Islands and Snakes: Isolation and Adaptive Evolution 0190676426, 9780190676414, 9780190676421

Citation preview

i

ISLANDS AND SNAKES

ii

iii

ISLANDS AND SNAKES Isolation and Adaptive Evolution Edited by Harvey B. Lillywhite Marcio Martins

1

iv

1 Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2019 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. CIP data is on file at the Library of Congress ISBN 978–​0–​19–​067641–​4 9 8 7 6 5 4 3 2 1 Printed by Sheridan Books, Inc., United States of America

v

I dedicate this book to my lovely wife Jamie, and I express my heartfelt appreciation for her love, encouragement and support for this work and throughout my career. —​Harvey B. Lillywhite   I dedicate this book to my lovely wife Eliana for having been a great source of inspiration and support throughout this work —​Marcio Martins

vi

vi

CONTENTS

Foreword

ix

Preface

xi

Acknowledgments List of Contributors 1. Ecology of Snakes on Islands—​M arcio Martins and Harvey B. Lillywhite

xiii xv

1

2. Isolation, Dispersal, and Changing Sea Levels: How Sea Kraits Spread to Far-​Flung Islands—​H arold Heatwole

45

3. Terrestrial Habitats Influence the Spatial Distribution and Abundance of Amphibious Sea Kraits: Implication for Conservation—​X avier Bonnet and Fran çois Brischoux

72

4. Physiological Ecology of Sea Kraits Inhabiting Orchid Island, Taiwan—​M ing-​C hung Tu and Harvey B. Lillywhite

96

5. The Queimada Grande Island and Its Biological Treasure: The Golden Lancehead—​M arcio Martins , Ricardo J. Sawaya , Selma Almeida-​S antos , and Otavio A. V. Marques

117

6. Pleasure and Pain: Insular Tiger Snakes and Seabirds in Australia—​ Fabien Aubret 138 7. The Eyes Have It: Watching Treeboas on the Grenada Bank—​ Robert W. Henderson

156

8. The Ecology and Conservation of the Milos Viper, Macrovipera schweizeri—​G öran Nilson

181

vi

v i i i   • 

Contents

  9. The Unique Insular Population of Cottonmouth Snakes at Seahorse Key—​H arvey B. Lillywhite and Coleman M. Sheehy , III

201

10. Living Without a Rattle: The Biology and Conservation of the Rattlesnake, Crotalus catalinensis, from Santa Catalina Island, Mexico—​G ustavo Arnaud and Marcio Martins

241

11. Decline and Recovery of the Lake Erie Watersnake: A Story of Success in Conservation—​R ichard B. King and Kristin M. Stanford

258

12. Defending Resources on Isolated Islands: Snakes Compete for Hatchling Sea Turtles—​A kira Mori , Hidetoshi Ota , and Koichi Hirate

288

13. Islands in the Sky: Snakes on South American Tepuis—​D . Bruce Means and C ésar Barrio-​A mor ós

310

Index

331

ix

FOREWORD

“Islands” and “snakes” are two words that evoke a powerful sense of discovery and adventure. Islands are realms of endemism and novelty, and their exploration has informed some of the most fundamental ideas in the history of biology. Snakes reflect the mystery and beauty in nature, and our hardwired fascination with them reminds us of our intimate connection with it. Editors Harvey Lillywhite and Marcio Martins reveal this world by drawing on their diverse collaborations and collective decades of scholarship and passion. Nineteen accomplished snake biologists, alongside Lillywhite and Martins, have contributed chapters that together cover the ecology, behavior, evolution, and conservation of snakes on islands. Each chapter is illustrated with color photographs of spectacular snakes and their island habitats. These range from bird-​eating tree boas in the Caribbean to amphibious sea kraits spanning Taiwan and New Caledonia and castaway Australian tiger snakes. The authors go far beyond the existing scientific literature by allowing the reader privileged insight into the passion and process behind their discoveries. The entertaining anecdotes shared in each chapter show that new questions and new knowledge are gained from lifelong curiosity and dedication, innovative thinking, and also some degree of risk-​taking. Furthermore, by reflecting on their closest professional collaborations and enriching interactions with local communities, the authors reveal the human side of scholarship. Books such as this are vital for stimulating public enthusiasm for science and conservation. Islands and Snakes:  Isolation and Adaptive Evolution is timely. Island environments are under threat from development, rising sea levels, and an increasing incidence of invasive species. And biologists are racing against time to discover what species live on islands, the crucial roles they play there, and how they have adapted to island life. By relating ecological and evolutionary insights gained from field studies of snakes on islands, this book will no doubt inspire numerous new research and conservation initiatives. Islands and Snakes:  Isolation and Adaptive Evolution is a must-​have for students,

