then processed the remainder (Thexton et al., 1980), suggesting that some acceptable food may be rejected because it is initially too large for intraoral management. Once food in the mouth is "acceptable for swallowing," stage II transport (i.e., passage through the fauces to the oropharynx) occurs. Bolus formation ensues and a swallow is initiated. Although experimental evidence shows a rhythm for ingestion to deglutition for liquids and soft semisolids (see Thexton and Crompton, 1998), the same data (Thexton et al., 1980) also show that there is essentially a one-to-one relationship between ingested "bite" and bolus formation. If true, and this is by no means clear for the broad range of mammals (the question was not asked in most studies), a given volume of solid food is ingested and processed until swallowed. While true for solid foods in opossum, cat, and possibly hyrax, it is not the case for the macaque and modern man. We have almost no data on other mammals. Does a cow swallow some portion of ingested grass and retain the inadequately triturated remainder in the oral cavity as do pigs or does it take a mouthful of silage and chew it all until swallowable? Why should this be an issue? The concern directly relates to our understanding of the neural control of the process. This must depend on sensory input to the central (hindbrain) effector motor systems. The nature of this input is currently disputed. The functional problem hinges on the issue of "segregation"—how can adequately triturated material be segregated from that which is inadequately reduced? Work in progress (human studies, see Hiiemae and Palmer, 1999, 2001) suggests a surprising answer, but one that may not be applicable to other mammals, as a single feeding sequence in humans may involve multiple swallows. The model posits that stage II transport moves food to the site of bolus formation. This aspect of the model is pivotal. Bolus formation does not occur in the oral cavity in most mammals studied. It occurs in the oropharynx around a larynx whose aditus (entrance or approach) is in the nasopharynx (see Thexton and Crompton, 1998). Triturated food is moved through
419
13. Feeding in Mammals the fauces (Fig. 13.13) and accumulates in the valleculae and piriform fossae. Swallowing (deglutition) occurs as a powerful ejection of the material from this perilaryngeal space into the esophagus. Unlike other mammals, the human larynx is positioned well below the soft palate in adults. The oropharynx is, therefore, much longer, and the larynx (and respiratory tract) is not shielded from the aspiration of food, except by the sphincter afforded by the epiglottis and vocal folds. Nevertheless, it has been determined (Palmer ei al., 1997; Hiiemae and Palmer, 1999) that bolus formation in humans, when fed solid foods, typically occurs in the oropharynx. This is the case regardless of whether subjects are in the upright (normal human position) or on "all fours" mimicking the feeding position of most mammals (Palmer, 1998). Beyond its implications for the management of patients with swallowing disorders, this finding is important because it suggests that H. sapiens retains a basic mammalian bolus formation mechanism, at least when solid foods in normal bite sizes are ingested. However, it is indisputable that liquid boli are formed in the oral cavity and swallowed by expulsion through the fauces. We therefore postulate that liquid bolus formation in the mouth is a uniquely human specialization (Palmer ei al., 1997; Fiiiemae and Palmer, 1999).
IV. MECHANICAL PROPERTIES A N D TEXTURES OF F O O D S The classification of mammals depends in large part, although not exclusively (e.g., differences be-
tween Perissodactyla and Artiodactyla) on the anatomy of the jaws and teeth. This practice follows from the fact that tooth and jaw form in mammals are highly sensitive indicators of diet, in a general sense (e.g., herbivory, carnivory, insectivory), and that diets are relatively stable at higher taxonomic levels (e.g., families and orders in some cases; see earlier discussion). Although food scientists have developed a large body of information on the rheology, mechanical properties, and texture perception of foods, this work is focused on natural and engineered components of the human diet. Lucas, with his colleagues and students, has contributed the most significant body of work on the properties of normal mammalian dietary items, with a particular emphasis on those consumed by cercopithecid primates [see Lucas and Corlett (1991) for an overview; Sibbing (1991), in the same volume, describes food processing by the pharyngeal jaws of cyprinid fish, which show analogous adaptations in tooth form related to diet]. Nonetheless, it remains true that, to date, we have a very poor understanding of the relationship between the mechanical properties of foods and tooth form. Figure 13.2 presents a broad-brush overview of food sources available to mammals. Table 13.2 summarizes the mechanical properties of common mammalian food items, coupled with the basic mechanical requirements for their reduction (processing). For food to be "reduced," its structure has to be disrupted. The most "effective" (energy conserving) method depends on the intrinsic mechanical structure of the particular food item. While the terms used in Table 13.2 have specific meanings for mechanical engineers and food
TABLE 13.2 The Mechanical Properties of Typical Mammalian Food Items/ the "Engineering" Required for Reduction (Appropriate Technique), and Its Morphological Translation into Tooth Form^ Appropriate ToughNotch sensitivity reduction ness
Tooth morphology
Found in
Type
Examples
Hard brittle
Seeds, nuts, unripe fruit, some tubers, roots, some adult insects
Stiff elastic
High
Low
High
Crushing, splitting to^ grinding
Mortar-pestle
Primitive mammals, insectivores, primates, omnivorous herbivores
Turgid
Ripe juicy fruits, some insect larvae
Plastic flow
Low
Variable
Variable
Crushing to^ grinding
Mortar-pestle
Primitive mammals, insectivores, primates, omnivorous herbivores
Soft tough
Animal soft tissues, some insects, young leaves (?), young grasses (?)
Viscoelastic/ pliant
Moderate
High
Low
Cutting, shearing, piercing
Blades
Carnivores (especially carnassials), premolars in insectivores, primitive mammals
Tough fibrous
Grass, fruit skin
Viscoelastic/ pliant
Variable, High depends on fiber direction
Low
Serial arrays of Cutting, lacerating, low-profile shearing blades
Deformability
"See text for explanation. ^After Lucas and Luke (1984) and Sibbing (1991).
strength
All facultative herbivores to varying degrees; extreme development in some groups
420
K a r e n M. H i i e m a e
scientists (see Jeronimidis, 1991; Purslow, 1991; Vincent, 1991; Rensberger, 1995), they can be briefly defined as follows: hardness (probably stiffness) is a measure of deformability (elastic or plastic) and is expressed as stress per unit of strain (i.e.. Young's modulus, E). These materials fail under load after a period in which they are loaded beyond recovery. Depending on the material, the required load may be relatively small or very large. A large load can be delivered most effectively by a sn\all, sharp tooth cusp moved by powerful adductor contraction, penetrating the material held between upper and lower teeth. (I once recorded a force of 17 kg in Didelphis produced at a premolar cusp tip over an area of about 1 mm^.) Brittle materials are readily cracked. For example, in eating a nut, compressive force is applied and the nut shatters. Hard and brittle materials break when a crack is created in the material by the application of external force and the crack is propagated throughout the material, resulting in breakage. A different problem is created by materials that are ductile, i.e., can flow or be deformed beyond their elastic limit and so resist crack propagation (viscoelastic materials). This makes such foods tough, i.e., more work is needed to generate fracture. Notch sensitivity refers to the likelihood of penetration leading to fracture. As Purslow (1991) shows, the fracture properties of meat are highly dependent on the orientation of the muscle fibers when subjected to occlusal forces. The external "working surfaces" of mammalian teeth, with a few exceptions (e.g., sloths, manatees), are covered (at least initially) with enamel, the hardest known biological tissue. Mammalian enamel is prismatic. As such, the enamel is organized microscopically into columns, which means that the orientation of the prisms relative to the direction of loading affects the strength of the enamel, and thus its resistance to fracture (see Rensberger, 1978,1995; von Keonigswald, 1982). Prisms are oriented perpendicular to the typical direction of loading. While enamel is highly resistant to wear, it is abraded in normal use. Thickness of crown enamel varies within groups and has been used as an indicator of likely wear stress in primates (Kay, 1975; Kay and Hylander, 1978; Spears and Macho, 1998) and ungulates (Janis and Fortelius, 1988). In some groups, the teeth reach full functionality once dentine is exposed (Fig. 13.4). It is alleged that mortality in wild mammals, such as shrews, is closely correlated with dental wear—once the teeth have lost their shearing surfaces, the animals can no longer process sufficient material to keep the "metabolic equation" in balance. It is also said that tigers in India become "man-eaters" when their teeth are so degraded that they are unable to bring down normal prey. Inspection of Anasazi skulls in southwestern U.S. mu-
1^ •m-wA
MORTAR-PESTLE
SHEARING BLADES
LTLJlA/
rmful
SERIAL ARRAYS -LOW PROFILE BLADES F I G U R E 13.4. Mechanical principles of tooth design in relation to food type. (Top) Pestle and mortar, cusp (pestle) acting against a mortar (fossa in upper tooth), and molars of a cercopithecoid frugivore. (Middle) Cutting and shearing blades, section through dental blades, and carnassial teeth found in felid carnivores. (Bottom) Grinding blades, cross section through lophodont cheek teeth (note alternating enamel and dentine), and molars of a selenodont artiodactyl. Enamel is shaded in all cross-sectional drawings. Note the direction of movement (heavy arrows) in each mechanical model and the degree to which this is mimicked in the dentitions shown.
seums provides a dramatic demonstration of the effect of a tough diet (Table 13.2) on human postcanines— and the Anasazi had fire! The bulk (crown and root) of the mammalian tooth is formed by dentine (see Chapter 2). In contrast to enamel, which is acellular, this tissue is supported by odontoblasts (analogous to osteocytes), which have their cell bodies in the pulp cavity, but long processes extending through the dentine. Dentine can be formed throughout life. It is softer than enamel and is therefore abraded more rapidly. This difference in wear resistance has been exploited by some groups of mammals, particularly herbivores and rodents. Both consume highly abrasive (siliceous) foods. The difference
13. Feeding in Mammals in hardness between enamel and dentine is used to enhance the efficiency of food reduction, as the enamel "edges" act as blades with the more rapidly abraded dentine as catchment areas for the product of toothtooth interaction (Fig. 13.4, and see Rensberger, 1995). Herbivorous mammals have evolved three approaches (often combined) to the challenge of tough, fibrous food sources. They can be characterized as (1) the fusion of discrete cusps into complex patterns of ridges or lophs (see Janis, 1995); (2) designs to preserve tooth function for as long as possible; and (3) designs to optimize the effectiveness of the teeth in accessing or processing available food. The herbivore solutions to the problem of maintaining tooth function have been (a) a greatly prolonged period of tooth development, i.e., hypsodonty ("high-crowned" teeth), which continue to form for years after initial eruption (e.g., Perissodactyla), but do have a finite life or (b) persistent growth (e.g., rodent incisors). Rodent incisors have an active tooth germ in the base of their alveolus that generates new enamel and dentine as the erupted tooth is worn away by normal gnawing activity. Some rodents (e.g., hystricomorphs) also have persistently growing cheek teeth and (c) "molarization" and fusion of teeth. If the "metabolic equation" requires the ingestion of very large volumes of vegetation (since its nutrient value is low), either (i) remodel the premolars into molariform teeth so as to increase the postcanine dental working surface for food reduction or (ii) make each tooth (premolars and molars) the equivalent of a complete cheek tooth row and erupt them seriatim over a very long lifetime. Fiorses, most ungulates, and some rodents have adopted the former approach; elephants, uniquely, the latter.
V. THE FEEDING APPARATUS The feeding apparatus can be divided, for convenience only, into two functional complexes: (1) the jaw complex, including jaws, teeth, and the muscles of mastication. The jaw complex is used primarily in food acquisition and processing. (2) The oropharyngeal complex, comprising mostly soft tissues. The oropharyngeal complex manages all intraoral food transport, bolus formation, and deglutition (and in some taxa, ingestion, as well). Clearly, more is known about the evolution of the former, given the nature of the fossil record. A. Jaw Complex 1, The Masticatory
Cycle
Using changes in the rate of jaw movement in the opossum as the criterion, the masticatory cycle was
421
divided into a number of stages (Hiiemae, 1976, 1978; see Chapter 2). Jaw closing was described as having fast close (FC) and slow close (SC), or power stroke (PS), phases; the latter being distinguished by a reduction in the rate of closure when teeth met the resistance of food. With more recent work, the splitting of jaw opening into slow open (SO) and fast open (FO) phases has proved a problem, as pointed out by Schwartz et al. (1989). In their study of feeding in the rabbit, they identified three phases in opening: O l , 0 2 , and 0 3 . 0 3 is not always present, but when it is, it is always rapid, ending at maximum gape (see Fig. 13.5). This four/ five-phase pattern in the masticatory cycle has been observed in most nonhuman mammals studied, including those that feed in unusual postures such as bats (de Greet and de Vree, 1984) and sloths (Naples, 1985). Originally, an intercuspal phase (IP), in which no discernible vertical movement occurs, was included in the power stroke. We are now treating this, at least for higher primates, as a distinct element in the feeding cycle (Fig. 13.5), not least because it occupies a significant proportion of total cycle time (Hiiemae et al, 1995, 1996; Palmer et al, 1997; Hiiemae and Palmer, 2001). Events in IP involve the transition from "jawbased" to "tongue-based" behaviors (see Section V,B). 2. Jaws and Teeth The shape and proportions of the skeletal elements of the feeding apparatus in mammals vary widely. The early evolution of the mammals dictated the fundamentals of their craniofacial anatomy. There is a large literature on mammalian cranial osteology (see Turnbull, 1970) and dental anatomy [see Peyer's (1968) classic text, which also includes a useful review of the early papers on the origin of the tribosphenic molar]. Other studies (e.g., Radinsky, 1981a, 1981b, and 1982 for carnivores) have examined the evolution of skull shape within major groups. The primitive mammalian dentition is highly differentiated, i.e., heterodont: the basic mammalian formula is three incisors, one canine, four premolars, and three molars in each jaw quadrant with each upper tooth occluding with its opposite lower jaw equivalent, but somewhat posterior to it (one-half tooth length behind in the case of the molars). There is general agreement that the molars of mammals were derived from an ancestral tribosphenic molar by selection for the features shown in Fig. 13.4 (Butler, 1952; Mills, 1955; 1966; Crompton, 1971; Hiiemae and Kay 1972; Kay and Hiiemae, 1974; Janis, 1979; Janis and Fortelius, 1988). The overall shape of the skull is affected by a wide variety of factors, including size and position of the orbits, the complexity of the nasal airway, the proportions of the muscles, and the size of the brain (e.g., van
422
K a r e n M. H i i e m a e
A. DIdelphls TIME
Tongue Reversal
Puncture Crushing: 315 ms.
Chewing: 400 ms
B. Macaca fascicularis GAPE
FIGURE 13.5. Gape-time plots for Didelphis and Macaca fascicularis. (A) Opossum. Mean values for the duration: total and included phases for the two cycle types identified in Hiiemae (1976). The first cycle shown is a puncture-crushing cycle. This occurs early in a feeding sequence and is used to crush/fracture the ingested food (N = 18). The teeth do not achieve occlusion but there is a clearly defined slow close phase in closing. Slow open (SO) and fast open (FO) phases are present in opening. In the longer chewing cycle (N = 24), the teeth reach occlusion. Arrows show the primary direction of jaw movement: vertical in puncture-crushing and lateral to medial in chewing. Tongue protraction occurs in the first part of opening, reversing at the transition between SO and FO. (B) Macaque. Analysis of jaw, tongue, and hyoid movement has shown that there are no "'stereotypicar' feeding cycles as measured by jaw movement. Cycle profiles are highly correlated with the type of food (initial consistency) and stage in the feeding sequence (Hiiemae et ah, 1995; Thexton and Hiiemae, 1997). Movements of a radio-opaque marker on the anterior tongue, and of the hyoid, measured from the anterosuperior aspect of the hyoid show that there are patterned movements of both tongue and hyoid during masticatory cycles. The one predictable relationship between jaw and tongue/hyoid movement is the reversal from tongue protraction to tongue retraction at the SO-FO transition. In macaque (Hiiemae, 1995), this linkage occurs within a 30-msec "window.'' Hyoid movement does not exactly parallel that of the tongue surface (see Section V,B).
der Klaauw, 1945; Schwenk, 2000b). Nonetheless, long tooth rows require long snouts and jaws. The relative proportions of braincase and snout (defined here as the combined anteroposterior length of the premaxilla and maxilla, inclusive of the tooth row, to the anterior margin of the orbit) are illustrated in Fig. 13.6 (Didelphis) where the ratio of braincase to snout length is approximately 1:2. Using slightly different criteria, Fearnhead et al. (1955) measured the same ratio using the glenoid (squamosal-dentary joint) as the reference point and showed that for insectivorous mammals (presumed to reflect the ancestral condition), the ratio averaged 1:37 for erinacids, 1:5 for tenrec, but only 1:1.4 overall for soricids, which have relatively expanded braincases.
As different mammal lineages evolved, the ratios changed. For example, Fearnhead et al. (1955) give the ratio as 1:1 in mustelids (Carnivora). Such changes reflect the characteristics of each order and the dietetic specializations within it (Fig. 13.6). It is important to note that the lateral surface of the braincase provides attachment for the temporalis muscle. Where the underlying brain is comparatively small, as in Didelphis, the presence of a large temporalis may lead to the development of sagittal crests. A combination of powerful temporalis and nuchal (neck) musculature can produce a T-shaped pattern of crests. This development occurs in primitive mammals, in many carnivores, and some primates. Similarly, the zygomatic arch may be
13. F e e d i n g in M a m m a l s
423
FIGURE 13.6. Simplified lateral views of the skulls of (A) Didelphis, (B) Cehus, (C) cat, (D) sheep, and (E) rat to illustrate the different proportions of the bony elements in examples of each of Turnbuirs ''dietetic groups/' All skulls are drawn to the same overall length to allow direct comparisons of the proportions of the cranium, the lower jaw, and the length of the tooth row. Areas of attachment for the major muscles groupings (see Fig. 13.7) are shaded (light stipple, temporalis complex; darker stipple, masseter complex). (A) Based on Hiiemae and Jenkins (1969), (B) on Le Gros Clark (1959), and (C-E) on TurnbuU (1970). Arrows are intended as a simplified indication of the overall line of action of each complex and are "weighted" in accordance with the proportions of the muscles shown in Fig. 13.7.
attenuated or missing in some insectivores, but very robust in herbivorous mammals with a powerfully developed masseter. Each half of the lower jaw (hemimandible) is composed of a single bone, the dentary. Embryologically, the dentary has three major elements (e.g., Atchley, 1993): (1) a basal component, the dentary proper, which encases the inferior dental nerve and extends posteriorly to form the mandibular condyle; (2) an alveolar component, i.e., the bone associated with tooth
attachment (above the nerve); and (3) posteriorly, the coronoid and angular processes of the mandibular ramus, which develop around the core dentary axis. The size and shape of these processes are dictated by the relative development and organization of the adductor muscles: the coronoid process serves primarily as a site of attachment for the temporalis muscle whereas the masseter and internal (medial) pterygoid muscles insert primarily on the angular process. It is important to note that experimentally impaired development or
424
Karen M. Hiiemae
ablation of developing temporalis or masseter/internal pterygoid muscles results in attenuated process development (Avis, 1961). The mandible develops as an intramembranous (dermal) ossification investing Meckel's cartilage. The latter is later resorbed anterior to the middle ear. Embryonically, the two halves are separated anteriorly by a fibrocartilaginous joint {mandibular symphysis). This joint may persist throughout life, allowing some independent movement of the two hemimandibles, or it may ossify. In the former case, the symphyseal joint may have a complex arrangement of the constituent fibers to regulate the range and direction of intramandibular movement [e.g., Didelphis, tenrec, some rodents, and herbivorous marsupials and placentals (Beecher, 1979)]. A moveable symphysis is often associated with the presence of a transversus mandibulae muscle. It follows that (a) the length of the upper and lower jaws is highly correlated with the length of the tooth row; (b) the vertical height of the body of the mandible is highly correlated with the length of the tooth roots; (c) the anatomy of the coronoid and angular processes reflects the relative development of the muscles attached to them (i.e., a large angular process is associated with a well-developed masseter and internal pterygoid and, conversely, a small coronoid process with a weakly developed temporalis); and (d) the morphology of the lateral margins of the premaxilla and maxilla is also affected by the alveolar bone associated with the anchorage of the upper dentition, although this bone may be "subsumed" into the lateral walls of the nasal cavity, thus masking the length of the tooth roots and creating a relatively flat hard palate from which only the crowns of the upper teeth appear to project. The mandible articulates with the skull through the squamosal-dentary joint (otherwise, the jaw-joint, cranio-mandibular joint, CMJ, temporo-mandibular joint, or TMJ—the latter term most often applied to humans). The squamosal, or the bone forming the cranial articular surface for the CMJ/TMJ, is fused with other ossifications during development of the mammalian skull to form a composite temporal bone, hence the mandible articulates with the "squamous portion" of the temporal bone. Movement of the lower jaw relative to the upper is produced by integrated actions of the jaw muscles acting to move the dentary condyle on the squamosal fossa. For a mammal to ingest and process food, it must be possible for the jaws to open as wide as needed to acquire or ingest the food item and then to close, bringing the entire tooth row into full occlusion, while at the same time facilitating the movements of the teeth into and out of occlusion (arrows in Fig. 13.4). In primitive mammals, this is achieved by a CMJ positioned just above the level of the occluded tooth row. The joint position is higher in other mam-
malian groups (see Fig. 13.6). The elevated position of the jaw joint relative to the bite point permits upper and lower teeth to engage more or less simultaneously along the entire tooth row during occlusion (as opposed to a pure "scissors" action in which the teeth would engage sequentially from posterior to anterior; see Greaves, 1995). The evolution of the squamosal-dentary joint (see Crompton, 1995) created the potential for mandibular movement in the mediolateral (coronal, M-L,) and anteroposterior (sagittal, A-P) planes in addition to the vertical movement common to all vertebrates. (In much of the older literature the following terms are used: orthal for vertical, ectental for mediolateral, and propalinal for anteroposterior movement.) Didelphis and its placental equivalents with basically tribosphenic molars utilize movement in all three axes coupled with hemimandibular rotation. To optimize the shearing capacity of their cheek teeth, the lower jaw on the working side (the side on which food is positioned; see later) is moved upward, medially, and forward. In addition, each hemimandible can rotate about its longitudinal (A-P) axis. Morphology of the putatively ancestral mammalian CMJ, which facilitated movement in all three planes, was modified in the various ordinal lineages to optimize movement in one or another of the primary directions. For example, in carnivorans, especially some felids and mustelids, movements are essentially restricted to the vertical and M-L movement is variably constrained, sometimes extremely so; in most rodents the capacity for A-P movement is greatly increased, whereas chewing in most other herbivores uses a primarily transverse (ML) jaw stroke, during which the condyles must pivot about a vertical axis on the glenoid fossa (Table 13.3, Fig. 13.8). Although there are exceptions, most mammals chew food on only one side of the mouth at a given time. This side is designated the working side, the other side, the balancing side. Experimental and theoretical (modeling) work on the biomechanics of the jaws has shown both how forces are transmitted between the two sides and how patterns of muscle activity not only produce the compressive or shearing forces needed to disrupt the integrity of food items, but also maintain the integrity of the CMJ (Greaves, 1978, 1995; Hylander, 1979). 3. Muscles of
Mastication
As shown in Fig. 13.6, the areas and shapes of the major adductor muscle attachment sites differ among the dietetic groups. The muscles of mastication can be divided into jaw closers {adductors) and jaw openers (or depressors) {abductors). The digastric muscle is generally considered the principal abductor, although this
TABLE 13.3 Primary Anatomical and Dental Features of the Craniofacial Complex in Terrestrial Mammalsa Skull Primitive mammals, e.g. Didelphis marsupialis
Lower iaw
Squamo-dentarv ioint
Dentition
Long snout, complete zygomatic arch, sagittal and nuchal crests
Well-developed coronoid Inflected angular process Shallow body Mobile symphysis
Transverse, postglenoid process Condyle convex A-P Joint just above occlusal plane
Is 4, C, PMs 4, Ms 3 Projecting canine Tribosphenic molars
Placental insectivores
Braincase expanded (Soricids) Some: incomplete zygomatic arch (Soricids, tenrec)
Big coronoid, hooked angular process Shallow body Mobile symphysis
As for Didelphis in most Soricids: double joint
Incisors may be reduced or modified depending on habit Tribosphenic or bunodont molars
Bats
Generally as for insectivores
Primates Prosimians
Anthropoids
Carnivores Ancestral, and canids, mustelids, many viverrids
Highly modified anterior dentition in many forms Molars tribosphenic with heavy emphasis on connecting ridges ('W' shape) Bunodont in fruit bats
Primitive (e.g., Tupaia)-like insectivores Progressive expansion of braincase, especially orbits Relative reduction in snout length
Primitive, as for insectivores Coronoid and angular processes well developed Shallow body Mobile symphysis
Glenoid fairly flat, condyle rounded both axes
Tends to reduce number of Is, PMs Molars tribosphenic or with shearing planes or bunodont derivative (four cusps)
Progressive migration of foramen magnum toward skull base Highly variable snout length, even within related groups Relatively flat palates
Well-developed processes Angles grossly enlarged in some cebids Simian shelf with recessed attachments for genial muscles Fused symphysis
Transversely oriented condyle Convex A-P Postglenoid process Well above occlusal plane
Spatulate incisors, diastema upper lateral I, and canine for occlusion lower C Canine sexual dimorphism Two PMs, lower first bladelike Molars generally four-cusped (uppers), sometimes five Molars tend to increase in size distally
Expanded cranial volume Expanded orbits Variable snout length
Large coronoid, smaller projecting angular process Mobile symphysis (but fused larger forms, e.a., " ursids)
Pre and postglenoid processes, glenoid sharply concave Condyle transverse and sharply convex Joint at level occlusal plane
Reduced in total number Powerful canines Last upper premolar, first lower molar modified to form carnassial shearing blades
Omnivorous, herbivorous ursids, procyonids, many viverrids Felids (pure carnivores) Ungulates Hyracoidea, Perissodactyla, Artiodactyla (Lagomorphsfi)
Rodents Sciuromorpha, Myomorpha, Hystricomorpha
Molars may be bunodont for puncture-crushing/ grinding Expanded cranium Short snout, heavy zygomatic arch palate
Short, robust Large coronoid, small angular process Deep masseteric fossa
As described above
Reduced molar series Camassials dominate, are very large
Skull dominated by large snout, making cranium appear small Reduced temporal fossa Short stout zygomatic arch
Long, body fairly deep Slender, often recurved coronoid process Masseteric area on mandibular ramus (including angle) expanded greatly Mobile symphysis in most (fused adult horses, swine)
Glenoid relatively flat Retroarticular process big artiodactyls Condyle small, flattened Condyle rotates horizontally (lateral movement jaws) Joint close to occlusal plane in primitve forms, rises significantly in advanced forms
Upper incisors replaced by keratinous pad (artiodactyls) Some large canines (pigs, peccaries, hippos) Canine dimvorohism
Small coronoid, projecting welldeveloped angular process Lower border curved anteriorly (lower incisor alveolus) Mobile symphysis
Glenoid oriented anteroposteriorly Concave in M-L section Condyle elongated A-P, convex M-L May have two articular facets (one for anterior occlusion = incisors; second for posterior occlusion (molars)
Reduced greatly One I each quadrant, persistent growth No canines Reduced premolars (usually 1) Lophodont molars, can be very complex (hystricomorphs) and of persistent growth Molar rows parallel or diverge posteriorly Isognathy Two separate occlusal planes
Rounded braincase (most) Large orbits Snout dominated by upper incisor alveolus Zygomatic arch expanded anteriorly for part masseter (sciuromorphs, myomorphs) Infra-orbital foramen for part massetel (myomorphs, hystricomorphs)
(b) hypsodonty Molarisation of premolars Absolute increase in molar size
aBased on Turnbull (1970), Weijs (1994), and basic texts such as James (1960) and Peyer (1968). For comparison with Fig. 13.7. This table is not intended to illustrate phylogeny or primitive and derived characteristics. b ~ l t h o u g lagomorphs h have rodent-like incisors, Weijs (1994) considers rabbits and hares functional ungulates.
426
Karen M. Hiiemae
may not be the case when the jaw is depressed in normal feeding (see later). The adductors can be broadly identified as temporalis, masseter, internal, and external pterygoid muscles. However, I concur with Herring (1993) when she states "unfortunately, jaw-closing muscles are not only heterogeneous internally but are also linked to each other externally, so their homologies, while undoubted, are imprecise." In fact, nomenclature for the various "parts" of the major muscle complexes is overly complex; same-named parts may have very different actions and mechanical effects (see Turnbull, 1970; Weijs, 1994, and literature therein). Herring continues by arguing that "for functional purposes, it might be better to designate muscles as vectors rather than as named parts." This approach was developed by some of the earlier comparative anatomists (Arendson, 1951; Becht, 1953, Maynard Smith and Savage, 1959), used by Hiiemae and Jenkins (1969), and was the foundation of later biomechanical modeling of muscle action and forces on the jaws (e.g., Scapino, 1972; Greaves, 1978, 1988; Smith, 1978; Bramble, 1978; see also review by Hylander et al, 1992). That said, mammalian jaw adductors have probably received more attention than perhaps any other muscles in the body. Their relative mass and fiber orientation (Turnbull, 1970) and internal architecture, as well as their motor unit organization and firing patterns, have all been examined (see reviews by Herring, 1992,1993,1994). In an exhaustive review, Weijs (1994) synthesized all the information then available to provide an invaluable discussion of the behavior of the muscles and the mandibular movements they produce based on Turnbull's classification. The following is, therefore, a brief of summary based on these reviews. Differences in the proportions of the major muscles shown in Fig. 13.7 reflect the linkage among jaw movement, tooth form, and diet. The dominance of the temporalis in Didelphis and Echinosorex is associated with powerful jaw-closing movements for crushing and piercing of prey followed by crushing and cutting. A similar pattern in the cat reflects the dominance of the canines and carnassials in felids, which use their adductors during both prey capture and ingestion. In carnivorans, generally, the line of action of the temporalis is directed so posteriorly that a large, recurved postglenoid flange has developed in the upper jaw joint to prevent posterior dislocation of the condyle. In felids and some mustelids, the development of an additional preglenoid flange has produced a deeply concave upper articular surface into which a nearly cylindrical mandibular condyle fits snugly, restricting jaw movement to the vertical with only the slightest "play" in the transverse plane permitted to adjust carnassial shearing (this jaw joint is often so tight that even when a prepared skull is picked up the mandible remains "attached"). A
sizable masseter adds additional power but with a slightly anteriorly directed line of action. Although the rodents and the large herbivores shown in Fig. 13.7 appear to have very similar percentages of niasseteric muscle mass, this masks a real difference: in the former, the superficial masseter is very large, whereas in the latter, the deep masseter and zygomaticomandibularis dominate. To engage their incisors, rodents protract the lower jaw. The power stroke during mastication also has a large anterior component and molar grinding of food is mostly A-P (Hiiemae and Ardran, 1968; Weijs, 1974; Weijs and Dantuma, 1975). These movements are powered by a large and nearly horizontal superficial masseter and a complex, anteriorly placed "deep" masseter, with some assistance from the internal pterygoid. In an earlier review (Hiiemae, 1978), I argued that the "pattern of EMG activity was broadly similar in all the mammals studied so far (to 1976) despite the differences in their profile of jaw movement," implying some generic (mammalian) pattern of central motor control. Given the differing muscle architectures in ordinal groups of mammals, I did not suggest that this "generic" pattern of activity would produce the same amplitude or direction of movement. This proposition was interpreted, not without justification, as embodying an hypothesis, i.e., the existence of a "mammalian effector Bauplan" for the activity pattern of the jaw adductor muscles emanating from a masticatory CPG central pattern generator in the hindbrain. This idea was followed by a paper by Bramble and Wake (1985) and a response from Smith (1994). Both papers are important, given the fact that mammals have evolved from vertebrates relying on a hyobranchial feeding apparatus. Weijs (1994) has systematically reviewed all available data, including his and other's work post-1976, and has developed an important synthesis, which demonstrates adjustments in the "basic mammalian pattern" associated with dietetic adaptations. Weijs (1994) examined electromyographic (EMG) data for the numerous species now studied, arguing that it warrants a "functional activity" approach to the interpretation of jaw adductor motor patterns during the closing (power) stroke of the jaw. His conclusions are based on motor activity patterns in several generalist mammal species presumed to represent the primitive condition. He proposes that there are muscle groupings that function as "triplets," in effect a " p u l l pull" system, to generate the laterally directed mandibular movement, which positions the teeth for the medially directed power stroke on the active side. The jaws are closed from initial maximum gape by the symmetrical action of vertically oriented muscle fibers (e.g., zygomaticomandibularis and allied fibers such as those in the most anterior and deep parts of tempo-
427
13. F e e d i n g in M a m m a l s PRIMITIVE MAMMAL
Didelphis
INSECTIVORA
Echlnosorex
CARNIVORA
Felis
PERISSODACTYLA Equus i
ARTIODACTVLA
Odocolleus
RODENTIA
Sculrus
10
20
30
40
50
60
70
80
90
100
Percentage Total Jaw Muscle Mass (Adductors and Digastric) Masseter Complex
Temporalis Complex
Internal Pterygoid
I '
I External ' Pterygoid
Digastric
FIGURE 13.7. Relative proportions of the adductor musculature, with digastric, expressed as a percentage of the total muscle mass (data from Turnbull, 1970). Relative proportions of the masseteric complex (superficial, deep, and zygomatico- mandibularis) in herbivores and rodents as compared with the temporalis (anterior, posterior, deep) in carnivores and insectivores is clearly shown. It should be noted that the proportions of the muscle complexes identified vary within the dietetic groups such that, for example, the actual percentage of temporalis/masseter muscle mass may differ substantially but the dominance of one or other complexes remains. Interestingly, in marine carnivores the proportion of digastric may rise to as much as 25% of the total.
ralis). As closing proceeds, the jaw is moved laterally by the combined activity of the working side, posteriorly directed fibers of the temporalis coupled with anteriorly directed fibers of the masseter and medial pterygoid on the balancing side. This grouping is Weijs' trip-
let 1. As the teeth approach tooth-food-tooth contact and occlusion, a second reciprocal grouping of the balancing side, posteriorly directed fibers and the active side, anteriorly directed fibers (triplet II) come into play carrying the teeth through to centric occlusion.
428
K a r e n M. H i i e m a e
The asynchrony in the onset of activity in triplet II varies within the generalized mammal group depending on the degree of transverse movement during the power stroke. This is small in opossum but large in tenrec and the little brown bat. Contraction of muscle fibers involved in both triplets contributes both to movement direction (until restricted by occlusal contacts) and to masticatory force.
Four variants of this basic pattern are identified (Weijs, 1994). In mammals where the transverse (L-M) movement is severely constrained either by carnassials (felid and mustelid carnivorans) or by interlocking canines (e.g., fruit-eating bats and most carnivores), triplets I and II fire synchronously. This is Weijs' carnivore symmetric pattern (Fig. 13.8.) A similar, but not identical, symmetrical pattern is found in at least three
I Fast Close i Slow Close Early Open PRIMITIVE
L
• ws+bs vertical fibers
M
)W////////////////A
WAm^Mmm ;sssssssssssssssssssss
V///////////////////A
L
CARNIVORE SYMMETRIC
• ws posterior fibers (TEM)" bs anterior fibers (MAS) • bs anterior fibers (MPT) _J ' bs posterior fibers (TEM) - ws anterior fibers (MAS) II • ws anterior fibers (MPT) _J
M
%mmw/////M y///////////////////A -1-
RODENT SYMMETRIC P^
p
^A
A )%^
Chewing
Gnawing
j^^^;^^^^?;?^^^ W///////////////A
low level or inconsistent activity "^-ws lateral pterygoid (LPT)
7/////////////////A
I |
TRANSVERSE L
I
RODENT ALTERNATE
r
^cis r^
Chewing
^"Vl
& 1
E55555555555^^^^
7///////////////////////^^^^
^m '
FIGURE 13.8. Patterns of adductor activity during the closing stroke of masticatory cycles in mammals (modified from Fig. 1 in Weijs, 1994). To illustrate the path of jaw movement during the cycle, typical, if somewhat diagrammatic, jaw movement "profiles'' are shown. The ''primitive,'' "carnivore symmetric," and "transverse" groups are based on data from Hiiemae (1978); for the "rodent symmetric" group from Hiiemae and Ardran (1968) and Weijs (1975); and for the "rodent alternate" group from Byrd (1981). In all cases (except rodent gnawing where the occlusal profile of an upper incisor is shown), the horizontal line represents the upper occlusal plane. The vertical dashed line for nonrodent groups represents the position of centric occlusion (maximal intercuspation) on the working side (L, lateral; M, medial with respect to centric occlusion). The direction of chewing and gnawing strokes in the "rodent symmetric" group is shown in the sagittal projection (P, posterior; A, anterior). The alternating profile for Cavia, "rodent alternate" group, is drawn in the coronal plane, ws, working side;bs, balancing side; TEM, temporalis (posteriorly directed fibers); MAS, masseter (anteriorly directed fibers, usually referred to as "superficial masseter"); MPT, medial (internal) pterygoid. Shading is intended to indicate the line of fiber pull, with anterior to the right of the figure. Triplets are indicated by brackets.
429
13. Feeding in Mammals rodent lineages (murids, pedetids, and some hystricomorphs), all of which have isognathous dentitions (Table 13.3) and chew simultaneously on both sides of the jaws. The rodent symmetric pattern involves early firing of the posteriorly directed fibers to retract the jaw, followed by anteriorly directed fibers, which protract it and power the anteriorly directed chewing stroke. In this case (Fig. 13.8), activity within both triplets is temporally spaced. Conversely, in those mammals with flattened (lophodont, bunodont) occlusal planes and anisognathous tooth rows (e.g., ungulates, lagomorphs, and higher primates), the power stroke is heavily transverse. In the transverse pattern, triplet I acts both to close the jaws and to swing the mandible out toward the working side. In humans, this lateral swing often occurs at maximum gape (Hiiemae and Palmer, 2001). Weijs suggests that activity in triplet II may be triggered by tooth-food-tooth contact at the beginning of the power stroke, as these muscles often show greater EMG activity than those of triplet I. Pterygoids on the working side are the last to stop firing; their activity may extend into the intercuspal phase or early opening (Ol, Weijs, 1994), facilitating the disengagement of the teeth on the working side. A few mammals (two rodent lineages, e.g., Cavia, Myocaster, Hydrochoerus, and the mole rat, Tachyoryctes), as well as the camel (Hiiemae, 1978), the warthog (Ewer, 1958), and other suids (Fierring and Scapino, 1973; FFerring, personal communication), chew by alternating working and balancing sides. The activity pattern of the muscles in the guinea pig, Cavia, is known (Byrd, 1981). Symmetrical closer and triplet I activity is reduced to very low levels, with dominant triplet II activity (Fig. 13.8). In his review, Weijs (1994) focuses on muscle activity during jaw closure and the power stroke, commenting that there may be some adductor activity during opening, as well as some digastric activity during closing. Hiiemae et al (1995) and Sher et al (1997) suggest that the explanation for these patterns is associated with tongue function during processing and intraoral transport. 4. Regulation of Adductor Muscle
Activity
Food processing in mammals, i.e., chewing sensu strictu, is rhythmic. The lower jaw opens and closes, changing gape. Gape change may also be augmented by movement of the cranium on the vertebral column (Hiiemae, 1978), as often occurs in nonmammalian tetrapods (Chapter 2). Considerable attention has been given to the neural control of this process [see Taylor (1990) and other papers in that symposium volume]. Given the general agreement that the process of
chewing is maintained, once initiated, by a CPG in the hindbrain (Bremer, 1923; Dellow and Lund, 1971), there is a serious question as to the sources of the sensory input that regulate motor outflow to the adductor and depressor musculature (Hiiemae, 1993). Two such sources have been identified: (1) spindles in the adductor muscles and (2) receptors in the periodontal ligaments of the teeth. Both clearly play a significant role, the former monitoring stretch in the muscles and the latter occlusal load on the teeth. It is easy to understand how these receptors can provide sensory input regulating jaw closing, especially a response to the resistance afforded by food between the teeth. A problem arises, however, when the regulation of jaw opening is considered (see Section VII; Thexton and Crompton, 1989), given the presumption that the periodontal receptors are "off loaded" as the tongue cycles relative to the palate as food is transported or repositioned. It has to be said that the complexity of the behaviors intrinsic to normal feeding in mammals has only become apparent over the last two decades and well documented in recent years. It is no longer sufficient to address the regulation of orofacial mechanisms in feeding solely in terms of jaw movement cycles. Activities of the hyoid, the tongue, and the soft palate, some of these also highly rhythmic and linked to jaw movements, must also be considered.
B. Oropharyngeal Complex 1.
Overview
In 1978,1 and my colleagues published a paper entitled, 'Tntra-oral food transport—a fundamental mechanism of feeding?" We had previously observed cycling of food within the mouth in the opossum (Hiiemae, 1976), but only then began to focus on the behavior of the tongue, hyolaryngeal complex, and pharynx. The oropharyngeal complex can be defined as two spaces, separated by a sphincter-like mechanism (see also Chapter 2): (1) the oral cavity extending from the lips and incisors anteriorly to the fauces posteriorly; roofed by the hard palate, walled laterally by the teeth and cheeks, and with a muscular mobile floor; largely, but not completely, occupied by a highly mobile tongue; and (2) the oropharynx, extending from the fauces back to the esophagus; with the soft palate and nasopharynx above, v/alls of pharyngeal musculature above or behind the hyoid, with the epiglottis, larynx and esophageal aditus, behind or below. It must be emphasized that the larynx and the esophageal aditus are intimately associated with this complex. Indeed, its physiology is designed to assure the passage of (a) air into the larynx and (b) swallowable food into the esophagus. The sphincter mechanism consists of the
430
Karen M. Hiiemae
palatoglossal arches, with their included muscle (otherwise, faucial arches or pillars), the soft palate (also muscular), and the posterior surface of the tongue. Not only is the tongue an essential component of both parts, so is the hyoid complex. However, as far as is known, there appears to be much less gross variation in the function of this complex (although substantial anatomical variation within each component) than in the jaws, teeth, and muscles of mastication (with the possible exception of intrinsic tongue anatomy). This is hardly surprising as the core functions of the oropharyngeal complex—food transport, bolus formation, and swallowing—are common to all mammals, in contrast to the variety of mechanisms used to process different types of foods by the teeth and jaws (see earlier discussion). However, other oral and perioral soft tissues play an important role in the management of food within the oral cavity and the processes illustrated in Fig. 13.3. These are the cheeks, lips, and the epithelium of the hard palate (Chapter 2). Epithelial surfaces of the oropharyngeal complex are lubricated by widely distributed mucus-producing glands. Serous, salivary glands generate copious volumes of saliva in response to the presence of food in the mouth (Hector and Linden, 1987) or in response to odors or other stimuli (e.g., Ludwig, 1850; Bernard, 1852; Pavlov, 1910). These glands are anatomically discrete, extraoral, and, in the case of the parotid and submandibular glands, have well-defined ducts. Secretions often include some enzymes, such as salivary amylase, which begin chemical breakdown of starches to simpler sugars. A discussion of salivary gland anatomy, variations in the chemistry of saliva, and the mechanisms by which salivation is controlled is beyond the scope of this chapter. The last topic has been reviewed by Garrett and Proctor (1998). 2. Oral Cavity The cheeks form the outer walls of the oral cavity, lateral to the tooth rows and attached to the alveolar bone above and below the teeth, creating vestibules within which food can be held or stored during food processing. In many mammals, especially rodents and some primates, the vestibules are highly extensible (e.g., the chipmunk, whose expanded cheeks are wider than its thorax when full of nuts) or have diverticuli to form extraoral pouches (e.g., Brylski and Patton, 1988). These are "food-gathering/hoarding" extensions of the vestibular spaces. The muscle of the cheek is the buccinator, whose fibers decussate around a modiolus ("hub") to pass into either the upper or the lower lip (Fig. 13.9). The anteroposterior position of the modiolus dictates the extent to which the cheeks can
control food position (Heath, 1991). In many mammals, e.g., Didelphis (Hiiemae and Jenkins, 1969), the modiolus is lateral to the first molar, limiting the area of the tooth row in which tongue and cheeks can work to control food position. The inability of the cat to "manage" 1-cm^ cubes of hard liver (see earlier discussion; Thexton et ah, 1980) is almost certainly a function of limited vestibular space due to a posteriorly placed modiolus. Fibers of the buccinator muscle merge anteriorly with those of the lip muscles, specifically the orbicularis oris. Mammalian lips can be highly mobile, as in anthropoid primates and humans, but more typically the upper lip is "tied down" anteriorly by the rhinarium. A complete anterior oral seal can, nevertheless, be formed, as its creation depends on the approximation of upper and lower lip margins, even if the modiolus is positioned posteriorly as in Didelphis. Intriguingly, rats have been observed to "close off" the mouth behind the incisors; indigestible, gnawed material then spills laterally behind the incisors, never entering the oral cavity (personal observation). There are little understood functional variations in the organization of the cheeks and lips in mammals. Many mammals, for example, are unable to appose mobile upper and lower lips in front of the incisors, as do higher primates and ungulates. This special facility allows for a novel range of facial expression and communication (see any film of a distainful camel or baboons or chimpanzees, where the eversion of the upper lip is a definite, and now well-documented, social message). It is worth noting, especially regarding the evolution of primates, that as the modiolus has moved anteriorly, the palatal rugae have become attenuated. Rugae are ridges on the palatal surface of the oral cavity, formed by dense accretions of connective tissue below the epithelium, which are often keratinized (Chapter 2). In most mammals the entire surface of the hard palate is rugose. The ridges are positioned relative to the postcanines so as to optimize their ability to collect triturated food at the end of each power stroke (see Crompton and Hiiemae, 1970; Fig. 13.13). The asymmetric architecture of rugae creates asymmetric, concave catchment areas on the hard palate. The anterior slope of each ruga is shallower than the posterior to prevent spillage of liquid, or semisolid foods, as the jaws open (see later). In higher primates, e.g., the macaque, rugae are reduced, although they still contribute to the management of food during stage I transport (German ei ah, 1989). In humans, they are represented by only a few small mediolateral elevations of the palatal epithelium behind the upper incisors. This pattern suggests an, as yet, untested evolutionary hypothesis for anthropoid primates: the
431
13. F e e d i n g in M a m m a l s
development of a peripheral oral muscular wall (cheeks and apposable lips) with an anteriorly positioned modiolus has provided an expanded containment mechanism for food in the oral cavity, which has replaced that offered by the rugose hard palate.