x

x   • 

Foreword

biologists, geographers, and anyone who values fragile island environments and their unique biodiversity. Kate Sanders Adelaide, Australia August 2018

xi

P R E FA C E

During the fall of 2013, we taught a seminar course together on island ecology at the University of Florida, and the genesis of ideas for this book was in large measure an outgrowth of this class. At the beginning, we had intended to write a review paper that covered both historical and contemporary aspects of the biology of snakes on islands. Then, we scrapped this idea and replaced it with a goal of producing an academic book that focused on aspects of island ecology and biogeography as viewed through the lens of the many studies in which snakes have been a biological focus of such investigation. Islands have been appreciated as natural “laboratories” for investigations of ecology, biogeography, and evolutionary biology since the time of Wallace, who dedicated a large amount of his writings to islands, including his seminal work Island Life, and Darwin, who was profoundly affected by his observations in the Galapagos Islands. In the 1960s, MacArthur and Wilson produced an important and influential theoretic framework for subsequent investigations of biodiversity and dynamics of insular biogeography. Since MacArthur and Wilson’s pioneering efforts, a robust literature on insular ecology and biogeography has continued to grow, and understanding the successful existence and adaptations to conditions on islands has advanced. Various investigators have extended earlier theoretic studies to increase understanding of important phenomena such as adaptive radiation, energetics, paleogeography, plasticity of colonizing biota, trophic changes, morphological evolution, and climate change. For reasons of practicality and personal interests, various specific elements of biota have been investigated as model organisms for clarifying insights regarding particular features of island ecology. Reptiles on various islands have replaced endotherms as primary herbivores and top carnivores. Because of ectothermy, reptiles have advantages over endotherms in exploiting scarce resources in circumstances that are challenging to the success of birds and mammals. Snakes are known to be very successful colonizers of islands, and

xi

x i i   • 

Preface

roughly 60% of literature on insular squamate reptiles deals with snakes. Indeed, studies of snakes have contributed much to our understanding of insular ecology, and these vertebrates are important subjects for investigating questions that might be difficult to approach in other systems. Details concerning the reasons why snakes have been successful in living on islands may be found in Chapter 1. We have been fortunate to observe snake populations on many islands throughout the world, including key locations in South America, the Gulf of California, Taiwan, Australia, and the tropical Pacific. In many instances, snakes on islands occur in amazing numbers and are often a dominant aspect of the local fauna. To produce this book, we have recruited authors from among authorities throughout the world who have focused influential studies of snakes that occupy interesting and important systems on various islands or archipelagos. As a concluding chapter, we have included studies of the spectacular tepuis that comprise an exceptional example of “ecological islands” in South America. All of the various authors provide entertaining narratives of the system they studied, woven as a fabric with solid empirical information, scientific theory, and personal insights regarding ecological and evolutionary principles as revealed by spectacular snakes and their adaptations to living on islands. Harvey B. Lillywhite Marcio Martins

xi

ACKNOWLEDGMENTS

We are grateful to the many persons who have made this book possible, especially our families, the various authors who have contributed thoughtful and stimulating chapters, numerous colleagues, reviewers, and others who have encouraged our adventures in science, including numerous visits to exciting islands. We also thank the editors and production staff at Oxford University Press for their professional guidance and assistance throughout this project. We hope that readers will find the enjoyment and satisfaction of reading this book that we have intended for them to discover.

xvi

xv

CONTRIBUTORS

Selma Almeida-​Santos Laboratório de Ecologia e Evolução Instituto Butantan São Paulo, Brazil Gustavo Arnaud CIBNOR La Paz, Mexico Fabien Aubret Station d’Ecologie Théorique et Expérimentale de Moulis CNRS Moulis, France School of Molecular and Life Sciences Curtin University, Australia  César Barrio-​Amorós Doc Frogs Expeditions San Isidro del General, Costa Rica Xavier Bonnet CEBC-​CNRS Villiers en Bois, France François Brischoux CEBC-​CNRS Villiers en Bois, France