D.Mass
3. Hyoid Complex and Tongue Base
S. Mass
Dig (pb)
The hyoid is a branchial arch derivative: the medioventral portion of the hyoid apparatus (basihyoid) and first branchial (thyrohyoid) arches form a C-shaped structure that cradles the larynx and is connected to it by muscles and ligaments. In all mammals, it forms a rigid arch buttressing the anteroinferior pharynx. In opossum, tenrec, lagomorphs, and primates, it has no connection to other skeletal elements. In ungulates, carnivores, and some rodents, it may be connected to the tympanic area of the cranial base by the remainder of the hyoid arch, which forms a jointed series of bony or cartilaginous rods. Since Edgeworth's (1914) classic paper, the comparative anatomy of the mammalian hyoid has been discussed by a number of authors; Hilloowala (1975) for primates; Woods (1975), who investigated the hyoid, laryngeal, and pharyngeal region in selected rodents; and Anapol (1988), who describes its anatomy in the rabbit and examined its movements. Hyoid movement patterns, based on Xray studies [cinefluorographic (CFG) or videofluorographic (VFG)], have been studied in opossum (Crompton et al, 1977), cat (Thexton et al, 1980; Thexton and McGarrick, 1988,1989), tenrec (Oron and Crompton, 1985), rabbit (Anapol, 1988), hyrax (German and Franks, 1991), macaque (Hiiemae et al, 1995; Hiiemae and Crompton, 1985; Thexton and Crompton, 1998), and human (Palmer et ah, 1997; Hiiemae and Palmer, 1999, 2000). In addition to contributing to our understanding of hyoid function during feeding in mammals, this work is proving to be relevant to questions regarding the capacity for speech in Neanderthal man. Comparative data provide a context in which attempts to extrapolate from the form of the only known fossilized human hyoid (the Kebara hyoid) to its functions have been made (e.g., Reidenberg and Laitman, 1992; Arensburg et al, 1989; Arensburg and Tillier, 1991; Laitman et al, 1990; Lieberman et al, 1992). As shown in Fig. 13.9, the hyoid is linked to the symphyseal region of the mandible by the geniohyoid muscle. In some mammals, the ventral fibers of genioglossus may also reach the hyoid (e.g., rodents; Woods, 1975). The anterior belly of the digastric is usually connected to the basihyoid bone through its tendon and the fascial plane overlying the hyoid periosteum. Posterodorsally, the stylohyoid connects the basicranium to the ventral surface of the hyoid, whereas
modiolus
Dig (ab)
SupC StyPii IVlidC
InfC GHy
HyG
Trachea
RQ
StHy FIGURE 13.9. (Top) Lateral view of the head of Didelphis to show the external anatomy of adductors (temporalis and masseter) as well as buccinator (with the modiolus, see text), and the arrangement of the superficial muscles attaching to the hyoid. (Bottom) A schematic sagittal hemisection to showing the deep hyoid muscles and the muscles of the pharynx. The symphysis and hyoid are shaded. Genioglossus radiates into the body of the tongue (lightly shaded) from the genial area on the symphysis and deep to the fibers of Hyoglossus, which radiate anteriorly from their origin on the hyoid (see also Fig. 13.12). Biomechanics of the hyoid musculature in the opossum are described in Crompton et al. (1977). [Redrawn from Fig. 14-4 in Hiiemae and Crompton (1985).] Tem, temporalis; D. Mass, deep masseter; S. Mass, superficial masseter; Bucc, buccinator; Dig (ab) and Dig (pb), anterior and posterior bellies of digastric; OmHy, Omohyoid; StThy, sternothyroid; StHy, sternohyoid; StyPh, stylopharyngeus; PC, palatoglossus; HyG, hyoglossus; GHy, geniohyoid; Sup C, Mid C, and Inf C, superior, middle, and inferior constrictors of the pharynx, respectively.
posteroventrally, the sternohyoid attaches to the sternum and the omohyoid to the scapula. These muscles form an adjustable suspensory mechanism (Crompton et al, 1977). Anapol (1988) modeled the possible range of theoretical hyoid movement using values of 30% shortening over resting length for muscle contraction and 130% length for maximum stretching. He calculated theoretical "domains" within which hyoid movement could occur and then recorded actual hyoid movement in feeding, mapped it onto his predicted domains, and found that observed movement domains correspond closely to the central and posterior sagittal
432
K a r e n M. H i i e m a e
et al. (1977) for Didelphis]. During normal feeding, the hyoid is in continuous motion, although its amplitude of movement may vary among feeding stages and among taxa (Figs. 13.5 and 13.10). Importantly, the hyoid moves upward and forward during the early part of opening (Ol and 02) and then downward and backward during 0 3 and closing in opossum, hyrax, cat, and human. It is also noteworthy that, with some slight delay in the macaque, the reversal of hyoid movement from forward to backward occurs at, or just after, the start of 0 3 (see Fig. 13.5B). Two issues immediately arise: (1) why this pattern and (2) what is the muscle activity responsible, or rather, what are the functions of the jaw abductors?
areas of the predicted horizontal domain. This approach is potentially valuable for studies in other mammals. However, as Anapol points out, hyoid movement during pellet feeding in rabbits is of very low amplitude. It would be worth applying such methods to mammals feeding on more natural foods. Classically, human anatomists taught that the jaw was opened by contraction of the digastrics, with geniohyoids acting against a "fixed'' or stabilized hyoid, i.e., activity in the posterior suprahyoids and the infrahyoid muscles prevented geniohyoid (and anterior digastric) contraction from pulling the hyoid forward as the jaw opened. EMG and movement studies have shown that this view is wrong [see Crompton
-10
t%»
I
«
*
' *'
*i
.A*
•* V
GO
70
80
90
HYOYLT -H*H
1
1
H
1
H
-f—I
1
H
h-
sec FIGURE 13.10. Gape-time plot for jaw and hyoid movement in a complete feeding sequence in human (chicken spread). Hyoid movement is shown in the horizontal axis (X) and the vertical (Y) relative to the upper (HYOU prefix) or the lower (HYOL prefix) occlusal planes. The fine dotted line marks the end of stage I and the thick dotted lines the start of stage II, i.e., the period in which food is (intermittently) moved through the fauces for bolus formation (oropharyngeal bolus accumulation time or OPAT). Fine arrows indicate cycles in which stage II transport occurred: the thick arrow represents an attenuated cycle in which a major food collection and transport movement took place. Swallows are shown as shaded bars. The continuous movement of the hyoid is clearly shown. The short period of clearance (with irregular jaw movement) is indicated by the horizontal bar between the two swallows. The horizontal axis shows time in seconds. The vertical axis gives a scale for the amplitude of each movement in millimeters but does not represent the actual range of movement overall as traces have been separated for clarity. Reproduced with permission from Hiiemae and Palmer (1999).
1
\r
13. Feeding in Mammals The answers hinge on the dual role of the hyoid: it is the posterior anchor of the floor of the mouth and the mobile but rigid arch to which the hyoid muscles are attached. In addition to anteroposterior^ oriented fibers of the geniohyoid, the floor of the mouth is generally considered to be formed by obliquely posteriorly directed fibers of mylohyoid, which link the hemimandibles to each other through a median raphe and to the hyoid. Contraction of the mylohyoid can pull the hyoid forward, as does the geniohyoid, and also elevate it. This action then changes the relationship of the body of the tongue to its base, shortening it, thus affecting the orientation of the lingual muscles, genioglossus (fanning into the tongue from the symphyseal region), and hyoglossus (fanning into the tongue from the hyoid). We have argued (Hiiemae and Palmer, 2000) that this shortening of the tongue base facilitates movements of the tongue associated with food transport mechanisms (see later). It is also important to note that the observed pattern of hyoid movement explains the recordings of adductor EMG activity in jaw opening and depressor activity in closing (Weijs, 1994; Hiiemae et al, 1995; Hiiemae and Palmer, 2000). Anterior digastric, posterior digastric, and geniohyoid have multiple bursts of activity in the macaque (Sher and Hiiemae, 1998) and human (Palmer et al, 1992). Activity during SC and IP is associated with tongue movement and then the forward movement of the hyoid in IP, O l , and 0 2 . The second major burst in 0 3 depresses the jaw to maximum gape for that cycle. In the macaque, where low-level activity in the superficial masseter or the internal pterygoid has been recorded in opening, it occurs during late O l and throughout a prolonged 0 2 . This strongly suggests that these muscles are acting to restrain any tendency for the jaw to open beyond the narrow range of gape (approximately equivalent to the maximum gape during lapping) associated with the 0 2 - 0 3 transition, given powerful activity in geniohyoid. It also suggests that large protrusive transport movements of the tongue can only occur within a narrow range of gape. Although anthropoid primates have shown a general tendency for the foramen magnum to migrate from the back of the skull to its base, such that in the macaque, the oropharynx lies somewhat below the oral cavity as opposed to behind it as in opossum (Thexton and Crompton, 1998); humans are the only mammals in which the long axes of the oral cavity and oropharynx are essentially at right angles (Thexton and Crompton, 1998). This shift, coupled with facial shortening and an increase in height of the oral cavity, has resulted in (a) a hyoid complex below, rather than behind, the tongue; (b) a lengthening of the oropharynx; and (c) an increase in tongue height and reduction in
433
length (Fig. 13.13). It is generally believed that these changes are associated with changes in primate vocalizations such that humans could, biomechanically, develop these sounds into the complex phonemes of speech (Lieberman et at, 1992, and earlier papers). It has long been thought that the human uniquely long separation of the soft palate-tongue sphincter (commonly called the posterior oral seal in humans) from the laryngeal aditus required that human bolus formation occurs in the oral cavity and, that once formed, it must be explosively propelled across the oropharynx while the vocal folds were closed and the laryngeal aditus protected by a lowered epiglottis. However, as described later, bolus formation for solid foods occurs in the oropharynx (not the oral cavity) in humans. Even more intriguing are observations (manuscript in preparation) on the domains of hyoid movement during feeding and speech. Clearly, given the data in Fig. 13.11, we can draw the following inferences: (1) the hyoid moves in both feeding and speech with a smaller amplitude in the latter and (2) to produce the range of movement seen (and with minimal overlap in the two functional domains), there must be different patterns of activity in the hyoid musculature, such that geniohyoid and mylohyoid are changing length around a shorter modal length in speech, while the posterior suprahyoids must be exhibiting the reverse pattern. This raises interesting questions about the motor control system for these muscles in an activity for which no one, as yet, has suggested a CPG. It is interesting to note that in our Xray studies of human speech, in the very few cases in which a subject swallowed while reading, hyoid movements were those of a conventional swallow, suggesting a capacity for instantaneous transition between functional domains (speech to feeding). 4. Tongue The mammalian tongue is an extraordinary organ. It has at least three basic functions: food collection, intraoral food management, and as an organ subserving the "special sense" of taste. In an important paper, Kier and Smith (1985) described the tongue as a "muscular hydrostat," characterized by constant volume such that a change in one dimension produces compensatory change in another (further developed in Smith and Kier, 1989) (see Chapter 2). While some have argued that the details of their model are unproven, the principle appears to hold for all mammals with muscular tongues (Doran, 1975). A simple, if deficient, analogy for the muscular hydrostat model is to see the tongue as a water balloon: decrease its diameter by squeezing and it elongates; squash it and it
434
K a r e n M. H i i e m a e -20
2b
25
3P
35
45
m
sp
56
60
65
-25 5
0.0000001. Data for eight subjects (both sexes) show the same pattern and no sexual dimorphism. From Hiiemae et al. (1999, 2001).
spreads, changing shape in all three dimensions. Compensatory shape changes are required by the fact that the tongue maintains a constant volume at all times due to the incompressibility of its constituents (primarily muscle tissue). In practice, the behavior of the tongue is far more complex than this simple analogy. Most studies of tongue movement reported in Section VI,A have used implanted, radio-opaque markers to provide consistent (within subject) reference points: just under the dorsal (gustatory) epithelium with surgically inserted tongue markers (opossum, tenrec, cat, macaque, the bat Pteropus) or by gluing markers to the tongue surface (human). These studies have shown that the tongue is capable of intrinsic expansion and contraction {Pteropus, de Greet and de Vree, 1984; cat, Thexton and McGarrick, 1989; rabbit, Cortopassi and Muhl, 1990; macaque, Hiiemae et al, 1995). The tongue
can be protruded extensively and rapidly retracted (as in lapping). Shortening of the tongue base (cf. contraction of geniohyoid) facilitates extensive protrusion of the intraoral part of the tongue. Hyoid retraction and elongation of the tongue base augment its retraction. A highly elastic connective tissue tunic surrounding the tongue permits such extensive shape changes and possibly helps restore the tongue to its resting shape (Schwenkeffl/., 1989). 5. Tongue as a Sensory Organ Doran (1975) described two types of mammalian tongues: type I, whose function is primarily intraoral, found in most mammals, and type II, found in mammals such as anteaters (Chapter 15), where tongue function is primarily food gathering. All mammalian
13. Feeding in Mammals tongues have four types of papillae on their oral surface: filiform, foliate, fungiform, and circumvallate. Doran argues that the filiform and foliate papillae have primarily mechanical functions, whereas the fungiform and circumvallate papillae are primarily gustatory. Taste and smell are the primary determinants of "palatability" (gate I, process model. Fig. 13.3). Taste receptors are organized in various areas of the mouth, but famously in the gustatory surface of the tongue. Gilbertson (1998) argues that the ''taste system'' in mammals can be considered as having two vital roles: (1) serving to identify the presence of essential nutrients such as minerals (e.g., salts), carbohydrates, proteins (amino acids), and fats and (2) identifying harmful or potentially toxic substances before they are ingested. Taste transduction mechanisms have been reported for all the essential dietary constituents in mammals. Gilbertson (1998) reports that many toxic substances, such as plant alkaloids and animal venoms, have an intensely bitter taste, and without the ability to detect such substances, the consuming animal might suffer fatal consequences. He argues that, despite an approach that has historically devalued the importance of taste perception as a regulator of nutritional state, new research is giving support to the hypothesis that taste receptor cells are not merely transducers of transient chemical signals, but may play a role in the longer term management of dietary intake via nutritional cues. Similarly, smell plays a vital role in feeding. "The olfactory system is a critical component in the location, identification, and hedonic perception of food" (Wilson and Sullivan, 1998). In mammals, odor molecules must reach the olfactory receptors in the cribriform plate of the ethmoid (the interface between the cranium and nasal cavity) and pass through a thin mucus layer before reaching them. Clearly, there are two possible routes for such molecules: either directly through the nares during inspiration or via the nasopharynx by means of retronasal airflow. Wilson and Sullivan (1998) suggest a third possible route through the circulation of blood-borne odorants (e.g., garlic?). While there is general agreement that respiration is suspended during swallowing (swallowing apnea), there is little useful information in the literature on the relationship between respiration and mastication during intraoral food processing. Triggered by our observation (Palmer and Hiiemae, 1997; Fiiiemae and Palmer, 1999) that bolus formation can occur in the oropharynx in humans, we have embarked on a series of experimental studies to investigate the relationship between respiration and mastication using a nasal cannula and pressure transducer (Arvas et ah, 1999). While very preliminary, and restricted to human subjects, our results
435
indicate an oscillation in basal respiratory rate directly correlated with adductor activity in humans, i.e., jaw closing during the feeding cycle is correlated with a pressure wave passing subposteroanteriorly through the nasal cavity. One can hypothesize that food processing may release important odorants not detectable before food breakdown. If food processing releases such odor molecules, then a retronasal route for their detection may have an important biological role. Although further experimentation is needed, this preliminary finding not only suggests how odor molecules may reach the olfactory epithelium during feeding through retronasal airflow, but may also explain our surprising finding that there is no predictable complete posterior oral seal in human (see later). Given the "intranarial" position of the larynx in all but hominids, and the possibility of reverse airflow round the larynx into the nasopharynx, it is possible that the valleculae in nonhuman mammals provide an "odor transmission route" to the olfactory epithelium when the airspace in the oral cavity is reduced by jaw closure and tongue elevation, creating a pressure wave such as recorded in humans. Chemoreception issues are not usually considered by those interested in explaining the functional morphology and evolution of feeding behavior. Nevertheless, they should not be neglected. It remains the case, however, that this is an area easiest to study in humans—human subjects can be asked to eat and report on their perceptions while eating (Mioche and Martin, 1998). 6. Tongue Musculature and Tongue Function Almost all mammals have type I tongues (Doran, 1975). Such tongues are spatulate and can protrude to a maximum of 50% resting length. The tongue is anchored to the mandibular symphysis, the hyoid, and, more remotely, the skull base via so-called "extrinsic muscles" (genioglossus, hyoglossus, and styloglossus, respectively). The palatoglossus is also included in this group, although it has functions associated with the posterior oral seal and the oropharyngeal sphincter. Within the body of the tongue, there are so-called "intrinsic muscles," longitudinal, transverse, and vertical, which have their attachments within the tongue and its included aponeuroses and tendinous septa. Livingstone (1956) argues that movement of the tongue depends largely on movement of the hyoid and that change of overall tongue position results from extrinsic muscle action, whereas intrinsic muscles provide for changes in shape and tongue mobility. Studies using radio-opaque markers to map change in tongue shape show, unequivocally, that the tongue not only changes its overall position relative to the hard palate during
436
Karen M. Hiiemae
feeding as a function of change in tongue base length (see earlier discussion), but that it normatively changes its shape (expands/contracts) within its intraoral body as processing proceeds (Thexton and McGarrick, 1989, Hiiemae et al., 1995). These results strongly argue for "revisiting" the traditional view that "intrinsic tongue muscles" (longitudinal, transverse, and vertical) function "independently" of the "extrinsic" muscles. All the experimental evidence to date indicates that the tongue functions in three dimensions as an integrated whole and that its high-speed changes in shape, both intrinsic and measured relative to the hard palate or the lower jaw, are the result of complex neuromuscular integration in the hindbrain of which we have essentially no understanding. It follows that the traditional distinction between "intrinsic" and "extrinsic" muscles is an "anatomical convenience" and has little to do with physiological reality. Parallel work (Schwenk, 2000a) confirms this view. There have been few studies of the tongue muscles addressing organization, fiber type, and contraction times. Hellstrand (1980, 1981) carried out an exhaustive study of both extrinsic and intrinsic muscles in the cat using anatomical and histochemical methods and used a novel technique (light reflection) to measure contraction times. Contraction times of intrinsic and extrinsic muscles averaged 20 and 33 msec, respectively. These values fall within the range of some fast muscles of the limbs and some laryngeal and facial muscles, but are longer than that reported for digastric, although shorter than for parts of the adductors (Thexton and Hiiemae, 1975). Tongue movements, in vivo, are incredibly complex. To attempt their description, four levels of analysis are required: (1) position of the lower jaw relative to the upper jaw and palate (change in gape), (2) position of the hyoid relative to the mandibular symphysis (length and orientation of the tongue base), (3) the gross relationship of the body of the tongue to its base (A-P and vertical relationships relative to hard and soft palates), (4) the actual shape of the tongue surface (intrinsic expansion, contraction, and rotation). In normal feeding, all four levels of activity occur concurrently, although experimental evidence to date suggests that some tongue movements are not only "patterned" but occur in relation to specific phases in the jaw movement cycle. It has, so far, proved impossible to "dissect" the totality of tongue muscle activity using EMG and to precisely delineate/ascribe specific roles for any muscle in vivo, despite reliable electrode placement in the extrinsic muscles. We cannot yet model the contributions to activity producing the changes in tongue shape fundamental to its intraoral food management and transport functions. However, the insertion/attachment of radio-opaque markers below or on the gustatory epi-
thelium has allowed the measurement of movements of its surface during feeding (albeit largely in the lateral projection). These findings further suggest that a functional separation of intrinsic vs extrinsic muscles is arbitrary and may be unhelpful (see also Schwenk, 2000a). Although the intrinsic biomechanics of the tongue remain poorly understood in all mammals, its general role in feeding is now well documented. The tongue muscles are innervated by cranial nerve XII, the hypoglossal. Sensory innervation of the tongue reflects its complex embryological development [see Grays Anatomy (1995) for a detailed description]. What is becoming clear, however, is the inadequacy of our knowledge of the relationship between oral sensation and tongue movement. Although the neurophysiological literature includes some very simplistic explanations for the mechanisms of, for example, tongue tip deviation to one or the other side, we have essentially no information as to the neural circuits that govern such functionally important behaviors as intrinsic contraction and expansion. The fact that these shape changes occur primarily during the IP and opening phases of the chewing cycle, when one might expect the periodontal receptors to be progressively "off loaded," raises intriguing questions. Furthermore, such intrinsic behaviors vary with feeding stage and food condition (cat, Thexton and McGarrick, 1989; rabbit, Cortopassi and Muhl, 1990; macaque, Hiiemae et ah, 1995). The tongue is not simply an intraoral organ—it has two parts, as conventionally described: (1) within the oral cavity and (2) in the pharynx (Fig. 13.12). Physiologically, these form a functional continuum; the oral part has a fundamental role in the management of food within the oral cavity, but when the food has reached a swallowable consistency, it is moved posteriorly onto the pharyngeal surface. The spatial relationships and functional implications thereof are very different between most mammals and humans. These differences hinge on the anatomy of the oropharynx and the sphincter separating the oropharynx from the oral cavity. 7.
Oropharynx
In anatomical terms, the oropharynx is the functional space between the soft palate and tongue anteriorly, in which bolus formation occurs, and the esophageal aditus (upper esophageal sphincter) posteriorly, through which the bolus enters the digestive tract (sensu strictu). In most mammals, but not humans, the oropharynx is separated from the airway such that the tidal airflow passes directly from the trachea, through the larynx to the nasopharynx. Oropharyngeal bolus accumulation involves (1) the
13. F e e d i n g in M a m m a l s
^^
GHy
Trans M.
HyG
StyG
GHy Longitudinal muscle fibers Densely interwoven intrinsic fibers Indication of transverse and vertical intrinsic fibers i^^j,^^^^ Tongue and oropharyngeal epithelium (in contact with food).
FIGURE 13.12. Diagrammatic transverse sections of the tongue and pharyngeal regions in a 1-day-old rat (drawn from photomicrographs in Fig. 11, Smith, 1992). (Top) Anterior section through the tongue at the level of the premolar tooth germs. While the bulk of the fasciculi of hyoglossus and genioglossus are clearly identifiable, their fibers are merging with those of the "intrinsic" muscles. The central region below the gustatory epithelium consists of a dense mesh of transverse and vertical fibers originating from fibrous septa and intermingling with "intrinsic" longitudinal fibers peripherally and with those of hyoglossus and genioglossus (see Fig. 13.9 for the orientation of those fibers in the sagittal plane). (Bottom) Posterior section from the back of the tongue, just posterior to the hard palate. Hyoglossus is fanning forward into the oral part of the tongue and geniohyoid is close to its hyoid insertion. Styloglossus (not shown in Fig. 13.9) is passing into the tongue, whereas Palatoglossus is still above the oropharynx, although it will pass into the tongue to form the "pillars of the fauces" (see text and Fig. 13.13). NC, nasal cavity; OC, oral cavity; OP, oropharynx; PT, pterygoid bone; UPM, upper premolar tooth germ; LPM, lower premolar tooth germ; LI, lower incisor tooth germ; MyH, mylohyoid; Trans. M, transversus mandibulae; TVP, tensor veli palatini; StyG, styloglossus. Otherwise as listed for Fig. 13.9.
437
movement of food through the posterior oral seal/pillars of the fauces (stage II transport, see later) and (2) to be collected in the valleculae and piriform fossae, lateral to the larynx. The bolus is then propelled into the esophagus in a single "swallowing action/' The anatomy and evolution of the oropharynx have been described by Smith (1992), and its anatomy and role in swallowing by Thexton and Crompton (1998). Figure 13.13 shows its basic organization in the opossum and human. The structures involved in each species—posterior (pharyngeal) surface of the tongue, the soft palate, the pharyngeal constrictors posteriorly and laterally, and the hyolaryngeal complex anteroinferiorly—are the same, but they differ in proportion and details of arrangement. It is important to emphasize that the oropharynx is a dynamic space: the tongue, soft palate, and pharyngeal constrictors are all muscular and all move, especially during feeding. Spatial relationships can change dramatically with wide jaw opening: for example, the hyoid can be pulled down and back and, with it, the larynx and the epiglottis; when pigs vocalize, the epiglottis can have an "intraoral'' position (Herring, personal communication). Postmortem specimens, the basis for most anatomical drawings, can give a distorted impression of the functional relationships of these structures. While difficult to visualize, even with drawings such as Fig. 13.13, the pharynx in Didelphis can best be described as a tube within a tube. The inner tube consists of the larynx with the trachea below. The entry to this tube lies in the posterior nasopharynx, a condition termed the "intranarial larynx." The outer tube is formed by pharyngeal constrictors, which originate above from the cranial base, and anteriorly from the pterygomandibular raphe, as well as the hyoid and its associated cartilages (thyroid, cricoid), descending to merge with the muscular esophagus. This tube is, therefore, technically open to the oral cavity anteriorly. In practice, that opening is gated by the soft palate, above, and the oropharyngeal surface of the tongue and the epiglottis, ventrally and laterally. Dimensions of the space between then\ are regulated by the position of the soft palate, but also activity in the palatoglossus muscles, forming the palatoglossal arch and defining the "pillars of the fauces." Spaces anterior to the fauces are considered to be within the oral cavity, posterior to the fauces, within the oropharynx. In Didelphis, there is a distinct lateral separation between the outer wall of the inner tube and the inner wall of the outer. The most anterior part (walled by the tongue, soft palate, and epiglottis) forms the valleculae, whereas the more posterior (lateral to the larynx) are the piriform fossae. It follows that food moves through the fauces (stage II transport), enters the valleculae, and, as volume builds, fills the expandable
skuii base esophagus B
nasopharynx
soft palate
j^y^j^ I larynx epiglottis palatal rugae
valleculae palatoglossus
nasopharynx palatoglossus soft palate genloglossus
pharyngeal constrictors valleculae epiglottis
geniohyoid mylohyoid upper esophageal sphincter
trachea F I G U R E 13.13. Midsagittal sections showing the anatomy of the oropharynx in the opossum and modern man. (A) Diagrammatic sagittal section through the head of the opossum to show the position of the oropharyngeal complex relative to the oral cavity in a normal feeding position. The area included within the rectangle is enlarged in B. (B) The airway is shown by the thicker arrow: Air enters the nasopharynx and passes directly into the larynx, given that the epiglottis projects above the soft palate. In contrast, food accumulating to form a bolus in the valleculae passes to the esophagus by going ''round'' the larynx through the piriform fossae (dotted section fine arrow). See text and Thexton and Crompton (1998). (C) While the same anatomical elements are present (e.g., palatoglossus, which forms the "pillars of the fauces"), their spatial relationships are very different. The tongue has an expanded oropharyngeal surface (shaded), and the soft palate and epiglottis are well separated. In normal respiration, air passes through the nasopharynx, across the oropharynx, and into the larynx (heavy arrow). Three basic bolus formation and swallowing mechanisms have been described: (a) small volumes of liquids are accumulated in the oral cavity anterior to palatoglossus, which contracts to form a posterior "oral seal" between the tongue and the soft palate; the bolus is then propelled through the oropharynx and into the esophagus with the hyoid elevated and the epiglottis depressed over the laryngeal aditus; (b) semisolid and sometimes processed solid food may pass into the oropharynx, reaching the valleculae in one chewing cycle; in the immediately following cycle, a bolus is expelled from the oral cavity, "collects" the material in the valleculae, and all is swallowed; and (c) bolus accumulation occurs on the oropharyngeal surface of the tongue, with the bolus moving into the valleculae before propulsion into the esophagus (see text and Hiiemae and Palmer, 1999).
439
13. Feeding in Mammals piriform fossae from which the bolus is propelled into the esophagus during swallowing. Similar arrangements are found in cats (Thexton and McGarrick, 1988, 1989) and pigs (Herring and Scapino, 1973), as well as goats and the tenrec (see Thexton and Crompton, 1998). It n\ust be emphasized that the spatial relationships of the larynx, epiglottis, and soft palate are dynamic in all mammals. In the macaque, the larynx is still intranarial, but its position is controlled more loosely and is somewhat lower relative to the soft palate, as compared to Didelphis (Thexton and Crompton, 1998). The important difference from the opossum is the reduction in the size of the vallecular space and the narrowing of the piriform fossae, which form channels rather than spaces. These fossae do not serve as bolus accumulation sites as in Didelphis, although some material can accumulate in the valleculae during mastication. Consequently, while a small volume of triturated food can move through the fauces directly to the valleculae, the bulk of the bolus is propelled through the fauces into the oropharynx, which then "collects'' the material in the valleculae from where it is moved into the esophagus during a swallow. Ontogenetically, H. sapiens matures from a neonatal oropharyngeal anatomy comparable to that seen in the macaque to the unique morphology shown in Fig. 13.13. Much has been made of the vertical separation between the posterior oral seal and the esophageal aditus. There is no question that, uniquely among mammals, the airway and foodway share a common passage—the oropharynx—in humans. Until recently, it was assumed that this arrangement required the bolus to be accumulated in the posterior oral cavity and then propelled directly into the esophagus by a coordinated activity of tongue, soft palate, and pharyngeal musculature, as well as by hyoid and epiglottal movement. As described earlier, evidence now suggests that bolus formation in humans can occur (1) in the oropharynx, analogous to the mechanism in Didelphis; (2) in both the oropharynx and the oral cavity with fusion of the two masses to form a single bolus for swallowing, as in the macaque; and (c) in the oral cavity as traditionally described. The movement of triturated solid food into the oropharynx for bolus accumulation or deglutition is generally regarded as controlled by the posterior oral sphincter, i.e., the "gate" afforded by activity in the muscles of the soft palate and the palatoglossus muscles. The latter muscles lift the back of the tongue, ensuring a tight contact between the tongue and the soft palate to form the seal. Food meeting the "swallowable" criterion (Fig. 13.3) is allowed passage to the oropharynx. Unfortunately, there is a problem with this interpretation. There is no study of the relation-
ship between the soft palate and the tongue during feeding reported in any mammal other than humans. Furthermore, data supporting the mechanical definition of a posterior oral seal are wholly dependent on human studies (e.g., Dantas et ah, 1990). Even more provocative, work on feeding in normal human subjects (Palmer and Hiiemae, 1997; Hiiemae and Palmer, 1999) has shown that the "seal" in humans is largely inoperative during mastication of solid food; rather, there is a wide-open, two-way channel. This is not to say that H. sapiens has no seal. It is vital to the management of ingested liquids. It may well prove that these observations on humans are irrelevant to the functions of the oropharynx in nonhuman mammals. Although the larynx is allegedly intranarial in most mammals, Larson and Herring (1996) report that the epiglottis descends to cover the laryngeal aditus during swallowing in pigs and ferrets. Mention of this issue is intended to draw attention to the hazards of extrapolating data from a peculiar, but much studied, species (H. sapiens) to other mammals.
VL FEEDING FUNCTION A. T o n g u e - J a w Linkages During the feeding sequence, and in each constituent cycle, three patterns of movement are occurring, linked, but not necessarily in complete synchrony: (1) the jaws are moving through the gape cycle (FC, SC, IP, and opening), (2) the hyoid is traveling upward and forward, and downward and backward, and (3) the tongue surface is both expanding and contracting, as well as cycling in an orbit that carries its surface downward and backward, then upward and forward (Fig. 13.14). Experimental evidence suggests that the jaw movement cycle can be considered as involving two discrete processes: (1) food processing, which occurs during the SC or PS phases of closing (Fig. 13.5); and (2) intraoral food management, which occurs primarily during IP and early opening (Ol, 02). It can be argued that the FO (03) and FC phases of the jaw movement cycle are for "repositioning," i.e., to return either the jaws and teeth to a position in readiness for the next chewing stroke or the tongue to a position from which it can collect and organize the food particles in the mouth. It is important to note that tongue surface movement, especially in its anterior and middle parts (see Hiiemae et ah, 1995), often precedes movement of its posterior (oral) surface and the hyoid may "lag" behind that. The tongue is not behaving as a monolithic unit, but rather one capable of subtle, instantaneous, and localized intrinsic responses to changes in the
440
Karen M. Hiiemae
Hard food - shortbread cookie
Soft food - chicken spread
Upper Occlusal Plane ANTEfitOR
1 cm
JAW OPEN
1 cm
FIGURE 13.14. Orbits of tongue surface markers during the feeding sequence when feeding on foods of different consistency (redrawn from Fig. 4 in Palmer et. ah, 1997). Positions of the anterior tongue marker relative to the upper occlusal plane at its maximum down (TD), back (TB), upward (TU), and forward (TF) positions are shown for early (bottom orbits), middle, and late sequence cycles and for swallows (uppermost orbits). The direction of tongue marker movement is shown by arrows. As is the case for hyoid, the tongue rises as the sequence progresses. This distinct vertical progression has not been observed in other mammals (e.g., opossum, cat, macaque) where the tongue has a much longer anteroposterior dimension than vertical (see Fig. 13.13). However, the direction of movement and its temporal relationship to jaw movement are homologous with that in other mammals studied (Hiiemae and Palmer, 2000).
intraoral environment. This pattern is interrupted only when a swallow occurs. Swallows are discrete but intercalated events: the IP and early SO phases of the masticatory cycle are prolonged as the swallowing CPG is activated (see Thexton and Crompton, 1998). Experimental data for Didelphis and hyrax (see Hiiemae and Crompton, 1985) suggest that the switch from processing (jaw-based behavior) to food management (tongue-based behavior) occurs at minimum gape. Analysis of all available published data (Hiiemae and Palmer, 2000) suggests that it would be more accurate to state that the changeover occurs after the teeth have reached maximal approximation in each cycle, and always within the IP period. Although not readily evident from published, time-compressed jaw movement plots, there is no precise moment within the jaw movement cycle, recorded with CFG or VFG, that can be unequivocally identified as minimum gape, or the transition point between the two behaviors. That said, the application of computer analysis to fully digitized data for which the Cartesian coordinates of pivotal reference points have been established and manipulated does allow such a point to be calculated. Such electronic precision may create an artificially precise time point, but render the result biologically questionable. When data for macaques (Hiiemae et ah, 1995) and humans (Palmer et ah, 1997) are examined, the j a w tongue linkage is present, but with slippage. None-
theless, "turn points" in the tongue movement cycle occur predictably in relation to defined jaw movement events. Accumulating evidence shows that tongue movements may, in certain circumstances, regulate jaw movements. Lapping, for example, occurs within a very narrow range of gapes (Thexton and McGarrick, 1988, for cats). Extensive tongue protrusion appears not to occur outside that gape range. In the macaque (Hiiemae et al, 1995; unpublished data), 0 2 always occurred within a narrow gape range (macaques do not, typically, lap), but within 0 2 extensive tongue protrusion could occur. Such protrusion never occurred at wider gapes. This suggests that the biomechanics of the hyoid musculature (with contributions from the adductors, see earlier discussion) are such that significant tongue base shortening, carrying the tongue base forward so that the body of the tongue has a positional advantage, may depend on the control of gape. This hypothesis needs testing. However, there is evidence (Palmer et al, 1992, 1997; Hiiemae et al, 1996) that in the apparently unique human behavior we have termed "clearance" (Fig. 13.10), movements of the tongue in collecting and aggregating food for bolus formation occur independently of jaw movement. Therefore, it appears that the jaw-tongue linkage found in regular processing cycles can be decoupled.
13. Feeding in Mammals If jaw and tongue movements are linked in normal masticatory cycles, but can be decoupled, then, ineluctably, an important question follows. It is generally agreed that jaw movement cycles are regulated by a CPG in the hindbrain. The concomitant occurrence of rhythmic tongue movements, despite a decade in which their occurrence has been reported, has not yet been considered in the context of central nervous system control mechanisms. The currently accepted models for the control of jaw movement (see Taylor, 1990) do not seem applicable to the tongue. Such models rely on sensory feedback from the periodontal ligament receptors and the spindles in the adductor muscles. Given what we now know about masticatory behavior, to attribute complex tongue behaviors to those sensory inputs is naive. Nevertheless, as should be clear from the foregoing, the behavior of the tongue is not readily analyzed using conventional neurophysiological techniques. The contraction pattern of the tongue muscles is clearly governed by hypoglossal nerve activity. How this activity pattern is generated remains to be determined. B. Food Manipulation and Movement Five discrete processes are involved in the movement of food into the mouth, within the oral cavity, through the fauces, and then into the esophagus (see Chapter 2): (1) ingestion, i.e., placement of food into the anterior oral cavity; (2) stage I transport, i.e., movement of the ingested material from the anterior oral cavity to the cheek tooth region; (3) manipulation, i.e., the cycling of food within the postcanine region as it is reduced; (4) stage II transport, i.e., movement through the fauces to the oropharynx; and (5) deglutition {swallowing). All involve the tongue and palate. As shown in Fig. 13.3 and thoroughly described in previous papers (e.g., Hiiemae and Crompton, 1985; Thexton and McGarrick, 1988; Thexton and Crompton, 1998) and in Chapter 14, liquids are ingested, transported, and swallowed in a continuous process in which the tongue acts as a conveyor belt, moving successive small volumes from the external source to the oropharynx in a definite rhythm. In contrast, processes used for solid foods are n\ore complex and may be temporally segregated. 1. Food Acquisition and Ingestion Mammals use a wide variety of techniques to acquire their food. Many, including primitive mammals, use a forefoot or hand to pick up food items and bring them to the mouth. The squirrel gnawing at a nut held in the forepaws or the anthropoid primate tearing at a hand-held, hard-skinned fruit are using their ante-
441
rior teeth to access the digestible material. Artiodactyls have no such option. Some use the tongue to wrap round leaves and then strip them; others crop vegetation using the lower incisors against a keratinized pad, which has replaced the upper incisors. The most dramatic use of teeth and jaws in food acquisition is seen in felid carnivores, where powerful canines and adductor muscles are used to bring down prey. Any avid watcher of natural history documentaries knows that the process of food acquisition in mammals can involve a complex pattern of limb, head, lip, tongue, and jaw miovements, and even those of an elongated airway, the highly specialized elephant trunk! Moreover, a single species may use more than one method of ingestion depending on the nature of the food. For example, grizzly bears catch spawning salmon in their forepaws or in their mouth and use yet another technique when it comes to obtaining honey. Regardless of how food is primarily acquired (a topic worthy of a separate review), it is first placed in the anterior oral cavity. The only exception occurs when the cheek teeth are actually used to separate pieces (bites) of food from their matrix using the postcanines. Many carnivores do this, as does Didelphis, a behavior described as ''ingestion by mastication" (Hiiemae, 1976). Although not strictly within the scope of this chapter, attention should be drawn to the important ''ancillary" uses to which the jaw apparatus is put in many mammals. We are all familiar with the beaver dam, made from logs trimmed by powerful, gnawing incisors. We know that the canine sexual dimorphism found in many mammals, as well as in primates, has a social function and that elephant trunks are used for digging and allied activities. Such behaviors can be assumed to have influenced the evolution of craniofacial and dental anatomy. 2. Stage I Transport In most mammals for which we have data, food deposited in the front of the mouth is moved to the postcanine region using patterned movements of the tongue and jaws. The alternative, seen in carnivores, primitive mammals such as Didelphis, tenrec, and the prosimian Tupaia, is inertial transport—the jaws are opened and the head is moved rapidly forward to surround the food item (see Fiiiemae and Crompton, 1985). Stage I transport and inertial transport are not mutually exclusive, but rather are two methods of dealing with solid material. The former is always used for liquids, however. The basic mechanism of stage I transport depends on the protraction of the tongue surface below the solid bite as the jaw opens (Ol and 02), its cradling on the tongue surface, and then posterior movement as the
442
Karen M. Hiiemae
tongue retracts in jaw closing. (It is important to note that CFG/VFG evidence for this process as described here for solid foods is based on data for only three species: cat, hyrax, and macaque.) As the jaw opens again, the bite is held against the palate in the concavity formed between rugae as the tongue again moves forward beneath it. This ratchet-like, pull-back mechanism may continue for several low-amplitude cycles at the beginning of the feeding sequence (see Hiiemae and Crompton, 1985; Franks et ah, 1985; German et al, 1989). In hyrax (Franks et al, 1985), the food may not be in contact with the rugae as the tongue protracts and retracts. The same principle is used in lapping (cats, opossum, and other species), although in this case the tongue is protruded into the liquid and rapidly retracted, carrying an aliquot of fluid on its surface. The posterior three-quarters of the tongue continue to retract, carrying it distally, while the anterior quarter remains a constant length, cradling the aliquot. Figure 14-11 in Hiiemae and Crompton (1985) details this process. The number of cycles needed to move food from the incisors to the postcanines appears correlated with snout (and, therefore, tongue) length. Even in the comparatively short-snouted macaque, two or three low-amplitude cycles are used (German et al, 1989). Little attention has been paid to the possibility of stage I transport in humans, probably because the shape of the dental arcade suggested that minimal distal transport would be needed. In fact (Hiiemae et al, 1996; manuscript in preparation), a distinct transport mechanism has been observed. After ingestion, the jaw is opened wide and the tongue surface is depressed below the occlusal plane of the lower molars. The food rests on the tongue surface even though it is posteriorly heaped. The body of the tongue is then sharply retracted with the hyoid moving both downward and backward. This retraction, or pull back, occurs while the jaws are held open. As they close, the tongue begins to rise and the food is brought into contact with the molars by elevation and axial rotation of the tongue to "tip" the bite onto the working side teeth. The whole process takes about 280 msec. 3.
Manipulation
The cycling movements of the tongue described earlier (Fig. 13.14) reposition food within the oral cavity by carrying material backward and downward (FO and FC phases), forward and upward, and forward and downward (IP and SO phases). This appears to be a common pattern in all mammals studied (see Fig. 1 4 13 in Hiiemae and Crompton, 1985; Hiiemae et al, 1995; Palmer et al, 1997). In many lateral projection CFG or VFG records of
feeding sequences, occasional attenuated cycles with long opening phases are seen in which the teeth do not reach occlusion. Such a cycle is shown (heavy arrow) in Figure 13.10. These "manipulation cycles" have been interpreted as associated with intraoral food positioning or side changing. In the case of the cycle shown in Fig. 13.10, the manipulation involved a complex tongue "sweeping" movement, which collected triturated material from the anterior hard palate where it had accumulated as a function of the forward thrusts of the tongue during preceding cycles (Fig. 13.14). The intriguing questions are (1) how does the tongue reposition inadequately triturated food on the postcanine occlusal table and (2) segregate triturated material from that requiring further processing? Dorsoventral CFG or VFG records are more difficult to interpret than those taken in the lateral projection. Cortopassi and Muhl (1990) carefully examined tongue behavior in the rabbit in the D-V projection. Rabbits have an intermolar eminence on the tongue, which Ardran and Kemp (1958) had argued prevented the movement of food across the midline. Cortopassi and Muhl (1990) found that the tongue, particularly the intermolar eminence, twisted toward the working side and suggest that this movement functions to keep food between the cheek teeth. Posteroanterior projection VFG in humans also shows rotation of the tongue, as measured by the change in shape and marker position, about its anteroposterior long axis. Originally described by Malik (1955), this movement certainly repositions food onto the working side occlusal table. However, work in progress (Palmer, Hiiemae, Mioche) suggests that, at least in humans, natural sized bites of solid food are managed in an unanticipated way. After the initial one or two puncture-crushing cycles, the bite is separated into two or sometimes three "lumps." All but one, which is retained on the working side, are transferred to the balancing side and held primarily in the vestibule on that side. The tongue affects this transfer. When the lump on the working side is reduced, the triturated product is moved posteriorly for stage II transport and bolus formation. The remaining lumps are then chewed seriatim. This new finding goes a long way in explaining the pattern of intermittent swallows during feeding sequences in humans (Hiiemae et al, 1996) and provides a simple explanation for intraoral segregation of triturated and barely triturated material. There is some consensus that mammals with isognathous postcanine occlusions, e.g., rodents, could either chew on both sides of the mouth simultaneously or at least use both occlusal planes alternately (Fig. 13.8; Byrd, 1981). It may be that many anisognathous mammals eating softer foods use a combination of tongue-palate compression coupled with a chewing
443
13. Feeding in Mammals stroke on one side. The net effect, at least for food such as bananas in humans, is to spread the material to both sides of the mouth, with incidental balancing side compression as the working side teeth leave occlusion and the jaw moves medially. 4. Stage 11 Transport Liquids and triturated (swallowable) food are moved posteriorly through the fauces, either for oropharyngeal bolus formation or as the first stage of the actual swallow. Two distinct stage II transport mechanisms have been identified in mammals for which CFG or VFG data are available. In primitive mammals, e.g., Didelphis, triturated food is accumulated in a hollow on the posterior surface of the tongue, anterior to the soft palate-tongue seal. In SO (Ol, 02) the tongue expands around this aggregate, compressing it against the hard palate. As the jaws open in FO (03), the middle part of the tongue rises in front of the material, cradling it within a surface depression while carrying it downward and backward away from the hard palate. At the same time, the soft palate flattens out, opening up the fauces. As the jaws begin to close in FC, the tongue rises and continues to move backward, pushing the food through the fauces into the oropharynx. We described this process as a ''squeeze wedge" (for a detailed description, see Hiiemae and Crompton, 1985). In the opossum, several stage II transport cycles are needed to form a bolus in the valleculae and piriform fossae, with the number of cycles depending on the food. The mechanism of stage II transport in macaque and human is rather different in that it uses a different part of the tongue movement cycle to propel material through the fauces. In both anthropoid species, the forward movement of the tongue during IP and SO (Ol) creates a contact zone between the tongue surface and the hard palate, which travels backward along the tongue as the tongue surface rises and moves anteriorly (see Fig. 13.14). Because the contact zone moves progressively backward, food between the tongue and the hard palate is squeezed posteriorly along the tongue. The triturated food is then pushed through the (opened) fauces into the oropharynx. It is important to note that the description of this process in human (Hiiemae and Crompton, 1985) is only partially correct. Liquid boli are assembled within the oral cavity, propelled through the fauces, and, by a sequence of highly coordinated tongue, epiglottal, and pharyngeal mechanisms, cross the oropharynx and enter the opened esophageal aditus (Fig. 13.3). The same can occur for soft or semiliquid foods. However, when natural-sized bites of soft or hard foods are consumed, several stage II transport cycles (small ar-
rows. Fig. 13.10) transfer material to the oropharynx, where the bolus is assembled (Hiiemae and Palmer, 1999; see earlier discussion). 5. Deglutition
(Swallowing)
The major issues involved in mammalian swallowing have already been addressed (e.g., in discussion of the oropharynx), albeit parenthetically. The mechanisms involved, and their control, have been exhaustively reviewed by Thexton and Crompton (1998). In part because of the clinical significance of swallowing in human, it has received more attention from the full panoply of basic and clinical scientists than any other component of the feeding process. They have used animal and human subjects in hundreds of studies. Given this background, and the availability of a very current review, the following treatment only briefly summarizes the process. Importantly, swallow cycles in all mammals studied are "intercalated'' among processing and transport cycles (in contrast to nonmammalian tetrapods; see Chapter 2). The IP and SO (Ol) phases of the chewing cycle are prolonged, but once the bolus has started to enter the esophagus, cyclical jaw movement resumes. The only exception to this, in feeding, is the terminal swallow, when the last "bits" of food are gathered and swallowed, clearing the mouth. The jaws then return to the resting position. The actual swallow, i.e., bolus propulsion from the oropharynx (hypopharynx in human), involves (1) upward and forward movement of the hyoid, which assists in opening the upper esophageal sphincter; (2) depression of the epiglottis; (3) a propulsive, posterior movement of the oropharyngeal surface of the tongue; and (4) a peristaltic wave of contraction in the pharyngeal constrictors beginning above, and traveling toward, the esophageal aditus. In mammals for which we have data, the bolus is propelled from the valleculae and piriform fossae. In the macaque, the process may start in the oral cavity with a stage II transport movement, which continues seamlessly into the pharyngeal process. The same may be the case in human, where the movement of the soft palate to close off the nasopharynx and the depression of the epiglottis are clear indications of a swallow. When the bolus has formed in the oropharynx, the same pattern of tongue, palate, and epiglottal movement is seen. There has been discussion as to whether, in human, swallowing is a truly "active" process or is at least assisted by gravity (see Palmer, 1998). This again illustrates the dichotomy between comparative and clinically focused approaches! In almost all mammals, the orientation of the oral cavity and the oropharynx is much as shown for Didelphis in Fig. 13.13. The bolus
444
K a r e n M. H i i e m a e
has to be propelled posteriorly, and somewhat ventrally. In megachiropteran bats, the bolus has to be propelled upward, given its "upside-down" feeding posture (de Greet and de Vree, 1984). In none of these mammals could gravity assist deglutition! However, to test this assertion in humans. Palmer (1998) had human subjects feed and swallow in the normal, upright position and then on "all fours," simulating the general mammalian posture. The process was the same regardless of posture. It is not unreasonable to conclude that the highly "programmed" or CPG-controUed process of deglutition is much the same in all mammals.