Harold Heatwole University of New England Armidale, New South Wales, Australia North Carolina State University Raleigh, North Carolina, USA Robert W. Henderson Section of Vertebrate Zoology Milwaukee Public Museum Milwaukee, Wisconsin, USA Koichi Hirate Okinawa Prefectural Sea Farming Center Okinawa, Japan Richard B. King Department of Biological Sciences and Institute for the Study of the Environment, Sustainability, and Energy Northern Illinois University DeKalb, Illinois, USA Harvey B. Lillywhite Department of Biology University of Florida Gainesville, Florida, USA

xvi

x v i   • 

Contributors

Otavio A. V. Marques Laboratório de Ecologia e Evolução Instituto Butantan São Paulo, Brazil Marcio Martins University of São Paulo São Paulo, Brazil D. Bruce Means Coastal Plains Institute and Land Conservancy and Florida State University Tallahassee, Florida, USA Akira Mori Department of Zoology Graduate School of Science Kyoto University Kyoto, Japan Göran Nilson Göteborg Natural History Museum Göteborg, Sweden Hidetoshi Ota Institute of Natural and Environmental Sciences University of Hyogo Museum of Nature and Human Activities Hyogo, Japan

Ricardo J. Sawaya Centro de Ciências Naturais e Humanas Universidade Federal do ABC São Bernardo do Campo, Brazil Coleman M. Sheehy, III Division of Herpetology Florida Museum of Natural History University of Florida Gainesville, Florida, USA Kristin M. Stanford F. T. Stone Laboratory The Ohio State University Put-​in-​Bay, Ohio,  USA Ming-​Chung  Tu Department of Life Science National Taiwan Normal University Taipei, Taiwan

1

1

ECOLOGY OF SNAKES ON ISLANDS

Marcio Martins and Harvey B. Lillywhite

Introduction Islands have been the subject of intense investigation—​biologically and ecologically—​since the time of Darwin and Wallace. Much research has focused on species assemblages and the dynamics of species richness on islands as well as other systems having geographic isolation and other characteristics similar to those of islands surrounded by water. An important theoretic model for insular biogeography was produced by MacArthur and Wilson (1963, 1967), and their work created a useful framework for subsequent investigations of biodiversity and its dynamics on islands (see reviews in Whittaker and Fernández-​Palacios 2007; Warren et  al. 2015; Santos et  al. 2016; Patiño et  al. 2017). Previously, insular faunas were regarded generally as either static or changing slowly and unpredictably due to environmental and climatic changes (Dexter 1978; Heaney 2000). However, the influence of more modern biogeographic theory enabled sometimes robust predictions of species richness in relation to the area of an island and its distance from a source of colonizing biota (MacArthur and Wilson 1967). Whether or not the biota of a given island or insular system is at “equilibrium” often remains debatable (Lomolino 2000; Warren et al. 2015). Since the 1960s, a robust literature on insular biology and ecology has continued to grow, with numerous investigators focusing on a variety of insular systems with attention to increasing detail concerning the requirements and dynamics of ecological factors that favor adaptation and successful existence on islands. Numerous questions are being addressed that cannot be explained by the existing theoretic models of insular biogeography (Gillespie

2

2   • 

I s l a n d s a n d  S n a k e s

and Roderick 2002). Examples include (but are not limited to) questions regarding adaptive radiation (Losos 1998; Gavrilets and Losos 2009), paleogeography (Iturralde-​Vinent and MacPhee 1999), energetics and energy resources (McNab 1994a, 1994b; Polis and Hurd 1996; Bonnet et al. 2002), climatic change (Bellard et al. 2013a, 2013b; Wetzel et al. 2013), life history characteristics of insular biota (Foufopoulos and Ives 1999), and anthropogenic influence (Steadman et al. 2005). For many practical reasons (often with advantages), various investigators of the ecology of islands focus on taxonomic elements of fauna or flora with which they have familiarity or interest or that represent “model” systems due to a prior database of information or suites of characteristics appropriate to particular questions. Such approaches have intrinsic value and also provide important data for more inclusive investigations of biodiversity (Myers et al. 2000). Reptiles have been the frequent subject of investigations of vertebrate faunas on islands, where they are often “replacements” of endotherms as primary herbivores and top carnivores. The favorable circumstances for ectotherms versus endotherms on islands include low rates of energy expenditure relative to resources that are often scarce or limited on islands, dispersal abilities, and superior colonizing abilities. Rates of energy expenditure in terms of field metabolic rates of endothermic mammals and birds are approximately 12 and 20 times higher, respectively, than those of equivalent size, ectothermic reptiles (Nagy 2005). Reptiles, therefore, have a crucial advantage in exploiting scarce resources and building populations in circumstances that preclude, or severely challenge, the success of birds and mammals. Snakes are very successful inhabitants of islands, and there is a very rich literature concerning the insular ecology of this group of vertebrates (Figure 1.1). An accounting of scientific articles using Google Scholar indicates that roughly 60% of the literature on islands and squamate reptiles (including Tuatara) deals with snakes. Thus, studies of snakes have contributed much to our understanding of insular ecology and are important subjects for attention to questions that might be difficult to investigate in other systems (discussed later). We emphasize that there are numerous reasons why snakes are important elements of insular biotas and play critical roles on numerous islands that can offer further insights for understanding ecology of islands. Some of the more important attributes of snakes related to successful “island living” are (1)  ectothermy and comparatively low energy requirements; (2)  attributes favoring abilities for overwater dispersal; (3)  life history features favoring comparatively rapid population growth; (4) range of body sizes favorable for inhabiting even very small islands; (5) breadth and plasticity of diet, including

3

Figure 1.1  Examples of snakes having high significance with respect to insular ecology and biogeography. (a)  Blue-​banded Sea Krait (Laticauda laticaudata) from Orchid Island, Taiwan. Sea kraits are amphibious and spend variable amounts of time secluded on numerous small islands of the Indo-​Pacific oceans (see Chapter  2). (b)  Southeast Asian Bockadam (Cerberus schneideri) is an amphibious snake associated with estuarine habitats in Southeast Asia, including many areas of the Philippines. (c) Brown Tree Snake (Boiga irregularis) is widespread in the Oriental and Oceanian regions, occurring naturally on more than 50 islands and accidentally introduced to Guam and other islands. See implications for conservation in Rodda and Savidge (2007). (d) The Oriental Blind Snake (Indotyphlops braminus) is native to almost 60 different islands and was accidentally introduced in dozens of other islands throughout the world (Wallach 2009). This specimen was found on Boca Chica Key, Florida. (e) The Solomon Island Ground Boa (Candoia paulsoni) is perhaps the most widespread snake on islands, occurring in more than 60 islands in the Oriental region. (f ) Feick’s Dwarf Boa (Tropidophis feicki) is representative of dwarf boas that occur throughout numerous islands of the Caribbean. This species occurs in Cuba, where the genus and family (Trophidophiidae) reach their highest diversity. There are 32 species of the genus Tropidophis, found in areas of South America and the West Indies, where this genus has more successfully speciated. Sources: Photographs by Ming-​Chung Tu (a), Mark O’Shea (b and e), Coleman Sheehy III (c), Jonathan Mays (d), and Javier Torres Lopez (f ), all reproduced with permission.

4

4   • 

I s l a n d s a n d  S n a k e s

scavenging; (6) effective means of prey acquisition; (7) infrequent feeding on relatively large prey and “slow” digestive physiology; (8) secretive behaviors and cryptic morphologies; (9) special scansorial capabilities in many species; and (10) thermal plasticity. In this chapter, we review much of the key literature relating to islands inhabited by snakes and the insular ecology of snakes. Using a database on the occurrence of snakes on islands, we also provide broad patterns of snake diversity. Our goals are to illustrate how studies of snakes on islands can inform general principles related to biology of islands or fragmented habitat, evaluate what are considered to be novel features related to the ecology of snakes on islands that can serve to enhance understanding of complex situations, and stimulate future research as a result of forward-​looking perspectives that emerge from the robust literature on snakes and islands.

Geography and Features of Islands Inhabited by Snakes The main island types regarding their origin are oceanic islands (originated over oceanic plates), continental fragments (portions of continental rock originated by plate tectonic processes), and continental—​or land-​bridge—​ islands (located on continental shelves; Whittaker and Fernández-​Palacios 2007). For practical reasons, here we combine continental fragments (e.g., Madagascar and New Caledonia) and oceanic islands (e.g., Hawaii and Canaries) into a single category, herein called oceanic islands (Pyron and Burbrink 2014; Figure 1.2). This review is concerned exclusively with islands in the sea—​that is, we do not deal with lake, river, and estuary islands, including those separated from mainland by a narrow (22,000 observations) offered us an opportunity to assess the relationships between terrestrial habitats, specific habitat requirements, and the spatial distribution of these snakes. Sea kraits are very abundant in New Caledonia, where the lagoon stretches approximately 600 km along a north–​south gradient. This large marine area of 24,000 km² is divided by an approximately 400-​km mountain chain, whereas the distance from land to the barrier reef ranges from a few hundred meters to >60 km (Bonnet et al. 2015). Numerous colonies of sea kraits are distributed across this very wide range of geographical and climatic conditions. Most colonies are found on small coralline islets (Figure 3.3, inset), especially in the widest parts of the lagoon, but sea kraits also colonize shorelines of the mainland. Each site exhibits a singular physiognomy characterized by the total surface available to the snakes (sea kraits do not penetrate far inland), surrounding seafloor, exposure to dominant winds and currents, elevation, geological substrate, climate, vegetation, animal communities, and so on (Figure 3.3). The distribution of microhabitats (e.g., beach rocks, sandy beaches, trees, and artificial constructions) is consequently heterogeneous among sites and also within each terrestrial site (see Figure 3.3). Due to the relatively small size and isolation of most islets, and due to the peculiar requirements of snakes, each rock, tree, log, or burrow may represent a potential refuge and thus creates critical fine-​scale heterogeneity (Brischoux et  al. 2009b; Lane and Shine 2011). We begin with a short review of the activities of the sea kraits when they are on land. For each of these activities, we derive the associated essential eco-​physiological requirements or constraints to examine the specific microhabitats selected by snakes and for what duration. Then, we examine the distribution of sea kraits over more than 30 sites in relation to available microhabitats to highlight the effects of variation in microhabitat and to recommend conservation strategies.