VII. CONTROL OF FEEDING BEHAVIORS The feeding process involves a variety of behaviors: some rhythmic (chewing, probably gnawing), some apparently "packaged" (swallowing), and others dictated by instantaneous circumstance. Clearly, the CNS control mechanisms involved in behaviors such as the capture of live prey by felids are very different from those required for the management of gum chewing in humans—an apparently absurd comparison, but one that highlights the issues: the former behavior requires the integration of the whole body's sensorimotor systems, whereas the latter is a mechanistic, virtually subconscious, use of the oropharyngeal complex's capacity for subcortically maintained rhythmic behavior. While the neurophysiology of mastication (chewing sensu strictu) is reasonably well understood, the linkage between rhythmic tongue movement and jaw movement is not. Similarly, Jean (1984, 1990) and others have argued for the existence of a swallowing CPG, but how this would be triggered in the IP phase of a chewing cycle is not clear. The source(s) of the sensory inputs regulating the complex tongue movements involved in feeding, beyond the circuits for the special senses, is also a matter of speculation. How is swallowing triggered? Clearly, the argument, for humans, that the trigger site is pillars of the fauces cannot apply to boli, which have accumulated in the oropharynx. As for other species, we know even less. What we do know beyond any doubt is that brain stem lesions can disrupt these behaviors, especially the ability to swallow. Far more research is needed before we will have a full understanding of the central control of this vitally important process. Acknowledgments I especially thank Drs. Susan Herring and Kurt Schwenk for reading this manuscript and for their helpful advice and suggestions. Drs. Allan Thexton
and Jeffrey Palmer also reviewed earlier drafts. Much of the research reported here was funded by a series of USPHS NIH Awards. Continuing support from the National Institute for Dental Research and the National Institute for Communications Disorders and Stroke has made possible many of the experimental studies of the last 20 years. References Anapol, R (1988) Morphological and videographic study of the hyoid apparatus and its function in the rabbit (Oryctolagus cuniculus) J. Morphol. 195:141-157. Ardran, G., F. Kemp et al. (1958) A radiographic analysis of mastication and swallowing in the domestic rabbit {Oryctolagus cuniculus L.) Proc. Zool. Soc. Lond. 130:257-274. Arendsen, d. W-E. (1951) On differences in the lower jaw of animalivorous and herbivorous mammals. Proc. Koninklijke Nederlandse Akad. van Wetenschappen Ser. C 54:237-246, 405-410. Arensburg, B. (1989) The hyoid bone from the Kebara 2 hominid. Investigation in South Levantine Prehistory. O. B-Y. a. B. Vandermeersch. Oxford, BAR International Series. No. 497:337-342. Arensburg, B., and A. M. Tillier (1991) Speech and the Neanderthals. Endeavour 15:26-28. Arvas, T., K. M. Hiiemae, J. B. Palmer, and L. Mioche (1999) How mastication affects respiration. J. Dent. Res. Special Issue lADR/ AADR/CADR Meeting, March 1999, Abstract. Atchley, W. R. (1993) Genetic and developmental aspects of variability in the mammalian mandible. In: The Skull, Vol. 1. J. Hanken and B. K. Hall (eds.). University of Chicago Press, Chicago. Avis, A. (1961) The significance of the angle of the mandible: an experimental and comparative study. Am. J. Phys. Anthropol. 19: 55-61. Basmajian, J., and G. Stecko (1957) A new bipolar electrode for electromyography. J. Appl. Physiol. 17:849. Becht, G. (1953) Comparative biological-anatomical researches on mastication in some mammals. I and II. Proc. Netherlands Acad. Sci. Ser. C 56:508-527. Beecher, R. (1979) Functional significance of the mandibular symphysis. J. Morphol. 159:117-130. Bernard, C. (1858) Legons de Physiologie Experimentale. Appliquee a la Medecine. Paris, J. B. Bailliere. Birt, P, L. S. Hall, and G. C. Smith (1997) Ecomorphology of the tongues of Australian Megachiroptera (Chiroptera: Pteropodidae). Aust. J. Zool. 45:369-384. Bramble, D. (1980) Feeding in tortoises and mammals: why so similar? Am. Zool. 20:931. Bramble, D., and D. Wake (1985) Feeding mechanisms of lower vertebrates. Pp. 230-261. In: Functional Vertebrate Morphology. M. Hildebrand, D. Bramble, D. Wake, and K. Liem (eds.). Belknapp Press of Harvard Univ. Press, Cambridge, MA. Bramble, D. M. (1978) Origin of the mammalian feeding complex: models and mechanisms. Paleobiology 4:271-301. Bremer, F. (1923) Physiologie nerveuse de la mastication chez la chat et lapin. Arch. Intl. Physiol. 21:309-352. Brylski, P., and J. Patton (1988) Ontogeny of a macroevolutionary phenotype: the external cheek pouches of geomyoid rodents. Evolution 42:391-394 Butler, P. (1952) Some functional aspects of molar evolution. Evolution 26:474-483. Byrd, K. (1981) Mandibular movement and muscle activity during
13. F e e d i n g in M a m m a l s mastication in the guinea pig (Cavia porcellus). J. Morphol. 170: 147-169. Clark, W. Le. Gros. (1959) The Antecedents of Man. The University Press, Edinburgh. Cortopassi, D., and Z. Muhl (1990) Videofluorographic analysis of tongue movement in the rabbit (Oryctolagus cuniculus). J. Morphol. 204:139-146. Crompton, A. (1971) The origin of the tribosphenic molar. In Early Mammals. D. Kermack and K. Kermack (eds.). Academic Press, London. Crompton, A. (1995) Masticatory function in nonmammalian cynodonts and early mammals. Pp 55-75. Functional Morphology in Vertebrate Paleontology. J. Thomason (ed.). Cambridge, Cambridge Univ. Press, Cambridge. Crompton, A., and K. Hiiemae (1969) Functional occlusion in tribosphenic molars. Nature 222:678-679. Crompton, A., and K. Hiiemae (1970) Molar occlusion and mandibular movements during occlusion, Didelphis marsupialis L. Zool. J. Linn. Soc. 49:21-47. Crompton, A., A. Thexton et al. (1977) The activity of the hyoid and jaw muscles during chewing of soft food in the opossum. Pp. 287-305. The Biology of Marsupials, Vol. 2. B. Stonehouse and D. Gillmore, MacMillan, London. Dantas, R., W. Dodds et al. (1990) Manometric characteristics of the glossopalatal sphincter. Digest. Dis. Sci. 35:161-166. Davis, D. D. (1964) The giant panda: a morphological study of evolutionary mechanisms. Fieldiana: Zool. Mem. 3:1-339. Dellow, P., and J. Lund (1971) Evidence for the central timing of rhythmical mastication. J. Physiol. (Lond.) 215:1-13. De Vree, P., and C. Cans (1976) Mastication in pygmy goats {Capra hircus). Ann. Roy. Belg. Zool. Soc. 105:255-306. Dobson, G. (1882) On the digastric muscle, its modifications and functions. Trans. Linn. Soc. Lond. Zool. Ser. 2:259-264. Doran, G. (1975) Review of the evolution and phylogeny of the mammalian tongue. Acta Anat. 91:118-129. Doran, G., and H. Baggett (1971) A structural and functional classification of mammalian tongues. J. Mammal. 52:427-429. Edgeworth, R (1911) On the morphology of the cranial muscles in some vertebrates. Quart. J. Microsc. Sci. 56:167-316 (N). Edgeworth, F. (1914) On the development and morphology of the mandibular and hyoid muscles of mammals. Quart. J. Microsc. Sci. 59:573-645 (N). Ewer, R. (1958) Adaptive features of the skulls of African Suidae. Proc. Zool. Soc. Lond. 131:135-155. Fearnhead, R. W., C. C. D. Shute et al. (1955) The temporo-mandibular joint of shrews. Proc. Zool. Soc. Lond. 125:795-806. Fortelius, M. (1985) Ungulate cheek teeth: developmental, functional and evolutionary interrelations. Acta Zool. Perm. 180:1-76. Franks, H. A., A. W. Crompton, and R. Z. German (1984) Mechanism of intraoral food transport in macaques. Am. J. Phys. Anthro. 65: 275-282. Franks, H. A., R. Z. German, A. W. Crompton, and K. M. Hiiemae (1985) Mechanism of intra-oral transport in a herbivore, the hyrax {Procavia syriacus). Arch. Oral Biol. 30:539-544. Garrett, J., and G. Proctor (1998) Control of salivation. Pp. 135-155. In: The Scientific Basis of Eating. R. W. A. Linden (ed.). Karger, Basel. German, R., and H. Franks (1991) Timing in the movements of the jaws, tongue and hyoid during feeding in the hyrax {Procavia syriacus). J. Exp. Zool. 257:34-42. German, R., S. Saxe, A. W. Crompton, and K. M. Hiiemae (1989) Mechanism of food movement through the anterior oral cavity in anthropoid primates. Am. J. Phys. Anthropol. 80:765-775. Gilbertson, T. (1998) Peripheral mechanisms of taste. Pp. 1-28. In: The Scientific Basis of Eating. R. W. A. Linden (ed.). Karger, Basel.
445
Gorniak, G. (1977) Feeding in golden hamsters, Mesocritus auratus. J. Morphol. 154:427-458. Grays Anatomy (1995) Churchill Livingstone, London. Greaves, W. (1974) Functional implications of mammalian jaw joint position. Form. Funct. 7:363-376. Greaves, W. (1978) The jaw lever system in ungulates: a new model. J. Zool. 184:271-278. Greaves, W. (1988) The maximum average bite force for a given jaw length. J. Zool. 214:295-306. Greaves, W. (1995) Functional predictions from theoretical models of the skull and jaws in reptiles and mammals. Pp. 99-115. In: Functional Morphology in Vertebrate Paleontology. P. Thomason (ed.). Cambridge Univ. Press, Cambridge. Gueldre, G. de., and F. de. Vree (1984) Movements of the mandible and tongue during mastication and swallowing in Pteropus giganteus (Megachiroptera). J. Morphol. 179:95-114. Hamlett, S. (1989) Dynamic aspects of lingual propulsive activity in swallowing. Dysphagia 4:136-145. Hart, R. T. V, V. V. Hennebel, N. Throngpreda, W. C. van Buskirk, and R. C. Anderson (1992) Modelling the biomechanics of the mandible: a three dimensional finite element study. J. Biomech. 25: 262-268. Heath, M. (1991) The basic mechanics of mastication: man's adaptive success. Pp. 144-166. In: Feeding and the Texture of Food. J. Vincent and P. Lillford (eds.) Cambridge Univ. Press, Cambridge. Hector, M., and R. Linden (1987) The possible roles of periodontal mechanoreceptors in the control of parotid secretion in man. Quart. J. Exp. Physiol. 72:285-301. Hedges, S. B., P H. Parker, C. G. Sibley, and S. Kumar (1996) Continental breakup and the ordinal diversification of birds and mammals. Nature 381:226-229 Hellstrand, E. (1980) Morphological and histochemical properties of tongue muscles in cat. Acta Physiol. Scand. 110:187-198. Hellstrand, E. (1981) Contraction tinies of the cat's tongue muscles measured by light reflection: innervation of individual tongue muscles. Acta Physiol. Scand. Ill:417-423. Herring, S. (1992) Muscles of mastication: architecture and functional organisation. Pp. 541-548. In: The Biological Mechanisms of Tooth Movement and Craniofacial Adaptation. Z. Davidovitch (ed.). Ohio State University, Columbus, OH. Herring, S. W. (1993) Functional morphology of mammalian mastication. Am. Zool. 33:289-299. Herring, S. (1994) Functional properties of the feeding musculature. Pp. 5-30. In: Biomechanics of Feeding in Vertebrates, Vol. 18. V. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Herring, S. W., and R. P. Scapino (1973) Physiology of feeding in miniature pigs. J. Morphol. 141:427-460. Hiiemae, K. (1967) Masticatory function in mammals. J. Dent. Res. 46:883-893. Hiiemae, K. (1976) Masticatory movements in primitive mammals. Pp. 105-117. In: Mastication. D. Anderson and B. Matthew (eds.). John Wright, Bristol. Hiiemae, K. (1978) Mammalian mastication: a review of the activity of the jaw muscles and the movements they produce in chewing. Pp. 359-398. In: Development, Function and Evolution of Teeth. P. Butler and K. Joysey (eds.). Academic Press, London. Hiiemae, K. (1993) Process and mechanism: mechanoreceptors in the mouth as primary modulators of rhythmic behaviors in feeding? Pp. 263-284. In: Sensory Research: Multimodal Perspectives. R. Verillo (ed.). Lawrence Erlbaum, Hillsdale, NJ. Hiiemae, K., and G. Ardran (1968) Mandibular movement in the rat: a cineradiographic study of mastication in Rattus norvegicus (L). J. Zool. 154:139-154. Hiiemae, K., and A. Crompton (1971) A cinefluorographic study of feeding in in the American opossum (Didelphiis marsupialis L.)
446
K a r e n M. H i i e m a e
Pp. 299-334. In: Dental Morphol and Evolution. A. Dahlberg (ed.). University of Chicago Press, Chicago. Hiiemae, K., and A. Crompton (1985) Mastication, food transport and swallowing. Pp. 262-290. In: Functional Vertebrate Morphology. M. Hildebrand, D. Bramble, K. Liem, and D. Wake (eds.). Belknapp Press, Harvard Univ. Press, Cambridge MA. Hiiemae, K., M. R. Heath, G. Heath, E. Kazazoglu, J. Murrray, D. Sapper, and K. Hamblett (1996) Natural bites, food consistency and feeding behavior in man. Arch. Oral Biol. 41:175-189. Hiiemae, K., and F. Jenkins (1969) The anatomy and internal architecture of the muscles of mastication in Didelphis marsupialis L. Postilla 140:1-49. Hiiemae, K., and R. Kay (1972) Trends in the evolution of primate mastication. Nature 240:486-487. Hiiemae, K., A. Reese, and S. Hayenga (1995) Patterns of tongue and jaw movement: a cinefluorographic study of feeding in the macaque. Arch. Oral Biol. 40:229-246. Hiiemae, K. M., and J. B. Palmer (1999) Food transport and bolus formation during complete feeding sequences on foods of different initial consistency. Dysphagia 14:31-42. Hiiemae, K. M., and J. B. Palmer (2000) Tongue-jaw linkages in feeding. Bull. Mus. Comp. Zool. Harvard University. Hiiemae, K. M., A. J. Thexton, and A. W. Crompton (1978) Intra-oral food transport: a fundamental mechanism of feeding? Pp. 181208. In: Muscle Adaptation in the Craniofacial Region, Vol. 8. D. Carlson and J. McNamara (eds.). University of Michigan, Ann Arbor. Hiiemae, K. M., A. J. Thexton et al. (1981) The movement of the cat hyoid during feeding. Arch. Oral Biol. 26:65-81. Hiiemae, K. M., S. W. Medicis, J. B. Palmer, B. S. Jackson, and J. Hegener (1999) How we can eat and speak: hyoid movements in feeding and speech are different. Am. J. Phys. Anthropol. Suppl. 28:151. (Abstract) Hilloowala, R. (1975) Comparative anatomical study of the hyoid apparatus in selected primates. Am. J. Anat. 142:367-384. Honeycutt, R. L., and R. M. Adkins (1993) Higher level systematics of eutherian mammals: an assessment of molecular characters and phylogenetic hypotheses. Annu. Rev. Ecol. System. 24:279305. Hylander, W. (1977) In vivo bone strain in the mandible of Galago crassicaudatus. Am. J. Phys. Anthropol. 46:309-326. Hylander, W. (1979) Mandibular function in Galago crassicaudatus and Macaca fascicularis: an in vivo approach to stress analysis of the mandible. J. Morphol. 159:253-296. Hylander, W. (1984) Stress and strain in the mandibular symphysis of primates: a test of competing hypotheses. Am. J. Phys. Anthropol. 64:1-46. Hylander, W. L., K. R. Johnson, and A. W. Crompton (1992) Muscle force recruitment and biomechanical modelling: an analysis of masseter muscle function during mastication in Macaca fascicularis. Am. J. Phys. Anthropol. 72:287-314 James, W. (1960) The Jaws and Teeth of Primates. Pitman Medical Publishing, London. Janis, C. (1979) Mastication in the hyrax and its relevance to ungulate dental evolution. Paleobiology 5:50-59. Janis, C. (1995) Correlations between craniodental morphology and feeding behavior in ungulates: reciprocal illumination between living and fossil taxa. Pp. 76-98. In: Functional Morphology in Vertebrate Paleontology. T. Thomason (ed.). Cambridge Univ. Press,
Jean, A. (1990) Brainstem control of swallowing. Pp. 294-321. In: Neurophysiology of the Jaws and Teeth. A. Taylor (ed.). MacMillan, London. Jeronimidis, G. (1991) Mechanical and fracture properties of cellular and fibrous materials. Pp. 1-17. In: Feeding and the Texture of Food. J. Vincent and P. Lillford (eds.). Cambridge University Press, Cambridge. Jongh, H. de., R. Dantuma, and H. J. M. Sluijsmans (1989) The shape of the mandible in the domestic sheep: a biomechanical analysis using EMG as an estimator of muscle force. Acta Morphol. NeerlScand. 27:63-73. Kallen, F , and C. Cans (1972) Mastication in the little brown bat, Myotis lucifugus. J. Morphol. 136:385-420. Kay, R. (1975) The functional adaptations of primate molar teeth. Am. J. Phys. Anthropol. 43:195-216. Kay, R., and K. Hiiemae (1974) Jaw movement and tooth use in recent and fossil primates. Am. J. Phys. Anthropol. 40:227-256. Kay, R., and W. Hylander (1978) The dental structure of mammalian folivores with special reference to Primates and Phalangeroidea. Pp. 173-191. In: The Ecology of Arboreal Folivores. G. Montgomery (ed.). Smithsonian Institution Press, Washington DC. Kier, W., and K. Smith (1985) Tongues, tentacles and trunks: the biomechanics of movement in muscular hydrostats. Zool. J. Linn. Soc. 83:307-324. Laitman, J., J. Reidenberg et al. (1990) The Kebara hyoid: what can it tell us about the evolution of the hominid vocal tract? Am. J. Phys. Anthropol. 81:254. Larson, J., and S. Herring (1996) Movement of the epiglottis in mammals. Am. J. Phys. Anthropol. 100:71-82. Lavigne, G., J. Kim et al. (1987) Evidence that periodontal pressoreceptors provide positive feedback to jaw closing muscles during changes in the pattern brought about by objects between the teeth. J. Neurophysiol. 58:342-358. Lieberman, P. (1992) On Neanderthal speech and Neanderthal extinction. Curr. Anthropol. 33:409-410. Lieberman, P., J. Laitman et al. (1992) The anatomy, physiology, acoustics and perception of speech: essential elements in analysis of the evolution of human speech. J. Hum. Evol. 23:447-467. Livingstone, R. (1956) Some observations on the natural history of the tongue. Ann. Roy. College Surg. Engl. 19:185-200. Lord, F. (1913) Observations on the temporo-mandibular articulation. Anat. Rec. 7:355-367. Lucas, P. (1982) Basic principles of tooth design. Pp. 154-162. In: Teeth: Form, Function and Evolution. B. Kurten (ed.). Columbia Univ. Press, New York. Lucas, P., and R. Corlett (1991) Quantitative aspects of the relationship between dentitions and diets. Pp. 93-121. In: Feeding and the Texture of Food, Vol. 44. J. Vincent and P. Lillford (eds.). Cambridge Univ. Press, Cambridge. Lucas, P., and D. Luke (1984) Optimum mouthful for food comminution in human mastication. Arch. Oral Biol. 29:205-210. Ludwig, C. (1850) Neue Versuche iiber die Beihiilfe der Nerven zu der Speichelsekreiten. Naturforsch Ges Zurich 53/54:210-239. Magendie, F. (1825) Precis Elementaire de Physiologic. MequignonMarvis, Paris. Malik, A. E. (1955) The part played by the tongue in mastication and deglutition. J. Anat. 89:250-254. Maynard Smith, J., and R. Savage (1959) The mechanics of mamma-
13. F e e d i n g in M a m m a l s Mills, J. (1966) The functional occlusion of the teeth of Insectivora. J. Linn. Soc. 46:1-25. Mioche, L., and J. Martin (1998) Training and sensory judgement effects on mastication as studied by electromyography. J. Food Sci. 63:1-5. Naples, V. (1985) Form and function of the masticatory musculatures in the tree sloths, Bradypus and Choloepus. J. Morphol. 183:25-50. Naples, V. (1986) The morphology and function of the hyoid region in the tree sloths, Bradypus and Choloepus. J. Mammal. 67:712724. Novacek, M. (1992) Mammalian phylogeny: shaking the tree. Nature 356:121-125 Orchardson, R., and S. Cadden (1998) Mastication. Pp. 76-121. In: The Scientific Basis of Eating. R. W. A. Linden (ed.). Karger, Basel. Oron, U., and A. Crompton (1985) A cineradiographic and electromyographic study of mastication in Tenrec ecaudatus. J. Morphol. 185:155-182. Palmer, J. (1998) Bolus aggregation in the oropharynx does not depend on gravity. Arch. Phvs. Med. Rehabilitation 79:691-696. Palmer, J., K. M. Hiiemae, and J. Lui (1997) Tongue-jaw linkages in feeding: a preliminary videofluorographic study. Arch. Oral Biol. 42: 429- 441. Palmer, J., N. Rudin, G. Lara, and A. W. Crompton (1992) Coordination of mastication and swallowing. Dysphagia 7:187-200. Pavlov, I. (1910) Lectures on the Work qf the Digestive Glands. Charles Griffen, London. Peyer, B. (1968) Comparative Odontology. University of Chicago Press, Chicago. Prinz, J., and P. Lucas (1997) An optimization model for mastication and swallowing in mammals. Proc. Roy. Soc. Lond. Ser. B 264: 1715-1721. Purslow, P. (1991) Measuring meat texture and understanding its structural basis. Pp. 35-56. In: Feeding and the Texture qf Food. J. Vincent and P. Lillford (eds.). Cambridge Univ. Press, Cambridge. Radinsky, L. (1981a) Evolution of skull shape in carnivores. 1. Representative modern carnivores. Biol. J. Linn. Soc. 15:369-388. Radinsky, L. (1981b) Evolution of skull shape in carnivores. 2. Additional modern carnivores. Biol. J. Linn. Soc. 16:337-355. Radinsky, L. (1982) Evolution of skull shape in carnivores. 3. The origin and early radiation of modern carnivores. Paleobiology 8: 177-195. Reidenberg, J., and J. Laitman (1994) Anatomy of the hyoid apparatus in Odentoceti (toothed whales): specializations of their skeleton and musculature compared with those of terrestrial mammals. Anat. Rec. 240: 598- 624. Rensberger, J. (1973) An occlusal model for mastication and dental wear in herbivorous mammals. J. Paleontol. 47:515-528. Rensberger, J. (1978) Scanning electron microscopy of wear and occlusal events in some small herbivores. Pp. 415-438. In: Development, Function and Evolution of Teeth. P. Butler and K. Joysey (eds.). Academic Press, London. Rensberger, J. (1995) Determination of stresses in mammalian dental enamel and their relevance to the interpretation of feeding behaviors in extinct taxa. Pp. 151-172. In: FunctionaI Morphol. in Vertebrate Paleontology. J. Thomason (ed.). Cambridge Univ. Press, Cambridge. Romer, A. (1974) Vertebrate Paleontology. University of Chicago Press, Chicago. Scapino, R. (1.972) Adaptive radiation of mammalian jaws. Pp. 3339. In: Morphology of the Maxillo-Mandibu]ar Apparatus. G. Schumacher (ed.). Leipzig, G. Thieme. IX International Congress of Anatomists. Schwenk, K. (2000a) Intrinsic versus extrinsic lingual muscles: a false dichotomy? Bull. Mus. Comp. Zool. Harvard Univ.
447
Schwenk, K. (2000b) Functional units and their evolution. In: The Character Concept in Evolutionary Biology. G. P. Wagner (ed.). Academic Press, San Diego. Schwenk, K., E. Brainerd, and A. W. Crompton (1989) Biomechanics of the mammal tongue: role of the lingual tunic during hydrostatic deformation. Am. Zoot. 29:35A. Sher L., K. M. Hiiemae, and J. B. Palmer (1997) Activity in the suprahyoid muscles of Macaca fascicularis during normal feeding. J. Dent. Res. Special Issue 77. (Abstract # 1512) Sibbing, F. (1991) Food processing by mastication in cyprinid fish. Pp. 57-92. In: Feeding and the Texture of Food. J. Vincent and P. Lillford (eds.). Cambridge Univ. Press, Cambridge. Smith, K. (1978) Electromyographic and cineradiographic studies of feeding in lizards. Am. Zool. 18 : 623. Smith, K. (1992) The evolution of the mammalian pharynx. Zool. J. Linn. Soc. 104: 313- 349. Smith, K. (1994) Are neuromotor systems conserved in evolution? Brain Behav. Evol. 43:293-305. Smith, K. K., and W. M. Kier (1989) Trunks, tongues and tentacles: moving with skeletons of muscle. Am. Sci. 77:29-35. Spears, I., and G. Macho (1998) Functional adaptations of modern human molars: implications for the fossil record. Am. J. Phys. Anthropol. 106: 467-482. Springer, M. S., G. C. Cleven, O. Madsen, W. W. de Jong, V. G. Waddell, H. M. Amrine, and M. J. Stanhope (1997) Endemic African mammals shake the phylogenetic tree. Nature 388:61-64. Taylor, A. (1990) Proprioceptive control of jaw movement. Pp. 237267. In: Neurophysiology of the Jaws and Teeth. A. Taylor (ed.). Macmillan, London. Teaford, M., and J. Runestad (1992) Dental microwear and diet in Venezuelan primates. Am. J. Phys. Anthropol. 88:347-364. Teaford, M., and A. Walker (1984) Quantitative differences in dental microwear between primate species with different diets and a comment on the presumed diet of Sivapithecus. Am. J. Phys. Anthropol. 64:191-200. Thexton, A. J. (1982) Mastication and swallowing: an overview. Br. Dent. J. 173:197-206. Thexton, A., and A. Crompton (1989) Effect of sensory input from the tongue on jaw movement in normal feeding in the opossum. J. Exp. Zool. 250:233-243. Thexton, A., and A. Crompton (1998) The control of swallowing. Pp. 168-222. In: The Scientific Basis of Eating. R. W. A. Linden (ed.). Karger, Basel. Thexton, A., and K. Hiiemae (1975) The twitch-tension characteristics of opossum jaw musculature. Arch. Oral Biol. 20:743-748. Thexton, A., and K. Hiiemae (1997) The effect of food consistency on jaw movement in the macaque: a cineradiographic study. J. Dent. Res. 76: 552-560. Thexton, A., K. Hiiemae et aI. (1980) Food consistency and bite size as regulators of jaw movement during feeding in the cat. J. Neurophysiol. 44: 456-474. Thexton, A., and J. McGarrick (1988) Tongue movement of the cat during lapping. Arch. Oral Biol. 33:331-339. Thexton, A., and J. D. McGarrick (1989) Tongue movement in the cat during the intake of solid food. Arch. Oral Biol. 34:239-248. Thexton, A., J. McGarrick, K. M. Hiiemae, and A. W. Crompton (1982) Hyo-mandibular relationships during feeding in the cat. Arch. Oral Biol. 27:793-801. Turnbull, W. (1970) Mammalian masticatory apparatus. Fieldiana: Geology 18(2): 153-356. van der Klaauw, C. J. (1945) Cerebral skull and facial skull: a contribution to the knowledge of skull-structure. Arch. Neerl. Zool. 7: 16-37. Vincent, J. (1991) Textures of plants and fruits. Pp. 19-33. In: Feeding
448
Karen M. Hiiemae
and the Texture of Foods. J. Vincent and P. Lillford (eds.). Cambridge Univ. Press, Cambridge. von Koenigswald, W. (1982) Zum verstandnis der Morphologie der Wiihlmausmolaren (Arvicolae, Rodentia, Mammalia). Zool. Geol. Wittenschaft Berlin 10:951-852. Weijs, W. (1994) Evolutionary approach of masticatory motor patterns in mammals. In: Biomechanics of Feeding in Vertebrates. V. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Weijs, W., and R. Dantuma (1975) Electromyography and mechanics of mastication in the albino rat. J. Morphol. 146:1-34.
Weijs, W., and R. Dantuma (1981) Functional anatomy of the masticatory apparatus in the rabbit {Oryctolagus cuniculus L.). Netherlands J. Zool. 31:99-147. Wilson, D., and R. Sullivan (1998) Peripheral mechanisms of smell. Pp. 29-39. In: The Scientific Basis of Eating, Vol. 9. R. W. A. Linden (ed.). Karger, Basal. Woods, C. (1975) The hyoid, laryngeal and pharyngeal regions of bathygerid and other selected rodents. J. Morphol. 147:229250.
C H A P T E R
14 The Ontogeny of Feeding in Mammals A. W. CROMPTON Museum of Comparative Zoology Harvard University Cambridge, Massachusetts 02138
R. Z. GERMAN Department of Biological Sciences University of Cincinnati Cincinnati, Ohio 45221
the laryngeal opening, both a bony hard and a muscular soft palate, and a large muscular tongue capable of complex motions. However, the significant differences in anatomy and the constraints suggest that the mechanisms used for feeding in infants will differ from those of adults (Ardran et al, 1958). Despite the differences between infant and adult feeding, and the significant variation in adult feeding mechanisms, there are many similarities in suckling across species of mammals. The transition to drinking from sources other than a teat and eating solid food, termed weaning, occurs at different points of development for different species, but presents similar functional challenges to the young animal. While there is a great deal known about suckling in infant humans from a clinical perspective (Logemann, 1983; Koenig et al, 1990; Jones and Conner, 1991), much less is known about variation among species of mammals, and even less about the evolution of suckling. This chapter begins with a review of the relevant morphology and then discusses what is known about the mechanics and function in a comparative framework. The next sections focuses on three related issues: the rythmicity of suckling; the coordination of respiration and swallowing in infants; and the maturation of feeding and the transition to eating solid food. The chapter concludes with some brief speculation on the evolution of suckling in mammals.
I. INTRODUCTION 11. MORPHOLOGY III. FUNCTION AND MECHANICS OF SUCKLING A. General Features B. Suckling in Infant Pigs C. Suckling in Infant Macaques D. Suckling in Infant Opossums IV. RHYTHMICITY AND CONTROL OF SUCKLING V. COORDINATION OF SWALLOWING AND RESPIRATION VI. TRANSITION FROM SUCKLING TO DRINKING AT WEANING VII. EVOLUTIONARY CONSIDERATIONS References
L INTRODUCTION One of the defining characteristics of mammals is the presence of specialized mammary glands in females that produce milk for feeding infants (Eisenberg, 1981). Infant mammals, including all marsupials and placentals, suckle on the maternal teat for some period of time after birth. Several constraints on suckling make it different from adult feeding: delivery from the teat and a purely liquid food source. These functional differences are associated with differences in infant anatomy, including lack of teeth, relative size of the tongue, and pharyngeal morphology (Bosma, 1985; Crelin, 1987; German et al, 1992). Both functional and morphological differences place constraints on feeding. Some aspects of basic mammalian oropharyngeal anatomy are constant through development, including relative posterior position of the esophageal opening to
FEEDING (K. Schwenk, ed.)
IL MORPHOLOGY Infant mammals are born at varying degrees of maturity, but all have the ability to suckle within hours, if
449
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
450
R. Z. German and A. W. Crompton
not minutes, of birth. Altricial animals, including marsupials, carnivores, and many rodents, are incapable of even immature locomotion and their eyes are not open. More precocious mammals, including ungulates, are alert and capable of locomotion immediately. However, most infant mammals are edentulous. Some taxa are born with teeth that are not functional for feeding. Infant pigs, for example, have "needle teeth" that are used for fighting with littermates (Fraser and Thompson, 1991). All infant mammals are characterized by heads and tongues that are large relative to adult proportions. The structure of the oral cavity and oropharynx differs from adults mainly in relative proportion, with a few additional differences. The viscerocranium, relative to the neurocranium, is smaller in infants than in adults. The face tends to be shorter in the anterior/ posterior dimension and the tongue larger (DuBrul, 1988). Rugae found on the roof of the mouth, critical for adult feeding, are either reduced or nonexistent in infants. Finally, the relative position of the airway to the esophagus is different in young mammals. In all infants the epiglottis is relatively high and is located internarially, i.e., reaching into the nasopharynx (Crelin, 1987; Crompton et ah, 1997). This provides a continuous, patent airway from the nares through to the larynx. This arrangement has significance for the relationship between respiration and swallowing, and whether the two processes can occur simultaneously in infants, but not adults.
III. FUNCTION A N D MECHANICS OF SUCKLING A. General Features The basic mechanism of suckling, in all infant mammals studied to date, is a pumping mechanism using dorsoventral movements of the tongue body. There is some debate concerning this mechanism. Early cineradiographic studies by Ardran and colleagues (1958) on infant lambs suggest that milk is obtained by expression, mechanical pressure on the teat by the tongue, and is accompanied by strong jaw movements. Ardran and Kemp (1959) also measured pressure changes in the nipple and oral cavity during suckling in infant lambs and humans to support their hypothesis that suckling was accomplished by positive pressure. They considered that suction, or relative negative pressure, was important only for holding the nipple in the oral cavity and refilling the artificial teat. One difficulty with these early results may have been a reliance on the outline of the milk containing barium to characterize
the tongue position; because of intense longitudinal furrowing, the outline corresponding to the middle of the tongue can be confused with the outline of the raised edges in the absence of markers (Thexton and McGarrick, 1988; German et al, 1992). More recent studies describe a different mechanism of suckling, but one that is consistent across a number of different species. For the initial acquisition of milk, similar mechanisms were found in marsupials (German and Crompton, 1996), ungulates and primates (German et al, 1992), and carnivores (unpublished data). In each of these species, suckling was characterized by a number of distinctive kinematic patterns of tongue movement. In all cases, the anterior portions of the tongue moved very little. Usually, the tip of the tongue was wrapped around the nipple. The middle of the tongue, lying beneath the hard palate and hard palate/soft palate junction, moved in a dorsal/ventral pathway, with little or no anterior/posterior movement. Posterior portions of the tongue tended to move in larger circular patterns. The orientation of this portion of the tongue is not anterior/posterior, but slopes downward. Thus, tongue movements are still orthogonal to the direction of milk flow. B. Suckling i n Infant Pigs The species with the most complete description of suckling is the domestic pig {Sus scrofa). Infant pigs weighed approximately 1.0-2.0 kg during the time these data were collected. This species uses a two-step mechanism, with each step involving pumping (German et al, 1992). In the first step, milk is pumped out of the nipple and into the oral cavity. The second step transports milk from the oral cavity and the oral pharynx and into the esophagus. Initially, the infant would wrap its tongue around the tip of the nipple, and the tip of the tongue curled around the nipple was visible outside the oral cavity. The tongue sealed against the hard palate and the anterior end of the soft palate (Fig. 14.1, frame 78). A small space existed in the anterior oral cavity, immediately posterior to the nipple. Between frames 80 and 87 the jaw opened slightly (Fig. 14.2), and the space between the tongue and the hard palate expanded as follows: (a) the middle of the tongue, near marker three, moved strongly downward and (b) the posterior contact between tongue and soft palate, as well as the positions of markers four and five, shifted in a posterior direction. These movements leave two seals, or potential seals, between tongue and palate: (i) an anterior seal between the hard palate and the anterior tongue and (ii) a posterior seal between the soft palate and the posterior tongue. Initially there was an increase in the space between the tongue and that
14. T h e O n t o g e n y of F e e d i n g in M a m m a l s
I^PPIQ
451
appearance on X-ray, suggesting air mixed in with the milk. By the time the jaw began to close and marker one (lying underneath the nipple) began to rise, most of the space between palate and tongue was filled with milk, and little, if any, more was seen to enter the cavity (Fig. 14.1, frame 92). The tongue reformed the anterior seal with the hard palate, just posterior to the nipple. The second stage began as the jaw opened between frames 92 and 98, and the middle of the tongue, near marker three, rose to contact the hard palate (Fig. 14.1, frame 98). The posterior tongue, near markers four and five, moved anteriorly and ventrally or downward (Fig. 14.2). This broke the posterior seal and formed a large space beneath the soft palate. At this time, the aliquot of liquid, which had been lying in the space between the anterior and the posterior seals underneath the hard palate, was moved through the pillars of the fauces and into the vallecular space above markers four and five (Fig. 14.1, frame 98). The jaw continued to open between frames 98 and 107 (Fig. 14.2). The posterior tongue markers moved upward and backward and the liquid moved out of the valleculae. By frame 113, the liquid had passed through the piriform recesses and into the esophagus. These swallows were regular, occurring in every second cycle. Rarely, a swallow occurred in the third cycle, but this only happened when little or no milk was acquired in one of the preceding cycles. C. Suckling in Infant Macaques
FIGURE 14.1. Lateral view of infant pig suckling by pumping mechanism. Each of five frames are taken from cine X-ray film of an infant pig sucking. Filming was at 100 fps, so each frame was 10 msec apart. Frame numbers are indicated at upper left. Anterior is to the right, posterior to the left. Five radio-opaque markers are in the body of the tongue and are numbered on the first frame. The pumping movements of the tongue are indicated with arrows and, given the seal of the tongue to the palate, produce a reduced pressure inside the oral cavity.
palate, which would imply the existence of a reduced pressure if the seals were complete. As marker three moved downward, the anterior seal broke in front of the nipple, and milk moved into the space between the middle tongue and the hard palate. Milk that entered the oral cavity as the space was created had a frothy
Suckling in macaques is also based on a dorsal/ventral pumping mechanism, although there were some differences from the infant pigs (German et ah, 1992). These animals were smaller than the infant pigs, weighing approximately 500 g. The position of the infant macaque tongue during suckling was similar to that of the infant pig, despite the macaque face and tongue being shorter in an anterior/posterior direction. The tongue was elevated and sealed against a large area of the hard and soft palates, leaving a small space in front of the nipple at the anterior end of the oral cavity (Fig. 14.3, frame 254). Over the next 500 msec the jaw opened and the middle of the tongue moved strongly downward, breaking its seal with the hard palate (Fig. 14.4). A small portion of the anterior tongue remained elevated without contacting the palate, leaving a smiall gape through which liquid flowed (Fig. 14.3, frame 261). The posterior of the tongue remained firmly sealed to the posterior region of the hard palate and most of the soft palate. As milk passed through the now incomplete anterior seal into the oral cavity, it did not have a uniform radiodensity but appeared to contain bubbles. At frame 288 (Fig. 14.4), the jaw began to close, and milk
452
R. Z. G e r m a n a n d A. W. C r o m p t o n
FIGURE 14.2. Movement of jaw and tongue markers from Fig. 14.1 vs time, in frame number. On the left is anterior/posterior movement, on the right is dorsal/ventral movement. Jaw movement, or gape, and movement of the anterior tongue are minimal. Movement of the posterior markers is large.
no longer left the nipple. The anterior tongue moved upward, and the posterior tongue downward and forward (Fig. 14.3, frame 282). Liquid transport continued with these movements for several frames as the jaw continued closing (Fig. 14.3, frame 296). At frame 300 (Fig. 14.3), the anterior of the tongue was still rising and resealing with the hard palate. The posterior tongue changed direction and began to move upward and somewhat backward, moving the contact between it and the palate (Fig. 14.3, frames 300 and 303). Through these frames, and to the end of this sequence, the posterior tongue established a seal with the hard palate initially, and subsequently with the soft palate, suggestive of a squeezing action. At the same time the liquid moved into the valleculae and then through the piriform recesses, culminating in a swallow. When tongue movement is measured relative to mandible movement, anterior tongue markers, lying underneath the nipple, moved very little relative to the lower jaw. However, marker 3 (Fig. 14.3), nearer the middle of the tongue, moved downward (dorsally) relative to the lower jaw, at the same time the jaw moved downward. The extent of upward movement of the anterior tongue was the extent of jaw movement, although again, marker three had a large excursion of movement in addition to jaw movement. Pressure change in the nipple during a single cycle
was a cyclic rise and fall that was correlated with tongue movement and milk flow. Change in pressure closely followed change in gape and middle tongue movement, with pressure relatively low at maximum gape, after liquid has been acquired from the previous cycle, and rising as the jaw moves upward and the milk is transported through the oral cavity. D . Suckling in Infant O p o s s u m s Suckling in infant opossums is also characterized by a dorsal/ventral pumping mechanism (German and Crompton, 1996). A single cycle of tongue movement indicated several similarities to the kinematics of suckling in other mammals. These infants were the smallest studied, less than 50 g, and thus had only two tongue markers. Initially the tongue formed a seal with the hard palate, just anterior to the hard palate/soft palate junction (Fig. 14.5, frames 00-10). In front of this seal, the tongue moved downward, as milk flowed from the nipple into the oral cavity (Fig. 14.5, frames 00-20). Between frames 00 and 10, the anterior tongue moved upward, maintaining the tongue to palate seal. The seal moved posteriorly between frames 10 and 20, as a slightly more posterior portion of the tongue contacted the soft palate. Next, the tongue near the first marker moved downward, increasing the space inside the oral
14. T h e O n t o g e n y of F e e d i n g in M a m m a l s 254
296
261
300
side the posterior region of the oral cavity. Finally, the tongue posterior to the seal, near marker two, moved upward to reach first the posterior-most edge of the hard palate and a large portion of the soft palate, forcing the milk in the anterior oral cavity into the oropharynx (Fig. 14.5, frames 49-63). The main movement of the tongue during suckling was in a dorsal/ventral direction (Fig. 14.6A).