75

Figure  3.3  Habitats of sea kraits in New Caledonia. (a)  Typical sandy beach at low tide on the western tip of Amérée islet (reserve Merlet, fully protected area, 22°26′40″S–​ 167°05′46″E). Dense vegetation provides easy access to shelters for sea kraits. (b) Western shore of Signal islet (protected area, 22°17′48″S–​166°17′34″E), also at low tide. Blocks of limestone—​deposited on the shore and used for lime production ages ago—​offer snakes excellent shelters. (Inset) Amérée Island is densely populated by sea kraits, especially Laticauda saintgironsi, which easily cross the narrow beaches. The habitat here is less favorable for Laticauda laticaudata due to the lack of abundant large rocks at the beach. Source: Photographs by Xavier Bonnet.

76

7 6   • 

I s l a n d s a n d  S n a k e s

Why Do Sea Kraits Come on Land Periodically? It is well known that sea kraits forage on the seafloor of coral reefs and then return to land for digestion or to lay their eggs (see Chapters 2 and 4, this volume). However, these two activities are not the only reasons why sea kraits utilize terrestrial habitats, and thus, they may not accurately account for the time and energy budgets of these amphibious snakes. In addition, the importance of the periods when snakes are secluded and resting is insufficiently recognized. Many crucial physiological or behavioral tasks do not require movement or displacements. For example, to optimize the digestion and assimilation of large prey, individuals should minimize their expenditure of energy during digestive periods and thus should remain motionless. When on land, breathing does not require locomotor movement and merely follows regular inspiration/​expiration cycles (Bartlett et al. 1986) that are not interrupted by relatively long periods of apnea associated with diving and swimming (Cook et  al. 2016; see Chapter  2, this volume). Thus, although not measured, it is likely that digestion is optimized on land but not at sea, particularly for tropical reptiles. Indeed, temperature rarely constrains digestive physiology in tropical reptiles (Shine and Madsen 1996). This reasoning may also apply to vitellogenesis, ecdysis, and resting. Significant displacements and elevated expenditures of energy are required during foraging, mating, and dispersal. Consequently, spending long periods while relatively motionless in appropriate terrestrial refuges is a well-​suited strategy that optimizes energy budgets in air-​breathing marine species. Here, we briefly discuss the main terrestrial activities of the sea kraits and attempt to provide a crude estimate of the relative importance of each of these activities. Digestion Digestion is often associated with a strong behavioral shift to inactivity for several reasons (Brischoux et al. 2011). First, availability of assimilated energy is likely to be maximized if other competing activities are reduced. Second, rapid digestion often requires the selection of a precise thermal range that optimizes the digestive activity, and this involves selection of a specific microhabitat during the digestive process (Greenwald and Kanter 1979). Third, a digesting snake usually has a very large prey item (see Figure 3.2) in the stomach for several days following ingestion, and this impedes locomotion (Shine and Shetty 2001). The last two points are particularly true for sea kraits that need to return to land to digest and thus to select optimal thermal

7

Sea Kraits: Implication for Conservation  • 

77

conditions for digestion that may not be available at sea and also because the presence of a large prey in the stomach impedes efficient swimming (Webb 2004, Winne and Hopkins 2006). For these reasons, digestion is probably one of the more important and frequent requirements that sea kraits perform on land. More than one-​third of the snakes we captured on land were digesting (31.6%, n = 19,404 observations), and this fraction reached almost 50% for some categories (e.g., Laticauda saintgironsi females; Table 3.1). Digestion is one of the terrestrial activities for which we found the largest numbers of snakes, with the proportions for other activities ranging between 2% and 47% (see Table 3.1). During the austral summer, most snakes were digesting while secluded beneath a shelter, probably because air and surface temperatures were high during the day (Brischoux et al. 2009a; Bonnet et al. 2009). On the other hand, many digesting individuals were found basking during the austral winter. Interestingly, the two species differ in this respect. During the winter, Laticauda laticaudata was still digesting under shelters, whereas L. saintgironsi was basking conspicuously during the day. This interspecific difference is perhaps related to differences in thermal ecology, sensitivity to transcutaneous evaporative loss of water (e.g., see Lillywhite et al. 2009), terrestrial tendencies (Bonnet et al. 2005), or a combination of these factors. Ecdysis Snakes periodically slough the epidermis of their skin (Figure 3.4), which permits growth, repairs injuries, and renews the permeability barrier, in addition to other possible functions. Each shedding cycle can be divided into several phases (Maderson 1965). During non-​shedding periods (sometimes weeks or months), the epidermis is relatively quiescent. During the pre-​ shedding phase (2 or 3 weeks on average), cells proliferate from the stratum germinativum and form a new inner epidermal generation beneath the older outer epidermal generation of stratum corneum. During this phase, the epidermis is metabolically very active. Several days before the completion of this phase, a separating layer develops between the inner and outer epidermal generations, dulling the color of the skin and eyes (see Figure 3.4). This opacity degrades vision, impacts defensive behaviors, and can be associated with anorexia (Aubret and Bonnet 2005). Finally, the snake sheds the old outer generation of stratum corneum (“skin”) by employing specific movements of the body (ecdysis). Starting from the head, the old “skin” is generally removed

78

Table 3.1  Summary of the Numbers of Observations of Sea Kraits Recorded During Their Main Activitiesa N (%)

Activity

Species

Sex

Digestion

Ls

F

1,314 (49.2)

M

2,303 (34.4)

Ll

F M

Sloughing skin

Ls Ll

Recovering

Ls Ll

Resting

Ls

329 (11.7)

M

651 (9.2)

F

262 (13.6)

M

750 (11.8)

F

114 (4.1)

M

160 (2.3)

F

92 (4.8)

M

269 (4.3)

F

509 (41.2)

F M

Reproduction

Ls Ll

1,209 (43.1) 479 (37.0) 1,282 (34.3)

F

224 (18.2)

M

62 (13.1)

F

202 (19.5)

M

Individuals found with prey item and/​or highly digested material in the stomach

2,374 (29.6)

F

M Ll

811 (39.5)

Additional Information

53 (3.2)

Individuals presenting at least one of the following characteristics: opaque eyes, milky skin, and sloughing Individuals presenting recent, sometimes severe, wound(s)

Individuals observed motionless in a refuge, without any material in the stomach, not wounded, and not (pre)shedding Males observed courting (e.g., male jerking on a female), or mating, or gravid females (e.g., follicles or developed eggs in the abdomen)

Sample sizes (N) indicate the number of individuals observed performing a given activity for each category of species and sex. Proportions (%) were calculated based on the total number of observations for which all the relevant information was available (e.g., 2,671 female Ls and 6,703 male Ls were palpated to assess their digestive status). Variations in sample size occur due to incomplete information (e.g., a snake observed with opaque eyes might not have been palpated). Several individuals were observed performing more than one activity (e.g., recovering from a wound and digesting), which explains why the sum of percentages does not equal 1. a

F, female; Ll, Laticauda laticaudata; Ls, Laticauda saintgironsi; M, male.

79

Sea Kraits: Implication for Conservation  • 

79

Figure 3.4  A male Yellow Sea Krait (Laticauda saintgironsi) in the pre-​shedding phase with shaded color pattern, sheltering in a log. A shed skin from another Yellow Sea Krait is visible on the left. The inset photographs feature a pre-​shedding sea krait with opaque eyes (left) and another in the process of shedding (right). Source: Photographs by Xavier Bonnet.

whole (see Figure 3.4). This process is facilitated if the snake can anchor the old “skin” onto a rough substrate. Terrestrial refuges offer the potential of precise thermoregulation needed to meet the increasing metabolic demands of ecdysis (Gibson et  al. 1989). Shelters also reduce the risks of predation (while opacity of the eyes obscures vision) and provide humid conditions that are essential for reducing evaporative loss of water during shedding. Because foraging and possibly transcutaneous exchanges of gases are also impeded during pre-​shedding, immobility is profitable and more accessible on land than at sea. The removal of “old skin” is physically easier and can be performed unseen on land. It is thus expected that amphibious snakes should preferably come on land during the pre-​shedding and shedding phases of ecdysis. We used visible indicators of pre-​shedding (e.g., opacity of the eye and whitish belly) and ecdysis to examine the extent to which sea kraits come on land to shed. It is likely that we misclassified some pre-​shedding snakes into the non-​shedding category when no obvious signs were visible. On average, 12.3% of the snakes on land were observed in the process of ecdysis (n = 18,161 observations). Most of these observations correspond to