^O
269
303
282
309
453
FIGURE 14.3. Lateral view of infant macaque suckling by pumping mechanism. Each of eight frames are taken from cine X-ray film of an infant macaque sucking. Filming was at 100 fps, so each frame was 10 msec apart. Frame numbers are indicated at upper left. Anterior is to the right, posterior to the left. Six radio-opaque markers are in the body of the tongue and are numbered on the first frame. The pumping movements of the tongue are indicated with arrows and, given the seal of the tongue to the palate, produce a reduced pressure inside the oral cavity. The final swallow is accomplished through a squeezing mechanism, rather than a pumping one. Stippled area is milk.
cavity. Between frames 20 and 34 (Fig. 14.5) the tongue near the first marker continued to move downward and slightly backward. There was considerable movement in the tongue that was not reflected in marker movement. The seal between palate and tongue shifted further in a posterior direction along the soft palate (Fig. 14.5, frames 24-34), and a portion of the tongue anterior to the first marker moved dorsally, upward toward the hard palate. These movements created a second space within the oral cavity, and milk flowed backward into this space. The rising portion of the tongue continued to rise, and by frame 49 again contacted the hard palate, sealing this bolus of milk in-
IV. RHYTHMICITY A N D CONTROL OF SUCKLING Suckling, as is true of drinking and mastication in adult mammals, is a rhythmic activity. In older animals, movement of jaws, tongue, and hyoid is regular and cyclic (German and Franks, 1991). In infants the behavior of tongue and jaws is also cyclic. The rate of this cyclicity varies with species, and there is significant variation in the rate among individuals of a single species (German et al, 1992,1997). The nature of neural control of rhythmic behavior has been studied extensively in adult mammals (Lund and Dellow, 1971; Lund and Olsson, 1983; Lambert et al, 1986; Goldberg and Chandler, 1990), as well as the extent to which peripheral sensory information can modify the ongoing rhythms of feeding. Little is known about the ontogeny of neural control of feeding. However, some information exists as to the rhythmic nature of infant feeding, in particular, what peripheral stimulation is necessary to elicit rhythmic activity and to what extent that rhythm is flexible and malleable. Mechanical stimulus by a nipple is not sufficient to elicit rhythmic jaw and tongue activity by infant pigs (German et ah, 1997). However, delivery of milk, often a single drop, or delivery at a low rate (an order of magnitude less than normal feeding frequency) was sufficient to elicit rhythmic movement. The frequency of the rhythmic jaw/tongue movement in this situation was significantly higher than when milk was delivered near the preferred rate of approximately 4 Hz (German et al, 1997). When the rate of delivery is varied from 50 to 150% of preferred suckling rate, there was no effect on the animal's suckling frequency. Individuals had a preferred suckling frequency independent of the rate of milk delivery. Given that milk delivery in these experiments was fixed over a long period of time, milk was inevitably delivered at all phases of the jaw cycle. This strongly suggests that subsequent transport and swallowing of the aliquot of milk are decoupled from the movements, and rhythm, of jaw and tongue movements involved in the acquisition of milk.
454
R. Z. G e r m a n a n d A. W. C r o m p t o n
t6
t6 t4 t3
t2 t1
gape 260
260
1
270
310
280
posterior
dorsal { anterior FIGURE 14.4. Movement of jaw and tongue markers from Fig. 14.3 vs time, in frame number. On the left is anterior/posterior movement, on the right is dorsal/ventral movement. Movement of the anterior tongue is minimal. However, jaw movement, measured through gape, and movement of the posterior markers are large.
o cycle start # cycle end 0.5 mm
FIGURE 14.5. Lateral view of suckling in an infant opossum. Ten frames are taken from cine X-ray film, at 100 fps, so each frame is 10 msec apart. Anterior is to the left, posterior to the right. There were two radio-opaque markers in the tongue indicated by filled-in circles. Open circles indicate the position of the markers in the above frame. The stippled area is milk. In infant opossums, the tongue forms a seal with the hard palate. The tongue pumps primarily in a dorsal/ventral direction. This creates a reduction in intraoral pressure, which in turn moves the milk through the oral cavity.
(
dorsal anterior
FIGURE 14.6. (A) Two-dimensional movement of radio-opaque markers in an infant opossum tongue while suckling. (B) Twodimensional movement of radio-opaque markers in an infant opossum who has been removed from the mother and consequently laps instead of sucks. Dorsal is up, and anterior is the right. Each curve traces the movement in a lateral plane. The unfilled circle is the beginning of the cycle, the filled circle indicates the end of the cycle. (A) Cycles are dominated by dorsal/ventral movements, associated with the pumping mechanism seen in other species. (B) Movements are entirely anterior/posterior characteristic of lapping in other adult mammals.
14. The Ontogeny of Feeding in Mammals V. C O O R D I N A T I O N OF SWALLOWING A N D RESPIRATION Feeding, particularly swallowing, is linked to breathing because both functions utilize common anatomical spaces. In mammals of all ages, including humans, inspired air must pass through the pharynx into the glottis. Swallowed food also passes through the pharynx, crosses the laryngeal opening, and must be excluded from the glottis. Mammals have devised various anatomical structures and neuromuscular mechanisms to minimize the disruption of air flow during mastication. All mammals, excluding adult humans, can lock the larynx (including the epiglottis) into the nasopharynx so that an airway extends from the external nares to the trachea, bypassing the oral cavity (Wood Jones, 1940; Negus, 1949). This arrangement permits suckling, and mastication and food manipulation (in nonhuman adults), to take place within the oral cavity, without interfering with rhythmic breathing. A great deal is known about the mechanics of swallowing in all mammals, especially humans (see Logemann, 1983; Hiiemae and Crompton, 1985; Jones and Dormer, 1991), but the relationship between swallowing and respiration is more contentious. Negus (1949) suggested that breathing need not necessarily be interrupted by swallowing because of the continuous airway described earlier. Negus's thesis that breathing and swallowing could and did occur simultaneously generated controversy and extensive contradictory data (Ardran et aZ., 1958; Lieberman, 1984; Crelin, 1987; Laitmann and Reidenberg, 1988,1993; Larson and Herring, 1996). Adult opossums are in the curious position of swallowing both ways; the airway is clearly interrupted when they swallow solid food, but it remains open when they swallow liquids. The relationship between respiration and swallowing is best known in adult humans, where the situation is different from all other mammals, given the unique, low position of the adult human larynx. These data suggest a complex, linked relationship between the two behaviors (Selley et ah, 1989,1990; Logemann et ah, 1992; Maddock and Gilbert, 1993; Martin et al, 1994; Paydarfar et al, 1995). A number of factors complicate the relationship between swallowing and respiration in infant mammals. The position of the epiglottis during swallowing depends on the liquid the animal is drinking (Laitman et ah, 1977). Several aspects of respiration increase with ontogeny, including respiratory drive (Farber, 1988; Barrington and Finer, 1991) and regularity of respiration during feeding (German et ah, 1992). Coordination between feeding and respiration also changes during ontogeny, as well as the pathway of the swallow through the oropharynx and the timing of swallows
455
relative to the respiratory cycles (German et ah, 1996; Crompton et ah, 1997). One interpretation of these ontogenetic changes is that, initially, breathing and swallowing are not correlated but that a linkage develops early in infant maturation. If this is the case, the question arises of why breathing and swallowing should need to be linked, given that infants maintain an intranarial larynx during a swallow. Some of the changes documented in respiration and swallowing may explain the discrepancies reported in the literature in relation to breathing and swallowing patterns in human infants. Laitman et ah (1977) claimed that, in human infants, a patent airway could be retained during a saliva swallow and that respiration was only briefly interrupted when milk or a barium milk mixture was swallowed. However, several authors (Ardran et ah, 1958; Wilson et ah, 1981; Koenig et ah, 1990; Selley et ah, 1990; Medoff-Cooper et ah, 1993) all agree that swallowing disrupts breathing patterns in human infants. Swallows can, nevertheless, occur at various times during the breathing cycle of human infants and we have found the same to be true in infant macaques. Bamford et ah (1992) found that swallows could occur at any time in the respiratory cycle of the human infant, whereas Selley et ah (1990) found that it could only occur (a) between inspiration and expiration or (b) during the middle of expiration. Most of these human studies, however, are limited to a small sample of infants of a particular age. Given the great variations in the methodologies employed, it is difficult to integrate the results obtained by one research group working on one age group with any other set of data.
VI. T R A N S I T I O N FROM SUCKLING TO D R I N K I N G AT W E A N I N G With maturation, infant mammals stop suckling and begin ingestion of solid food and drinking from sources other than their mothers. This change occurs as teeth erupt and is associated with morphological changes in the face, particularly an increase in the anterior/posterior direction (Helm and German, 1996). Little is known about initial consumption of solid food, but some data exist on the transition from suckling to juvenile drinking (Thexton, et ah, 1998). Longitudinal data, showing the change in tongue movements in the same individual, indicate significant differences between suckling and drinking. In suckling, movements of any magnitude are restricted to the posterior twothirds of the tongue. This region is involved in the aquisition of milk from the nipple and the transport into the vallecular recess. In drinking, the whole tongue is involved in the ingestive movements, which
456
R. Z. G e r m a n a n d A. W. C r o m p t o n
reflect a combination of lapping and sucking movements. Jaw and hyoid movements are minimal in both activities. Consequently, the different movements of the tongue during the two activities are generated by changes in the contraction patterns of muscle within the tongue, i.e., they are produced by the contraction of different groups of intrinsic and extrinsic tongue muscles and not by changes in the movement of the tongue base (hyoid). In suckling, the marked cyclical depression in the midregion of the tongue anterior to the fauces lowers intraoral pressure at that point and assists in drawing milk from the teat. This movement is initially coupled with high levels of electromyographic (EMG) activity in those muscles forming a base to the tongue (digastric, geniohyoid); subsequently, the genioglossus and hyoglossus increase in activity and then the styloglossus, omohyoid, and sternohyoid show activity before the cycle repeats. Swallows occur during the jaw-opening phase of every other cycle so that a suckling sequence can be divided into suckling and suckle/swallow cycles. The swallow is characterized by increased levels of activity in the hyoglossus, styloglossus, and omohyoid muscles, which pull the tongue back to force milk out of the valleculae and into the esophagus. In contrast, when a pig drinks from a bowl there is a reduced sucking action combined with lapping movements and the swallows occur every third or fourth cycle. Activity levels of the hyoglossus and omohyoid, during both drinking cycles and drink/swallow cycles, are increased relative to suckling while the activity of the other monitored muscles decreased significantly. It would appear that the generation of negative pressure is less important for ingestion by drinking whereas the anterior/posterior movement of the tongue surface is more important. Differences in the activity levels of the monitored muscles during suckling and drinking may consequently be explained on the basis of the work necessary to generate the negative pressures while keeping the tongue in a stable overall position. Recordings from a larger suite of muscles are required to confirm or reject this suggestion. The timing of the transition from suckling to drinking may not be a strict function of age (German and Crompton, 1996). Evidence from infant opposums suggests that suckling and drinking can exist contemporaneously in littermates in different environments. Infants that are not yet mature enough to exist independent of n\aternal care, i.e., incomplete fur, inability to thermoregulate, and endentulous, will no longer suckle if they are removed from maternal care. Instead they exhibit only adult drinking mechanisms, characterized by the anterior/posterior movements of the tongue (Fig. 14.6B). Littermates remaining in maternal
care continue to suckle using a dorsal/ventral pumping mechanism that characterizes other infant mammals.
VII. EVOLUTIONARY CONSIDERATIONS Suckling is clearly a mammalian synapomorphy. Only mammals suckle, and all extant mammals suckle as infants. Even monotremes, which lack true mammary glands, suckle as infants (Griffths, 1978). Furthermore, suckling mechanisms appear to be homologous in widely separated taxa. In all species studied to date, suckling is characterized by dorsal/ventral movement of the tongue, and at least one seal between tongue and palate to isolate the oral cavity from the the oropharynx. This common suckling mechanism is correlated with a number of anatomical traits common to infant mammals: a relatively short face or viscerocranium, especially relative to the neurocranium, and relatively large tongues (Clark and Smith, 1993; Maunz and German, 1996; Helm and German, 1996). Despite these similarities in infant form and function, considerable variation exists in adult feeding mechanisms. Differences in feeding structures, particularly teeth, associated with different feeding mechanics are remarkable and are one of the hallmarks of mammalian evolution (Crompton and Jenkins, 1979; Eisenberg, 1981; Hiiemae and Crompton, 1985). Suction feeding occurs in nonmammalian groups, particularly snakes (Kardong and Haverly, 1993; Cundall, 1995). In these animals, the floor of the oropharyngeal cavity rises and falls, functioning as a pump that aspirates water into the cavity and forces it posteriorly into the esophagus. The expansion and compression of the buccal cavity are achieved through n\ovement of the oral floor, and a "bellows-like" displacement of the mandible (Kardong and Haverly, 1993). The tongue functions as a guide for the water, but is not an active part of a pumping mechanism. Cundall (1995) cautions that a more complex pumping mechanism is likely, given that water appears to be moving not only during depression of the buccal floor, but also during elevation, when compression within the cavity must be occurring. It is clear, however, that these mechanisms are used only by infant mammals in suckling. Whereas both snakes and infant mammals produce a pressure gradient to move liquid, the function of the mammalian tongue is unique to mammals. References Ardran, G. M., and R H. Kemp (1959) A correlation between sucking pressures and movement of the tongue. Acta Paediatr. 48:261272.
14. T h e O n t o g e n y of F e e d i n g in M a m m a l s Ardran, G. M., R H. Kemp, and J. Lind (1958) A cineradiographic study of bottle feeding. Br. J. Radiol. 31:11-22. Bamford, O., V. Taciak, and I. H. Gewolb (1992) The relationship between rhythmic swallowing and breathing during suckle feeding in term neonates. Pediatr Res. 31:619-624. Barrington, K., and N. Finer (1991) The natural history of the appearance of apnea of prematurity. Pediatr. Res. 29(4 Ft 1): 372-375. Bosma, J. (1985) Postnatal ontogeny of performances of the pharynx, larynx and mouth. Ann. Rev. Respir. Dis. 131: Supp. S10-S15. Clark, C. T, and K. K. Smith (1993) Cranial osteogenesis in Mondelphis domestica (Didelphidae) and Macropus eugenii (Macropodidae). J. M o r p h 215:103-114. Crelin, E. S. (1987) The Human Vocal Tract: Anatomy, Function, Development and Evolution. Vantage Press, New York. Crompton, A. W., and F. A. Jenkins (1979) Origin of mammals. In: Mesozoic Mammals: The First Two-Thirds of Mammalian History. J. Lillegraven, Z. Kielan-Jaworoska, and W. A. Clemens (eds). University of California Press, Berkeley, CA. Crompton, A. W., R. Z. German, and A. J. Thexton (1997) Protection of the airway during swallowing in infant mammals. J. Zool. Lond. 241:89-102. Cundell, D. (1995) Drinking in snakes. Am. Zool. 35:106A. DuBrul, E. L. (1988) Oral Anatomy. Ishiyaku EuroAmerica, St. Louis. Eisenberg, J. F. (1981) The Mammalian Radiations: An Analysis of Trends in Evolution, Adaptation, and Behavior. University of Chicago Press, Chicago. Farber, J. P. (1988) Medullary inspiratory activity during opossum development. Am. J. Physiol. 254(4 Ft 2) :R578-584. Eraser, D., and B. K. Thompson (1991) Armed sibling rivalry among sickling piglets. Behav. Ecol. Sociobiol. 29:9-15. German, R. Z., and H. A. Franks (1991) Timing in the movement of jaws, tongue and hyoid during feeding in the Hyrax, Procavia syriacus. J. Exp. Zool. 257:34-42. German, R. Z., A. W. Crompton, L. C. Levitch, and A. Thexton (1992) The mechanism of suckling in two species of infant mammal: miniature pigs and long-tailed macaques. J. Exp. Zool. 261:322330. German, R. Z., and A. W. Crompton (1996) Ontogeny of suckling mechanisms in opossums. Brain Behav. Evol. 48:157-164. German, R. Z., A. W. Crompton, C. McCluskey, and A. J. Thexton (1996) Coordination between respiration and deglutition in a preterm infant mammal. Arch. Oral. Biol. 41(6): 619-622. German, R. Z., A. W. Crompton, and A. J. Thexton (1997) Determinants of rhythm and rate in suckling. J. Exp. Zool. 278:1-8. Goldberg, L. J., and Chandler, S. H. (1990) Central mechanisms of rhythmical trigeminal activity. Pp. 268-289. In: Neurophysiology of the Jaws and Teeth. A. Taylor (ed.). Macmillan Press, London. Griffths, M. (1978) The Biology of the Monotremes. Academic Press, New York. Helm, J. W., and R. Z. German (1996) The epigenetic impact of weaning on craniofacial morphology during growth. J. Exp. Zool. 276: 243-253. Hiiemae, K. M., and A. W. Crompton (1985) Mastication, food transport and swallowing. Pp. 262-290. Functional Vertebrate Morphology. M. E. Hildebrand et al. (eds.). Harvard Univ. Press, Cambridge. Jones, B., and M. W. Dormer (1991) Normal and Abnormal Swallowing. Springer-Verlag, New York. Kardong, K. V., and J. E. Haverly (1993) Drinking by the common boa. Boa constrictor. Copeia. 1993(3):808-818. Koenig, J. S., A. M. Davies, and B. T Thach (1990) Coordination of breathing, sucking, and swallowing during bottle feedings in human infants. J. Appl. Physiol. 69:1623-1629. Laitman, J. T, E. S. Crelin, and G. J. Conlogue (1977) The function of the epiglottis in monkey and man. Yale J. Biol. Med. 50:43-48.
457
Laitman, J. T, and J. S. Reidenberg (1988) Advances in understanding the relationship between the skull base and larynx with comments on the origins of speech. Hum. Evol. 3:99-109. Laitman, J. T, and J. S. Reydenberg (1993) Specializations of the human upper respiratory and upper digestive systems as seen through comparative and developmental anatomy. Dysphagia 8: 318-325. Lambert, R. W., L. J. Goldberg, and S. H. Chandler (1986) Comparison of mandibular movement trajectories and associated patterns of oral muscle electomyographic activity during spontaneous and apomorphine-tnduced rhythmic jaw movements in the guinea pig. J. Neurophysiol. 55:301-315. Larson, J. E., and S. W. Herring (1996) Movement of the epiglottis in mammals. Am. J. Phys. Anthrop. 100:71-82. Lieberman, P. (1984) The Biology and Evolution of Language. Harvard Univ. Press, Cambridge. Logemann, J. A. (1983) Evaluation and Treatment of Swallowing Disorders. College-Hill Press, San Diego. Logemann, J. A., P. J. Kahrilas, J. Cheng, B. R. Pauloski, P. J. Gibbons, A. W. Rademaker, and S. Lin (1992) Closure mechanisms of the laryngeal vestibule during swallow. Am. J. Physiol. 262 (Gastrointest. Liver Physiol. 25), G338-G344. Lund, J. P., and P. G. Dellow (1971) The influence of interactive stimuli on rhythmical masticatory movements in rabbits. Arch. Oral. Biol. 16:215-223. Lund, J. P., and K. A. Olsson (1983) The importance of reflexes and their control during jaw movement. Trends Neurosci. 6:458-463. Maddock, D. J., and R. J. Gilbert (1993) Quantitative relationship between liquid bolus flow and laryngeal closure during deglutition. Am. J. Physiol. 265 (Gastrointest. Liver Phsyiol. 28), G704-G711. Martin, B. J. W., J. A. Logemann, R. Shaker, and W. J. Dodds (1994) Coordination between respiration and swallowing: respiratory phase relationships and temporal integration. J. Appl. Physiol. 76:714-723. Maunz, M., and R. Z. German (1996) Craniofacial heterochrony and sexual dimorphism in the short-tailed opossum {Monodelphis domestica). J. Mammol. 77:992-1005. McFarland, D. H., J. P. Lund, and M. Gagner (1994) Effects of posture on the coordination of respiration and swallowing. J. Neurophysiol. 75(5): 2431-2437. Medoff-Cooper, B., T. Verklan, and S. Carlson (1993) The development of sucking patterns and physiologic correlates in very-lowbirth-weight infants. Nurs. Res. 42(2): 100-105. Negus, V. E. (1949) The Comparative Anatomy and Physiology of the Larynx. Heinmann, London. Paydarfar, D, R. J. Gilbert, C. S. Poppel, and R F Nassab (1995) Respiratory phase resetting and airflow changes induce by swallowing in humans. J. Physiol. 483:273-288. Selley, W. G., F C. Flack, R. E. Ellis, and W. A. Brooks (1989) Respiratory patterns associated with swallowing. 1. The normal adult pattern and changes with age. Age Aging. 18:168-172. Selley W. G., R. E. EUis, F C. Flack, and W. A. Brooks (1990) Coordination of sucking, swallowing and breathing in the newborn, its relationship to infant feeding and normal development. Bri. J. Disorders Commun. 25:311-327. Thexton, A. J., A. W. Crompton, and R. Z. German (1998) Transition from suckling to drinking at weaning: a kinematic and electromyographic study in miniature pigs. J. Exp. Zool. 280:327-343. Thexton, A. J., and J. D. McGarrick (1988) Tongue movement of the cat during lapping. Arch. Oral. Biol. 39:599-612. Wilson, S. L., B. T Thach, R. T Brouillette, and Y. K. Abu-Osba (1981) Coordination of breathing and swallowing in human infants. J. Appl. Physiol. 50:851-858. Wood Jones, F (1940) The nature of the soft palate. J. Anat. 74: 147-170.
This Page Intentionally Left Blank
C H A P T E R
15 Feeding in Myrmecophagous Mammals KAREN ZICHREISS^ Section of Ecology and Systematics Cornell University Ithaca, New York 14853
I. INTRODUCTION A. Defining the Problem B. Taxonomy and Phylogeny of Mammalian Myrmecophages II. FORAGING ECOLOGY A. Prey Characteristics B. Myrmecophage Foraging Ecology and Behavior III. MORPHOLOGY OF THE FEEDING APPARATUS A. The Myrmecophagous Morphotype B. Exceptions to the Morphotype IV. FUNCTIONAL MORPHOLOGY A. Jaw Movements B. Tongue Movements C. Pharynx and Soft Palate Movements D. Feeding Stages in Myrmecophages V. EVOLUTION OF MYRMECOPHAGOUS SPECIALIZATIONS A. Phylogenetic Pathways to Myrmecophagy B. Structural Pathways to Myrmecophagy C. Primitive and Derived Features in the Myrmecophagous Feeding Apparatus VI. DIRECTIONS FOR FUTURE RESEARCH A. Form B. Function C. Evolution References L INTRODUCTION A. Defining tlie Problem Many ant- and termite-eating mammals share a set of anatomical features, including a long, thin, and 1. Present address: Department of Biological Sciences, Humboldt State University, Areata, CA 95521, e-mail:
[email protected].
FEEDING (K Schwenk, ed.)
459
highly extensible tongue, viscous saliva produced by hypertrophied salivary glands, reduction or loss of the teeth, v^ell-developed olfactory structures, a muscular and gizzard-like stomach, and forelimbs with robust flexor musculature and enlarged claws (Griffiths, 1968). Similarities in feeding structures prompted Cuvier (1817) to classify echidnas, anteaters, armadillos, pangolins, and aardvarks, as well as platypuses and sloths, as the Edentata. From an evolutionary perspective, however, this is not a natural group. Even though the phylogenetic interrelationships of some of its members are uncertain, Cuvier's Edentata contains four modern mammalian orders and at least five separate radiations of ant and/or termite specialists. Because the similarities seen in these ant and termite eaters are thought to be independently derived, they are frequently used as textbook examples of convergent evolution (Simpson and Beck, 1965; Eisenberg, 1981; Savage and Long, 1986; Pough et al, 1989). Knowing when convergence is an appropriate label for patterns in morphological evolution requires both robust phylogenetic hypotheses and close study of morphology. The supposed convergence in the feeding apparatus of ant-eating mammals has never been examined explicitly and in detail. Relatively little is known about the natural history of most ant-eating species, few comparative studies of the relevant anatomy exist, functional studies are altogether lacking, and the phylogeny of the taxa under consideration is debated. These are serious deficits, but ant-eating mammals present an interesting set of problems in both evolutionary and functional morphology, and a critical review of the available literature is overdue.
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
460
Karen Zich Reiss B. Taxonomy and Phylogeny of Mammalian Myrmecophages
Many mammals eat ants and/or termites (myrmecophagy) but far fewer subsist exclusively on this restricted diet. Redford (1987) reviewed the literature on mammalian stomach and fecal contents extensively and found a wide-ranging proportion of ants and termites in mammal diets. Committed myrmecophages are those mammals in which at least 90% of the diet is ants and/or termites (Table 15.1) and their phylogenetic distribution among mammals is summarized in Fig. 15.1. Myrmecophages are found in each of the three major groups of mammals: the egg-laying Monotremata, the pouched Marsupialia, and the placental Eutheria. This tripartite division on the basis of reproductive characters (Blainville, 1816) is supported by cladistic studies, although molecular studies emphasize the closeness of the three groups (Westerman and Edwards, 1992; Gemmel and Westerman, 1994; Retief et ah, 1994) whereas morphological studies suggest that the Marsupialia and Eutheria are sister taxa (Marshall, 1979; Rowe, 1988; 1991; 1993; Wible and Hopson, 1993; Zeller, 1993) that together comprise the Theria. TABLE 15.1 Myrmecophagous Mammals'' Taxon
Diet"
Monotremata
Tachyglossidae
Tachyglossus aculeatus
4A,T
Marsupialia
Myrmecobiidae Didelphidae Thylacomidae
Myrmecobius fasciatus Metachirus nudicaudatis Macrotis lagotis
2A,3T 3T 3A,T
Manidae Myrmecophagidae
Manis sp. Cyclopes didactylus Tamandua sp. Myrmecophaga jubata Dasypus sp. Cabassous sp. Tolypeutes tricinctus Priodontes maximus Parascalops breweri Emballonura nigrescens Perodictus potto Cercopithecus sp. Otocyon megalotis Bdeogale cmssicauda Cynictis pencillata Mungos mungo Rhynchogale melleri Proteles cristatus Orycteropus afer Funisciurus sp. Prionomys batesi Oxymycterus rufus Praomys erythroleucus
4A,T 4A,2T 4A,T 4A,T 2-3A, T 4A,T 4A,T 4A,T 3A 3A, IT 3A 3 - 4 A , IT 2A,3T 3A,T 2A,3T 3A,2T 4T 2A,4T 4A,T 3-4A,T 4A 4A 1-4A,T
Eutheria Pholidota Xenarthra
Dasypodidae
Insectivora Chiroptera Primates Camivora
Tubulidenatata Rodentia
Talpidae Vespertilionidae Lorisidae Cercopithecidae Canidae Herpestidae
Protelidae Orycteropidae Sciuridae Cricetidae Muridae
^After Redford (1987). ^A, ants; T, termites; numbers indicate proportion ants or termites in diet: 1, 90%.
There is one monotreme myrmecophage, the echidna (spiny anteater) Tachyglossus aculeatus, found in Australia and Tasmania. The only other extant echidna is Zaglossus bruijni, from New Guinea, but it appears to be an earthworm specialist (Griffiths, 1968; Nowak, 1991). Monotremes are considered highly autapomorphic descendents of the common ancestor of extant mammals. Echidnas are thought to be derived with respect to the only other extant monotremes, the platypuses, mainly because the oldest known fossil monotremes are more platypus-like than echidna-like (Woodburne and Tedford, 1975; Archer et al, 1985; 1992; Pascual et al, 1992). Among marsupial mammals, the numbat Myrmecobius fasciatus, is the most myrmecophagous, but it is considered to be an "amateur" ant eater (Griffiths, 1968) along with aardvarks and aardwolves (discussed later). Numbats are terrestrial squirrel-sized inhabitants of wandoo eucalyptus forests of southwestern Australia, where they are sympatric with echidnas (Fleay, 1942; Calaby, 1960). They are monotypic members of the Myrmecobiidae, which are most closely related to either the carnivorous Dasyuridae or Thylacinidae (the marsupial wolves) (Luckett, 1994). Among the placental mammals, or Eutheria, most myrmecophages are in the orders Xenarthra (anteaters, sloths, and armadillos; synonomous with the Edentata of some recent authors) and Pholidota (pangolins, or scaly anteaters). Extant xenarthrans are found only in the New World. There are four species of anteaters in the Myrmecophagidae. Cyclopes didactylus, the silky anteater, is a small (about 250 g) tropical forest canopy dweller. The two species of Tamandua, the vested or collared anteaters, are about the size of a fox terrier and semiarboreal. Myrmecophaga jubata, the giant anteater, is large (about 30 kg) and predominantly terrestrial. Myrmecophaga and Tamandua are sister taxa that are derived with respect to Cyclopes (Engelmann, 1978; 1985; Patterson et al, 1992; Gaudin and Branham, 1998). Sloths (Tardigrada) are the likely sister group of anteaters, although the extant sloths (Bradypodidae) are arboreal folivores only distantly related through numerous fossil sloth taxa (Patterson and Pascual, 1972; Webb, 1985; Gaudin, 1990; 1992; 1995). Sloths and anteaters together form the Pilosa, which is the sister group to the other major xenarthran radiation, the Dasypodidae (or Cingulata, armadillos) (Engelmann, 1978, 1985; Gaudin, 1990). Some armadillos are myrmecophagous: Priodontes maximus is the rare giant armadillo; Cabassous contains four species of naked-tailed armadillos Tolypeutes contains two species of three-banded armadillos, and Dasypus contains five species of long-nosed armadillos. These myrmecophagous armadillos com-
461
15. Feeding in Myrmecophagous Mammals
4?
^y'
.^^ J^^
^^ :^ ^
^
,W'
>^
^^ J>'
.cS' A>
°*
Xenarthra
pitheria Eutheria Mammalia FIGURE 15.1. Phylogenetic distribution of mammalian myrmecophages. Interrelationships of the extant eutherian orders are based on a 50% majority rule consensus of published analyses of both morphological and molecular data sets (from Honeycutt and Adkins, 1993); interrelationships of xenarthran families follow Engelmann (1978) and Gaudin (1995). Bolded clades contain committed myrmecophages with feeding apparatus specializations, and number of myrmecophagous species and common names are in parentheses. Asterisked clades also contain myrmecophages, but no anatomical specializations have been reported in these taxa. There appear to have been two independent origins of myrmecophagy in the Xenarthra, and despite controversy over the phylogenetic position of Pholidota, there have been at least six independent origins of anatomically specialized myrmecophages among extant mammals.
prise the tribe Dasypodini, within which Dasypus is probably the most derived taxon (Engelmann, 1978; 1985). The sister group of the Dasypodini is the Euphractini, which contains all the nonmyrmecophagous armadillos (Redford, 1985, 1987; Smith and Redford, 1990). All armadillos are terrestrial and fossorial. In contrast to the Xenarthra, the extant Pholidota are found only in the Old World. The Pholidota consists of the single family Manidae, which contains seven species of pangolin, all in the genus Manis. The Asian species {M.crassicaudata, M.pentadactyla, M.javanica) can be distinguished from the African species (M. gigantea, M. temmincki, M. tricuspis, M. longicaudata) on the basis of morphology (Pocock, 1924; Emry, 1970; Patterson, 1978). The monophyly of order and the Asian subgroup have recently been confirmed, but the African species appear to be polyphyletic (Gaudin and Wible,
1999). Within each geographic grouping there is a predominantly arboreal species, a semiarboreal species and a fully terrestrial species (Pages, 1970; Kingdon, 1974; Patterson, 1978) paralleling the situation in the anteaters. The phylogenetic positions of Xenarthra and Pholidota are unclear. An analysis based on morphological data that has dominated higher-level mammalian systematics for a decade (Novacek and Wyss, 1986) places Xenarthra and Pholidota as sister taxa, which together form the basal clade of extant eutherian mammals. The idea of a particularly close relationship between these orders is an old one (Storr, 1780; Blainville, 1816; Cuvier, 1817; Gill, 1870; Gregory, 1910) and there has been some molecular support for this alliance (McKenna, 1992; Norman and Ashley, 1994). Moreover, cladistic studies that investigate relationships among
462
Karen Zich Reiss
xenarthran and pholidotan taxa without constraining the monophyly of the orders have resulted in the inclusion of pangolins within Xenarthra (Engelmann, 1978; MacPhee, 1994; Norman and Ashley, 1994; Reiss, 1997a). However, there are vociferous opponents to these hypotheses. Simpson (1945, 1978) believed throughout his lifetime that Xenarthra was an endemic South American group and that common ancestry with Pholidota was so remote as to be irrelevant. Rose and Emry (1993) reviewed morphological evidence for the alliance and found it unconvincing due to polarity uncertainties, uneven within-taxon character distributions, and the likelihood of adaptive convergence. Some molecular studies place pholidotes with carnivores or a carnivore-ungulate assemblage (dejong, 1982; Miyamoto and Goodman, 1986; Shoshani, 1986; McKenna, 1987; Ohnishi, 1991) and there is some morphological support for this idea (Novacek and Wyss, 1986; Szalay and Schrenk, 1998). Aside from some xenarthrans and all pholidotans the only other eutherians that are committed myrmecophages are the sole extant member of the Tubulidentata (the aardvark Orycteropus afer), a single hyaenid in the Carnivora (the aardwolf Proteles cristatus), and scattered rodents and primates. These myrmecophagous taxa are found only in Africa. Affinities of aardvarks are controversial and range from fairly unresolved (Novacek and Wyss, 1986; 1988) to ungulate (Shoshani, 1993; Shoshani and McKenna, 1995) or paenungulate (dejong, 1982; Miyamoto and Goodman, 1986) alliances, and some cranial similarities with lipotyphlan insectivores have been noted (Thewissen, 1985; Novacek, 1986). The aardwolf is clearly a feloid carnivore in the family Hyaenidae, although feloid interrelationships are unresolved (Flynn et al, 1988; Wayne et ah, 1989; Wozencraft, 1989). Despite these controversies regarding eutherian interrelationships, outgroup criteria make it clear that myrmecophagy has evolved several times (see Fig. 15.1). The only taxa that might have inherited myrmecophagy from a common ancestor are anteaters and pangolins (Reiss, 1997a,b), and this only if the Xenarthra are not monophyletic. Myrmecophagy arose at least once in Monotremata, once in Marsupialia, twice in Xenarthra, once in Tubulidentata, once in Carnivora, and perhaps one more time in the Pholidota. All these myrmecophages, defined on the basis of trophic ecology, can be further subdivided into those that are anatomically specialized (i.e., have many of the features that were cited earlier) and those that are not. As we will see, the most anatomically specialized myrmecophages are echidnas, pangolins, anteaters, and, to a lesser degree, the armadillos. These taxa are the core of this review. Aardvarks, aardwolves, and numbats
are less anatomically distinct, but each has a few interesting anatomical features that will be discussed.
IL FORAGING ECOLOGY A. Prey Characteristics Ants and termites are ubiquitous, abundant, and live at high densities. They are most concentrated pantropically, but also extend into temperate regions, ants more so than termites. Ants and termites are found throughout all vertical levels of tropical forests, although ants are more common high in the canopy and termites are more common on the ground (Redford, 1987). Also, most species are colonial and thus, conveniently for their predators, exist at high local densities. Finally, ants and termites are geologically old. Ants are first known from the late Cretaceous, and termites are thought to be older. Their ubiquity and abundance appear to have been long-standing. Thus, ants and termites are likely to have been an abundant food source throughout the entire history of mammalian myrmecophagy. Unfortunately, ants and termites are also aggressive, noxious, nutritionally poor, and small. Aggression may involve stinging, biting, and spraying with endogenous irritants. Kinds and levels of aggression depend both on the species and on the status of an individual in a species' social system. However, because ant and termite sociality frequently involves a caste system with specialized soldiers, even milder individuals can have formidable armies protecting them. Soldiers are generally mobilized in response to perceived threats, either at the nest site or along frequently used paths to and from the nest. The time it takes for soldiers to be mobilized varies, but it is finite and characteristic for each species. Ant and termite aggressive/defensive behaviors are either based on or complemented by allelochemicals, whether poison in an ant's sting or the spray dispensed by some termites (genus Nasutitermes). The high concentrations of these various secondary compounds must be ingested and metabolized by a predator, along with the structural compounds that make up ant and termite bodies. Most nitrogen is in the chitinous exoskeleton and the presence of chitinases in mammals is only beginning to be explored. Also, with the exception of certain reproductive and larval stages, ants and termites have virtually no fat (Redford and Dorea, 1984). In addition to being nutritionally poor, ants and termites are quite small compared with the relatively large mammals that feed on them, thus predator/prey size ratios are large.
15. Feeding in Myrmecophagous Mammals B. Myrmecophage Foraging Ecology and Behavior Prey distribution and abundance, prey aggressive/ defensive strategies, the prey's low nutritional value, and the large predator/prey size ratio are all factors one expects to affect myrmecophage foraging. Not surprisingly, the global distribution of mammalian myrmecophages correlates with prey distribution and abundance. Moreover, sympatry of myrmecophagous species occurs in both New and Old World tropics and in Australia (Pages, 1965,; 1970; Lubin et al, 1977; Montgomery, 1985). Stomach content analyses show that diets of sympatric species may overlap partially but never entirely. This prey resource partitioning is, in part, passively dependent on predator microhabitat associations. In general, arboreal myrmecophages eat more ants than terrestrial myrmecophages, and even among Tamandua, individuals that are niore arboreal ingest more ants than terrestrial individuals in the same habitat (Pages, 1970; Montgomery, 1985). Passive partitioning also results from diurnal and seasonal variation in prey activity patterns (Montgomery, 1985; Redford, 1987). Additionally, there is some evidence that myrmecophages actively select preferred prey species and avoid others. Certain ant taxa, ponerines (Ponerinae), army ants (Dorylinae), and leaf cutters (Myrmicinae), are conspicuously absent from the diets of myrmecophagid anteaters, despite their abundance within feeding territories (Montgomery and Lubin, 1977). Where echidnas and numbats are sympatric, echidnas eat a higher percentage of ants, whereas numbats prefer termites, although away from numbats echidna preference for termites correlates with termite biomass (Abensperg-Traun and De Boer, 1992). This prey species selectivity appears to be olfaction mediated. Nasal turbinates are particularly well developed in myrmecophages, and there are numerous field observations that foraging myrmecophages will periodically stop, sniff in different directions, and then walk directly toward a particular prey species' nest where feeding is resumed (e.g., Sweeney, 1956; Montgomery and Lubin, 1977). Maze experiments have demonstrated that captive giant anteaters will preferentially walk toward food and away from other volatile compounds (McAdam and Way, 1967). Electroreception may also play a role in some myrmecophages. It has been suggested that the fleshy tentacles on the aardvark snout are electrosensory (Kingdon, 1974), and the echidna also has electroreceptors in its snout (Andres et a/., 1991; Augee and Gooden, 1992; Proske et al, 1998), which may assist in the detection of subterranean prey. With respect to prey aggressive/defensive strate-
463
gies, we would expect that myrmecophages are either resistant or use behavioral means to minimize the impact. Resistance could be conferred by integumental specializations, such as the scales of pangolins and the carapace of armadillos. Some anecdotal evidence suggests that pangolins are unperturbed by prey attacks. They have been seen rolling in prey nests with their scales elevated, after which they clamp down the scales, move to a body of water, elevate their scales again, and finally dart their tongue over the water's surface, collecting the now floating prey (Hatt, 1934; Nowak, 1991). This is an amusing feeding strategy that could serve other purposes as well, e.g., prey allelochemicals may offer integumental parasite protection. There have also been observations that pangolins shake their scales when swarmed over, sending prey flying (Hatt, 1934; Kingdon, 1974). I know of no reports on armadillo behavior in the face of angry prey, but one would expect their carapace to be particularly effective protection. The other myrmecophages, all of whom are described as thick skinned, show obvious displeasure during prey attacks. I have seen Tamandua feeding furiously as their forelimbs expose more and more of a nest, while a hind limb scratches at the nape and shoulders just as furiously. On the assumption that myrmecophages are bothered by prey defenses, patterns of foraging behavior in myrmecophagid anteaters have been interpreted as adaptations for avoiding prey defenses (Montgomery and Lubin, 1977; Montgomery, 1985). Foraging patterns in anteaters are usually described as "farming." A nest is broken into and fed upon for a short period of time, and then the anteater moves on, taking isolated prey on the way to another nest. Foraging grounds are generally large, many nests are visited in a single activity cycle, and damaged nests are not returned to for days. A similar pattern is seen in three species of African pangolins and the aardvark (all sympatric in West Africa) and in numbats (Calaby, 1960; Pages, 1970). The presumed advantages of this kind of foraging pattern are twofold: exposure to prey soldier defenses are minimized by short feeding bouts and prey colonies are maintained as long-term food resources. There is both interspecific and individual variation in the duration of feeding bouts at nest sites and this variation has been ascribed both to the predator's sensitivity to prey attacks and to seasonal variation in prey nutritional content (Kruuk and Sands, 1972; Lubin and Montgomery, 1981; Redford and Dorea, 1984). Pangolins feed from termite nests for much longer than Tamandua, perhaps because the pangolin's scaly covering affords it some protection. Tamandua does, however, spend an unusually long time at termite nests when fat-rich alates are present (Lubin and Montgomery,
464
Karen Zich Reiss
1981), perhaps providing incentive for Tamandua to suffer more annoyance than usual. The echidna Tachyglossus also shows a preference for alates. Indeed, the only time they feed on Iridomyrmex ants is when alates are present, and this is correlated with the echidna's emergence from hibernation and presumably higher energy needs (Griffiths and Simpson, 1966; Griffiths, 1968). Among myrmecophages, aardwolves and numbats are the main exception to nest-based feeding. Aardwolves specialize on a single genus of a surfaceforaging termite {Trinivertermes) and usually feed from aggregations of termites on the veld floor and only sometimes lick termites off the surface of a nest. They will not (and perhaps cannot) break into nests (Richardson, 1987,1990). Numbats feed mostly from the termite-infested bases of wandoo trees, as well as from the forest floor generally (Calaby, 1960). Prey size, indigestibility and noxiousness all suggest that myrmecophages should have unusual metabolic physiology and biochemistry. However, aside from comparative studies of mammalian basal metabolic rates, virtually nothing is known. Eutherian myrmecophages do have unusually low metabolic rates. A low metabolic rate may be a retained primitive character, but it has also been suggested that a low metabolic rate is an adaptation for myrmecophagy or, alternatively, is the result of the ingestion of concentrated allelochemicals (McNab, 1984). Regardless of controversies over its causal basis, a low metabolic rate probably facilitates subsistence on a small and nutritionally poor food source. Other physiological adaptations for myrmecophagy are likely, e.g., specialized metabolic pathways may also be present. However, in a strict sense, myrmecophages are committed but not obligate trophic specialists. Several species are successfully raised in captivity on a mash diet that does not include ants and termites, although this diet can lead to obesity (Heath and Vanderlip, 1988).
III. MORPHOLOGY OF THE FEEDING APPARATUS The preceding section discussed the exigencies imposed by prey characteristics on mammalian myrmecophages. There is some indication that myrmecophage foraging behavior is structured to cope with prey aggressive/defensive strategies. Also, mammalian myrmecophages may have physiological mechanisms that allow them to cope with their prey's small size, poor nutritional value, and perhaps even toxicity. These ecological, behavioral, and physiological issues should all be explored further. However, the most
obvious features that set myrmecophages apart from other mammals are in the feeding apparatus. A. The Myrmecophagous Morphotype The frequently cited features in which myrmecophages resemble one another are often superficially defined. It remains to be seen how closely they resemble one another if the anatomy is scrutinized. The following sections review osteology and the muscles of the jaws, tongue, pharynx, and soft palate. Each section begins with an overview of the anatomy of the region and is followed by a comparative review of the morphology in echidnas, pangolins, anteaters, and dasypodine armadillos. The goal is to discuss the comparative anatomy with particular emphasis on issues of homology and to provide a basis for a subsequent discussion of function. 1.
Osteology
Studies of the echidna skull (Griffiths et al, 1991), pholidotan skull (Emry, 1970; JoUie, 1968), edentate jaw joint (Lubosch, 1908), xenarthran auditory region (Gaudin, 1990, 1992; Patterson et al, 1989), xenarthran temporal region (Guth, 1961), hyoid apparatus of echidnas, anteaters, and pangolins (Gasc, 1967; Schneider, 1964), and the mammalian skull generally (Novacek, 1986, 1993) are available. This section is based on these accounts and my own observations and summarizes features that are directly relevant to feeding apparatus anatomy and function: the shape of the skull and dentary, their muscular processes, the jaw joint, dentition, the hyoid apparatus, and, for some species, the xiphoid processes of the sternum to which tongue muscles are attached. The myrmecophage skull is typically long, with a small braincase and relatively elongated rostrum (Fig. 15.2). The zygomatic arch is incomplete in anteaters and in most of the pangolins {Manis javanica and M. pentadactyla are variable. Rose and Emry, 1993), whereas echidnas and armadillos have a complete arch. The dorsal surface of the skull is smooth and without pronounced sagittal crests, and the temporal fossa is always reduced (Rose and Emry, 1993). The paraoccipital and styloid processes are either reduced or absent, and nuchal crests and mastoid processes are well developed only in armadillos (Gaudin, 1990). A long hard palate extends beneath the elongate rostrum. Palatine shelves extend nearly to the jaw joint, and the pterygoids extend to or beyond the otic region. In the anteaters Tamandua and Myrmecophaga, the pterygoids have medially expanded laminae that meet in the midline, prolonging the hard palate to the level of
465
15. Feeding in Myrmecophagous Mammals
iof
FIGURE 15.2. Comparison of skulls and dentaries of myrmecophages. Lateral views are to the left and palatal views are to the right. (A) echidna, Tachyglossus aculeatus; (B) pangolin, Manis pentadactyla; (C) anteater, Myrmecophaga tridactyla; (D) anteater, Cyclopes didactylus; and (E) armadillo, Dasypus novemcinctus. Skull sizes have been normalized to equivalent cranial lengths despite a 20-fold difference between C, the largest, and D, the smallest. In A through D the skulls and dentaries are streamlined, muscular processes are weak, and teeth are absent. (E) A myrmecophagous armadillo stands out in being relatively robust, but its skull and dentary are lightly built compared with nonmyrmecophagous armadillos.
the caudal bulla and eliminating the interpterygoid vacuity across which the soft palate usually lies (see Fig. 15.3 for a more detailed illustration of the skull of Tamandua). In the anteater Cyclopes, pangolins, and armadillos the pterygoids are similarly long but do not meet in the midline; only pangolins and Cyclopes have a defined pterygoid hamulus. Teeth are either reduced or absent altogether. Reduction, seen in the armadillos, typically consists of the loss of incisors and canines and the conversion of cheek teeth into homodont pegs that lack enamel. The dentary of the myrmecophagous armadillos is weak compared with nonmyrmecophagous relatives but remains L-shaped with prominent muscular processes
fo
at
pif
he
FIGURE 15.3. Skull and dentary of Tamandua mexicana (traced from photographs of CU 17743, Cornell Vertebrate Collections, Ithaca, NY). Dorsal view is at the top; palatal view in the middle; and lateral view of skull and dentary at the bottom. Arrowhead indicates the coronoid process of the dentary. In particular, note the streamlined appearance of the skull and dentary, indicative of the absence of robust processes for the muscle attachment. On the lateral view, also note the incomplete zygomatic arch and the unusual caudoventral position of the auditory foramen, from which the soft tissue auditory tube runs caudally. On the ventral view, note the absence of teeth and the hard palate that extends nearly the entire length of the skull. Scale: 1 cm. Abbreviations used in this figure and in Figs. 15.4 and 15.9. a, arytenoid; ac, anterior cornu; as, alisphenoid; at, auditory tube; bo, basioccipital; c, cricoid, ch, ceratohyal; co, corpus hyoideum; cs, constrictor salivaris; d, dentary; eam, external auditory meatus; eh, epihyal; es, esophagus; f, frontal; flp, foramen lacerum posterior; fo, foramen rotundum; gg, genioglossus; gh, geniohyoideus; he, hypoglossal canal; hg, hyoglossus; ih, interhyoideus; iof, infraorbital foramen; j , jugal; 1, lacrimal; Ivp, levator veli palatini; mh, mylohyoideus; mvp, medialis veli palatini; mx, maxilla; n, nasal; of, optic foramen; om, oral mucosa; p, parietal; pc, posterior cornu; pg, palatoglossus; pi, palatine; plf, posterior lacerate foramen; pm, premaxilla; pp, palatopharyngeus; pt, pterygoid; sd, salivary duct; sg, sternoglossus; sh, stylohyal; smf, stylomastoid foramen; so, supraoccipital; sof, sphenorbital fissure; sp, stylopharnygeus; sq, squamosal; sta, sternal aponeurosis; t, thyroid bone; to, tongue; and ty, tympanic bulla.
(Smith and Redford, 1990). Echidnas, anteaters, and pangolins are all toothless. In all but the anteater Cyclopes, the dentary is an elongate and bowed splint that
466
K a r e n Zich Reiss
has no prominent coronoid, angular, or condylar processes. The dentary of Cyclopes is similarly slender but has a peg-like coronoid and a defined angular process. The jaw joint in most myrmecophages is relatively low on the skull, at the same level as the hard palate (Rose and Emry, 1993). Among armadillos, this characteristic is most pronounced in Dasypus, and the jaw joint of other armadillos is higher (Patterson et al, 1989; Smith and Redford, 1990). The articular surfaces of the jaw joint are relatively flat in all taxa. The anteaters Tamandua and Myrmecophaga have flat, narrow, and rostrocaudally elongate facets, whereas other myrmecophages have oval to triangular facets that are slightly dished. The hyoid apparatus in most myrmecophages is well ossified and placed close to the sternum. The echidna rostral cornu (cornu hyale) is composed of the usual three elements, hypohyal, ceratohyal (= epihyal), and stylohyal, and is distally in contact with the auditory region (Gasc, 1967; Schneider, 1964). The pangolin hyoid apparatus has short rostral comua containing only the hypohyal, and a stylohyal ligament continues to the skull (Gasc, 1967). The anteaters are similar to one another (Fig. 15.4 illustrates the hyoid apparatus of T. mexicana): the elongate stylohyal lies superficially along the side of the neck, nearly reaching the skull in Tamandua and Cyclopes but not Myrmecophaga (Reiss, 1997b). The usual proximal articulation between the rostral cornua and the basihyal is absent;
instead, contact is at the epihyal-ceratohyal junction with the thyrohyal (caudal cornu or cornu branchiale) (Schneider, 1964). This condition is also seen in the armadillo Cabassous (Starck, 1967). The xiphoid processes of the sternum give rise to some extrinsic tongue muscles in echidnas, anteaters, and pangolins. In African pangolins these processes are elongated magnificently (Fig. 15.5). They extend from the sternum down the ventral abdominal wall, follow the curve of the pelvic basin to turn cranially, and end freely at the level of the diaphragm (Ehlers, 1894). In Asian pangolins the xiphoids are shorter but
B
B FIGURE 15.4. Hyoid apparatus of Tamandua mexicana (traced from photographs of CU 17743, Cornell Vertebrate Collections, Ithaca, NY): (A) dorsal view and (B) ventral view. Anterior cornua are especially elongate and, in situ, give skeletal support to the long postcranial pharynx. The proximal ends of the epihyals sit freely on the surface of the corpus hyoideum and the anterior cornu is attached to the hyoid apparatus mainly via the thyrohyal. Scale: 1 cm. See Fig. 15.3 legend for abbreviations.
FIGURE 15.5. Xiphoid process of the sternum in pangolins. (Top) Ventral views of an isolated xiphoid process of (A) an Asian pangolin and (B) an African pangolin. Notice the dramatic difference in the length of the process. (C) The bottom is a right lateral view of the abdomen of an African pangolin lying on its back; the abdominal wall has been cut away to reveal the position of the xiphoid process in situ (drawn from Manis tricuspis, AMNH 86944, American Museum of Natural History, New York, NY). Notice that the xiphoid process runs down the abdominal wall, curves around the pelvic basin, and terminates dorsal to the abdominal viscera; i, intestines; k, kidney; 1, liver; arrow denotes gall bladder.
15. Feeding in Myrmecophagous Mammals variable (Chan, 1995). A synovial joint capsule is present between the caudal sternebra and the xiphoid process (Doran and AUbrook, 1973; Sikes, 1966). 2. Jaw Musculature Mammalian jaw-closing muscles primitively include a masseter, a temporalis, and a pterygoid complex (Crompton, 1989; Davis, 1961; Hiiemae, 1976). The masseter typically has superficial and deep portions, and the deep portion may be further subdivided. The temporalis is usually a large and architecturally complex muscle. The pterygoideus complex is composed of a pterygoideus externus and a pterygoideus internus. All these muscles are derived from the embryonic mandibular muscle plate and are innervated by the mandibular branch of the trigeminal nerve (V3). In monotremes, jaw opening is accomplished by another trigeminal-innervated muscle, the detrahens mandibulae, which has no clear homologue in therians (Edgeworth, 1931; Schulman, 1906). Jaw opening in therians is accomplished by the digastricus, which, as its name suggests, has two bellies. The anterior belly is derived from the embryonic mandibular muscle plate, is innervated by the trigeminal, and is homologous to the monotreme intermandibularis (along with the therian mylohyoideus). The posterior belly is derived from the hyoid muscle plate, is innervated by the facial nerve (VII), and is homologous to the monotreme interhyoideus (Edgeworth, 1931). The adductors of myrmecophages are small and architecturally simple compared with nonmyrmecophagous close relatives. Echidnas and pangolins have superficial and deep components to their small masseter whereas anteaters have only a superficial layer (Edgeworth, 1923; Naples, 1999; Reiss, 1997a,b; Schulman, 1906). The echidna temporalis has a three-part structure that is characteristic of monotremes, but its main portion is small and architecturally simple, as is the temporalis of pangolins and anteaters (Reiss, 1997a; Schulman, 1906). Dasypodine armadillos have robust and architecturally complex adductors for myrmecophages, although they are reduced and simple compared with nonmyrmecophagous armadillos (Edgeworth, 1923; Lubosch, 1908; Reiss, 1997a; Smith and Redford, 1990). Pterygoideus muscles vary among myrmecophages. In the echidna, the pterygoideus internus and externus fuse during development (Edgeworth, 1931). In pangolins, the pterygoideus externus is present, although Edgeworth (1923) claims that the pterygoideus internus atrophies in the embryo. I interpret this muscle to be present and partially merged with the mylohyoideus (Reiss, 1997a). In anteaters and dasypodine arma-
467
dillos, both pterygoideus muscles are well developed and the pterygoideus externus has two distinct heads (Edgeworth, 1923; Naples, 1999; Reiss, 1997a,b; Kiihlhorn, 1939). Jaw-opening muscles are also variable. As mentioned earlier, the echidna has a detrahens mandibulae that runs from in front of the external auditory meatus to the angle of the mandible (Saban, 1971). In pangolins, the anterior portion of the digastric originates from the mandible and splits distally to insert on a salivary gland, the deep surface of the rostral cornu of the hyoid, and the fascia overlying the scalenus anterior (Edgeworth, 1923; Chan, 1995; Reiss, 1997a). The homologue of the posterior digastric forms the outermost layer of the pangolin's glossal tube (Chan, 1995; see Section III,A,3,b). Anteaters also lack the typical compound digastric. Instead, a sternomandibularis muscle runs from the sternum to the dentaries. The compound innervation of this muscle suggests that its cranial portion is homologous to the anterior digastric, and its caudal portion is homologous to the sternohyoideus (Reiss, 1997b). The posterior digastric of anteaters forms an interhyoideus (see Section III,A,3,b). Armadillos have variable digastric anatomy: Dasypus has a sternomandibularis and an interhyoideus like anteaters (Reiss, 1997a,b), whereas Tolypeutes has an anterior digastric that attaches to the transverse tendon of the interhyoideus (Edgeworth, 1923,1935). This apparent interspecific variation could represent individual variation (Reiss, 1997a,b). Finally, all myrmecophages have an unusually well-developed mandibulo-auricularis, a facial muscle that extends from the angle of the mandible (or the squamosal in pangolins) to the tympanic bulla, rostral to the external auditory meatus (Edgeworth, 1923; Reiss, 1997a,b). 3. Tongue a. External Features Tetrapod tongues are epithelium-covered muscular bags, usually anchored proximally to the hyoid apparatus. Most mammals have chunky, dexterous, and moderately extensible tongues (Doran and Baggett, 1971). The dorsal surface is typically covered with numerous specialized epithelial papillae. Conical papillae (filiform and foliate) are the simplest and most numerous and probably increase surface area and act as mechanoreceptors. Fungiform papillae bear specialized taste buds and are interspersed among the conical papillae. Vallate papillae, unique to mammals, are found at the junction of the oral and pharyngeal parts of the tongue and are the most specialized morphologically. They sit in taste bud-littered epithelial depressions into which both mucous and serous glands empty. Much effort has been made to characterize and
468
K a r e n Zich Reiss
classify the distribution of papillae types in mammalian tongues (see references in Doran, 1975). The tongues of echidnas and armadillos, both of which maintain hyoid attachments (Fig. 15.6), are long and thin rostrally, but widen substantially toward the root. The tongues of pangolins and anteaters, free of hyoid attachments, are long and thin throughout, and the diameter increases only slightly as one proceeds from rostral to caudal. The stratum corneum of the dorsal tongue epithelium is unusually thick in myrmecophages (Sonntag, 1925), and echidnas have toothlike keratinized papillae on the base of the tongue (Griffiths, 1968; Sonntag, 1923). Myrmecophages also have what appear to be reservoirs of extra tongue mucosa surface area. The mucosa is folded into numerous
transverse sulci at the base of the tongue of echidnas, pangolins, and anteaters (personal observations; Oppel, 1899). Anteaters and pangolins also have a mucosal hood that overlies the root of the tongue, and in pangolins, tongue mucosa is reflected onto the lining of their unique glossal tube. All types of tongue papillae are sparse in myrmecophages. Echidnas have two vallate papillae and small foliate papillae on the base of the tongue, whereas the free portion of the tongue is bare (Doran and Baggett, 1972; Griffiths, 1968; Oppel, 1899; Sonntag, 1925). Pangolins have three vallate papillae, and sparse fungiform and filiform papillae rostral to these (Kubota et al, 1962b; Oppel, 1899; Sonntag, 1923). All xenarthrans have only two vallate papillae (Sonntag, 1923). In anteaters, the tongue is otherwise bare. Taste buds are found only on the walls of the vallate papillae, and the bases of the vallate papillae bear the openings of serous gland ducts (Kubota et ah, 1962a; Sonntag, 1923). Armadillos have filiform and fungiform papillae scattered all over the tongue and particularly concentrated at its tip. Reports on the distribution of other papillae types and taste buds in armadillos are conflicting (Sonntag, 1923). In some myrmecophages there are specialized structures on the tip of the tongue. The tip of the pangolin tongue bears two well-innervated tactile organs that resemble Pacinian corpuscles (Kubota et ah, 1962b; Doran and AUbrook, 1973; Cheng, 1986). In xenarthrans, the tip of the tongue is richly innervated but no differentiated tactile organs are seen (Sonntag, 1923). In Dasypus, two small muscularized projections extend beneath the tip of the tongue but their histological structure is unknown (Sonntag, 1923). b. Extrinsic Musculature
F I G U R E 15.6. The tongue, soft palate, and oropharyngeal specializations of myrmecophages. Schematic illustration comparing (A) the echidna Tachyglossus, (B) the pangolin Manis, (C) the armadillo Dasypus, (D) the anteater Cyclopes, and (E) the anteater Myrmecophaga. The skull, hyoid, and laryngeal elements are stippled, the soft palate is hatched, and the tongue is bold. The esophagus (e) and the trachea (t) are illustrated in A through E but labeled only in A. A sternoglossus muscle, formed from fused hyoglossus and sternohyoideus muscles, makes up the tongue in all but the armadillo, but note that the echidna retains remnants of the primitive muscle attachments to the hyoid. The sternoglossus is entirely free from the hyoid in the other taxa. Also notice that the ventral muscular sheet, indicated by the large arrowhead, terminates on the hyoid in the echidna and armadillo, but continues caudally in the other taxa. This sheet contains the mylohyoideus and interhyoideus, which contribute to a muscular sheath that surrounds the tongue and pharynx of anteaters and pangolins. Finally, anteaters and pangolins possess a specialization of the oral muscosa called the prehyoid pouch, indicated by the asterisk, that overlies the root of the tongue, but that only pangolins have a glossal tube (shaded) that completely surrounds the caudal portions of the tongue.
The extrinsic musculature of the mammalian tongue includes muscles that originate from outside the tongue and terminate in the tongue (the -glossus muscles) and muscles that insert on the hyoid and indirectly affect movements of the tongue (the -hyoid muscles). The first group includes the genioglossus, hyoglossus, and styloglossus, all derived from embryonic hypobranchial musculature and innervated by the hypoglossal nerve (XII). Also in this group is the palatoglossus, derived from caudal arch branchiomeric musculature and innervated by the vagus nerve (X). Muscles of the second group have diverse embryonic origins and innervations. The mylohyoideus is derived from the embryonic mandibular muscle plate (which also gives rise to the anterior digastric) and is innervated by the trigeminal (V3). The stylohyoideus is derived from the hyoid muscle plate (which also gives rise to the posterior digastric) and is innervated by the facial nerve
469
15. Feeding in Myrmecophagous Mammals
1973), but in pangolins and anteaters the sternoglossus bypasses the hyoid apparatus entirely (Chan, 1995; Doran and AUbrook, 1973; Owen, 1862; Pouchet, 1867; Naples, 1999; Reiss, 1997a,b; Sikes, 1966). The sternoglossus is probably derived from a fusion of hyoglossus and sternohyoideus: it is innervated by the descending branch of the hypoglossal nerve, the presumed ancestral muscles appear to be missing (except see the anteater hyoglossus discussed later), and the morphology in echidnas suggests an intermediate condition. Myrmecophages with a sternoglossus muscle have modified the rest of their feeding apparatus so that most of the primitive extrinsic tongue muscles have little to do with the tongue. There are two common patterns to the modifications seen. First, a muscular sheath formed in part by homologues of the primitive intermandibularis and interhyoideus surrounds the tongue and may contribute circumferential fibers to its base (see Section III,A,3,c). Second, although some of the extrinsic tongue and hyoid muscles are unusually modified, others appear to be altogether absent. In echidnas (Fig. 15.7) the intermandibularis (myloglossus of Doran, 1973; mylohyoid of Duvernoy, 1830) arises from the dentaries, the hard palate, and distal tip of the rostral cornua of the hyoid apparatus (Duvernoy, 1830; Edgeworth, 1935). It is bilaminar (the deep layer is probably the annulus inferior of Doran), reminiscent of the differentiation of its therian homologue into the anterior digastric and mylohyoideus. It is continuous caudally with the interhyoideus (styloglossus of Doran), which arises from the skull and the rostral cornua (Doran, 1973; Edgeworth, 1931). The superficial
(VII). The geniohyoideus, sternohyoideus, sternothyroideus, and thyrohyoideus are hypobranchial muscles that are innervated by the descending branch of the hypoglossal nerve (XII). There is much conflict an\ong reports on the extrinsic musculature of echidna tongues (Doran, 1973; Duvernoy, 1830; Edgeworth, 1931,1935; Schulman, 1906), pangolin tongues (Chan, 1995; Cheng, 1986; Doran and AUbrook, 1973; Edgeworth, 1923; Ehlers, 1894; Nene, 1978; Reiss, 1997a; Sikes, 1966; Sonntag, 1923), and anteater tongues (Duvernoy, 1830; Owen, 1862; Pouchet, 1867; Naples, 1999; Reiss, 1997b). My own dissections of anteaters and pangolins, along with careful scrutiny of the literature, have suggested that at least some of the conflict is attributable to differing assessments of muscle homologies and varying anatomical nomenclature. To avoid compounding the confusion I make homology arguments explicit wherever there are points of contention and give muscle synonymies throughout the following text. Table 15.2 contains a summary of the conditions of the extrinsic tongue musculature in echidnas, anteaters, pangolins, and armadillos. The bulk of the tongue of echidnas, pangolins, and anteaters is an unusual muscle called the sternoglossus (see Fig. 15.6), also present in some nectarivous bats (Griffiths, 1982). The sternoglossus arises from the xiphoid process of the sternum, passes through the thorax on the deep surface of the sternum, passes through the neck ventral to the larynx and hyoid apparatus, and ultimately rises up into the oral cavity to form the free part of the tongue. In echidnas, some sternoglossus muscle fibers attach to the basihyal (Doran,
TABLE 15.2 Homologies of Tongue and Hyoid Muscles of Echidnas, Pangolins, Anteaters, and Armadillos^ Anteaters
Armadillos
modified +, in soft palate
sternomandibularis +, in soft palate
+ /sternomandibularis
interhyoideus
In glossal tube
interhyoideus
+ /interhyoideus
Genioglossus Hyoglossus Styloglossus Palatoglossus
+ +, sternoglossus
+, in glossal tube iternoglossus
+ +, sternoglossus
+, not in tongue
+, not in tongue
Geniohyoideus Sternohyoideus Sternothyroideus Thyrohyoideus
+ +, as sternoglossus +
+, glossal tube sternoglossus
+ sernoglossus + +
Muscle
Echidnas
Anterior digastric Mylohyoideus
intermandibularis intermandibularis
Posterior digastric Stylohyoideus
Sternoglossus
+
Pangolins
+ + +
+
+
+ + + + + + + ~
^A plus sign indicates the muscle is present and unmodified from the primitive condition; modifications from the primitive condition are described; a slash indicates anatomical variation within the taxon; and a minus sign indicates that no homolog is present.
470
K a r e n Zich Reiss
FIGURE 15.7. Ventral view of echidna cranial musculature (A and B from Schulman, 1906; C from Edgeworth, 1945). Progressively deeper dissections of echidna throat muscualture are shown. Abbreviations follow the nomenclature of the original authors: a, annulis intimus; b, buccinator; ch, ceratohyal; dm, detrahens mandibulae; gge, genioglossus extemus; ggl, genioglossus lateralis; gh, geniohyoideus; ih, interhyoideus; ims, intermandibularis; m, masseter; md, mandible; oh, omohyoideus; pi, pterygoideus lateralis; sg, stemoglossus; sm, stemomastoideus; st, stemothyroideus; t, temporalis; zm, zygomaticomandibularis. Notice, in particular, the contributions of the intermandibularis and interhyoideus to a sheath that surrounds the entire tongue as well as to the circumferential muscualture that is applied to the surface of the stemoglossus muscles. These are structural characteristics common to echidnas, anteaters, and pangolins.
portions of both these muscles insert onto a ventral midline raphe and the basihyal, whereas deeper fibers wrap around the lateral edges of the sternoglossi and insert on the ventral surface of the base of the tongue (Doran, 1973; Duvemoy, 1830). The genioglossus arises from the mandible; its median portion attaches directly to the ventral tongue base, but lateral fasiculi wrap around the lateral edge of each stemoglossus before entering the base of the tongue (Duvemoy, 1830; Saban, 1971). The geniohyoideus is not unusual and both styloglossus and palatoglossus are absent (Edgeworth, 1931,1935; Saban, 1971). The sheath surrounding the tongue of pangolins (Ehlers, 1894; Sikes, 1966) has been dubbed the glossal tube (Chan, 1995) (Fig. 15.8; also see Fig. 15.6). It is formed, in part, by contributions from the posterior digastric, geniohyoideus, and genioglossus (Chan, 1995; Edgeworth, 1923, 1935; Reiss, 1997a; Sikes, 1966). The glossal tube begins in the oropharynx and extends into the neck ventral to the hyoid apparatus. It terminates by attaching to the surface of the paired stemoglossus muscles at a level reported to be species specific (Chan, 1995) but that I have found to vary among individuals of the same species (personal observations). The lumen of the oral cavity extends into the glossal tube for its
full extent, so the tongue lies first within the oral cavity and then within the glossal tube. The glossal tube is made up of four concentric layers that are muscle ventrally and fascia dorsally and an inner layer of connective tissue and epithelium that is continuous with the oral mucosa (Chan, 1995). The outermost muscular layer (hyoglossus of Cheng, 1986; glossovaginalis superficialis of Ehlers, 1894) arises from the rostral cornu of the hyoid and is cranially continuous with an extensive mylohyoideus that arises from the dentaries, palatines, pterygoids, and, remarkably, the soft palate (Reiss, 1997b). This outer layer has transverse fibers, is innervated by the facial nerve (VII) and is the likely homologue of the posterior digastric. The next two layers of the glossal tube (glossovaginalis externum stratum externum and medium of Ehlers) are formed by the geniohyoideus, which arises from the mandible and splits distally into layers of longitudinal muscle fibers. The innermost muscular layer (glossovaginalis externum stratum internum of Ehlers) is formed by the genioglossus, which arises from the mandible and diverges so its muscle fibers lie transversely. There is some question as to the ultimate insertion of the geniohyoideus- and genioglossus-derived layers of the glossal tube. Doran describes a disk of dense
471
15. Feeding in Myrmecophagous Mammals
ghl(II) ghm(II) ghm(II)
gg(IV) fp
ghl(IIl) V FIGURE 15.8. Ventral view of pangolin cranial musculature (from Chan, 1995). The illustration on the left depicts the caudal extent of the mylohyoideus and its continuity with the outer layer of the glossal tube, derived from the posterior digastric. These muscles are reflected in the illustration on the right revealing the deeper layers of the glossal tube that surround the sternoglossus muscles. Abbreviations follow the nomenclature of the original author: fp, free portion of tongue; gg, genioglossus; gh, geniohyoideus; gh, geniohyoideus lateralis; ghm, geniohyoideus medialis; gpn, glossopharyngeal nerve; gt, glossal tube layers I-IV; he, hyoid cartilage; hn, hypoglossal nerve; hnd, descending branch of hypoglossal nerve; la, lingual artery; In, lingual nerve; mh, mylohyoideus; mha, mylohyoideus aponeurosis; mhp, mylohyoideus, pterygoid portion; mht, mylohyoideus, tympanic portion; nmh, nerve to mylohyoid and anterior digastric; sm, sternomastoideus; smd, submandibular duct; smg, submandibular gland; slg, sublingual gland; st, sternothyroideus; th, thyrohyoideus; tmh, tympanohyoid; vh, ventral hiatus.
connective tissue interrupting the fiber continuity of the sternoglossus and to which the caudal ends of the geniohyoideus and genioglossus are attached (Doran and AUbrook, 1973). Chan (1995) did not see this structure and claims that the geniohyoideus dissipates in loose connective tissue on the surface of the sternoglossus whereas the genioglossus inserts on the ventral midline. My dissections of three species of pangolin are in accordance with Chan (Reiss, 1997a). Edgeworth (1923, 1935) claims that the palatoglossus is absent in pangolins but that they have a unique pterygohyoideus muscle. I disagree with his interpretation and view his pterygohyoideus as a palatoglossus with an unusual origin from the hyoid apparatus resembling the conditon seen in anteaters (Reiss, 1997a,b). The styloglossus and stylohyoideus are absent in pangolins (Edgeworth, 1923; Reiss, 1997a). The muscular sheath formed by the mylohyoideus and interhyoideus that surrounds the anteater tongue also lines the oropharynx [Fig. 15.9; also see Reiss (1997b) for additional figures of Tamandua cranial my-
ology]. As in pangolins, the mylohyoideus arises from the dentaries, the hard palate, and the soft palate and is continuous with the interhyoideus (interstylohyoideus of Naples, 1999; ceratohyoideus of Owen, 1862). The interhyoideus arises from the rostral cornua of the hyoid apparatus and runs ventrally to meet its antimer via a stout tendon that crosses the ventral midline. It is innervated by the facial nerve and is the homologue of the posterior digastric (and the outer layer of the pangolin glossal tube). The geniohyoideus, which has the usual attachments, and sternomandibularis (discussed earlier) are closely applied to the sheet of muscle formed by the mylohyoideus and interhyoideus (Reiss, 1997b). Thus, in the neck, the sternoglossus lies in a space that is defined dorsally by pharyngeal mucosa and the hyoid and laryngeal elements, and laterally and ventrally by the multilayered sheet that consists of mylohyoideus, geniohyoideus, and interhyoideus. The genioglossus lies on the dorsal surface of this sheet and rises up between the paired sternoglossus muscles to form the frenulum (Reiss, 1997b, musculaire du frein
472
K a r e n Zich Reiss
F I G U R E 15.9. Schematic parasagittal view of the interior of the oropharynx of Tamandua mexicana. Rostral is to the left. The vertebral column and associated musculature, the dorsal wall of the pharynx, and the ventral stemomandibularis have been removed. The skull is tilted away from the page and the soft palate is pulled dorsally to afford view of the ventral surfaces of the hard and soft palate; the asterisk indicates where the left stemoglossus was cut and removed. Note the extensive contributions of mylohyoideus to the wall of the oropharynx and the mylohyoideus and stylopharyngeus to the soft palate. Also note that the hyoglossus and the palatoglossus terminate in the oropharyngeal walls.
of Pouchet, 1867), more lateral fibers bind the sternoglossi together where they rise up into the oropharynx (musculaire annulaire of Pouchet, 1867) and envelop each stemoglossus caudal to the root of the tongue (ringmuskel of Dingier, 1964a; musculaire spiraux of Pouchet, 1867). The hyoglossus is present as a distinct muscle (epihyoglossus of Owen, 1862; part of the hyobuccalis of Pouchet, 1867), even though its anlage also contributes to the stemoglossus. It originates from the free proximal end of the epihyal, runs forward in the floor of the pharynx dorsal to the still submerged stemoglossus, and dissipates in the mucosal hood overlying the root of the tongue. This hood forms the roof of a space that Owen calls the prehyoid pouch (Owen, 1862), present in both anteaters and pangolins (see Fig. 15.6). The styloglossus and stylohyoideus are absent in anteaters, and the palatoglossus is modified like that of pangolins (although it does not reach the palate in Tamandua). Armadillos do not have a stemoglossus and their extrinsic tongue muscles are fairly typical for eutherians (Doran, 1975; Edgeworth, 1923; Saban, 1968; Smith and Redford, 1990; Sonntag, 1923; Windle and Parsons, 1899). However, they do lack a palatoglossus and a stylohyoideus (Edgeworth, 1923; 1935; Reiss, 1997a). c. Internal Structure of the Tongue In addition to the extrinsic muscles that originate outside the tongue, the mammalian tongue contains in-
trinsic muscle fibers that both begin and end in the tongue. These intrinsic fibers, the lingualis propria, usually consist of transverse and vertical bundles of muscle. These, along with the longitudinal fibers provided by extrinsic muscles, are complexly organized by tongue connective tissues and give the mammalian tongue its notable dexterity. Primitively, a median connective tissue septum separates the right and left extrinsic muscles, and within this connective tissue usually lies the lingual artery and vein. The tongue is permeated by nerve fibers serving motor and general and special (taste) sensory functions. This general description is fairly representative of the armadillo tongue because armadillos do not have a stemoglossus (Edgeworth, 1923). In contrast, the internal structure of the tongue of echidnas, pangolins, and anteaters is comparatively simpler (Fig. 15.10). In general, the paired sternoglossi are dominant and the lingualis propria is sparse at the base of the free portion of the tongue. As one proceeds toward the tip of the tongue, the sternoglossi subdivide into a variable number of longitudinal muscle bundles and the intrinsic fibers become progressively more abundant. Myrmecophages differ in the source and geometry of the circumferential fibers, the geometry of the subdividing sternoglossi, and in the presence of specialized connective tissue structures (lyttae) in the tongue's tip. In the echidna, a layer of spiraling muscle fibers, the annulis intimis (annular muscle of Edgeworth, 1931; not the annulus inferior of Doran, 1973), is closely apposed to each stemoglossus (Duvernoy, 1830; Griffiths, 1968; Saban, 1971). These fibers appear to originate and insert on tongue connective tissues, although this is not explicitly stated by any author. In the free portion of the tongue, a tortuous artery lies between the sternoglossi and communicates with extensive vascular cavernous sinuses that lie in both the dorsal and the ventral midline of the free portion of the tongue. Some circumferential muscle fibers attach to the walls of the sinuses, and transverse and vertical fibers are sparse (Doran and AUbrook, 1973; Doran and Baggett, 1970). The origin of the circumferential fibers at the base of the pangolin tongue is siniilarly unclear, although at least some appear to have extrinsic origin (personal observations). The circumferential fibers attenuate rostrally and the sternoglossi subdivide into 20 peripherally placed bundles between which lie interwoven vertical and transverse fibers of the lingualis propria (Dingier, 1964a; Doran and AUbrook, 1973). Further rostrad, the tongue flattens and 10 bundles of longitudinal fibers lie on each side of a central connective tissue lytta. Vertical and transverse fibers are sparse and lie deep to the longitudinal muscles. The lytta surrounds some loosely packed longitudinal muscle fibers, a central artery (Cheng, 1986; Kubota et ah,
473
15. Feeding in Myrmecophagous Mammals
B
echidna
pangolin
anteater
FIGURE 15.10. Schematic transverse sections through the tongue of echidnas, anteaters, and pangolins. All illustrations are schematics drawn from published micrographs: echidna after Griffiths (1968) and Doran (1970); pangolin after Doran (1973) and Dingier (1964b); and anteater after Dingier (1964a). (A) A section near the root of the tongue, (B) near the middle of the free portion of the tongue, and (C) a section from the tip of the tongue. Longitudinal muscle bundles are shaded darkly; lingual nerves are solid circles; wavy circles represent vascular sinuses; and all other internal lines represent muscle fiber directions. Myrmecophages have a tongue muscle architecture that is simpler than most mammals. Longitudinal muscle fibers derived from the paired sternoglossus muscles are dominant, circumferential muscle fibers derived in part from extrinsic muscles surround these, and the transverse and vertical intrinsic fibers are sparse throughout but become more abundant toward the tip of the tongue. Section A of the echidna tongue shows the dual contribution of sternoglossus and hyoglossus (see Fig. 15.6 for a lateral view); other notable differences between taxa include the geometry of the subdividing bundles of longitudinal fibers and the presence of possibly erectile vascular sinuses in the tip of the tongue of echidnas and pangolins but not anteaters.
1962b), and small vascular sinuses (Doran and Allbrook, 1973). The fibers that initially envelop the anteater sternoglossi are clearly derived from the genioglossus (Dingier, 1964; Pouchet, 1867; Reiss, 1997b). Rostrally, the sternoglossi subdivide repeatedly into peripheral longitudinal muscle bundles, and transverse and vertical fibers are interwoven throughout the central portion of the tongue (Dingier, 1964a). These bundles extend the length of the tongue, although their number progressively diminishes. Toward the tip of the tongue, radial fibers dominate the tongue's core and there is no centrally placed connective tissue lytta or vascular sinus (Kubota et al, 1962a). 4. Pharynx and Soft Palate The pharynx is a tube of variable length that is continuous with the oral and nasal cavities. Its lateral and dorsal walls are muscularized and its floor is formed by the root of the tongue and the hyolaryngeal struc-
tures. Pharyngeal muscles consist of a set of transversely oriented constrictors and a set of retractors or dilatators that tend to be more longitudinal in orientation. Eutherians typically have a superior constrictor that originates from the skull and various cranial fascia and ligaments, a middle constrictor that originates from hyoid elements, and an inferior constrictor that originates from laryngeal elements. Each of these is often subdivided further into several muscles named for the precise site of origin. The retractor/dilatators originate from various places on the skull and run more or less longitudinally to insert on the dorsolateral pharynx. A stylopharyngeus and a palatopharyngeus are fairly consistently present. Marsupials differ from eutherians in that they lack a superior pharyngeal constrictor and their stylopharyngeus has both transverse and longitudinally oriented fibers (Edgeworth, 1916,1935; Smith, 1992,1994). The stylopharyngeus is innervated by the glossopharyngeal nerve (IX). All other pharyngeal muscles are innervated by the vagus nerve (X). The soft palate is a muscular structure that divides
474
Karen Zich Reiss
the rostral pharynx into ventral oral and dorsal nasal portions. It originates from the caudal hard palate and is usually suspended between the caudally extending pterygoid processes. Eutherian palatal muscles include the tensor veli palatini (innervated by the mandibular branch of the trigeminal nerve, V3) and levator veli palatini (whose innervation is controversial, but is probably from both facial and vagus nerves, VII and X) (Dellon, 1989; Furusawa et ah, 1991a,b; Hecht et al, 1993; Ibuki et al, 1978; Keller et al, 1984; Nishio et ah, 1976b; Van Loveren et al, 1983). Both of these muscles arise from the pterygoids and the auditory tube and insert on the connective tissue of the soft palate. Marsupials do not have a levator veli palatini (Edgeworth, 1935; Saban, 1968). The medialis veli palatini (= uvularis, palatinus) arises from the midline of the hard palate and runs caudally in the midline of the soft palate. The palatopharyngeus (mentioned earlier) spans the wall between the palate and the pharynx, and finally, a pterygopalatinus sometimes joins the other palatal muscles arising from the pterygoids. These muscles are all innervated by the vagus nerve (X). The pharynx and soft palate of echidnas are notably simpler than the preceding description based on therians. There is no superior constrictor and the stylopharyngeus is wholly transverse in orientation. An undifferentiated constrictor pharyngeus lies caudally and the palatopharyngeus is the sole longitudinal muscle (Edgeworth, 1935). The soft palate is composed mainly of connective tissue (Edgeworth, 1916). Edgeworth describes the pangolin pharynx to be typically eutherian with a full set of pharyngeal constrictors and dilatators that include a wholly longitudinal stylopharyngeus and a palatopharyngeus (Edgeworth, 1923) but my results differ. I could not identify a palatopharyngeus and found the pangolin stylopharyngeus to have both transverse and longitudinal bundles characteristic of the marsupial condition (Reiss, 1997a). Also, Edgeworth noted only the presence of a tensor veli palatini in pangolins (Edgeworth, 1923), whereas I found that the mylohyoideus and the transverse component of stylopharyngeus also enter the soft palate (Reiss, 1997a). Perhaps the messy interdigitation of palatal and pharyngeal muscle fibers and the deep position of the palatal portions of mylohyoideus and stylopharyngeus account for our differing opinions. Anteaters also have a two-part stylopharyngeus, and its transverse portion joins the mylohyoideus in the soft palate (Reiss, 1997b), as in pangolins. A superior pharyngeal constrictor is present (although weak in some species) palatopharyngeus is present, and the tensor and levator veli palatini vary among species (Reiss, 1997b).
Armadillos, like anteaters and pangolins, have a two-part stylopharyngeus (Edgeworth, 1923; Reiss, 1997a). It does not enter the soft palate, however, and in all other ways, armadillos have characteristic eutherian pharyngeal and palatal anatomy (Reiss, 1997a). B. Exceptions to the Morphotype The feeding apparatus of the more derived eutherian myrmecophages is generally indicative of ancestry, e.g., the aardwolf looks more or less like a hyena. Myrmecophagous squirrels and primates have no unusual anatomical features, whereas numbats, aardvarks, and aardwolves each have a few tell-tale features that are uncharacteristic of their lineage and probably represent myrmecophagous specializations. Numbats have in common with the specialized myrmecophages a long snout and hard palate, as well as simple cheek teeth that are variable in number. The mandible is reduced only slightly and the forelimbs are not particularly robust (Calaby, 1960; Fleay, 1942). Their tongue is long and flattened and the epithelium is highly keratinized. They have no sternoglossus, but cross sections of the tongue show numerous peripheral bundles of longitudinal muscle between which lie a dense array of vertical and transverse fibers (Griffiths, 1968). Aardwolf skulls are generally hyena-like, but both the skull and the hard palate are significantly wider than similar-sized relatives (Anderson et al, 1992). The cheek teeth are peg-like and do not occlude when the jaws are closed, although the jaw musculature is nevertheless robust. The aardwolf tongue is somewhat longer than that of its close relatives, but is mostly distinguished by having numerous highly cornified papillae (Anderson et al, 1992). Aardvark skulls have an elongate rostrum, but the mandible is L-shaped and has typical muscular processes. The dentition is reduced and similar to myrmecophagous armadillos in that cheek teeth are homodont, peg-like, and lack enamel. However, aardvark teeth are unique among Mammalia in having parallel pulp-filled tubules of dentine that give the group its ordinal name, Tubulidentata. Jaw musculature is robust and well differentiated, and a digastric is present (Frick, 1951). The aardvark tongue is thick at the root and pointed at the apex, and it bears filiform, fungiform, and three vallate papillae, but no terminal papillary concentrations or specializations (Sonntag, 1923). The hyoid musculature is typical for Epitheria, e.g., hypobranchial muscles are all present and attach to the hyoid, and a styloglossus and stylohyoideus are both present (Edgeworth, 1924; Frick, 1951; Shoshani, 1993).
15. Feeding in Myrmecophagous Mammals
475
IV. FUNCTIONAL MORPHOLOGY
B. Tongue Movements
Patterson (1975) suggested that the main functional requirement of the myrmecophage feeding apparatus is speed of ingestion. Myrmecophages must ingest a large number of individual prey daily, and if myrmecophages are sensitive to prey defenses then consuming prey quickly before colony defenses ensue will also be advantageous. Testing the hypothesis that the myrmecophage feeding apparatus is designed for speed (or any other adaptationist hypothesis) requires a basic understanding of the functional correlates of structure. Unfortunately, dynamic studies of myrmecophage feeding are lacking except for a single study that compares feeding in a dasypodine and a euphractine armadillo (Smith and Redford, 1990). This study's singularity gives it importance, but because armadillos lack a sternoglossus muscle it tells us little about echidnas, anteaters, or pangolins. The absence of functional morphological analyses of feeding in these latter taxa requires that the following section be based largely on inference from structure. It begins with a consideration of the possible functional consequences of the unusual structural features of the jaws, tongue, pharynx, and soft palate. This is followed by a brief consideration of how the basic stages of tetrapod feeding—lingual ingestion, prey processing, intraoral food transport, and swallowing—might be modified in myrmecophages.
Myrmecophages are noted for their long tongues capable of extreme protrusion. Two issues arise: (1) How is it protruded so far? and (2) how is it retracted? Tongue protrusion in mammals results, in part, from shape changes affected by the intrinsic musculature. Tongues can be viewed as constant volume sacs of muscle that function as a muscular hydrostat (Kier and Smith, 1985). Intrinsic muscle fiber orientation determines the kinds of shape changes that can occur, and the resting size and shape of the tongue determine the magnitude of shape changes that can occur. In the case of myrmecophage tongues, contraction of circumferential muscle, both intrinsic and extrinsic, will decrease the cross-sectional area of the tongue, which will be compensated for by an increase in length such that the total volume of the tongue remains constant. Because of geometric considerations, a tongue that is initially long and thin will get much longer in response to a relatively little shortening of circumferential fibers than would a shorter tongue under the same circumstances (Kier and Smith, 1985,1992). In myrmecophages, the level at which extrinsic muscles apply circumferential fibers to the sternoglossi, which is always caudal to the first appearance of intrinsic circumferential fibers, should determine the length of the tongue acting as a muscular hydrostat. Tongue protrusion in mammals is also partially dependent on the extrinsic muscles genioglossus and geniohyoideus pulling the base of the tongue forward (Crompton et al, 1977). A tongue free of hyoid attachments should have a more mobile tongue base than one that retains hyoid attachments. The long skull base, the caudal position of the root of the tongue, and reports that genioglossus fibers run caudally after entering the tongue in anteaters and pangolins (Dingier, 1964b) should all contribute to the absolute value of this displacement. The effect of extrinsic muscle contraction in echidnas and armadillos will depend largely on hyoid mobility, and because the echidna hyoid is anchored to the skull, their extrinsic muscles may not have much of an effect on tongue protrusion. The pangolin glossal tube is unique and the role of extrinsic musculature in pangolin tongue protrusion may be more complex. By manipulating fixed specimens, Chan (1995) observed that the glossal tube is everted and obliterated when the tongue protrudes, and he proposed that its function was to provide extra mucosa to the lengthening tongue. This may indeed be a role of the structure but it does not explain why the tube is so well muscularized. The orthogonal muscle fiber orientations in different layers of the glossal tube and the attachment of the longitudinal layer to caudal
A. Jaw Movements Myrmecophages have a small gape, relatively flat jaw joint facets, tooth reduction, and an overall reduction in jaw muscle mass compared with close relatives. These features, along with evidence from anteater stomach contents (Lubin et al, 1977; Montgomery, 1985), suggest that mastication is absent. At the same time, the degree of jaw muscle differentiation suggests that some mandibular movements do occur. The shape of the jaw joint, the fiber direction of the masticatory muscles, and the robust pterygoid muscles suggest that mandibular movements are primarily parallel to the skull base and may include protraction/retraction, medial/lateral movement, and/or rotation. In echidnas, manipulation of fresh specimens has led to the hypothesis that mouth opening is the result of the rotation of the dentary along its long axis (Murray, 1981). Due to the curvature of the dentaries, rotation causes the tips of the dentaries to move away from one another, thus opening the mouth. This hypothesis has recently been advanced for the anteater Myrmecophaga as well (Naples, 1999).
476
Karen Zich Reiss
portions of the sternoglossi suggest that contraction of the glossal tube may actively extrude the sternoglossi. Assuming that the different muscles of the tube can contract independently, contraction of layers with circumferential fibers and the attending sternoglossus shape change would affect partial tongue protrusion, whereas full tongue extension would require the added contraction of the longitudinal layer, which would evert the tube. The portion of the sternoglossus caudal to the attachment of extrinsic fibers presumably functions as a tongue retractor, as does its primitive homologue, the sternohyoideus (Crompton et al, 1977). Because a contracting muscle can shorten to only a finite fraction of its resting length (Gans and Bock, 1965), greater absolute shortening requires a longer muscle. In addition, the tongue is at a mechanical disadvantage when fully protruded due to the relationship between muscle fiber length and half-sarcomere overlap (Gans, 1982; Gordon et al, 1966). Any added tongue length not involved in the muscular hydrostat (i.e., not surrounded by circumferential muscle) reduces the average extension experienced over the entire muscle, which allows more forceful contraction upon retraction. These arguments parallel Smith's discussions of the biomechanical implications of tongue muscle length in lizards with protrusible tongues (e.g.. Smith, 1986) and are based on a simple model in which the sternoglossus is made up of long and parallel muscle fibers. Muscle fiber lengths, fiber type, filament and fiber packing, connective tissue architecture, and motor unit structure have complex and not fully understood effects on whole muscle behavior (Gans and Gaunt, 1991, 1992, and references contained therein) and not one of these factors has been investigated for the sternoglossus of any myrmecophage. Most myrmecophagous species have extra lingual mucosa in the form of mucosal pleats or the glossal tube so it is unlikely that the lingual tunic mechanically limits absolute displacement. Also, both pangolins and anteaters exhibit a behavior that appears somewhat like a yawn and involves a very slow protrusion of the tongue to a distance that is roughly twice that observed when wild Tamandua are feeding on surface-foraging termites, captive Manis and Myrmecophaga are feeding from a bowl of mush, or captive Myrmecophaga are feeding through clear, flexible, plastic tubing (personal observations). This suggests that mechanical limits may be irrelevant to normal feeding behavior. Unfortunately, when wild myrmecophages are feeding from nests, tongue displacement is obscured by the nest itself. The myrmecophage tongue appears designed for large excursions but excursion has its costs. For one.
excursion takes time [see Kier and Smith (1985) for a theoretical discussion of this issue in muscular hydrostats] and if speed is an important factor in myrmecophage feeding then the tongue should protrude as little as is sufficient in any given situation. That captive animals can protrude their tongue farther than they typically do while feeding is consistent with this expectation. However, there has been no empirical description of the kinematic relationship between displacement and speed in any myrmecophage with a sternoglossus. Discussions of muscular hydrostats emphasize the role of a structure's length/width ratio (e.g., Kier and Smith, 1985, 1992) in achieving displacement, but the thinness of myrmecophage tongues may also serve to minimize inertia, which would have effects on speed. Also, muscle fiber physiology and muscle architecture can affect muscle contraction velocity, but as mentioned earlier, these factors remain unexplored. Thin tongues are, in part, the result of fewer muscles in the tongue and this presumably has effects on tongue dexterity. In most mammals, the spatially diverse origins of the extrinsic muscles and the diversity of fiber orientations in the lingualis propria give remarkable dexterity to the tongue. In myrmecophages with a sternoglossus, the absence of any other -glossus muscle, the predominance of circumferential fibers, and sparse vertical and transverse intrinsic fibers all suggest that tongue dexterity is minimal. Popular accounts depict the myrmecophage tongue as prehensile and dexterous, snaking its way through labyrinthine tunnels in ant and termite colonies. However, such meandering could also result from the tongue passively following the contours of prey nests. My own preliminary experiments filming Myrmecophaga tongues in curved clear tubing suggest that this is the case. Moreover, it appears that one thing anteaters cannot do is lick their own snout deliberately (personal observations). C. Pharynx and Soft Palate Movements Echidnas, anteaters, and pangolins all have a muscle sheet derived from the intermandibularis (or mylohyoideus) that at least partially surrounds the oropharynx. This sheet of transverse muscle extends into the soft palate in anteaters and pangolins, forming a nearly complete ring of muscle. Its lateral portions are topographically reminiscent of the buccinator muscle and could function similarly, squeezing prey from the buccal chamber back into the lingual chamber of the caudal oral cavity. It is also reminiscent of reptilian pharyngeal constrictors, particularly in anteaters which have a long oropharynx, and could conceivably aid in prey transport. Finally, it may serve to elevate the tongue's base, a function usually carried out by
15. Feeding in Myrmecophagous Mammals muscles that are absent in anteaters and pangolins (styloglossus and palatoglossus). This has been proposed as the mechanism by which echidnas chew, apposing their robust tongue pad against keratinous palatal teeth (Griffiths, 1968). In pangolins, anteaters, and armadillos (as well as in echidnas) the stylopharyngeus has transverse fibers, which suggest function as a pharyngeal constrictor. In pangolins and anteaters, these transverse fibers also enter the soft palate and, along with palatopharyngeus, form a ring of musculature around the soft palate and nasopharynx. This well-developed palatopharyngeal sphincter would, at the very least, prevent live unmasticated prey from crawling up into the nasopharynx. Another unusual feature of anteaters and pangolins is the prehyoid pouch. In reference to Myrmecophaga, Owen (1862) suggested that it was a distensible holding chamber for the accumulation of prey items prior to swallowing. If this is true, the muscularization of the roof of this chamber in both Tamandua and Myrmecophaga could facilitate its eversion and emptying, but there is no muscle in the prehyoid pouch of Cyclopes or pangolins. D . Feeding Stages in Myrmecophages Mammalian feeding is generally broken down into four stages: lingual ingestion, prey processing, intraoral food transport, and swallowing. Myrmecophages differ from most other mammals in the relative importance of lingual ingestion and prey processing. 1. Lingual
Ingestion
Lingual ingestion involves tongue protrusion and retraction, which have been discussed earlier. Other important considerations are the roles of sensory cures in prey detection and the mechanics of prey acquisition. As discussed previously, initial orientation to prey appears to be olfactory mediated and, in some species, electroreception may also play a role. At closer range, the absence of taste buds or other chemoreceptors on the tongue's tip suggests that chemical clues are not a factor. However, the role of tactile organs in some species needs explanation. With respect to mechanisms of acquisition, prey are generally observed all over the length of the anteater tongue (personal observations), suggesting that local deformation of the tongue's surface, as is seen in fluid lapping in other mammals, is not occurring. Myrmecophage saliva seems the most likely n\eans of prey acquisition. It is extremely viscous and prey may simply stick to the tongue. Additionally, myrmecophage saliva could be chemoattractive to prey.