80

8 0   • 

I s l a n d s a n d  S n a k e s

pre-​shedding snakes (58.2%)—​an expected result because pre-​shedding lasts longer compared to shedding. Many shed skins, either compacted (if shedding occurred between rocks) or stretched (if shedding occurred in the vegetation), were observed in places where refuges are abundant, but never on open, flat surfaces such as sandy beaches. During summer rainfalls following droughts, many snakes leave their shelter to drink fresh water from temporary puddles (Bonnet and Brischoux 2008). The proportion of pre-​shedding snakes was greater during these observations, reinforcing the notion that well-​buffered refuges are intensively utilized by sea kraits for shedding. We also observed many pre-​shedding and digesting snakes that were sheltered in partly emerged shipwrecks. The skin of these snakes showed a marked red coloration, suggesting they had spent long periods in contact with rusted materials. Only 4% of the snakes were observed simultaneously with a prey in the stomach and in the process of ecdysis, but this reached 8.6% for snakes with well-​digested food items (very soft palpation) (n = 13,053 observations). These observations suggest that foraging and ecdysis tend to be temporally distinct. Clearly, ecdysis represents a major reason for sea kraits to come onto land, and during the pre-​shedding period they are usually found quiescent within well-​buffered shelters. Recovery from Injuries During the course of the population survey, we observed many snakes with wounds or significant scars indicated by a marked change in the color of the tegument (Brischoux and Bonnet 2009). In many cases, such wounds were inflicted by fish, as revealed by the typical V-​shape cuts caused by the teeth rows of anguilliform fish (Figure 3.5; Bonnet et al. 2010). We do not know how much time is necessary for individuals to recover from recent injuries, but the recovery time correlates with the severity of injury and is likely influenced by the general condition of the snake. We can broadly estimate that at least 1 week is necessary. Because swimming ability, osmotic balance, immunity, and possibly cutaneous respiration might be compromised by injury, sea kraits must shelter under well-​buffered terrestrial refuges during the recovery period. Our data suggest that injury and recovery are relatively frequent in sea kraits. We observed 635 cases in which at least one recent wound was visible (4.5%, n = 17,925 observations; see Table 3.1). Detailed information was available on 358 occasions. Although in 49.5% of the cases the skin was superficially damaged, in 14.4% of the cases the underlying tissues were also

81

Figure  3.5  A  Yellow Sea Krait (Laticauda saintgironsi) with deep cuts (top) likely inflicted by a retaliating prey (Moray Eel or Conger). The cuts were sewn (bottom), and the snake recovered as indicated by recaptures. Source: Photographs by Xavier Bonnet.

82

8 2   • 

I s l a n d s a n d  S n a k e s

visible (see Figure 3.5). In severe wounds, muscles and/​or the body cavity were exposed, and in the most extreme cases substantial portions of the skin and underlying tissues were missing. Snakes did not always recover from these deep wounds (e.g., some wounded individuals were found dead), but most sea kraits exhibit scars, suggesting that recovery on land is generally successful providing that the snake can reach a refuge. Overall, terrestrial episodes are essential for sea kraits recovering from wounds (and possibly from diseases). During these episodes, movements are likely to be minimized, thereby limiting accumulation of sand in the wounds and/​or encounters with ticks that selectively target damaged skin. Resting Resting enables individuals to recover from exhausting activities. It also permits individuals to optimize their time budget through the selection of the best time windows for engaging in energetically demanding or risky activities. Resting is thus an essential function in many animal species, but it has received little scientific attention. Theoretically, assessing resting is easy in snakes because the main activities are chronologically delimited. For example, foraging is generally separated from digestion, and in many instances, snakes do not engage simultaneously in multiple tasks or activities. Thus, the “resting status” can be relatively easily distinguished from other functions. Table 3.1 illustrates that the sums of the proportions for each activity per species or sex category are usually lower than 1, suggesting there is limited overlap among activities. Here, we considered that a snake observed to be motionless within a terrestrial refuge, without any obvious prey item in the stomach, without wounds, without any obvious signs of pre-​shedding, and not involved in reproduction (e.g., developing eggs in females and mating season in males) was truly a resting individual. Although arbitrary, this procedure limited the risk of mixing resting with other activities when individuals, although motionless in their refuge, achieve other functions. Indeed, an apparently resting female may well be involved in the extremely demanding process of vitellogenesis (Bonnet et al. 1994; Van Dyke and Beaupre 2011). In practice, results were very similar whether reproduction was factored out or not, and therefore our data include reproductive individuals. Our data suggest that a significant proportion of individuals (38.3%, n = 9,073 observations; see Table 3.1) were observed while resting on land. This proportion is likely overestimated because snakes recovering from a disease,