477
2. Prey Processing In most mammals, prey processing involves mastication with occluding dentition. Echidnas lack true teeth and so do not masticate in the usual sense, but can be observed chewing using cornified structures on their tongue base and hard palate (Crompton, personal communication). The absence of teeth in pangolins and anteaters, along with the presence of whole prey in the stomach, argues strongly for an absence of mastication in these species. Even in myrmecophages with teeth (armadillos, numbats, aardvarks, aardwolves, and Funisciurus squirrels), observations of feeding and stomach content analyses suggest that chewing occurs rarely. Northern populations of Dasypus novemcinctus are generalized insectivores that can be seen laboriously chewing insect prey both in the wild and in captivity (personal observations; Smith and Redford, 1990), whereas southern myrmecophagous conspecifics do not chew (Redford, 1987). Aardvarks eat and chew an African wild cucuniber {Cucumis humifructus) in what may be a symbiotic relationship (Meeuse, 1963), but they do not chew their usual majority diet of termites and some ants (Kingdon, 1974; Patterson, 1975). Aardwolves have not been observed chewing, but do use their teeth in intraspecific social displays and defense (Ewer, 1973; Richardson, 1985, 1991). Finally, numbats and Funisciurus squirrels do not chew when feeding on social insects (Emmons, 1975). These observations suggest that chewing is either disadvantageous or not necessary for myrmecophages. Chewing would slow down the rapid ingestion of prey (Emmons, 1975; Redford, 1987), prey chemical defenses might be released and irritate sensitive oral tissues (Anderson et ah, 1992), and these chemicals, as well as the large amounts of soil frequently ingested with ants and termites, could exacerbate tooth wear. Of course, mechanical breakdown is not the only way food can be processed in mammals. Chemical digestion is also a form of processing and occurs, in part, in the mouth. Despite a great deal of speculation, virtually nothing is known about myrmecophage saliva or other glandular secretions. Oral secretions may be important in neutralizing the effects of noxious prey secretions (Anderson et ah, 1992; Kratzing and Woodall, 1988), could assist in the digestion of chitin, and, as mentioned earlier, could be chemoattractive to prey species. 3. Intraoral Food Transport Before food can be swallowed it must be transported from the oral cavity to the oropharynx. Intraoral food transport mechanisms seem to correlate with the length of the oral cavity, surface texture of the hard
478
Karen Zich Reiss
palate, and size of the vallecular pouches (Franks et ah, 1984). While no myrmecophage has the pronounced rigid palatal rugae seen on the hard palate of many mammals, anteaters have epithelial papillae on the tongue and lining of the oral cavity that point toward one another (Dingier, 1964b; Kuhlhorn, 1939) and a ratcheting mechanism could work between these surfaces. This is supported by films of anteater feeding that show several tongue protrusion/retraction cycles after food is taken into the mouth but before it is swallowed (personal observations). At least one species of pangolin has caudally pointing fringed epithelial folds on the hard palate (personal observation), but whether these are coupled with surface irregularities on the tongue is unknown. It has been suggested that the basihyal might scrape prey off the protruding pangolin tongue (Doran and AUbrook, 1973) so that prey accumulate in the caudal oropharynx. Whether the tongue's excursion upon retraction is great enough to allow this to be an important mechanism of intraoral transport in any myrmecophage is uncertain. 4.
Swallowing
Mammalian swallowing is a reflex that involves the coordinated actions of the tongue, palatal, pharyngeal, and laryngeal striated muscle, and esophageal smooth muscle. The complexity of the behavior is related, in part, to the necessary coordination between feeding and respiration that results from the crossing of air and food passageways in the pharynx . The number and extent of oropharyngeal modifications in anteaters and pangolins suggest that mechanisms of swallowing may be substantially different from those reported for other mammals (e.g., De Gueldre and De Vree, 1984; German and Franks, 1991; Herring, 1993). However, the unknown functional consequences of each of these modifications make speculation futile. Mechanisms of bolus formation depend in part on mechanisms of intraoral transport, the prehyoid pouch and glossal tube could play a role both in bolus formation and in swallowing, and there may be coordination between the timing of tongue movements and the initiation of a swallow. 5. Neonate
Suckling
Few myrmecophages have bred successfully in captivity and nursing is rarely witnessed in the wild. It has been reported that echidna newborns do not suckle, but instead lap milk, which has been cited as evidence that lapping is a primitive mammalian feeding behavior (Crompton, 1989). In those mammals that do suckle, apposition of the tongue to the soft palate is an important component of forming a seal around the nipple. As mentioned in Section IV,C, the muscles that
usually cause this movement are absent or modified in anteaters and pangolins, but the unusual mylohyoideus may be a functional replacement. However, reports on captive-raised anteaters claim that the tongue hangs freely out of the mouth during nursing, and infants appear to be pumping rather than sucking fluids (M. Flint, personal communication). The generality of these phenomena for anteaters and other myrmecophages remains to be confirmed. V. EVOLUTION OF MYRMECOPHAGOUS SPECIALIZATIONS It is certain that myrmecophagy among extant mammals is the product of numerous independent evolutionary events. However, for none of these events do we have a good series of intermediates between a nonmyrmecophagous and a myrmecophagous form. This section reviews what we know about the history of myrmecophagy in diverse lineages, which suggests the phylogenetic pathways by which myrmecophagy can evolve. Following this, I construct a morphoseries of extant forms that suggest the pathways of structural change that lead to anatomically specialized myrmecophages. Finally, I discuss the implications of feeding apparatus character polarity and character state distributions in anteaters and pangolins. A. Phylogenetic Pathways to Myrmecophagy Patterson suggested that there are two evolutionary pathways to myrmecophagy: (1) from generalized fossorial insectivores and (2) from carnivores (Patterson, 1975). Consideration of the known fossils referred to each of the extant myrmecophagous taxa along with the phylogenetic structure of each taxon suggests that myrmecophagy is more easily achieved. Echidnas are not known as fossils. Moreover, the oldest known monotreme, the early Cretaceous Steropodon from Australia (Archer et ah, 1985), and the Miocene platypus-like Obdurodon (Archer et ah, 1992) are well toothed and are both more like platypuses than echidnas. This suggests that the immediate ancestors of echidnas may have been herbivorous. Myrmecobius, the sole marsupial myrmecophage, is itself known only from the Pleistocene. However, its phylogenetic relationships suggest a carnivorous ancestry. The other marsupials that have a high percentage of ants or termites in their diets (Table 15.1) are also marsupicarnivores. There are no fossil marsupials thought to be myrmecophagous. The history of myrmecophagy in the Pholidota is confused by uncertainties regarding the sister group to
15. Feeding in Myrmecophagous Mammals Pholidota (see Section I,B) and their uncertain relationship to the fossil taxon Palaeanodonta. Palaeanodonts are small fossorial insectivores known from the Late Paleocene and Early Oligocene. Some palaeanodonts have craniodental and limb modifications similar to those seen in extant myrmecophages and may themselves have been myrmecophagous (Rose and Emry, 1983; 1991). The oldest presumed pholidotan myrmecophage is the Eocene Eomanis from the Messel formation of Germany (Storch, 1978). Its morphology is similar to extant pangolins, including the presence of scales (Koenigswald et al, 1981), although some evidence points to affinities with the Palaeanodonta (Rose and Emry, 1993). The Oligocene Patriomanis from the Big Horn Basin of Wyoming (Emry, 1970) is also clearly a myrmecophage, although Simpson doubted its identification as a pangolin and suggested that it too was a palaeanodont (Simpson, 1978). The controversies surrounding pholidotan fossils, if anything, suggest a close relationship between palaeanodonts and pholidotes (Emry, 1970; Rose and Emry, 1993). This would imply insectivorous or perhaps myrmecophagous ancestry of Pholidota. The trophic diversity within extant Xenarthra and the affinities between armadillos and fossil glyptodonts on the one hand and anteaters and sloths on the other hand make it clear that myrmecophagy is not primitive for xenarthrans. The common ancestor of Xenarthra is debated and palaeanodonts have been suggested here as well (Simpson, 1931). Even so, myrmecophagy certainly evolved independently in armadillos and anteaters. At what point anatomical specializations arose in each of these two lineages is less certain. The fossil armadillo Stegotherium, from the Miocene of Argentina, is an undisputed myrmecophage on the basis of cranial anatomy, but this genus is generally placed as the sister group to Dasypus, the most derived of the extant dasypodines (Engelmann, 1978, 1985). The morphology of other fossil armadillos is not indicative of incipient myrmecophagy. Outgroup criteria applied to dasypodine armadillos suggest that ancestral armadillos were herbivorous (as are sloths, glyptodonts, and some other armadillos). The origin of anteater myrmecophagy is similarly unclear. Eurotamandua, also from the Messel formation (Storch, 1981), was originally described as a myrmecophagid. Radiographic analyses of the hard palate and middle ear show striking similarity to Tamandua (Storch and Habersetzer, 1991), but this fossil has been referred, alternatively, to the Pilosa (Gaudin and Branham, 1998) and the Pholidota (Rose, 1988, 1993; Shoshani et al, 1997). The Early Miocene Protamandua, from Argentina, is an undisputed myrmecophagid (Ameghino, 1904; Hirschfeld, 1976) and, like the extant Cyclo-
479
pes, its palatal shelves do not meet in the midline (Patterson et ah, 1992). Both these fossil taxa were clearly myrmecophagous, but a great deal of morphological change separates sloths from any of these anteaters (Reiss, 1997a) and the fossil taxa are of little value in reconstructing these events. The earliest aardvark, the Miocene Myorycteropus, as well as the Pliocene Leptorycteropus, both have more generalized anatomy than Orycteropus, suggesting that myrmecophagy was not primitive for that group. The Paenungulata, often suggested as close relatives of aardvarks, are all herbivorous. The Recent Plesiorycteropus is more specialized for myrmecophagy than Orycteropus (Patterson, 1975) and, as its name suggests, had been thought to be an aardvark. However, this genus is distinct from aardvarks in many ways and is not clearly allied with any other myrmecophagous eutherian. It has been diagnosed as a new eutherian order that evolved in isolation on Madagascar (MacPhee, 1994), suggesting yet one more independent origin of myrmecophagy in the Mammalia. Finally, while aardwolves are not known as fossils, other extant hyaenids are carnivorous. Clearly, we do not have all the desired evidence to confirm or refute Patterson's hypothesis, but it is likely that several of the extant myrmecophagous lineages arose from herbivorous ancestors. This suggests that the evolution of myrmecophagy is not constrained by ancestral trophic specializations. Moreover, the fact that the earliest pangolin and anteater fossils are so similar to extant forms suggests that whatever one's ancestry and whatever phylogenetic pathway is taken to niyrmecophagy, the resulting ecological niche and morphology are extremely stable. B. Structural Pathways to Myrmecophagy Unusual anatomical specializations are seen in the feeding apparatus of many myrmecophages. At the same time, there are several mammals that are ecologically specialized ant and termite feeders that have few anatomical specializations. The evidence that ancestral trophic specializations do not constrain the evolution of myrmecophagy suggests that a general morphoseries can be constructed from the least to the most specialized feeding apparatus. This morphoseries can give us insight as to how myrmecophagy arises in such diverse lineages. The most common feature shared by myrmecophages is the loss, reduction, or disuse of the teeth. Even those species that have well-developed teeth do not use them when eating ants or termites. That this feature is so commonly shared suggests that teeth are dispensable to a myrmecophage and that teeth are
480
Karen Zich Reiss
particularly evolutionarily malleable in mammals. This is supported by observations that tooth count is variable in toothed myrmecophages (Anderson et al., 1992; Calaby, 1960). The next most common feature shared by myrmecophages is an unusually long and thin tongue. A tongue's resting length confers a displacement advantage and a tongue's slenderness may confer a speed advantage. Even aardvarks have a tongue that is relatively long and slender despite the overall resemblance of their feeding apparatus to that of other ungulates. The repeated occurrence of this feature suggests that a quick and extensible tongue is advantageous if a mammal eats ants and termites or at least that a squat tongue is disadvantageous. The most anatomically specialized myrmecophages, echidnas, pangolins, and anteaters, have a long slender tongue that is composed primarily of the sternoglossus muscles. Other modifications of the feeding apparatus are also seen in taxa with a sternoglossus. Hyoid arch musculature is unusual and the palate and pharynx contain an unusual complement of muscles. Speculation regarding the functional significance of these anomalies was discussed in Section IV, but these shared features of anatomically specialized myrmecophages might also have phylogenetic implications.
blance has no basis in homology. Lyttae in mammalian tongues are remnants of the primitive reptilian entoglossum, but this potential homology has never been thoroughly explored. Most of the anatomiical features that characterize the feeding apparatus of anteaters and pangolins are derived features. Moreover, in most cases these derived features arise from an ancestral condition that is characteristically eutherian (the one exception is the stylopharyngeus, see Reiss, 1997a). While this is true regardless of where pangolins are placed with respect to other mammals, the homologous and derived similarities between anteaters and pangolins are so extensive that the most parsimonious interpretation of the anatomy is derivation from a common myrmecophagous ancestor within the Xenarthra (Reiss, 1997a). That this result is in keeping with some analyses based on broader data sets is encouraging (see Section I,B), but until there is more widespread consensus on pholidotan relationships, the evolutionary interpretation of shared derived features in anteater and pangolin anatomy will remain open.
C. Primitive and Derived Features in the Myrmecophagous Feeding Apparatus
A. Form
One result of the discussion just given is that the most specialized anatomy occurs in taxa that are relatively basal, particularly if one accepts the hypothesis of close relationship between Xenarthra and Pholidota. This raises the question of whether some of these anatomical features are, in fact, primitive characteristics retained from the ancestral mammalian feeding apparatus. This would lead to the hypothesis that the evolution of anatomically specialized myrmecophages in more derived mammals was, perhaps, constrained by the acquisition of a more derived feeding apparatus. Because it is very difficult to choose an appropriate outgroup with which to compare monotreme anatomy, the following discussion is restricted to the eutherian myrmecophages, anteaters and pangolins. Cladistic analysis of xenarthran and pangolin cranial muscles, using marsupials as an outgroup, demonstrates that only the absence of the stylohyoideus and the presence of a complex stylopharyngeus can be interpreted as retained primitive features (Reiss, 1997a). Other characteristics that are primitive in appearance are secondarily derived. Cross sections through a sternoglossus-based tongue shows marked resemblance to some reptilian tongues, but this resem-
VI. DIRECTIONS FOR FUTURE RESEARCH
A great deal is known about the structure of the feeding apparatus in myrmecophagous mammals, but there are still holes in data that derive both from insufficient taxon sampling and from relatively superficial analysis of a very complex set of structures. Some of the remaining conflict in published anatomical accounts may be due to misunderstanding the taxonomic level of patterns of variation, and there are always controversies in the determination of homology. Many of these animals are exotic, studies are often based on dissection of single specimen of a species, and the more speciose groups (armadillos and pangolins) have not been fully sampled. Knowledge of the full diversity of adults is important, but closely staged developmental series are also invaluable in homology discussions. Closely staged developmental series are likely to remain rare for many of the myrmecophages discussed here, but a series of Dasypus novemcinctus should be readily obtainable because Dasypus are easily bred and colonies are used extensively in leprosy research. Also, much can be learned from studies of mammals that have the primitive and generalized form of the mammalian feeding apparatus. The study on the development of craniofacial musculature in the marsupial Monodelphis (Smith, 1994) is incredibly valuable in
481
15. Feeding in Myrmecophagous Mammals
this regard, but many questions still remain. One important task that needs to be undertaken is an attempt to correlate this immunohistochemical analysis with Edgeworth's classic observations on mammalian cranial development that have formed the basis for many of the homology assessments in this review. B. Function With respect to function, organismal natural history is always a good place to start. Observations of what animals actually do with their anatomy helps define functionally and evolutionarily important features (Bock, 1979), which can guide functional investigations. In the laboratory, the potential value of X-ray cineradiographic studies of normal feeding in anatomically specialized myrmecophages cannot be overemphasized. Many of the competing hypotheses discussed earlier could be evaluated with simple, noninvasive X-ray cineradiographic observations employing radio-opaque food items. Basic patterns of food transport and swallowing could be easily observed, and some insight could be gained into the coordination of tongue movements and swallowing and respiration even without the surgical implant of markers into the tongue. More invasive techniques, such as electromyography, are difficult in exotic and endangered species, but the captive breeding of many of these species is a high priority at many zoos and surgical techniques may become feasible. Zoo animals have a great deal to offer even without resorting to invasive techniques. They provide ample opportunity to observe basic feeding behavior and also offer a means to test biomechanical predictions based on structure. Finally, there are some functional questions that require fixed specimens, specifically muscle fiber type histochemistry and muscle architecture. Museum specimens are sometimes too precious or inappropriately prepared for certain techniques, but there are other means of obtaining specimens. I have had good fortune collecting freshly road-killed Tamandua, and several zoos have given me former captives that died or required euthanasia. The functional morphology of myrmecophage feeding will never be as well understood as it is in Didelphis or macaques, but the problem is too interesting to be left untouched and persistence and ingenuity pay off. C. Evolution Our knowledge of mammalian structure and function is always enriched by an evolutionary perspective. Fiowever, in the case of myrmecophagous mammals, we are limited by unresolved questions regarding
mammalian phylogenetic history. This review has emphasized the importance of determining the phylogenetic positions and within-group relationships of both Xenarthra and Pholidota. This is a general problem. Numerous researchers have pointed out that modern mammalian groups were formed in an early and explosive radiation, the cladistic pattern of which may be irretrievable. Moreover, a great deal of subsequent anagenetic change may have obscured what little information existed. I believe that this is an excessively pessimistic viewpoint. Cladistic methodologies are still young, and new techniques (e.g., those devised to handle total evidence analyses) are continually being developed and explored. Moreover, slowly but surely, new comparative studies (both morphological and molecular) and new fossil finds are ever expanding the data set. Nevertheless, it cannot be denied that specialized basal groups such as monotremes and xenarthrans and (perhaps) pholidotans are particularly problematic. Abundant plesiomorphy, adaptation, and the consequent abundance of autapomorphies and the difficulty of choosing a suitable outgroup are all difficulties that must be reckoned with if we are to ever fully comprehend the evolution of mammalian myrmecophagy.
References Abensperg-Traun, M., and E. S. De Boer (1992) The foraging ecology of a termite- and ant-eating specialist, the echidna Tachyglossus aculeatus (Monotremata: Tachyglossidae). J. Zool. Lond. 226:243257. Ameghino, R (1904) Nuevas especies de mamiferos Cretaceos y Terciarios de la Republica Argentina. Ann. Soc. Cient. Argen. 58: 225-291. Anderson, M. D., P. R. K. Richardson, and R R Woodall (1992) Punctional analysis of the feeding apparatus and digestive tract anatomy of the aardwolf Proteles cristatus. J. Zool. Lond. 228:423-434. Andres, K. H., M. Von During, A. Iggo, and U. Proske (1991) The anatomy and fine structure of the echidna Tachyglossus aculeatus snout with respect to its different trigeminal sensory receptors including the electroreceptors. Anat. Embryol. 184(4): 371-393. Archer, M., T. R Flannery, A. Ritchie, and R. E. Molnar (1985) Pirst Mesozoic mammal from Australia: an early Cretaceous monotreme. Nature (Lond.) 318:363-366. Archer, M., P. A. Jenkins, S. J. Hand, P. Murray, and H. Godthelp (1992) Description of the skull and non-vestigial dentition of a Miocene platypus {Obdurodon dicksoni n.sp.) from Riversleigh, Queensland, Australia, and the problem of monotreme origins. Pp. 15-27. In: Platypus and Echidnas. M. L. Augee (ed.). Royal Zoological Society of New South Wales, Sydney. Augee, M. L., and B. A. Gooden (1992) Evidence for electroreception from field studies of the echidna, Tachyglossus aculeatus. In: Platypus and Echidnas. M. L. Augee (ed.). Royal Zoological Society of New South Wales, Mosman. Blainville, H. M. D. (1816) Prodrome d'une nouvelle distribution systematique de regne animal. Bull. Soc. Philomatique 1816: 105-124. Bock, W. (1979) The synthetic explanation of macroevolutionary
482
K a r e n Zich Reiss
change: a reductionistic approach. Bull. Carnegie Mus. Nat. Hist. 13:20-69. Calaby, J. H. (1960) Observations on the banded anteater Myrmecobius f. fasciatus Waterhouse (Marsupialia) with particular reference to its food habits. Proc. Zool. Soc. Lond. 135:183-207. Chan, L.-K. (1995) Extrinsic lingual musculature of two pangolins (Pholidota: Manidae). J. Mammal. 76(2): 472-480. Cheng, H. (1986) Anatomical studies of Manis pentadactyla's tongue and structures relative to the tongue. Mem. Beijing Nat. Hist. Mus. 36:1-9. Crompton, A. W. (1989) The evolution of mammalian mastication. Pp. 23-40. In: Complex Organismal Functions: Integration and Evolution in Vertebrates. D. B. Wake and G. Roth (eds.) Wiley, Dahlem. Crompton, A. W., A. Thexton, P. Parker, and K. Hiiemae (1977) The activity of the hyoid and jaw muscles during chewing of soft food in the opposum. Pp. 287-305. In: The Biology of Marsupials. B. Stonehouse and D. Gilmore (eds.). MacMillan, London. Cuvier, G. (1817). La Regne Animal. Deterville, Paris. Davis, D. D. (1961) Origin of the mammalian feeding mechanism. Am. Zool. 1:229-234. De Gueldre, G., and F. De Vree (1984) Movements of the mandibles and tongue during mastication and swallowing in Pteropus giganteus (Megachiroptera): a cineradiographical study. J. Morphol. 179:95-114. dejong, W. (1982) Eye lens protein and vertebrate phylogeny. Pp. 7 5 114. In: Macromolecular Sequences in Systematic and Evolutionary Biology. M. Goodman (ed.). Plenum Press, New York. Dellon, A. L. (1989) The levator veli palatini muscle arises from the second branchial arch: a hypothesis. Arm. Plast. Surg. 23:317319. Dingier, E. C. (1964a) Die konstructive Form der Zunge des Ameisenbaren. Verhandlungen Anat. Gesellschaft 58:405-416. Dingier, E. C. (1964b) Vergleichende funktionell-anatomischeStudien an Edentatenzungen {Dasypus sexcinctus, Smutsia temmincki, Tamandua tetradactyla und Myrmecophaga jubata). Gegenbaurs Morphologisches Jahrbuch 105(3): 375-395. Doran, G. A. (1973) The lingual musculature of the echidna, Tachyglossus aculeatus. Anat. Anzeiger 133:468-476. Doran, G. A. (1975) Review of the evolution and phylogeny of the mammalian tongue. Acta Anat. 91:118-129. Doran, G. A., and D. B. Allbrook (1973) The tongue and associated structures in two species of African pangolins, Manis gigantea and Manis tricuspis. J. Mammal. 54(4): 887-899. Doran, G. A., and H. Baggett (1970) The vascular stiffening mechanism in the tongue of the echidna. Anat. Rec. 167:197-204. Doran, G. A., and H. Baggett (1971) A structural and functional classification of mammalian tongues. J. Mammal. 52(2): 427-429. Doran, G. A., and H. Baggett (1972) The specialized lingual papillae of Tachyglossus aculeatus. I. Gross and light microscopic features. Anat. Rec. 172:157-166. Duvemoy, G. L. (1830) De la langue, consideree comme organe de prehension des alimens ou recherches anatomiques sur les mouvemens de la langue dans quelques animaux, particulierement de la classe des mammiferes et de celle des reptiles. Mem. Soc. Hist. Nat. Strasbourg 1:1-16. Edgeworth, F. H. (1916) On the development and morphology of the pharyngeal, laryngeal and hypobranchial muscles of mammals. Quart. J. Microscop. Sci. 61:383-432. Edgeworth, F. H. (1923) On the development of the cranial muscles of Tatusia and Manis. J. Anat. Physiol. Lond. 57:313-335. Edgeworth, F. H. (1924) On the masticatory, intermandibular and hyoid muscles of Orycteropus capensis. J. Anat. Physiol. Lond. 58: 134-139. Edgeworth, F. H. (1931) On the development of the external ocular.
masticatory, and hyoid muscles of Monotremata. Proc. Zool. Soc. Lond. 1931:809-815. Edgeworth, F. H. (1935) The Cranial Muscles of Vertebrates. Cambridge Univ. Press, Cambridge. Ehlers, E. (1894). Der Processus Xiphoideus und seine Muskulatur von Manis macrura Erxl. und Manis tricuspis Surdev. Zool. Miszellen I. Abhandlungen der Koniglighe Gesellschaft der Wissenschaften39:l-34. Eisenberg, J. F. (1981) The Mammalian Radiations: An Analysis of Trends in Evolution, Adaptation, and Behavior. University of Chicago Press, Chicago. Emmons, L. H. (1975) Ecology and Behavior of African Rainforest Squirrels. Doctoral Dissertation, Cornell University. Emry, R. J. (1970) A North American Oligocene pangolin and other additions to the Pholidota. Bull. Am. Mus. Nat. Hist. 142(6): 459-510. Engelmann, G. F. (1978) The Logic of Phylogenetic Analysis and the Phylogeny of the Xenarthra. Doctoral Dissertation, Columbia University. Engelmann, G. F. (1985) The phylogeny of the Xenarthra. Pp. 51-64. In: The Evolution and Ecology ofArmadillos, Sloths, and Vermilinguas. G. G. Montgomery (ed.). Smithsonian Institution Press, Washington, DC. Ewer, R. F. (1973) The Carnivores. Cornell Univ. Press, Ithaca, NY. Fleay, D. (1942) The numbat in Victoria. Vict. Nat. 59:3-7. Flynn, J. J., N. A. Neff, and R. H. Tedford (1988) Phylogeny of the Camivora. In: The Phylogeny and Classification of the Tetrapods, Vol. 2. M. J. Benton (ed.). Oxford Univ. Press, New York. Franks, H. A., A. W. Crompton, and R. Z. German (1984) Mechanisms of intraoral transport in macaques. Am. J. Phys. Anthropol. 65:275-282. Frick, H. (1951) Uber die Trigeminusmuskulatur und die tiefe Facialismuskulatur von Orycteropus aethiopicus. Zeithschrift Anat. Entwicklungsgeschichte 116:202-217. Furusawa, K., M. Yamaoka, M. Kogo, and T. Matsuya (1991a) The innervation of the levator veli palatini muslce by the glossopharyngeal nerve. Brain Res. Bull. 26:599-604. Furusawa, K., M. Yamaoka, and T. Kumai (1991b) Properties of the lingual and Ivp branches of the glossopharyngeal nerve. Brain Res. Bull. 28:1-7. Cans, C. (1982) Fiber architecture and muscle function. Exercise Sport Sci. Rev. 10:160-207. Gans, C , and W. J. Bock (1965) The functional significance of muscle architecture: a theoretical analysis. Adv. Anat. Embryol. Cell Biol. 38:115-142. Gans, C , and A. S. Gaunt (1991) Muscle architecture in relation to function. J. Biomech. 24, Supple. 1:53-65. Gans, C , and A. S. Gaunt (1992) Muscle architecture and control demands. Brain Behav. Evol. 40:70-81. Gasc, J.-P (1967) Squellete hyobranchial. Pp. 550-583. In: Traite de Zoologie, Vol. 16. P.-P Grasse (ed.). Masson et Cie, Paris. Gaudin, T. J. (1990) The ear region of the Pilosa (Mammalia, Xenarthra) and the phylogeny of the Tardigrada. J. Vertebr. Paleontol. 10(3, Supple.) :24A. Gaudin, T. J. (1995) The ear region of Edentates and the phylogeny of the Tardigrada (Mammalia, Xenarthra). J. Vertebr. Paleontol. 15(3): 672-705. Gaudin, T. J., and D. G. Branham (1998) The phylogeny of Myrmecophagidae (Mammalia, Xenarthra, Vermilingua) and the relationship of Eurotamandua to the Vermilingua. J. Mammal. Evol. 5(3): 237-265. Gaudin, T. J., and J. Wible (1999) The entotympanic of pangolins and the phylogeny of the Pholidota (Mammalia). J. Mammal. Evol. 6(1): 39-65. Gemmel, N. J., and M. Westerman (1994) Phylogenetic relationships
15. F e e d i n g in M y r m e c o p h a g o u s M a m m a l s within the class Mammalia: a study using 12s RNA sequences. J. Mammal. Evol. 2(1): 3-23. German, R. Z., and H. A. Franks (1991) Timing in the movement of jaws, tongue and hyoid during feeding in the hyrax, Procavia syriacus. J. Exp. Zool. 257:34-42. Gill, T. (1870) Arrangements of the families of mammals and synoptical table of characters of the subdivisions of mammals. Smith. Miscell. Collect. 11:1-98. Gordon, A. M., A. R Huxley, and F. Julian (1966) The variation in isometric tension with sarcomere length in vertebrate muscle fibers. J. Physiol. 184:170-190. Gregory, W. K. (1910) The orders of mammals. Bull. Am. Mus. Nat. Hist. 27:1-524. Griffiths, M. (1968) Echidnas. Pergamon Press, Oxford. Griffiths, M., and K. G. Simpson (1966) A seasonal feeding habit of spiny anteaters. CSIRO Wildlife Res. 11(2): 137-143. Griffiths, M., R. T. Wells, and D. J. Barrie (1991) Observations on the skulls of fossil and extant echidnas (Monotremata: Tachyglossidae). Aust. Mammal. 14:87-101. Griffiths, T. A. (1982) Systematics of the New World nectar-feeding bats (Mammalia: Phyllostomidae), based on the morphology of the hyoid and lingual regions. Am. Mus. Novit. 2742:1-45. Guth, C. (1961) La Region Temporale des tdentates. Le Puy, Haute Loire. Hatt, R. T. (1934) Pangolins. Nat. Hist. 34(8): 725-732. Heath, M. E., and S. L. Vanderlip (1988) Biology, husbandry and veterinary care of captive Chinese pangolins {Manis pentadactyla). Zoo Biol. 7:293-312. Hecht, C. S., P. J. Gannon, and A. R. Eden (1993) Motor innervation of the Eustachian tube muscles in the guinea pig. Laryngoscope 103:1218-1226. Herring, S. W. (1993) Functional properties of the feeding musculature. In: Biomechanics of Feeding in Vertebrates. V. L. Bels, M. Chardon, and P. Vandewalle (eds.). Springer-Verlag, Berlin. Hiiemae, K. (1976) Masticatory movements in primitive mammals. Pp. 105-118. In: Mastication. D. J. Anderson and B. Matthews (eds.). Wright and Sons, Bristol. Hirschfeld, S. E. (1976) A new fossil anteater (Edentata, Mammalia) from Colombia, S.A. and evolution of the Vermilingua. J. Vertebr. Paleontol. 50:419-432. Ibuki, K., T. Matsuya, J. Nishio, Y. Hamamura, and T. Miyazaki (1978) The course of the facial nerve innervation for the levator veli palatini. Cleft Palate J. 15:209-214. JoUie, M. (1968) The head skeleton of a new-born Manis javanica with comments on the ontogeny and phylogeny of the mammal head skeleton. Acta Zool. 49:227-305. Keller, J. T., M. C. Saunders, H. van Loveren, and M. T. Shipley (1984) Neuroanatomical considerations of palatal muscles: tensor and levator veli palatini. Cleft Palate J. 21:70-75. Kier, W. K. (1992) Hydrostatic skeletons and muscular hydrostats. Pp. 205-231. In: Biomechanics—Structures and Systems, a Practical Approach. A. A. Biewener (ed.). Oxford Univ. Press, New York. Kier, W. M., and K. K. Smith (1985) Tongues, tentacles and trunks: the biomechanics of movement in muscular-hydrostats. Zool. J. Linnaean Soc. 83:307-324. Kingdon, J. (1974) East African Mammals: An Atlas of Evolution in Africa. University of Chicago Press, Chicago. Koenigswald, W., G. Richter, and G. Storch (1981) Nachweis von Homschuppen bei Eomanis waldi aus der Grube Messel bei Darmstadt (Mammalia, Pholidota). Senckenbergiana lethaea 61(3): 291-298. Kratzing, J. E., and P. F Woodall (1988) The rostral anatomy of two elephant shrews. J. Anat. 157:135-143. Kruuk, H., and W. A. Sands (1972) The aardwolf (Proteles cristatus
483
Sparmann 1783) as a predator of termites. East Af. Wildlife J. 10: 211-227. Kubota, K., J. Kubota, N. Fukuda, and S. Asakura (1962a) Comparative anatomical and neurohistological observations on the tongue of the great anteater {Myrmecophaga jubata Linne). Anat. Rec. 143: 15-25. Kubota, K., J. Kubota, T. Nakamura, N. Fukuda, S. Asakura, S. Nakagawa, and M. Masui (1962b) Comparative anatomical and neurohistological observations on the tongue of the pangolin {Manis pentadactyla, Linneus). Anat. Rec. 144:43-55. Kiihlhorn, F. (1939) Beziehungen zwischen Ernahrungsweise und Bau des Kauapparates bei einigen Giirteltier und Ameisenbarenarten. Gegenbaurs Morphologisches Jahrbuch 84:55-85. Lubin, Y D., and G. G. Montgomery (1981) Defenses of Nasutitermes termites (Isoptera, Termitidae) against Tamandua anteaters (Edentata, Myrmecophagidae). Biotropica 13(1): 66-76. Lubin, Y D., G. G. Montgomery, and O. P Young (1977) Food resources of anteaters (Edentata: Myrmecophagidae) I. A year's census of arboreal nests of ants and termites on Barro Colorado Island, Panama Canal Zone. Biotropica 9(1):26-34. Lubosch, W. (1908) Das Kiefergelenk der Edentaten und Marsupialier. Nebst Mittheilungen iiber die Kaumuskulatur dieser Thiere. Semons Zoologische Forschungsreisen in Australien und dem Malayischen Archipel. 4(6): 520-555. Luckett, W. P. (1994) Suprafamilial relationships within Marsupialia: resolution and discordance from multidisciplinary data. J. Mammal. Evol. 2(4):255-283. MacPhee, R. D. E. (1994) Morphology, adaptations and relationships of Plesiorycteropus, and a diagnosis of a new order of eutherian mammals. Bull. Am. Mus. Nat. Hist. 220:1-214. Marshall, L. G. (1979) Evolution of metatherian and eutherian (mammlian) characters: a review based on cladistic methodology Zool. J. Linnaean Soc. 66:369-410. McAdam, D. W., and J. S. Way (1967) Olfactory discrimination in the giant anteater. Nature (Lond.) 214:316-317. McKenna, M. C. (1987) Molecular and morphological analysis of higher-level mammalian interrelationships. In: Molecules and Morphology in Evolution: Conflict or Compromise. C. Patterson (ed.). Cambridge Univ. Press, Cambridge. McKenna, M. C. (1992) The alpha crystallin A chain of the eye lens and mammalian phylogeny. Ann. Zool. Fennici 28:349-360. McNab, B. K. (1984) Physiological convergence amongst ant-eating and termite-eating mammals. J. Zool. Lond. 202:485-510. Meeuse, A. D. J. (1963) A possible case of interdependence between a mammal and a higher plant. Arch. Neerland. Zool. 13, Supple.(l): 314-318. Miyamoto, M. M., and M. Goodman (1986) Biomoloecular systematics of eutherian mammals: phylogenetic patterns and classification. Syst. Zool. 35:230-240. Montgomery, G. G. (1985) Movements, foraging and food habits of the four extant species of neotropical vermilinguas (Mammalia: Myrmecophagidae). Pp. 365-377. In: The Evolution and Ecology of Armadillos, Sloths, and Vermilinguas. G. G. Montgomery (ed.). Smithsonian Institution Press, Washington, DC. Montgomery, G. G., and Y D. Lubin (1977) Prey influences on movements of neotropical anteaters. Pp. 103-131. In: Proceedings of the 1975 Predator Symposium. R. L. Phillips and C. Jonkel (eds.), Montana Forest and Conservation Experiment Station, University of Montana, Missoula. Murray, P. F (1981) A unique jaw mechanism in the echidna, Tachyglossus aculeatus (Monotremata). Aust. J. Zool. 29:1-5. Naples, V. L. (1999) Morphology, evolution and function of feeding in the giant anteater (Myrmecophaga tridactula) J. Zool. Lond. 249: 19-41.
484
K a r e n Zich Reiss
Nene, R. V. (1978) A note on the tongue muscle of the scaly anteater, Manis gigantea. J. Biol. Sci. 21(1):48-49. Nishio, J., T. Matsuya, K. Ibuki, and T. Miyazaki (1976a) Roles of the facial, glossopharyngeal and vagus nerves in velopharyngeal movements. Cleft Palate J. 13:201-214. Nishio, J., T. Matsuya, J. Machida, and T. Miyazaki (1976b) The motor nerve supply of the velopharyngeal muscles. Cleft Palate J. 13: 20-30. Norman, J. E., and M. V. Ashley (1994) Molecular systematics of the xenarthrans. In: American Society of Mammalogists, Vol. 271 A, Washington, DC. Novacek, M. J. (1986) The skull of leptictid insectivores and the higher-level classification of eutherian mammals. Bull. Am. Mus. Nat. Hist. 183:1-111. Novacek, M. J. (1993) Patterns of diversity in the mammalian skull. Pp. 438-531. In: The Skull, Vol. 2. J. Hanken and B. K. Hall (eds.). University of Chicago Press, Chicago. Novacek, M. J., and A. R. Wyss (1986) Higher-level relationships of the Recent eutherian orders: morphological evidence. Cladistics 2:257-287. Novacek, M. J., A. R. Wyss, and M. C. McKenna (1988) The major groups of eutherian mammals. Pp. 31-71. In: The Phylogeny and Classification of the Tetrapods, Vol. 2. M. J. Benton (ed.). Clarendon Press, Oxford. Nov^ak, R. M. (1991) Walker's Mammals of the World. Johns Hopkins Univ. Press, Baltimore, MD. Ohnishi, K. (1991) A tentative evolutionary tree of mammalian orders constructed by Hennigian comparison of the amino acid sequences of alpha-crystallin A chain, myoglobin, and hemoglobin alpha chain. Sci. Rep. Niigata Univ. Ser. D (Biol.)(28): 19-31. Oppel, A. (1899) Ueber die Zunge der Monotremen, einiger Marsupialier und von Manis javanica. Zoologische Forschungsreisen in Australien und dem Malayischen Archipel. 4(2): 107-170. Owen, R. A., Sir. (1862) On the anatomy of the great anteater {Myrmecophaga jubata, Linn.). Trans. Zool. Soc. Lond. 4:117-140. Pages, E. (1965) Notes sur les pangolins du Gabon. Biol. Gabonica 1(3): 209-237. Pages, E. (1970) Sur Tecologie et les adaptations de Torycterope et des pangolins sympatriques du Gabon. Biol. Gabonica 6(1): 27-92. Pascual, R., M. Archer, E. O. Jaureguizar, J. L. Prado, H. Godthelp, and S. J. Hand (1992) The first non-Australian monotreme: an early Paleocene South American platypus (Monotremata, Ornithorhynchidae). Pp. 2-15. In: Platypus and Echidnas. M. L. Augee (ed.). Royal Zoological Society of New South Wales, Sydney. Patterson, B. (1978) Pholidota and Tubulidentata. In: Evolution of African Mammals. Maglio and Cooke (eds.). Patterson, B., and R. Pascual (1972) The fossil mammal fauna of South America. Pp. 247-309. In: Evolution, Mammals, and Southern Continents. A. Keast, F. C. Erk, and B. Glass (eds.). State University of New York Press, Albany, NY. Patterson, B., W. Segall, and W. D. Tumbull (1989) The ear region in Xenarthrans (= Edentata: Mammalia). I. Cingulates. Field. Geol. (18): 1-46. Patterson, B., W. Segall, W. D. Turnbull, and T. J. Gaudin (1992) The ear region in Xenarthrans (= Edentata: Mammalia). II. Pilosa (sloths, anteaters), palaenodonts, and a miscellany. Field. Geol. (24): 1-79. Patterson, B. K. (1975) The fossil aardvarks (Mammalia: TubuHdentata). Bull. Mus. Comp. Zool. 147:185-237. Pocock, R. I. (1924) The external characters of the pangolins (Manidae). Proc. Zool. Soc. Lond. 46:707-719. Pouchet, H. (1867) Memoires sur le grand Fourmilier (Myrmecophaga jubata). Paris.
Pough, H. P , J. B. Heiser, and W. D. McFarland (1989) Vertebrate Life. Macmillan, New York. Proske, U., J. E. Gregory, and A. Iggo (1998) Sensory receptors in monotremes. Phil. Trans. Roy. Soc. Lond. B 353:1187-1198. Redford, K. H. (1985) Food habits of armadillos. Pp. 429-437. In: The Evolution and Ecology of Armadillos, Sloths, and Vermilinguas. G. G. Montgomery (ed.). Smithsonian Institution Press, Washington, DC. Redford, K. H. (1987) Ants and termites as food: patterns of mammalian myrmecophagy. Pp. 349-399. In: Current Mammalogy, Vol. 1. H. H. Genoways (ed.). Plenum Press, New York. Redford, K. H., and J. G. Dorea (1984) The nutritional value of invertebrates with emphasis on ants and termites as food for mammals. J. Zool. Lond. 203: 385-395. Reiss, K. Z. (1997a) How the Anteater Got Its Tongue: Structure and Evolution of the Feeding Apparatus in Xenarthra and Pholidota (Mammalia: Eutheria). Ph.D. Dissertation, Cornell University. Reiss, K. Z. (1997b) Myology of the feeding apparatus of myrmecophagid anteaters (Xenarthra: Myrmecophagidae). J. Mammal. Evol. 4(2): 87-117. Retief, J. D., R. J. Winkfein, and G. H. Dixon (1994) Evolution of the monotremes: the sequences of protamine PI genes of platypus and echidna. Euro. J. Biochem. Richardson, P R. K. (1990) The lick of the aardwolf. Nat. Hist. (April): 78-84. Richardson, P. R. K. (1985) The Social Behavior and Ecology of the Aardwolf Proteles cristatus in Relation to Its Food Resources. Doctoral Dissertation, University of Oxford. Richardson, P. R. K. (1987) Aardwolf: the most specialized myrmecophagous mammal? South Af. J. Sci. 83:643-646. Richardson, P. R. K. (1991) Territorial significance of scent marking during the non-mating season in the aardwolf Proteles cristatus (Camivora: Protelidae). Ethology 87:9-27. Rose, K. D. (1988) Early Eocene mammal skeletons from the Bighorn Basin, Wyoming: significance to the Messel fauna. Cour. Forsch.Inst. Senkenberg 107:435-450. Rose, K. D., and R. J. Emry (1983) Extraordinary fossorial adaptations in the Oligocene palaeanodonts Epoicotherium and Xenocranium (Mammalia). J. Morphol. 175:33-56. Rose, K. D., and R. J. Emry (1993) Relationships of Xenarthra, Pholidota, and fossil ''edentates": the morphological evidence. Pp. 8 1 102. In: Mammal Phylogeny, Vol. 2. F. S. Szalay, M. J. Novacek, and M. C. McKenna (eds.). Springer-Verlag, New York. Rose, K. D., L. Krishtalka, and R. K. Stucky (1991) Revision of the Wind River faunas. Early Eocene of central Wyoming. II. Palaeanodonta (Mammalia). Ann. Carnegie Mus. 60(1): 63-82. Rowe, T. (1988) Definition, diagnosis, and origin of Mammalia. J. Vertebr. Paleontol. 8(3): 241-264. Rowe, T. (1993) Phylogenetic systematics and the early history of mammals. Pp. 129-145. In: Mammal Phylogeny, Vol. 1. F. S. Szalay, M. J. Novacek, and M. C. McKenna (eds.). Springer-Verlag, New York. Saban, R. (1968) Musculature de la Tete. Pp. 229-471. In: Traite de Zoologie, Vol. 16. P.-P Grasse (ed.). Masson et Cie, Paris. Saban, R. (1971) Particularites Musculaires des Monotremes: Musculature de la Tete. Pp. 681-732. In: Traite de Zoologie, Vol. 16. P.-P. Grasse (ed.). Masson et Cie, Paris. Savage, R. J. G., and M. R. Long (1986) Mammal Evolution: An Illustrated Guide. Facts On File Publications, New York/Oxford. Schneider, R. (1964) Der Larynx der Saugetiere. Pp. 1-128. In: Handbuch der Zoologie, Vol. 8. Helmcke, Lengerken, Starck, and Wermuth (eds.). deGruyter, Berlin. Schulman, H. (1906) Vergleichende Untersuchungen iiber die Tri-
15. Feeding in Myrmecophagous Mammals geminus-Musculatur der Monotremen, sowie dabei in Betracht kommenden Nerven und Knochen. Zool. Forschungsreisen Australien und dem Malayischen Archipel. 3(2): 303-400. Shoshani, J. (1986) Mammalian phylogeny: comparison of morphological and molecular results. Mol. Biol. Evol. 3:222-242. Shoshani, J. (1993) Hyracoidea-Tethytheria affinity based on myological data. Pp. 235-256. In: Mammal Phylogeny, Vol. 1. R S. Szalay, M. J. Novacek, and M. C. McKerma (eds.). Springer-Verlag, New York. Shoshani, J., and M. C. McKenna (1995) Mammalian evolution vis-avis Asian biogeography. J. Vertebr. Paleontol. 15(3): 53A. Shoshani, J., M. C. McKenna, K. D. Rose, and R. J. Emry (1997) Eurotamandua is a pholidotan not a xenarthran J. Vertebr. Paleontol. 17:76A. Sikes, S. K. (1966) The tricuspid tree pangolin, Manis tricuspis (Rafinesque): its remarkable tongue complex. The Nigerian Field 31(2): 99-110. Simpson, G. G. (1931) Metacheiromys and the Edentata. Bull. Am. Mus. Nat. Hist. 59:295-381. Simpson, G. G. (1945) A classification of the mammals. Bull. Am. Mus. Nat. Hist. 85:1-350. Simpson, G. G. (1978) Early mammals in South America: fact, controversy, and mystery. Proc. Am. Philosoph. Soc. 122(5):318-328. Simpson, G. G., and W. S. Beck (1965) Life: An Introduction to Biology. Harcourt, Brace and World, New York. Smith, K. K. (1986) Morphology and function of the tongue and hyoid apparatus in Varanus (Varanidae, Lacertilia). J. Morphol. 187: 261-287. Smith, K. K. (1992) The evolution of the mammalian pharynx. Zool. J. LinnaeanSoc. 104:313-349. Smith, K. K. (1994) Development of craniofacial musculature in Monodelphis domestica (Marsupialia, Didelphidae). J. Morphol. 222: 149-173. Smith, K. K., and K. H. Redford (1990) The anatomy and function of the feeding apparatus in two armadillos (Dasypoda): anatomy is not destiny. J. Zool. Lond. 222:27-47. Sonntag, C. F. (1923) The comparative anatomy of the tongues of Mammalia. IX. Edentata, Dermoptera, and Insectivora. Proc. Zool. Soc. Lond. XXXIV: 515-529. Sonntag, C. F. (1925) The comparative anatomy of the tongues of Mammalia. XII. Summary, classification and phylogeny. Proc. Zool. Soc. Lond. : 701-764. Starck, D. (1967) Le crane des mammiferes. Pp. 405-549. In: Traite de Zoologie, Vol. 16. P.-P. Grasse (ed.). Masson et Cie, Paris. Storch, G. (1978) Eomanis waldi, ein Schuppentier aus dem MittelEozan der Grube Messel bei Darmstadt (Mammalia: Pholidota). Senckenbergiana lethaea 59(4): 503-529. Storch, G. (1981) Eurotamandua joresi, ein Myrmecophagide aus dem
485
Eozan der Grube Messel bei Darmstadt (Mammalia, Xenarthra). Senckenbergiana lethaea 61(3): 247-289. Storch, G., and J. Habersetzer (1991) Riickverlagerte Choanen und akzessorische Bulla tympanica bei rezenten Vermilingua und Eurotamandua aus dem Eozan von Messel (Mammalia: Xenarthra). Zeitschrifte Saugetierkunde 56:257-271. Storr, G. C. C. (1780) Prodromus methodi Mammalium . . . inaugural disputationem propositus. Fridriech Wolffer, Tubingen. Szalay, R S., and R Schrenk (1998) The Middle Eocene Eurotamandua and a Darwinian phylogenetic analysis of "Edentates''. Kaupia 7: 97-185. Sweeney, R. C. H. (1956) Notes on Manis temmincki. Ann. Magazine Nat. Hist. Lond. Thewissen, J. G. M. (1985) Cephalic evidence for the affinties of Tubulidentata. Mammalia 49:257-284. Van Loveren, H., M. C. Saunders, and J. T. Keller (1983) Localization of motoneurons innervating the levator veli palatini muscle. Brain Res. Bull. 11:303-307. Wayne, R. K., R. E. Benveniste, D. N. Janczewski, and S. J. O'Brien (1989) Molecular and biochemical evolution of the carnivora. Pp. 465-494. In: Carnivore Behavior, Ecology, and Evolution. J. L. Gittleman (ed.). Cornell Univ. Press, Ithaca, NY. Webb, S. D. (1985) The interrelationships of tree sloths and ground sloths. Pp. 105-112. In: The Evolution and Ecology of Armadillos, Sloths, and Vermilinguas. G. G. Montgomery (ed.). Smithsonian Institution Press, Washington, DC. Westerman, M., and D. Edwards (1992) DNA hybridisation and the phylogeny of monotremes. Pp. 28-34. In: Platypus and Echidnas. M. L. Augee (ed.). Royal Zoological Society of New South Wales, Sydney. Wible, J. R. (1991) Origin of Mammalia: the craniodental evidence re-examined. J. Vertebr. Paleontol. 11:1-28. Wible, J. R., and J. A. Hopson (1993) Basicranial evidence for early mammal phylogeny. Pp. 45-61. Mammal Phylogeny, Vol. 1. R S. Szalay, M. J. Novacek, and M. C. McKenna (eds.). SpringerVerlag, New York. Windle, B. C. A., and R G. Parsons (1899) On the myology of the Edentata. I. Muscles of the head, neck and forelimb. Proc. Zool. Soc. Lond. 1899:314-339. Woodburne, M. O., and R. H. Tedford (1975) The first Tertiary monotreme from Australia. Am. Mus. Novitates 2588:1-11. Wozencraft, W. C. (1989) The phylogeny of the recent Carnivora. Pp. 495-535. In: Carnivore Behavior, Ecology, and Evolution. J. L. Gittleman (ed.). Cornell Univ. Press, Ithaca, NY. Zeller, U. (1993) Ontogenetic evidence for cranial homologies in monotremes and therians, with special reference to Ornithorhynchus. Pp. 95-107. In: Mammal Phylogeny, Vol. 1. R S. Szalay, M. J. Novacek, and M. C. McKenna (eds.). Springer-Verlag, New York.
This Page Intentionally Left Blank
C H A P T E R
16 Feeding in Marine Mammals ALEXANDER WERTH Department of Biology Hampden-Sydney College Hampden-Sydney, Virginia 23943 I. INTRODUCTION IL FEEDING STRATEGIES A. Filter Feeding B. Suction Feeding C. Raptorial Feeding D. Grazing III CONCLUSIONS References
or dolphin, is a full-fledged marine mammal by virtue of its subsistence entirely within the marine trophic web (nonetheless, it spends most of its life on and around sea ice and is a very impressive swimmer). Indeed, among the frequent suggestions to explain the reversion of mammals to the oceans, the promise of abundant food is regarded as a strong possibility, especially when considering critical near-shore eutrophication and other productivity changes in many Tertiary habitats (Lipps and Mitchell, 1976). Although different marine mammals variously evolved in the warm, shallow waters of the Tethys Sea during the Paleocene and Eocene (presumed to have been the case for cetaceans, sirenians, and perhaps phocid pinnipeds) or on the rocky, steeply sloping shores of the cool North Pacific beginning in the Miocene (in the case of otariid and odobenid pinnipeds, desmostylians, the polar bear, and marine mustelids), all would have avoided increasing competition from the intense radiations of contemporaneous terrestrial mammals. At the same time, these seagoing pioneers could fill niches vacated by the recent extinction of most marine reptiles. Just as feeding might have provided the original impetus for this major environmental transition, it was also a likely cause of further marine mammal diversification. For example, the cetacean suborder Mysticeti (baleen whales) is believed to have originated as a direct result of the Oligocene development of the Circum-Antarctic current, which created nutrient-rich upwelling and in turn led to huge shoals of zooplankton in the South Pacific (Fordyce, 1977,1980). Patterns of sirenian evolution show close correlation with the spread of temperate seagrasses in the Pacific (Domning, 1978b, 1982). Other examples of speciation and dispersal concurrent with changes in food supply
L INTRODUCTION The eutherians commonly known as "marine mammals" obviously do not constitute a valid systematic taxon, as they comprise diverse mammals of carnivoran, condylarth, or subungulate ancestry (approximately 120 extant species in three orders; Table 16.1, Fig. 16.1). However, like so many other living and extinct vertebrates that have, for various reasons, reverted to an aquatic habitat, all have independently evolved a suite of morphological and ecological characters that reflect their shared environment. Just as their locomotor and reproductive systems adapted to an aquatic lifestyle, so too has the feeding system been affected considerably by the new medium (see Chapter 1). This habitat shift necessitated many radical and often dramatic alterations in the feeding methods and mechanisms of their terrestrial ancestors. Marine mammals exhibit widely varying degrees of morphological and physiological adaptation for aquatic locomotion and reproduction, both because of their diverse ancestry and, perhaps more importantly, the great disparity in their times of origin. Cetaceans, for example, appeared long before marine carnivorans. However, the traditional criterion for inclusion in this polyphyletic assemblage concerns diet and foraging. Thus the polar bear, while a less adept diver than a seal FEEDING (K. Schwenk,ed.)
487
Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.
488
Alexander Werth
TABLE 16.1 Simplified Taxonomy of Marine M a m m a l s w i t h Estimated N u m b e r of Living G e n e r a / S p e c i e s Order Cetacea Suborder Archaeoceti'' Suborder Mysticeti Balaenidae: right whales (3/3) Balaenopteridae: fin or rorqual whales (2/6) Eschrichtiidae: gray whale (1/1) Suborder Odontoceti Physeteridae: sperm whales (2/3) Ziphiidae: beaked whales (5/19) Monodontidae: narwhal and beluga (2/2) Delphinidae: dolphins (17/34) Phocoenidae: porpoises (3/6) Platanistidae: river dolphins (4/5) Order Sirenia Trichechidae: manatees (1/3) Dugongidae: dugong (1/1) Order Desmostylia" Order Carnivora Suborder Pinnipedia Phocidae: true (earless) seals (10/19) Otariidae: sea lions and fur seals (7/14) Odobenidae: walrus (1/1) Suborder Fissipedia Mustelidae: sea otter (1/1) Ursidae: polar bear (1/1) ''Extinct.
(and, indirectly, to the development of ocean currents, thermal gradients, and upwelling processes) involve the Miocene radiation of pelagic odontocetes (toothed whales), origin of pinnipeds, and extinction of archeocetes (Lipps and Mitchell, 1976). Oceanographic conditions such as thermal stratification, salinity, depth, and currents play critical roles in marine mammal feeding (Murison and Gaskin, 1989; Simenstad et al, 1990; Smith and Whitehead, 1993). Marine trophic webs are deceptively complex. The critical ecological interaction between sea otters and algivorous sea urchins has been well studied (Estes and Steinberg, 1988; Estes et al, 1998), but it is possible that this relation in turn affected the historic distribution of sirenians (Dayton, 1975). Given the tacit importance of feeding and the urgent need for all marine mammals to adopt new feeding strategies rapidly, one might expect to find a small number of recurring solutions. The biomechanical problems associated with prey capture in an extremely dense and viscous medium should lead to rampant convergence and parallel evolution, both within this group as well as relative to other aquatic vertebrates. Thus, despite the broad taxonomic representation within marine mammals, a few basic feeding types predominate.
It is much more effective to clean small floating or suspended particles from a swimming pool with a dip net or vacuum tube than by attempting to grasp them individually. When attempting to grasp an object in the water, movement toward the object tends to push it away due to a preceding pressure wave of water. This problem is confronted by all aquatic predators who approach a food item jaws first. While many marine mammals rely on this raptorial seizing method and some have evolved elongate pincer-like jaws replete with elaborate dentition, many species, particularly within Cetacea (the oldest and perhaps most highly derived marine mammals), rely on the tried-and-true methods of suspension and suction feeding (dip net and vacuum tube approaches, respectively). These are the predominant modes of feeding among fishes and other aquatic vertebrates as well. It must be stressed, however, that although the simple solutions to these hydrodynamic difficulties are fairly universal, marine mammals (and indeed all marine tetrapods, particularly birds and reptiles) were obliged to remodel functional complexes evolved for feeding in air, and thus had a critical starting point far from that of distant piscine ancestors. In particular, the mammalian hyolingual apparatus is notably distinct from that of fishes in construction and proportions, as are the dentition, oral cavity, and pharynx (Chapter 2). Likewise it is important to remember that the feeding mechanics of fish and aquatic mammals are fundamentally different, even if basic principles remain the same. Thus, from a functional standpoint there are numerous underlying similarities but there are as many key differences. Filter feeding may be the primitive mode of prey capture for many lower vertebrates, but it is a secondary derivation for mysticete whales, as it was for certain marine reptiles of the Mesozoic Era. Because of marine mammals' relatively recent adaptation to a habitat immensely different from that of their immediate progenitors, it is often possible to separate the dual factors of internal and external evolutionary pressures—the competing influences of heredity and environment—and therefore to determine which of their features are functionally adaptive and which are merely vestiges of their terrestrial ancestry. Few organisms offer the opportunity to distinguish so clearly the plesiomorphic and apomorphic characters that create an evolutionary mosaic, and thus marine mammals constitute an ideal group with which to study the pattern and process of evolution. An example related to feeding is the compartmentalized cetacean stomach, cited by some as a practical (i.e., adaptive) derivation (Hosokawa and Kamiya, 1971) yet by others as a historical holdover from ruminant ancestry (Zhou, 1982). This chapter explores marine
MONOTREMATA MARSUPIALIA PHOLIDOTA XENARTHRA - CARNIVORA INSECTIVORA MACROSCELIDEA LAGOMORPHA RODENTIA PRIMATES SCANDENTIA DERMOPTERA CHIROPTERA TUBULIDENTATA ARTIODACTYLA CETACEA PERISSODACTYLA HYRACOIDEA SIRENIA PROBOSCIDEA tDESMOSTYLIA F I G U R E 16.1. A cladistic phylogeny of mammal orders based on Novacek (1992). The position of Desmostylia is based on McKenna and Bell (1997) (some consider them to be more closely related to sirenians). Orders containing marine mammals are in bold. Cetacea, Sirenia, and Desmostylia are entirely aquatic, whereas only some families within Carnivora contain marine taxa (Phocidae, Otariidae, Odobenidae, Mustelidae, and Ursidae). Among these families, the first three are obligate marine taxa, but ursids and mustelids have only one or a few (respectively) marine species. The Carnivoran famihes Phocidae, Otariidae, and Odobenidae have been regarded traditionally as a separate order, Pinnipedia, but current consensus places them within the Carnivora. To include them in a separate order would render the Carnivora a paraphyletic taxon. In any case, monophyly of 'Tinnipedia'' is in dispute; the families may have arisen from different carnivoran ancestors. Desmostylia is the only extinct taxon included in the cladogram.
490
Alexander Werth
mainraal feeding from a historical perspective, including consideration of the morphology and systematics of extinct taxa. However, marine mammal phylogenetics remains a highly controversial subject; even such simple presumptions as the monophyly of Cetacea and the pinniped carnivorans are hotly contested. Thus, relationships presented in Fig. 16.1 should be considered tentative. Despite the apparent advantages for preservation in aquatic environments, the fossil record of marine mammals is spotty and incomplete, particularly in early periods of evolution. Fortunately, ongoing molecular analyses may shed light on marine mammal phylogenies. Despite recent technological advances that have greatly improved our understanding of vertebrate feeding generally, much remains to be learned regarding marine mammals. This is not surprising for they are notoriously difficult animals to study in captivity, let alone in the wild. Research is hindered by obvious logistical and fiscal limitations (mostly stemming from their size and remote location) as well as by equally formidable legal restrictions. Many marine mammals remain endangered or severely threatened due to hunting and habitat destruction. Thus there is an inability to apply standard techniques of vertebrate morphology (e.g., electromyography via surface or surgically implanted electrodes, high speed video, cineradiography) because it is neither feasible to maintain marine mammals as laboratory subjects nor possible (if permissable) to use even mildly invasive or stressful procedures in the field. Nonetheless, marine mammals remain a perenially popular choice of research subject, and despite the challenges, biologists have made great advances in our knowledge of their feeding. Perhaps due to the paucity of direct experimental data, considerable attention has been paid to feeding behavior, such as the bubble netting of humpback whales {Megaptera novaeangliae) to corral prey, the remarkable diving capabilities of Weddell seals {Leptonychotes weddelli), and the ubiquitous tool use of sea otters {Enhydra lutris) to crack open hard-shelled molluscs and echinoderms. Research on captive animals is often limited to husbandry or behavioral studies. Indeed, vastly more is known about marine mammal diet and foraging ecology than of the actual mechanics of prey capture, ingestion, transport, and swallowing. Many details of marine mammal food and feeding have been gleaned from centuries-old lore of European and American whalers or from the Inuit, South Pacific islanders, and other native cultures—all venerable and generally reliable sources. Nevertheless, this information is a poor substitute for controlled experiment or observation as in most cases it has not been, nor can it be, independently substantiated.
Much of our understanding of the mechanical aspects of marine mammal feeding comes from speculative extension of anatomical knowledge, i.e., the inference of function from form (see Chapter 1). While in many cases this n\ay constitute a reasonable approximation of function, this information must be considered conjectural, for in the absence of experimental evidence (or even underwater observation of animals in natural or captive conditions) there is no way to verify it directly. However, despite the many, admittedly critical, differences between marine and terrestrial mammals, there is little reason to suspect that they differ in either the basic principles or the actual details of oral processing and swallowing. There do, however, appear to be major exceptions, such as the masticatory apparatus of sirenian jaws, which deviates markedly from that of other mammalian grazers. In cetaceans, the permanent intranarial larynx and attendant changes in pharyngeal constrictor musculature with a patent airway presumably lead to divergent mechanisms of deglutition, although this is not certain. The aim of this chapter is to review what is currently known about function and relevant structures of marine mammal feeding, but the reader should be forewarned that this information often consists purely of basic anatomical and ecological data (e.g., dental formulae, gross myology; stomach contents, captive foraging observations) or other nuggets of natural history. Ample published literature exists concerning marine mammal digestive systems, although morphological description here will largely be restricted to the oral region. Unless otherwise noted, all details of marine mammal ingestion, transport, processing, and deglutition are presumed to be similar to the basic mammalian conditions, as described elsewhere in this book (Chapters 2 and 13). Marine mammals differ from other aquatic animals in obvious yet critical ways that influence their feeding. As endotherms their feeding must be sufficient to support a high metabolic rate, yet although their elevated body temperature permits a high degree of activity and enables them to exploit many cold environments, it also constrains their morphology so as to prevent undue heat loss. As obligate air breathers, marine mammals must dwell near to the surface, or at least return there periodically. They cannot benefit from pharyngeal slits that ventilate the gills of fishes and larval amphibians and permit a unidirectional flow of prey-laden water into the oral cavity; in marine mammals, flow is necessarily bidirectional, i.e., all water that enters the mouth must be expelled from whence it came or be swallowed (excepting balaenid mysticetes, which circumvent this problem and achieve unidirectional flow because water enters the mouth anteriorly yet exits it posteriorly). Strong pharyngeal and laryngeal musculature prevents
16. Feeding in Marine Mammals
water from entering the trachea and filling the lungs when the mouth is open underwater. Several marine mammals undertake long migrations that carry them far from productive feeding grounds, necessitating long bouts of fasting. This means that certain species must not only build up plentiful nutritional stores for extended periods of time, but also maximize their intake during periodic bouts of feeding. Finally, because marine mammals must suckle (Chapter 14), their feeding apparatus is obliged to accommodate both neonate/juvenile suckling and adult feeding or else undergo some transformation between juvenile and adult phases of their life history. Not only does this affect the feeding strategies of marine mammals but also, of course, their suckling behavior and mammary morphology. In cetaceans, for example, milk is not so much pulled from the mammary gland via sucking action of the tongue as ejected by the contraction of smooth muscles surrounding the nipple (Arvy, 1974). Lactation itself also places many constraints on marine mammals: small litter size, a single pair of internal mammary glands, whose nipples are everted through slit-like openings (to maintain streamlined profile), and exceptionally fatty milk (to prevent water loss). Costa (1991) and Olesiuk (1993) discuss foraging behavior and energy budgets in relation to pinniped reproduction and life history patterns. Marine mammals occupy a unique and fascinating position, for while they are vastly unlike other aquatic vertebrates, they are at the same time highly divergent in comparison to other mammals, particularly in terms of that most celebrated mammalian feature, dentition. In numbers of teeth, marine mammals variously exceed or fall short of the typical eutherian pattern. Not only are teeth typically reduced in quantity, but they are also reduced in differentiation along the tooth row (reduc-
491
tion in heterodonty), reduced in generations, and reduced in number and complexity of cusps. Dental formulae of marine mammals cannot be considered diagnostic aids as individual variation is common. Mysticetes are edentulous as adults, and odontocetes may possess teeth far in excess of the usual eutherian maximum of 42. In the sperm whale {Physeter catodon) only mandibular teeth erupt, whereas in Risso's dolphin {Grampus griseus) there are no maxillary teeth. In many ziphiid whales teeth erupt only in males. Odontocete teeth are typically conical in shape, without cusps or shelves, and occur in just one generation. All of these circumstances—polyodonty, homodonty, and monophyodonty—diverge from standard eutherian conditions. Incisor and canine tusks are found in diverse marine mammals such as the narwhal, walrus, and dugong. Manatee teeth are replaced horizontally throughout life, as in proboscideans. In fact, aside from herbivorous sirenians, there is typically little if any reduction or other processing of food in the oral cavity, although teeth may be used to slash or tear pieces from large prey in some species. Mastication is the hallmark of mammals, yet few marine mammals chew their food. The mouth is specialized for food acquisition alone. The cetacean skull (Fig. 16.2) nicely illustrates some of the profound ways in which marine mammal cranial morphology departs from that of terrestrial mammals. In the absence of mastication the primary mammalian jaw adductors (masseter and temporalis) are reduced; pterygoids are now the dominant jaw-closing muscles. Consequently the zygoma are reduced to slender, thread-like or styliform rods in odontocetes, whereas jugal bones are likewise limited to small stubs in mysticetes; in neither case is this arch firmly fused to the squamosal. Jaws and braincase are greatly modified, as the posterior narial migrations form a long
a.
FIGURE 16.2. Skulls of mysticete and odontocete cetaceans: (a) right whale, Eubalaena glacialis, and (b) long-finned pilot whale, Globicephala melaena.
492
Alexander Werth
"telescoped" rostrum with concomitant elongation of numerous contributions of the dermatocranium andneurocranium (and asymmetrical elements in some species). The rami of the dentaries (straight and compressed in odontocetes, arched and rounded in n\ysticetes) bear rudimentary coronoid and condyloid processes, except in balaenopterid mysticetes, where the coronoid is associated with a ligamentous band that aids in mandibular kinesis. Mammalian carnivores typically exhibit a scissor-like jaw action with a jaw joint that permits no room for lateral motion, yet cetaceans have a rounded mandibular condyle instead of a hingelike one. Bite force is often weak in cetaceans, and except for mysticetes (particularly rorquals and balaenids), there is little "play" or mobility in the mandibular symphysis. The enlarged, flattened hyoid apparatus of cetaceans consists of several ossified and cartilaginous elements articulated by synovial joints. In the absence of a mobile soft palate and epiglottis, the palatal and pharyngeal musculature is significantly altered. Whereas other marine mammals do not demonstrate such extreme cranial modification, their musculoskeletal systems likewise reflect adaptations for feeding in water. The head is obviously an important functional complex. Numerous manifest activities and developmental processes affect marine mammal oral morphology, including modification of nares and nasal passages for air breathing at the surface and streamlining for improved locomotion. Hearing and other special senses are modified significantly in marine mammals, not only to suit their new environment, but as direct consequences of alterations in craniofacial anatomy for feeding. The marine mammal cochlea is attuned to infra- and ultrasonic frequencies that travel well in water. The odontocete "melon" used to focus sonic pulses greatly affects their cranial profile. While cetaceans are well known for their exceptional echolocation and other auditory abilities, they possess no olfactory sense and the presence of a gustatory sense is questionable. Vision is important for foraging, particularly in pinnipeds. Although marine mammals are an admittedly diverse group, several basic generalizations relevant to their feeding can be made. As noted, many fast for extended periods of time, up to half a year (e.g., summer in mysticetes, winter in polar bears). Sexual dimorphism is not uncommon; however, aside from dentition, this does not usually affect diet or feeding method. The stomach is often complex and compartmentalized. The alimentary canal is typically long, without clear transition between large and small intestines, and digestive organs such as the gall bladder may be absent. Several marine mammals, particularly pinnipeds and odontocetes, are known to ingest stones and pebbles inciden-
tally as well as intentionally. These gastroliths may serve as a grinding mill for mechanical digestion in the stomach, especially since prey are often swallowed whole or "bolted" (Owen, 1980), or they may aid in buoyancy regulation. It is not unusual for marine mammals to ingest a variety of other inorganic objects, especially man-made pollution. Concretions or pathological obstructions of the intestine (e.g., ambergris) are found occasionally. Cooperative foraging (with conspecifics, or other marine mammals or birds) is common, and a few odontocetes harken to their terrestrial ancestry by chasing fish onto river banks or surfing onto beaches to snatch sea lions. Pinnipeds and sea otters are so well adapted for aquatic locomotion that although they come ashore to mate and give birth, they are ungainly on land and ill-suited to catching prey there. Despite dwelling in saline habitats, some marine mammals (manatees and sea lions) demonstrate an occasional need for fresh drinking water; others gain sufficient water fron\ their animal or plant diet, from metabolic water production and blubber degradation, and from physiological or behavioral adaptations of digestion, excretion, reproduction, and thermoregulation.
IL FEEDING STRATEGIES Marine mammals employ a variety of strategies to capture and ingest prey. They subsist on virtually every multicellular form of life (including other marine mammals) found in every marine environment, from intertidal or littoral benthic zones to neritic, pelagic, and abyssal habitats of all ocean basins. Their food ranges from tiny zooplankters to large nekton (e.g., squid, sharks, bony fishes) and is obtained at all levels of the water column, including the surface and bottom. Their prey range from nonmotile to highly elusive and can be solitary or gregarious, soft, or covered with a hard shell or test (the latter often cracked or broken for feeding, but sometimes swallowed intact). As a group, marine mammals are predominantly piscivores, although most filter feeders prey on small arthropods such as euphausiids, amphipods, copepods, and mysids. Many odontocetes, especially suctionfeeding species, are almost exclusively teuthophagous (cephalopod-eating). With few exceptions (notably the killer whale, Orcinus orca), all carnivorous marine mammals eat prey whole. Baleen whales are often described in popular literature as grazers; however, this term is properly reserved for animals that consume herbage, algae, or phytoplankton rather than zooplankton, so that carnivory is an appropriate if unlikely term for mysticete feeding. However, the
493
16. Feeding in Marine Mammals presence of volatile fatty acids and symbiotic bacteria in mysticete forestomachs suggests microbial fermentation (Herwig et ah, 1984). Sirenians are the only herbivorous marine mammals; they may browse (i.e., eat leaves, stems, or other portions of plants), but they normally crop all surface growth or consume the entire plant. Catalogues of diet are commonplace and will not be reviewed here. However, there have been no previous attempts to compile a comprehensive, comparative survey of the methods and mechanisms of marine mammal feeding. Because of the tremendous diversity (taxonomic and ecological) within this artificial grouping, this chapter is outlined in a slightly different format, presenting information on structure, function, and evolution by feeding type rather than for marine mammals as a unified whole. Marine mammals can be categorized generally as filter feeders, suction feeders, raptorial ("predatory") feeders, and grazers. These groupings cross taxonomic boundaries. Aside from grazing (Sirenia alone), none of these four categories contains a single taxon exclusively. Cetaceans and pinnipeds have filter, suction, and raptorial representatives. The diversity of these taxa cannot be overemphasized. Still, the main phylogenetic division of Cetacea relates to feeding (hence the old subordinal names Filtrales and Raptoriales), and several researchers have attempted an ecological or systematic classification of cetaceans according to diet and feeding type (Tomilin, 1954). For example, Gaskin (1976) divides odontocetes into ichthyo-, teutho-, and sarcophagi, yet admits that such rigid classifications are impractical, as cetaceans, like all marine mammals, are opportunistic feeders whose preferences are dictated by seasonal migrations and circumstances of prey distribution, depth, and abundance (Frost and Lowry, 1981; Kim and Oliver, 1989; Goebel et a/., 1991; Boyd et ah, 1994). The four categories can be further subdivided into ecological guilds on the bases of habitat, prey size, and locomotion, but there is likely to be considerable overlap in these and other traits. It goes without saying that numerous mammals of various orders inhabit or visit freshwater habitats regularly. Fiowever, none demonstrate the level of separation from land seen in marine mammals, nor do any exhibit the extreme range of anatomical and physiological specialization resulting from tens of millions of years of adaptation to life in water. With few exceptions (e.g., the duck-billed platypus, Ornithorhynchus anatinus), and aside from rare differences in diet and foraging behavior, feeding in aquatic monotremes, marsupials, insectivorans, rodents, and ungulates does not differ notably from that of their terrestrial relatives. The same can be said for marine fissiped carnivorans, for even the
truly marine sea otter and polar bear do not typically ingest prey underwater. A. Filter Feeding Filter or suspension feeding (straining small prey items or particulate organic matter suspended in water) can claim a proud history among vertebrates, as it is not only presumed to be the primitive mode of chordate feeding but has typically been used by the largest aquatic animals, extant and extinct. Among fishes, the largest chondrichthyans (the manta ray and whale, basking, and megamouth sharks) are filter feeders, as are paddlefish, mackerel, and other large osteichthyans. Filter feeding has also been adopted by several tetrapod lineages that reverted independently to an aquatic heritage. Various reptiles from plesiosaurs to pterosaurs were filter feeders, as are some larval amphibians, several aquatic birds (e.g., ducks, flamingos), and some mammals. A filter feeding strategy allows animals to feed near the bottom of a trophic pyramid and thus to reap the rewards of greater biomass (and therefore more enegy) available for consumption, permitting filter feeders to attain giant body size (or, conversely, to support huge populations of smaller animals). Some of the earliest mysticete whales measured only 4 or 5 m (still rather large), yet Recent baleen whales are without question the largest animals that have ever lived. The blue whale (Balaenoptera musculus) is the terminal predator in an abbreviated food chain that often includes just diatoms (Fragilariopsis) and krill (Euphausia). Short trophic pyramids also permitted mysticetes' formerly extensive distributions prior to their decimation by hunting. The mammalian suborder Mysticeti ("mustached whales") comprises three families of baleen whales, each of which employs a distinct type of suspension feeding (Fig. 16.3). Like other tetrapods, most mammals are intermittent rather than continuous filter feeders, i.e., they engulf a single mouthful of water at a time and separate food from this water before expelling it, unlike lower vertebrates that generally pump or push water constantly and unidirectionally through the mouth (Sanderson and Wassersug, 1993). Preyladen water may be ingested either by ram filter feeding, as in balaenopterid mysticetes, or by suction filter feeding, as in the gray whale {Eschrichtius robustus). The distinction depends on how water is drawn into the mouth: by forward motion in the former case and rapid expansion of the buccal cavity to create negative intraoral pressure in the latter. In both cases the ultimate cause of water influx is muscular contraction, with energy supplied by the locomotor or feeding apparatus
494
Alexander Werth
FIGURE 16.3. Mysticete filtering methods: (a) balaerdd (right whale) surface skim feeding, (b) rorqual (humpback, Megaptera novaeangliae) lunge feeding (compare to Fig. 16.7), and (c) gray whale (Eschrichtius robustus) benthic suction feeding.
(tail or tongue, respectively), and both methods are effective at delivering water to the filtering device, the plates and filters of baleen. Members of Balaenopteridae, commonly called rorquals (from the Norwegian "furrowed whale," referring to their pleated throats), rely on ram "gulping" in which a single mouthful of prey-laden water is engulfed by rapid lunges of forward locomotion. As the mouth is closed to expel water, schooling fish, krill, or other micronekton are trapped by the baleen sieve. The gray whale (Eschrichtiidae) is also an intermittent filter feeder, yet it can feed while stationary; its mouth is filled by rapid tongue depression, enlarging the oral cavity to draw in benthic invertebrates, which again are filtered from water and sediment as water is forced from the mouth. A third family, Balaenidae, utilizes continuous ram filter feeding. Right whales slowly skim or "graze" copepods and other minute zooplankton that deposit on finely fringed baleen by a steady stream of water flowing through the mouth. With such substantial variation in mysticete diet and foraging behavior, it is not surprising that each family also possesses a distinct suite of morphological characters associated with feeding, including differences in the filter itself. Baleen is commonly known as whalebone, although it is not an ossified or mineralized tissue, but rather a keratinaceous integumentary product that develops
along the margins of the palate and is suspended from the maxillae in large, flexible laminae or plates (Fig. 16.4). Baleen ranges in color from creamy white or yellow to olive, gray, and black. Each side or "rack" may contain (depending on the species) from 100 to 400 triangular plates of baleen, ranging from 50 cm to nearly 5 m in length and only a few millimeters in thickness. The plates are arranged in transverse series, like the teeth of a comb, at intervals of roughly 1 cm. The base of the scalene triangle is rooted to a foundation layer in the gum, whereas the long outer edge is smooth and the medial side (the hypotenuse) is frayed with bristles or "hairs" that intertwine to form a fibrous mat. The continuously skim-feeding balaenids possess 250-350 plates of narrow, flexible, finely fringed baleen plates, up to 5 m long in bowheads and with 35-70 bristles per cm^ (Leatherwood et ah, 1983). The rorquals and gray whale have much wider, shorter baleen, 0.5-1 m long, with much coarser fringes. As might be expected, the porosity of the filter (number of plates and fringes per plate) correlates strongly with prey size. The finer toothed comb of balaenids captures tiny zooplankters that would pass easily through the filters of other whales, primarily piscivores or krill eaters (although the Sei whale, Balaenoptera horealis, also feeds on small zooplankton and possesses the finest fringes of any rorqual). However, the simple relation between strainer
495
16. Feeding in Marine Mammals
F I G U R E 16.4. Mouth of bowhead whale, Balaena mysticetus, showing arrangement of baleen plates.
and prey is in fact more complex, as the straight inner surface of balaenid baleen means that few fringes can be exposed here. Pivorunas (1976) suggested that smaller fringes alleviate the shortcomings of fewer fringes and
presented a mathematical model correlating plate features with fringe characteristics. Where the plates angle outward, as in rorquals, many fringes are exposed on the medial surface by friction (from water, tongue, and prey) so that the coarse fibers in this dense, tangled, interwoven mat need not have such a small diameter. Baleen is neither homologous nor histologically related to teeth. In fact, tiny tooth buds develop in most mysticetes, but these anlagen disappear before birth. Baleen appears to be homologous to the horny palatal ridges of artiodactyls, which is not surprising given their close phylogenetic affinity (Graur and Higgins, 1994). Nonetheless, teeth and baleen share some developmental similarities. Like teeth, baleen growth begins as a dermal/epidermal interaction (Fig. 16.5). Long, slender, conical papillae extend from an underlying
inner fringes keratin sheath outer fringes
papilla
baleen plate
horn tube (papilla and keratin) epithelium (gingiva) base of papillary process upper jaw bone intermediate layer (compacting horn) basal (foundation) plate F I G U R E 16.5. (1979).
Diagrammatic section of baleen developing from palatal dermal process. After Pivorunas
496
Alexander Werth
basal plate of connective tissue, the dermal process, which overlays the bones of the upper jaw and is in turn covered by a layer of epidermis in which the papillae are enveloped with a horny layer of keratin to form long tubular bristles or horn tubes. These tough, fibrous strands are likewise surrounded and cemented together by a layer of compacting horn, although the innermost tubes are loosely packed without connective tissue so that they can slide relative to one another. A softer, cushioning layer of intermediate horn substance provides a dense cortex covering the anterior and the posterior faces of the plate. The more lateral horn tubes grow faster than the medial ones, so that the long (smooth) side of the triangle faces the outside of the mouth, whereas friction abrades the matrix medially, wearing away compacting horn to reveal the hollow horn tubes, which remain as fine or coarse fringes to serve as the sieving apparatus. Cells in the rubbery, pliant epithelium of the gums anchor baleen to the palate and divide to replenish abraded gingival tissue, just as the base of each papilla grows to replace the worn-out horn tubes. Although the visible portions of baleen consist exclusively of dead, cornified cells, the dermal process remains a living tissue so that baleen is analogous to the part-living tissue in a horse's hoof (Slijper, 1962). The anisotropic nature of this growing tissue's construction, with a homogeneous cortical layer covering free, hollow, cylindrical tubes, affords maximal strength with minimal mass. Baleen is strong yet pliant and elastic, a suitable material to meet the demands of constant exposure to friction and abrasion. Unfortunately, the physical characteristics of this material also made it an extremely precious commodity that helped to fuel the whaling industry. Baleen whales eat whole organisms, whereas most tiny filter feeders, with correspondingly diminutive filters, feed on detritus (fragmented organic debris). However, like other strainers, mysticetes are not selective; they locate patchy sources of food and swallow whatever they trap. In fact, all marine mammals are opportunistic feeders and ingest their share of rubbish, but grazing, raptorial, and suction feeders exercise more discrimination in the items they ingest. Although baleen fringes probably capture prey simply by passive sieving, other mechanisms best known from studies of suspension feeding fishes—direct interception, inertial impaction, and gravitational deposition—maybe possible (Rubenstein and Koehl, 1977). Mysticetes do not rely on mucus or other sticky substances to trap prey. As noted, the porosity of the filter determines prey size, but fine filters catch large as well as small prey and hence may be more versatile. However, dietary studies indicate that coarse-fringed mysticetes are less specialized (Nemoto, 1959, 1970), ingesting large items ranging
from fish to hapless seabirds. Because baleen is not a rigid material, filter porosity varies according to hydrodynamic factors such as velocity of ram locomotion, size and density of prey, and direction and pressure of water flow (Sanderson and Wassersug, 1990). While very tiny filter feeders (such as many marine invertebrates) must contend with difficulties of moving in a dense, viscous fluid, mysticete kinematics depend more on inertial than on viscous forces. Mysticetes can exploit certain aspects of water flow and incompressibility, since by virtue of their size and speed they operate at the opposite end of the spectrum, with Reynolds numbers many orders of magnitude distant from invertebrate filter feeders. Balaenid whales (of which there at least three species: bowhead, northern/southern right, and pygmy right) are slow, rotund cruisers that filter a constant current of water through their mat of finely fringed baleen to skim dense swarms of zooplankton collectively designated by whalers as "brit" (Watkins and Schevill, 1979; Lowry and Frost, 1984, Mayo and Marx, 1990; Carroll et al, 1987). As in whale and basking sharks, the head of balaenids is enormous, measuring up to onethird of their 10- to 20-m length. The upper jaw is strongly curved (Fig. 16.2) to fit the exceptionally long, narrow baleen plates, which fold back when the scoopshaped lower jaw is closed. The giant head functions as a plankton tow net, although it is of course not pulled, but pushed while feeding at about 5 k m / h r so that these whales also share the gradual pace and gentle placidity of the largest sharks. Balaenids are the only mysticetes with a large anterior cleft between the left and the right baleen racks (Fig. 16.4). Water enters this subrostral gap, flows through the oral cavity and between baleen plates, and passes out a gutter-like orolabial sulcus at the rear of the lower lip (Fig. 16.3). It has been supposed that the great pressure drag during continuous filtration slows balaenids and prevents capture of large or evasive prey. Watkins and Schevill (1976) noted that water backs up into the mouth during surface feeding, with a level higher than that of the surrounding sea (although gravity would then force water through the baleen). Lambertsen et al. (1989) conducted a thorough photogrammetric study of bowhead baleen to document its curvature and speculate as to its role in establishing currents of water flow through the oral cavity. As in all mysticetes, plates curve somewhat from medial to lateral aspect so that they appear slightly C-shaped when viewed in transverse section, and the lateral edge of each rack bows outward as well. This creates a narrowing space between the baleen and lips that may be responsible for an inertial rather than passive flow of prey-laden water that enters the mouth, as the decreased space increases pressure and hence flow
16. Feeding in Marine Mammals velocity, setting up a Bernoulli effect (Lambertsen et a/., 1989). As this water flows posteriorly, water still in the center of the buccal cavity is induced to flow medially (perpendicularly) through the plates by a Venturi effect. Lambertsen et al (1989) also studied the morphology of the lower jaw that allows the lip to rotate outward, varying the width of this channel. Werth (1995) devised mathematical and physical models of the bowhead oral cavity to test these biomechanical predictions. Use of such hydrodynamic, rather than simple hydraulic forces, allows balaenids to reduce turbulent flow, improve filtering efficiency, and avoid creation of an anterior pressure wave so that they may in fact be able to capture elusive prey even at slow swimming speeds (Lambertsen et a/., 1989; Werth, 1995). Although these hydrodynamic effects actually draw water into the mouth, the negative intraoral pressures, if any, would be negligible and should not be considered akin to suction feeding by any means. Wiirsig (1988) suggested that foraging in tight formations also allows balaenids to consume evasive prey. Although bowhead and right whales are often observed feeding on the surface or lower in the water column, there is little doubt that they skim along the bottom as well, as evidenced by scratched, muddy heads and by the presence of benthic crustaceans in stomach contents. Right whales also feed while "tail sailing" (Hamner et al, 1988). Balaenids have an uncanny knack for locating dense slicks of zooplankton in which they turn back and forth like mowers cutting fields of grass (as beautifully described in Moby Dick; Melville, 1851). Presumably they find and judge the density of such swarms visually. Regrettably, extremely limited visibility precludes underwater filming of balaenid feeding, although films from clearer summer breeding grounds (Werth, 1990) suggest that the massive, firm, elevated tongue (the largest muscular organ in the world at lengths of nearly 5 m in large animals) sweeps laterally to deflect or channel water to the baleen racks, as suggested by Nemoto (1959) and Gaskin (1982). Rorquals (Balaenopteridae) display the most extreme morphological and mechanical specialization of any mysticete. This family includes the familiar humpback whale {Megaptera novaeangliae) and five species of the genus Balaenoptera (including blue and fin whales), all longer, slimmer, and faster than balaenids, enabling them to draw in mouthfuls of prey-laden water on an enormous scale through rapid gulps and lunges. This is accomplished by inward folding of the distinct, yet loosely organized and weakly muscular tongue into a ventral space with subsequent massive gular expansion. This pouch, which receives engulfed water and the displaced, bag-like tongue, also extends over the thorax to the umbilicus, momentarily giving the body a spec-
497
tacular bloated tadpole shape (Figs. 16.3 and 16.7). Storro-Patterson (1981) calculated that the total volume of a large, hypothetical blue whale (B. musculus) increases over 600% during feeding, from roughly 5550 to 35,700 cubic feet! He further speculated that as much as 1000 tons of water are engulfed in a single gulp. Pivorunas (1979) provided a more conservative estimate of at least 60 m^ (approximately 70 tons) of water ingested in one gulp, still an incredible amount of water equal to roughly 50% of a blue whale's total body volume. While gulping is not so pronounced in other rorquals, this impressive capacity for gular distention is seen in all balaenopterids. Despite trivial physical variation (apart from length) among species within the genus Balaenoptera, there appears to be considerable behavioral divergence and ecological partitioning. The sei whale (B. borealis) has finely fringed baleen and prefers to skim copepods (Kawamura, 1974), whereas Bryde's whale (B. edeni), another moderately sized rorqual, mainly eats fish (Nemoto, 1959). The versatile foraging oiMegaptera includes bottom feeding (Hain et al, 1995). Once depicted as stout dirigibles (based on animals stranded or hauled out on flensing platforms), underwater photographs now confirm balaenopterids' exceptionally streamlined form, particularly in Balaenoptera. Rorquals are built for speed, and rapid locomotion powers their feeding. Rather than expanding a space to create negative pressure and draw water in, rorquals unlock their jaws and relax adductor musculature to open the mouth (at least 30° and up to 90°), which suddenly fills with water in much the same way as a bag is opened by rapidly pulling it through air. Positive inertial pressure forces open the space into which the water and prey flow; seawater is passively enveloped rather than displaced forward or sucked internally (Orton and Brodie, 1987). However, smaller rorquals may also retract the tongue somewhat to pull in a stream of water that fills the oral sac toward the rear rather than the sides (Pivorunas, 1979). In any case, the goal is to avoid a wave of resistant pressure that would disperse prey at the entrance to the mouth and thus interfere with their capture. Because the expansive pouch allows huge quantities of water to enter, pressures build only when the incurrent stream slows (Pivorunas, 1979). Orton and Brodie's mathematical analysis (1987) confirmed that rorqual feeding can be powered solely by locomotion. Because gular distention depends on swimming speed, precise timing with respect to prey location is critical for successful engulfment. If jaws are opened a moment too soon the filled mouth pushes the intended prey ahead with a compressive bow wave (Brodie, 1977). Key anatomical innovations for feeding in this family include the loose tongue (distinct yet flaccid and deformable, with a central furrow) and floor of the mouth
498
Alexander Werth
(with broad, flattened mylohyoid and geniohyoid); accordion-like longitudinal elastic throat pleats; wideopening jaws with a locking joint that creates a hydrostatic seal to prevent unintentional jaw opening during rapid locomotion, and a frontomandibular stay that stores kinetic energy for jaw closure; unfused mandibular symphysis with Y-shaped fibrocartilage arms extending along the mandibular rami (Pivorunas, 1977); and flat, streamlined rostrum. The cavum ventrale, named and described by von Schulte (1916), is an intermuscular fascial cleft extending under the mouth, throat, and chest. The slick walls of this space slide freely over adjacent surfaces and the cavum enlarges to form a giant, bulging vestibule as it receives engulfed water and the displaced tongue and oral floor (Pivorunas, 1979). Dissection of fetal and adult rorquals (Pivorunas, 1979) reveals major ontogenetic changes in the tongue as it is transformed from a solid, muscular structure used in suckling to the deformable, flaccid sheet seen in adults. From around birth until weaning, the paired lingual muscles separate from the midline septum and muscle tissue is replaced with adipose and elastic connective tissues as the increasingly saccular
cavum ventrale
organ flattens and spreads laterally. Intrinsic lingual muscle fibers are scattered and poorly developed (Fig. 16.6). As in balaenids, the rorqual tongue may also serve as a seasonal store of adipose tissue (Howell, 1930; Tarpley, 1985). Lambertsen (1983) elucidated the anatomy of the balaenopterid tongue and its dynamic inversion into the intracaval position (Fig. 16.7). Experiments with the head of a minke whale (B. acutorostraia, a species with a more bulky tongue) demonstrated that when the mouth is filled with water to simulate engulfment, the loose tongue folds into the cranial portion of the cavum (between its inner and outer walls) to initiate distention of the capacious oral pouch. The inverted tongue acts as an elastic sac whose lumen is continuous with the buccal cavity and whose walls are formed by the nowinvaginated tongue and nonlingual intermandibular lining. The mouth balloons out in bullfrog or pelicanlike fashion as the distensible pleats expand. Lambertsen (1983) speculated that closure and expulsion occur by the active contraction of jaw adductors (especially the temporalis) and elastic recoil of the pouch. In their biomechanical study of the throat wall, Orton and
pouch musculature
gular pleats
F I G U R E 16.6. Diagrammatic transverse section through rorqual mouth showing relation of baleen to lips. Note the cavum ventrale (black space) separating the sheet-like lingual musculature from the longitudinal pouch muscle underlying ventral blubber.
16. Feeding in Marine Mammals
FIGURE 16.7. Rorqual engulfment mechanism, indicating normal streamlined body profile and extensive inversion of tongue and distension of cavum ventrale (dashed lines) during maximal pouch expansion. After Lambertsen (1983).
Brodie (1987) outlined three means of pouch deflation and water expulsion: dynamic pressure of water coming to a stop at the front of the mouth, stored elastic energy in the blubber and other tissues surrounding the cavum, and possibly active contraction of muscle underlying the blubber (although they noted that this has yet to be shown). They found large amounts of elastin in ventral pleat blubber and gular musculature and quantified limits of reversible deformation in these tissues (circumferential and longitudinal expansion of groove blubber up to 4 and 1.5 times resting length, respectively). The loose articulation of the mandibular rami by fibrous ligaments rather than a bony symphysis permits a wide gape and cushions against the stresses of water entry (Gaskin, 1976). Lambertsen et al (1995) synthesized anatomical study, behavioral observation, and biomechanical analysis to describe the structure and function of a fibrous frontomandibular "stay" apparatus extending (as an appendage of the temporalis) from the supraorbital process of the skull to the coronoid process of the jaw. This taut connection optimizes gape angle and mandibular rotation during engulfment and acts as a spring to store elastic energy to facilitate gape closure by kinematic reversal (transferring the enormous momentum of the engulfed water to the jaws) and perhaps to focus the flow of expulsion through baleen. Water may also cascade out of the mouth by gravity during lunges above the surface. The gray whale {Eschrichtius robustus) employs benthic suction feeding, turning on its side and skimming the bottom while rapidly depressing and retracting the tongue to stir up sediments and suck in prey (Fig. 16.3), primarily molluscs and gammaridean amphipods that are winnowed from a single mouthful of muddy water with stiff, short, coarse baleen (Murison et al., 1984; Nerini, 1984). Scores of suction-generated pits or depressions (1-5 m deep) scar the ocean bottom along gray whale feeding grounds (Oliver and Slattery, 1985; Nel-
499
son and Johnson, 1987; Weitkamp et al, 1992), and muddy snouts, scarred and abraded jaws, and unilaterally worn baleen are common in this monotypic family. Gray whales seem predominantly "right handed," as baleen on this side is often shorter, whereas barnacles collect on the left side of the head (Kasuya and Rice, 1970). Scarring is also asymmetrical. The tongue is firm and muscular and there is a small subrostral gap as in balaenids. Although Eschrichtius occasionally scrapes prey off strands of kelp or skims fish and squid suspended in open water (even gulping like rorquals; Sund, 1975), the "mussel grubber" (as Yankee whalers called it) is the only mysticete to ingest food with strong, internally generated suction pressures. This behavior was first documented in a young captive animal given blocks of frozen squid (not a natural diet), which, when thawed and separated, sank to the floor of the pool (Ray and Schevill, 1974). The yearling whale turned with her lip just above the bottom and opened her jaws slightly, sucking in water and expanding her throat so that the gular grooves (two to five deep creases) bulged out. Lips could be moved independently and curled away from the baleen. She ate 900 kg of squid daily but spat out fish that had been engulfed in this manner. Turbid water was also seen squirting from the mouth. Although this animal always turned to the left while feeding, this is probably because she was given food on this side when trained to accept squid. Ray and Schevill (1974) cautioned that suction feeding may not be normal behavior of a wild adult, but the abundance of pits and mud plumes trailing behind gray whales argues otherwise. It has been suggested that this gouging and plowing of the sea floor actually increases productivity (Nelson and Johnson, 1987; Klaus 6f al, 1990). By the Middle Oligocene all three mysticete lineages were present, and it is interesting to consider which of the feeding mechanisms described earlier, if any, was ancestral for Mysticeti, especially since Eschrichtius is often regarded as an archaic species. Were mysticetes originally suction feeders? Did gulpers and skimmers develop from whales that captured prey with suction? Whether the balaenid method of continuous ram filter feeding or the intermittent mechanisms of other mysticetes constitutes the plesiomorphic condition is a mystery that the fossil record cannot yet resolve. Although continuous filter feeding is generally presumed to be a primitive character for vertebrates, many believe that Recent gray and rorqual whales come closer in form and function to the earliest baleen whales, suggesting that balaenids diverged from the mysticete lineage later on and that their feeding is a derived condition. Although the rostral shape of the earliest known mysticetes (Aetiocetidae) is controversial and their soft tissue anatomy obviously unknown, it is possible that some
500
Alexander Werth
were rather blunt-headed suction feeders. However, Sanderson and Wassersug (1990, 1993) proposed that intermittent suction feeding occupies an intermediate functional stage between continuous ram filter feeding and capture of individual prey items via suction, which would suggest that balaenids employ the original mysticete feeding method. As Pivorunas (1979) noted, balaenids are the only whales in which the sides of the mouth are entirely blocked by baleen when the mouth is opened to full gape, for their "whalebone'' both catches and gathers prey, unlike rorqual and gray whales who employ a two-step process in which the oral cavity catches the prey and the baleen retains it as water is rapidly expelled. The latter whales do not capture their prey directly from water per se, but instead utilize lunge feeding or suction to engulf a mouthful of preyladen water, then take advantage of the baleen filter to separate their (typically larger) food from the exiting water. The straining can occur at leisure, without fear of prey escaping, so that the baleen can be said to be of secondary importance to prey capture, as it technically plays no role in the initial stages of ingestion (Pivorunas, 1979). Another interesting point concerns intermediate species during the evolution of baleen. Early mysticetes had small, multicusped teeth that may have acted as a filter (as in filter feeding seals), whereas later forms possessed tiny teeth and jaw adaptations similar to modern mysticetes, raising intriguing possibilities of transitional forms and the gradual evolution of this filtration mechanism from simple palatal papillae. Miller (1929) presented the elevated, rugose gums that largely obscure the teeth of Dall's porpoise (Phocoenoides) as an analogue for the origin of baleen. More recently, Mitchell (1989) hypothesized that filter feeding developed in mysticetes with widely spaced, notched or serrated teeth, creating a dental straining apparatus analogous to that of the crabeater seal, Lohodon carcinophagus (Fig. 16.15). Location of new fossil material should help resolve these issues. Mysticetes are surface, midwater, and bottom feeders, yet they depend heavily on producers and thus do not stray far from the photic zone (upper 200 m). Gray whales are not found beyond the continental shelf, and balaenids likewise have a coastal distribution, feeding in productive waters near upwelling zones or at high latitudes (where long periods of daylight spawn abundant phyto- and zooplankton). Rorquals are more pelagic and cross large stretches of open ocean, but like right whales, they feed in polar regions or near seamounts, escarpments, or other major oceanographic features that augment prey density. Mysticete zoogeography is well defined. Migration patterns are firmly established and like winter breeding grounds, summer feeding grounds vary little, if any, from year to
year. However, whales are known to follow shortterm fluctuations in prey distribution and concentration caused by temperature and currents. Pivorunas (1979) speculated that while pack ice in the Arctic Ocean and other northern seas limits phytoplankton growth, it also breaks up large stretches of water where storms would tend to develop, forming heavy seas and other disturbances, as in southern oceans. He proposed that the calmer, Arctic waters might allow indigenous cetaceans (bowhead, beluga, and narwhal) to remain yearround instead of remaining only for brief summer periods, as occurs around Antarctica. The lack of strong, disruptive waves or upwelling currents might have made continuous filter feeding (particularly at the surface) a more attractive prospect for bowhead and northern right whales. Curiously, while the southern right whale {Eubalaena glacialis australis) inhabits a notoriously stormy region, it prefers to spend much of its time in sheltered bays such as Golfo San Jose near Peninsula Valdez off Patagonia, at least during winter breeding and calving seasons, although in general right whales are less migratory than rorqual and gray whales. Specialized behaviors associated with lunge feeding are as remarkable as the mechanics. Rorquals employ numerous agile feeding behaviors such as stereotyped circular "pinwheel" or figure-eight swimming patterns, breaching and surface lunging (vertical, lateral, or upside down), and flipper and fluke splashing to smack and stun prey. Humpbacks use remarkable entrapment devices such as bubble nets, clouds, columns, rings, and curtains emitted by the blowholes (Jurasz and Jurasz, 1979; Watkins and Schevill, 1979; Hain et al, 1982; Gormley, 1983; Wiirsig, 1988). Side swimming is very common and seems to be an effective means of engulfing two-dimensional swarms of prey schooling at the surface. Side swimming is seen in raptorial odontocetes and other narrow-snouted, but wide-gaped predators and has been postulated for archaeocetes as well (Barnes and Mitchell, 1978). Other 'lateralized" foraging tendencies (e.g., breaching, spinning, flipper slapping) are well documented (Clapham et ah, 1995). Humpback "flick feeding" involves wave production by slapping the tail when diving; the whale then surfaces and swims through the wave (Evans, 1987). Hays et ah (1985) described a resourceful humpback that by slowly sinking created a low pressure zone and thus concentrated its prey. "Lobtailing," in which a whale slaps its tail on the surface to stun or congregate prey momentarily, is also common. Foraging costs in mysticetes (both solitary and group feeders) correlate with prey patch density and depth (Dolphin, 1988). As with other social behaviors, mysticete foraging yields clues to their ungulate ancestry (Wiirsig, 1988). Balaenid and balaenopterid foraging is often cooperative (although
16. Feeding in Marine Mammals this may in fact simply stem from locally abundant food supplies) and may involve interspecific groups (Whitehead and Carlson, 1988) or special swimming formations such as V-shaped echelons that apparently limit prey escape (Wiirsig, 1989). Wlirsig (1989) suggested that the nongregarious gray whale may establish and defend individual feeding ranges. Photoidentification confirms that individual humpback and minke whales develop distinctive, characteristic foraging behaviors (Bonner, 1989; Hoelzel et ah, 1989). Coloration apparently plays an important role in mysticete feeding: for concealment and camouflage (countershading to disrupt body outline or mottling to blend in with shoals of prey), for attraction, or for herding prey into denser concentrations (Mitchell, 1970). Two commonly cited examples are the long, white, knobby flippers of humpbacks and the asymmetrical pigmentation of fin whale jaws (white on right side, dark on left), both of which may startle and scare dense clouds of prey toward the mouth (Brodie, 1977). Megaptera relies on complex foraging maneuvers whereas Balaenoptera uses plain, rectilinear locomotion (and typically rolls to the right) while feeding. Brodie (1993) also proposed that noise generated by synovial joint cracking during mandibular realignment in fin whales might confuse and corral shoals of prey. Unlike odontocetes, mysticetes do not use sound production to locate prey. Regarding the timing of engulfment noted earlier, visual prey detection is possible but unlikely given poor visibility underwater. Tactile mechanoreceptionis more feasible considering the vibrissae scattered over the tip of the rorqual rostrum and mandible. The large "stovebolts" or mental/genial and maxillary tubercles of Megaptera each contain several sensory hairs; Brodie (1977) speculated that since this species is less streamlined than other rorquals, the nodes allow vibrissae to project beyond the thick boundary layer created by more turbulent flow. Lingual mechanoreceptors may also function in prey detection, although these would, of course, be useless when the mouth is closed. Mysticetes consume huge quantities of prey, with estimates ranging from 200 to 1000 kg per meal and 200 to 600,000 kg annually [Gaskin (1982) summarized information on mysticete food intake, metabolism, and energy budgets]. The retention of prey by baleen is straightforward, but the mechanism of food removal from this filter is not yet understood. Several competing theories have been proposed to explain this mystery, all loosely supported by anecdotal evidence (Werth, 1990). The most common idea is that the tongue is simply elevated and retracted to scrape entrapped items from the mat. Baleen is estimated to grow roughly 30 cm per year in balaenids, and its occasional presence in whale feces is offered as evidence that it regularly wears away. Lin-
501
gual depression and retraction would allow a bolus of prey collected on the central furrow to be swallowed. Potential disadvantages of this mechanism include rapid removal of baleen and inefficient clearing of prey from fringes that do not directly contact the tongue. Another hypothesis, also mechanical in nature, depends on observations of rapid head-shaking behavior in southern right whales. Whales have been sighted shaking their heads rapidly from side to side just above the surface, with a sound that can be heard from a great distance (not unlike the "baleen rattle" during right whale skim feeding, produced by lapping of water over partially submerged plates; Watkins and Schevill, 1976). Again, this method might not release prey adequately, although because balaenids sometimes skim for hours between apparent swallows, it is likely that they would release enough food to swallow. However, right whales often make short, lunging rushes during bouts of skim feeding, which Gaskin (1982) hypothesized could agitate and remove clinging food particles. A final plausible proposal involves "backwashing" by brief entry of water into the mouth to remove effectively items from the baleen, depositing them on the tongue for swallowing. Opening (and perhaps outward rotation) of the jaws coupled with rapid tongue depression might cause sufficient negative pressure to draw water in from all directions. The backwash current need not be especially strong merely to transport prey into the middle of the oral cavity. Although this method would be very effective, it requires lingual and labial mobility. Tongue movements in all three hypotheses involve changes in position rather than shape, which is consistent with myological findings in right and bowhead whales (Werth, 1993) for which extrinsic lingual muscles (originating outside the tongue and serving to move it) are far more prominent than intrinsic muscles (which exist solely within the tongue and presumed to be the main effectors of lingual shape change). Another filter-feeding marine mammal is the crabeating seal, Lobodon carcinophagus (Phocidae), which, in fact, preys not on crabs as its name implies, but on krill. Krill are sieved from the water with uniquely lobed, ornate cheek teeth, which, when interlocked, permit water to drain from the mouth. The incisors and canines are shaped as in other phocids (although the canines are notably small), yet the five upper and lower postcanines possess a complex array of well-defined, comb-like cusps (Fig. 16.15) that interdigitate when occluded to retain prey while water is expelled from the oral cavity (King, 1961). Prey are prevented from escaping posterior to the tooth rows by bony protuberances (covered with gingiva) of the dentary and maxilla. Lobodon has been observed swimming open mouthed through dense shoals of krill; however, observations of captive
502
Alexander Werth
animals suggest that its intermittent suspension feeding is more likely due to suction generation than locomotion-based ram intake. Although this species is rarely held in captivity, two juveniles showed an ability to suck small fish into the mouth rapidly and forcefully from distances as great as 50 cm. Prey were held in the mouth by the dentition and immobilized against the palate by the tongue as the jaws closed and the lips raised (along with lingual elevation) to expel water with a "lip-smacking" action prior to deglutition, although it is not known whether some water was swallowed as well (Ross et al, 1976). The elongated mandibular symphysis affords large surfaces for genioglossal and geniohyoid muscle origins, presumably the basis of powerful tongue movements (King, 1961). This reliance on suction (for prey ingestion), combined with a straining apparatus (to filter prey from the water), demonstrates once again the versatility of marine mammals and the incomplete division of their feeding strategies. Noting the much smaller size of the head of Lobodon relative to mysticetes, Bonner (1990) proposed that the lobulated teeth are an adaptation for expelling large amounts of water engulfed with individual prey items; because less water is ingested (and swallowed) with prey, a smaller filtration mechanism suffices. Thus, rather than engulfing a mouthful of prey (as in baleen whales), crabeater seals apparently suck prey items selectively, and once sufficient krill accumulate a bolus is swallowed. According to 0ritsland (1977), the average crabeater seal meal is roughly 8 kg of krill. As each krill weighs only 1 g, Lobodon spends a large proportion of its time filter feeding, made practicable only with the immense swarms of this crustacean in Antarctic seas (Bonner, 1990). Given such abundance, it seems likely that Lobodon ingests several krill at a time (King, 1961). Although Lobodon feeds exclusively on krill throughout most of its circumpolar distribution, it may also take small fish (as in captivity). Stomach contents and captive behavior also hint at benthic foraging (Ross et al, 1976). King (1983) speculates that Lobodon feeds mostly at night when krill and other crustaceans approach the surface; feeding dives are shallow (to 25 m) and short in duration, but may occupy the entire night (Bengtson and Stewart, 1992). The crabeater seal's suction/filter feeding is doubtless highly effective, for this Antarctic phocid is variously described as the world's most abundant seal (Reeves et al., 1992) or even the world's most abundant large mammal (Laws, 1984). With an estimated annual consumption of 63 million tons of krill (Reeves et ah, 1992), crabeater seals are strong competitors of other krill-eating species, especially mysticete whales, whose severe reduction after several decades of intense hunting around Antarctica has probably contributed greatly to the success of crabeater seal populations.
Some phocids, notably the leopard seal, Hydrurga leptonyx, have complex cusped postcanines reminiscent of (but not as pronounced as in) Lobodon (Fig. 16.15). These sharp points probably serve to immobilize struggling prey, although they may permit water to flow from the mouth, as zooplankton constitutes a small yet seasonally important component of the diet of ringed seals (Phoca hispida), Antarctic fur seals {Arctocephalus gazella), and other northern and southern phocids and otariids. Noting the presence of slight yet similar bony protuberances on dentaries of several phocids. King (1961) speculated that filter feeding of the crabeater seal may represent retention of a primitive condition, from which other seals later adapted to piscivory; alternatively, this habit may have arisen twice, but evolved to a far more specialized level in Lobodon. B. Suction Feeding Suction is used not only by select suspension feeders, of course, but also by many marine mammals that ingest and transport prey with negative pressures generated via oral and pharyngeal expansion. The difference is that while gray whales and crabeater seals consume vast quantities of aggregating organisms (hence the need for filtration mechanisms), other cetaceans and pinnipeds vacuum larger and ordinarily more evasive solitary prey; again, items are retained while engulfed water is expelled back through the mouth prior to swallowing. Although broad variation exists in prey size, type, and activity, mesopelagic cephalopods are by far the most common choice of marine mammal suction feeders, with bivalve molluscs and other benthic invertebrates second. This extremely effective and versatile solution to the problems of aquatic prey capture has been independently adopted by nearly all marine and freshwater vertebrates and has been extensively studied in elasmobranchs, actinopterygian fishes, lungfishes, and the coelacanth. Among tetrapods, suction feeding has been described in larval and adult salamanders, pipid frogs, caecilians, and turtles (Lauder, 1985). Although simple in principle, suction feeding often requires substantial modification of the skull, jaws, and hyolingual apparatus. Thus it is not surprising that this mode of feeding is practiced by some of the most derived and specialized marine mammals, including several odontocetes and pinnipeds. Among the former at least five families (delphinids, monodontids, phocoenids, physeterids, ziphiids) have definite or putative suctionfeeding members, whereas there are undisputed suction feeding representatives of the carnivoran families Odobenidae and Phocidae (walruses and seals, respectively). It must be emphasized, however, that no family of marine mammals is composed entirely of suction
16. Feeding in Marine Mammals feeders. Clearly this trait has evolved multiple times. Likewise, there appear to be no species that rely exclusively on suction feeding. While the walrus and narwhal are perhaps the most specialized suction feeders (in terms of morphology and foraging ecology), they are nonetheless extremely opportunistic mammals that may utilize other feeding methods when necessary. Unfortunately, while the kinematics, muscle mechanics, and pressure attributes of the discrete phases of suction feeding have been worked out in great detail for myriad suction feeding fishes and amphibians, this is not the case for marine mammals, as research is severely hampered by the constraints noted earlier. In fact, morphologists were reluctant to accept the feasibility of suction feeding in marine mammals given the anatomy of "typical" dolphins and seals (e.g., long jaws and open "notched" gape with numerous sharp teeth), particularly their lack of features seen in suction feeding fishes, which mediate precise, undirectional water flow (e.g., pipette-like mouth, opercular ostium). Despite the traditional view of long-snouted snappers with elongate jaws and prodigious teeth, many toothed whales and pinnipeds exhibit an entirely different profile and dentition (Fig. 16.14), far from that of the wellknown bottlenose dolphin and harbor seal. Their feeding behavior is likewise divergent from the customary picture. A cursory examination reveals major variation in characters linked to suction feeding, from oral openings to gular myology. The anatomical mechanism of suction feeding involves rapid, piston-like retraction of a flat, hemicylindrical tongue, creating a negative (less than ambient) pressure in the buccal cavity into which water and prey are drawn. Ingested water in this bidirectional flow system is momentarily accommodated by the expandable, elastic pharynx and (in many cases) by external throat grooves. It is possible that the stomach participates in some species as well, although influx this far posteriorly (and internally) is doubtful, as it would result in the consumption of undesirable quantities of seawater. However, Fiarrison et at. (1967) showed that the odontocete forestomach could forcibly eject ingested water, and the excretory capabilities of marine mammals are sufficient to counteract this osmotic load. Fay (1960) described the walrus's extensive and highly elastic pharyngeal pouches, which may be used to store water ingested via suction; he suggested buoyancy regulation, storage for food or air reserves, and sound production as alternative functional possibilities. In all cases, ingested water is forced back out through the mouth when the oropharyngeal cavity is compressed by jaw closure and tongue elevation. Generation of suction by pulmonary expansion and subsequent inhalation is untenable in odontocetes, as (like all cetaceans) they maintain a permanent patent
503
airway, effectively separating the oropharynx and trachea, nor is it feasible in pinnipeds, especially when considering the great depths at which many feed, where intense pressures inhibit thoracic expansion. Nonetheless, inspiration of water remains a hazard for walruses and other suction feeding pinnipeds (Gordon, 1984). Exceptionally deep feeders such as Physeter would not need to generate any more or less suction pressure; all that matters is the relative change from ambient pressure. Difficulties of oral/gular expansion against extreme hydrostatic pressure may nonetheless explain the massive nature of the sperm whale hyoid and sternum. Floating or otherwise stationary marine mammals (often at the bottom of the water column) suction feed as well as swimming ones, demonstrating that this is not merely "ram" feeding, in which a predator relies on rapid locomotion to overtake and engulf prey. Even raptorial odontocetes and pinnipeds that seize prey rely on suction to transport it to the rear of the oral cavity for swallowing. In suction feeding whales, dolphins, and porpoises, the hyoid skeleton consists of a large, flattened basihyal and thyrohyal body with robust stylohyal arches anchored to the skull just posterior to the otic complex via tympanohyals (Fig. 16.8). Bony and cartilaginous elements of the hyoid are joined by highly mobile, synovial joint capsules (Reidenberg and Laitman, 1994); the hyoid is also connected to the immobile larynx, which is bound in an intranarial position by a powerful sphincter of pharyngeal muscles originating from the pterygoids and palate. A massive sternohyoideus muscle is largely responsible for hyoid depression, whereas the large, paired mm. hyoglossus and styloglossus project anteriorly and insert into the tongue body. Along with the m. genioglossus (which has extensive origins along the rami of the mandibles), these extrinsic lingual muscles rapidly retract the tongue to expand the oral cavity and create a space into which water and prey are sucked. Comprehensive analysis of attachments, positions, and actions of lingual and hyoid musculature disclosed significant differences between suction feeding versus raptorial dolphins in the size, shape, thickness, and curvature of the hyoid, as well as in the proportional weight and cross-sectional area and relations of the sternohyoideus and hyoglossus (Werth, 1992). The tongue itself is a large, firm, muscular hydrostat (see Chapter 2) with only a tiny free anterior tip, but an extensive system of longitudinal folds or plicae indicating marked mobility, nonetheless. The tongue has a smooth, flat dorsum and does not taper but maintains uniform height and width, making it a perfect hemicylindrical piston. A fringe of marginal papillae ( 2 5 mm) is frequently seen along the anterior and lateral edges of the tongue body (Yamasaki et al, 1976; Kastelein and Dubbeldam, 1990). Found predominantly in
504
Alexander Werth
F I G U R E 16.8. Odontocete hyolingual specializations for suction feeding: (a) fringe of marginal papillae on tongue of adult Atlantic white-sided dolphin, Lagenorhynchus acutus, and (b) robust hyoid apparatus of pygmy sperm whale, Kogia breviceps, showing muscle origin. Note enlarged insertion of sternohyoid d, digastric; gh, geniohyoid; hg, hyoglossus; ih, interhyoid; mh, mylohyoid; oh, occipitohyoid; sg, styloglossus; sh, sternohyoid; th, thyrohyoid.
suction feeding genera, these structures (Fig. 16.8) appear late in development yet persist throughout life, unlike in other mammals where they create a tight seal between the tongue and the palate for improved neonate suckling. Retention of this juvenile feature is strongly suggestive of sucking activity by the tongue in certain adult odontocetes. Sokolov and Volkova (1973) proposed a receptor function for these "fimbriae," but histological analysis does not support their claim. Examination of intrinsic lingual musculature reveals a limited capacity for shape changes and protrusion yet great retractile capabilities (Werth, 1992). A ventral sublingual space of loose areolar tissues provides little resistance and allows for rapid, piston-like withdrawal
of the tongue. The palate is generally smooth, but in beaked whales bears many papillose rugosities (unlike the ridges of many terrestrial mammals), which may function to hold cephalopods or other slippery prey (Heyning and Mead, 1991,1996). Phocoenoides dalli, the virtually toothless Dall's porpoise, also has a ribbed palate. Many large odontocetes, especially sperm and beaked whales, possess several parallel or wishboneshaped external throat grooves (Hubbs, 1946; Clarke et al, 1968). Ross (1987) proposed that these creases are found in large suction feeders where gular expansion from the engulfment of large quantities of water would otherwise be prohibited by the thickness and rigidity of the overlying blubber. Heyning and Mead (1991) confirmed that these folds are distensible and controlled by the contraction of superficial ventral musculature in ziphiids, suggesting that they may serve as active rather than merely passive expanders of the pharynx. They also found that ziphiid intrinsic lingual musculature is quite complex and that certain gular muscles (such as the interhyoid and sternohyoid) are relatively larger in beaked whales than in other putative suctionfeeding odontocetes, correlating with their larger hyoid skeletons. Several lines of circumstantial and anecdotal evidence supported the idea of marine mammal suction feeding before its actual documentation. This accumulated evidence includes anatomical and ecological data as well as direct observations by marine mammal handlers. One of the first and most compelling pieces of information to support marine mammal suction feeding indirectly is the surprising reduction or total loss of teeth, especially in "toothed" whales, where the presence of teeth is a diagnostic character if a poor descriptive one. Some species are truly edentulous, such as the narwhal {Monodon monoceros; Fig. 16.14), whose sole apparent tooth is a long, spiraled tusk that normally erupts only in males and which is clearly not used in feeding (it would, in fact, hinder raptorial feeding). Many teeth remain obscured by gingiva, however, in all odontocetes. Where present, teeth are often partially or completely worn to the root (due to the thin cap of soft, aprismatic enamel) or covered with barnacles or other epizoic organisms (Clarke, 1966; Morris and Mowbray, 1966; Rice, 1989) that prevent occlusion and indicate the nonfunctional nature of the dentition, as these would be crushed if jaws were used in typical mammalian fashion. The strap-toothed whale, Mesoplodon layardi, represents an extreme case of nonfunctional dentition (Fig. 16.9): the pair of flat lower teeth completely encircle the rostrum, restricting gape to a small, round anterior opening analogous to a vacuum cleaner attachment or "slurp gun" used by divers to collect fish. As in other beaked whales, teeth are greatly reduced in
16. F e e d i n g in M a r i n e M a m m a l s
F I G U R E 16.9. Nonfunctional jaws and dentition in odontocetes: (a) head of male strap-toothed whale, Mesoplodon layardi, showing teeth protruding outside and encircling mouth; and (b) curved jaw of sperm whale, Physeter macrocephalus (ventral view).
number and project outside the oral cavity, where they would be useless for prey capture and processing. Oddly, these edentulous species feed on squid, the most slippery of odontocete prey. M. layardi also illustrates the standard ziphiid condition of tooth eruption only in males. Far from vestigial, however, these teeth and tusks undoubtedly perform a function, suggested by their unisexual distribution, as a secondary sexual character used like horns and antlers of terrestrial mammals in dominance displays, fighting, and courtship rituals (Caldwell and Caldwell, 1966; Kleinenberg et al, 1969). Scarring along the flanks and head is common in many odontocetes (McCann, 1974; Heyning, 1984), including sperm whales (Berzin, 1972; Best, 1979; Kato, 1984), whose plentiful scratches and marks are often attributed to squid, although their patterns perfectly match the spacing of sperm whale teeth (Boschma, 1938) and are rarely found on females, whose feeding (and diet) is identical (Kawakami, 1980). Scarring is particularly common in solitary bachelor bulls that engage in the most fighting. Even in species bearing respectable dentition, teeth may not erupt until sexual maturity (long after wean-
505
ing), suggesting that they are not necessary for feeding. The reduced complexity apparent in the gross and ultrastructural morphology of odontocete teeth also reflects their apparent nonutility. Enamel is often weak, poorly developed, and limited to a small area on the tip of the crown, representing another notable degeneration from the standard eutherian dental condition (Ishiyama, 1987; Werth and Stern, 1992). Pinniped teeth are less variable in size, shape, and num^ber, but the suction feeding walrus and elephant seal show marked reduction in the standard dental battery. Flattened or tiny peg-like teeth may be used to crush or grasp prey but certainly not to obtain it; canines or incisors are enlarged (e.g., walrus tusks) only for social purposes, as in odontocetes. Not only are teeth greatly reduced or totally lost, but evidence suggests that the jaws, themselves, are nonfunctional in some odontocetes. A common congenital defect in sperm whales causes the mandible to develop in a loose spiral (Murie, 1865; Nasu, 1958; Spaul, 1964; Nakamura, 1968), and broken jaws incurred by fighting, entanglement in submarine cables (Heezen, 1957), or collision with other undersea obstacles often heal at an improper angle that precludes jaw closure (Fig. 16.9). However, sperm whales with deformed or broken jaws (or without teeth) grow as large, live as long, and, most importantly, feed on the same size and type prey as animals with normal jaws (Clarke et ah, 1988). Equally convincing evidence for marine mammal suction feeding abounds in ecological data, especially information on prey condition showing that shrimp and other delicate prey items are found intact and unharmed, without bite marks or tooth holes (Gunter, 1951; Jones, 1981). Spanish whalers have even reported live squid flopping out of the stomachs of freshly caught sperm whales (Clarke, 1955; Norris and Mohl, 1983). Although stomach content data are understandably hard to obtain for marine mammals, the limited information available suggests that food is often swallowed whole and unmarked. Odontocetes and pinnipeds ingest a large variety and quantity of inorganic debris or exotic, nonfood items ranging from rocks and sticks to fishing gear and plastic objects (Nemoto and Nasu, 1963; Ridgway, 1965; Berzin, 1972). It is not unusual to find sand dollars, brittle stars, and other poorly digested benthic invertebrates in stomach contents of belugas and pilot whales, suggesting that these species often scour and vacuum clean the bottom for food. The underslung, shark-like jaw of Kogia is an apparent adaptation for benthic feeding. Many suction feeders demonstrate impressive diving capacity (depth and duration) for protracted feeding at the thermocline or sea floor (Gaskin, 1982). Norris and Mohl (1983) presented much of the
506
Alexander Werth
circumstantial evidence for suction feeding in their review of the prey-stunning hypothesis first proposed by Bel'kovich and Yablokov (1963). The hypothesis holds that certain odontocetes are able to debilitate prey acoustically with sonic pulses generated by nasal valves and sacs and focused by the melon, perhaps developed as an extension of echolocation for prey detection (Hult, 1982). Despite intense interest and scrutiny, the acoustic stunning theory has never been substantiated (Zagaeski, 1987; MacKay and Pegg, 1988) [some cetaceans, notably killer whales, stun prey via fluke slaps (Simila and Ugarte, 1993)]. Nonetheless, Norris and Mohl outlined two major trends in odontocete evolution: the reduction or loss of teeth and the progressive widening and shortening of the skull and jaws to create a more blunt head. While these obvious and indisputable trends demonstrate the nonutility of long jaws and dentition for feeding in many odontocetes, they do not appear to relate to the sound generation needed for prey stunning. It is clear that head shape is, however, a critical factor in suction generation. Manipulative experiments on severed heads of stranded odontocetes displaying a diversity of head forms demonstrated quantitatively the effect of head shape on water flow patterns (Werth, 1989,1992). Cannulae were threaded through successive holes drilled into rostra of heads ranging from a blunt-headed harbor porpoise {Phocoena phocoena) to a long-snouted common dolphin {Delphinus delphis). Transducers introduced into these cannulae measured pressures in the anterior, middle, and posterior thirds of the oral cavity as infra- and suprahyoid muscles of submerged heads were manually retracted and depressed with string handles; site, speed, and force of pull were varied. Dowels inserted between the rostrum and the mandibular symphysis controlled gape; the esophagus was sutured closed. Pressure drops, calibrated and measured in mm Hg, were converted into digital waveforms for analysis; the time course for each pulse of negative pressure was also measured to ensure that all pulls were consistent and thus comparable. Pressures of roughly - 4 5 mm Hg were measured, suitable for capture of small prey; certainly greater pressures could be attained by live animals. More importantly, these experiments showed that shortening and widening of the rostrum greatly improve suction generation. Species with a smaller, rounder mouth opening are better able to draw in prey anteriorly, as this reduces the "notched" gape of longsnouted forms that tends to draw water in laterally. However, experiments documenting suction feeding in captive juvenile long-finned pilot whales, Globicephala melaena, showed that the side of the mouth may also be used to ingest prey via suction, at least in this remarkably adaptable species (Werth, 1992, 2000). Three pilot whales held in a large pool during rehabilitation
following stranding (in preparation for their eventual return to the wild) were filmed feeding on frozen herring, mackerel, and squid at the surface, midwater, and especially off the bottom, from which they would rapidly "pull" their admittedly nonelusive food. This footage provided confirmation of suction feeding behaviors, such as expulsion of water from the mouth (from a whale that ascended nearly vertically and "missed" a prey item at the surface) and indicated the extent to which this species rolls or rotates while ingesting prey, presumably to engulf food in an orientation so that it may be swallowed without manipulation in the mouth. It is important that fish be swallowed head first and squid mantle first to prevent spines, fins, or other protruding structures from catching in the throat. [In this light, however, it is interesting that captive juvenile crabeater seals (Ross ei a\., 1976) had difficulty orienting food in the mouth but gradually learned to suck in or manipulate prey so as to facilitate deglutition.] While feeding on the bottom, however, pilot whales showed that prey could be as easily drawn in from the side of the mouth by using the buccinatorius and orbicularis oris muscles to close off the other side and restrict the opening through which water and prey are sucked. Like belugas (Brodie, 1989), pilot whales may purse their lips to reduce the oral opening or make it rounder to improve suction ingestion of prey, although this was not seen, nor was feeding observed with live prey. Osteological data on rostral and mandibular dimensions (Werth, 1992) reveal an impressive range of odontocete head profiles (Fig. 16.14) rivaling that of other aquatic tetrapods (e.g., gracile and robust crocodilians ranging from gavials to caimans), with extremely slender-snouted platanistids and oceanic delphinids (Stenella) as well as short, bulldog-like heads in dwarf and pygmy sperm whales (Kogia) and many dolphins and porpoises. Not unexpectedly, blunt heads correlate with other anatomical, ecological, and behavioral traits associated with suction feeding. A surprise, however, is that the species for which the most compelling circumstantial evidence for suction feeding has been adduced, the sperm whale, has long, narrow jaws. However, unlike pilot whales, sperm whales have no proper oral cavity according to the standard definition, and with its pendulous jaw hanging fully open (as is often seen) the oropharyngeal isthmus of the sperm whale presents a perfect circle (Fig. 16.10), which in the absence of an oral orifice is the opening through which prey are sucked directly into the oropharynx, thus eliminating the need for intraoral prey transport and combining ingestion and swallowing into a single feeding stage! The glistening white mouth of the sperm whale, standing in stark contrast to the remainder of its jet black or charcoalcolored body, has prompted researchers to propose a passive luring stratagem for attracting squid (Beale,
16. Feeding in Marine Mammals
F I G U R E 16.10. Lateral profile of sperm whale '"oral cavity" showing round oropharyngeal opening (dashed), throat grooves, and hyoid apparatus in extended position.
1839). Gaskin (1967) confirmed that a luminous substance rubs off the bodies of bioluminescent squid easily and documented instances of glowing sperm whale mouths. Heyning and Mead (1996) suggested that white-beaked whatle mouths may also serve as squid attractants. Clarke's (1979) observation that Physeter often sounds and surfaces in the same spot casts doubt on the belief that it actively chases prey, and Fristrup and Harbison (1993) thought that vision plays a critical role in passive foraging. Cooperative foraging may also allow sperm whales to catch elusive prey (Whitehead, 1989). The huge, barrel-shaped head of Physeter has been explained as a ballast tank (Clarke, 1970,1976) or as a sonic lens for echolocation (also stunning; Norris and Harvey, 1972; see earlier discussion). The potential oropharyngeal nature of odontocete suction feeding is unique relative to other vertebrates, but it also hints at a scenario for the evolution of suction feeding in marine mammals. Although suction feeders possess clear conformities in morphology and ecology, it is important to remember that this feeding method was independently adopted by several distinct lineages; clearly the plesiomorphic condition is raptorial feeding (seizing and snapping). The use of suction to transport grasped prey in gars and other long-snouted fishes suggests an identical function in marine mammals, especially odontocetes, which have no alternatives, as swallowing occurs entirely underwater (Pilleri et ah, 1970). Odontocetes do not shake or rotate the head, use inertial or gravity transport, or any cranial kinesis to transport prey with a racheting action. Lingual transport is also precluded by the extremely long mandibular symphysis of many long-snouted odontocetes, in which the tongue does not reach the anterior extent of the oral cavity. Gavials (long-snouted piscivorous croc-
507
odylids) are known to take prey above the surface, where they transport it inertially by lifting the head and throwing it back into the pharynx, but such behaviors have never been described in cetaceans despite years of intensive observation of morphologically similar river dolphins in captivity. Suction ingestion in marine mammals appears to have evolved from the initial use of suction for intraoral transport of prey grasped in the jaws (Fig. 16.11). The grasp and separate transport steps were later lost (along with elongate jaws and elaborate dentition), simplifying and expediting the feeding process and freeing teeth for adaptation to social functions. Evolution of suction feeding from neonate suckling is unlikely given the anomalous nature of odontocete lactation, in which milk is actively pumped into the calf's mouth by the contraction of smooth muscles surrounding the mammary glands (Slijper, 1962; Arvy, 1974). Werth (2000) described the kinematic sequence of suction feeding events in Globicephala (Fig. 16.12), including a preparatory phase with partial gape followed by jaw opening and rapid hyoid depression to suck prey in at a mean distance of 14 cm, although some prey were taken from much greater distances. The actual suction or gular expansion phase averages 90 msec in duration; the entire ingestion cycle, not including the initial approach, elapses in one-third of a second. Although prey items were no longer visible at the conclusion of most ingestion events, it was still seen at the rear of the buccal cavity in a few feeding sequences. Unfortunately, jaw closure prevents clear observation of prey transport. However, slight movements of the mouth floor and throat are evident, suggesting that lingual, hyoid, or mandibular motions (or perhaps some combination of the three) are responsible for prey transport to the faucial pillars prior to deglutition. Heyning and Mead (1996) reported that fish placed in the mouths of captive Cuvier's beaked whales {Ziphius cavirostris) disappeared in less than 0.08 sec. They also noted that when live-stranded Hubb's beaked whale {Mesoplodon carlhubbsi) calves suck on fingers, hyoid motion and suprahyoid muscle contraction can be palpated externally. Despite the lack of functional analysis for other suction feeding cetaceans, several observers offer unequivocal reports of this ability in captive odontocetes. While perhaps less instructive than experimental studies of structure and function, these accounts provide incontrovertible evidence of suction feeding in a wide range of marine mammals and offer clues as to their feeding behavior. Ray (1966) described belugas sucking coins from the bottom of their pool and spitting them out. Brown (1962) and Donaldson (1977) recount similar sucking and expulsion behaviors in short-firmed pilot whales {Globicephala macrorhynchus) and killer whales (Orcinus orca), respectively. Numerous unpublished anecdotes of captive marine mammals (especially
508
Alexander Werth
PREY CAPTURE IN INIA VS. GLOBICEPHALA
1. Ingestion
1. Ingestion (SUCTION)
2. Transport (SUCTION)
3. Deglutition
2. Deglutition
FIGURE 16.11. Evolution of odontocete suction ingestion (as in pilot whale, Globicephala spp.) from transport of grasped prey in long-snouted odontocetes (e.g., the boutu or Amazon river dolphin, Inia geoffrensis) by loss of transport step and attendant elongated jaws and elaborate dentition.
belugas and beaked whales) describe repeated sucking and blowing out of small fish, leaves, or other debris. Among pinnipeds there is compelling anatomical and ecological evidence for suction feeding in several large, predominantly teuthophagous phocids such as the Ross seal, Ommatophoca rossi, bearded seal, Erignathus harhatus, and elephant seals, Mirounga sp.; its use is also suspected in an otariid, the cape fur seal, Arctocephalus pusillus (King, 1983). The Ross seal possesses large, well-developed lingual, suprahyoid, and pharyngeal musculature to aid in holding and swallowing elusive, slippery cephalopods (King, 1964; Bryden and Felts, 1974). Jaw adductors are strong despite the weak dentition (Fig. 16.15), and the tongue and epiglottis are situated far posteriorly. Muscles involved in tongue elevation and deglutition (especially hyoglossus and styloglossus and pharyngeal constrictors) appear to be adapted to gripping and swallowing large squid (King, 1964), although they may also relate to the loud bellowing as the throat expands during sound production (Bryden and Felts, 1974). King (1964) speculated that the exceptionally long soft palate (over 60% of total palate
length) enlarges the oral capacity longitudinally and vertically, whereas the ventral position of tracheal cartilages aids in esophageal expansion; she suggested that similar palatal and tracheal modifications permit the leopard seal to swallow bulky prey. However, the champion pinniped suction feeder is unquestionably the walrus (Odobenus rosmarus), which produces powerful suction currents to dislodge softbodied prey (e.g., annelid, sipunculid, or priapulid worms) along with clams, cockles, mussels, and other benthic molluscs from the gravelly or muddy substrate of the sea floor. Walruses use suction not only for ingestion, but for processing. They are able to remove bivalve siphons and feet from their shells while rarely swallowing shell fragments, a feat (like that of bearded seals sucking live whelks from their shells; Bonner, 1990) requiring remarkable suction strength and control. The shells are presumed to be held in the lips and rejected once the meat is removed by suction. Natives often rinse and eat undigested clams taken from stomachs of hunted walruses (Kenyon, 1986b). Not only can walruses obtain six or more clams per minute (for a meal
509
16. F e e d i n g in M a r i n e M a m m a l s
TIME (msec)
£ o UJ
o z