83

Sea Kraits: Implication for Conservation  • 

83

with invisible internal wounds, during a non-​obvious part of pre-​shedding, or at the end of a digestive episode might have been wrongly counted as resting. Despite such potential imprecisions, however, it is noteworthy that sea kraits frequently come on land to rest, and this activity concerns a proportion of individuals similar to the cohort of digestive snakes. The definition of resting we adopted here is not applicable to all species of snakes. For example, Turtle-​headed Sea Snakes forage almost continuously, at least during daytime (Shine et al. 2004), and thus do not follow the sequential lifestyle for this major trait. Similarly, the notion that to be resting an animal must not be engaged in any other function (e.g., reproduction) is questionable. Nonetheless, our data clearly suggest that snakes spend considerable amounts of time resting and that in most cases they are not performing any other obvious physiological task. We emphasize that resting is a neglected, albeit important, behavior that deserves further investigation. Reproduction Sea kraits are oviparous, and thus gravid females are constrained to find suitable terrestrial nesting sites to lay their eggs. Gravid female sea kraits sometimes undertake long migrations to communal coastal oviposition sites, which is a classical observation for terrestrial species. Supposedly, specific sites offer appropriate conditions for incubation that are unavailable in remote offshore sandy islets (Bonnet et al. 2014). True sea snakes are viviparous and thus do not face these constraints. Discussing the respective advantages of oviparity versus viviparity is beyond the scope of this chapter (e.g., see Shine 2014). However, there are several potential benefits associated with oviparity in sea kraits. Copulation may be facilitated on land because the intromission of the hemipenis requires intimate contact between the partners and a position that appears much less stable under water (for pictures, see Heatwole 1999; Ineich and Dune 2013). Furthermore, we can hypothesize that it might be easier for a male to follow pheromonal trails deposited by reproductive females on land rather than those possibly dispersed and transported by sea currents (courtship appears highly visual in water; Shine 2005). Mating concentrates individuals in prescribed places such as small coralline islets, where rates of encounter increase; sexual selection is enhanced by the density of snakes because many males compete to access copulation and females can be choosy. Differential access to females may involve variable searching effort, ability to detect sheltered females, and jerking effort that can last for hours.

84

8 4   • 

I s l a n d s a n d  S n a k e s

The more terrestrial sea kraits (L. saintgironsi) form mating balls (Figure 3.6; see also Table 3.1) that are typically observed in near-​shore vegetation. Gravid females with one to five males, curled and jerking, are routinely observed under large trees, or in the cavity of fallen trees, and within the dense herbaceous vegetation close to the shore (see Figure 3.6; see also Chapter 4, this volume). Mating is frequently observed during the day. On the other hand, the more aquatic species (L. laticaudata) tends to mate in, or very close to, the intertidal zone, in the open on wet substrates, or in refuges very close to the water (e.g., logs, beach rocks, rocky jetties, or sea walls), and especially at night. Thus, mating of each species differs in time and specific microhabitats, respectively distributed from the intertidal zone to the vegetation above the berm of the shore. Although females appear highly selective in choosing oviposition sites, current information is indirect. Gravid females with large eggs in the abdomen have been intercepted moving from the sea to large rock formations, piles of large boulders, natural crevices, or burrows of seabirds. Female L. saintgironsi are able to climb steep cliffs to reach deep crevices that are well exposed to the sun. Finally, neonates were observed near these rocky sites and are absent from remote offshore sandy islets, generating a marked spatial age

Figure 3.6  A mating ball of Yellow Sea Kraits (Laticauda saintgironsi). A female (not visible) is covered by males. In this species, mating balls are often observed in the vegetation just above the berm of the shore. Source: Photograph by Xavier Bonnet.

85

Sea Kraits: Implication for Conservation  • 

85

structure of the population where the lagoon is large (Bonnet et  al. 2015). In L. saintgironsi, nurseries are associated with very thick herbaceous vegetation situated just above the shore bank (Figure 3.7), where neonates spend the first weeks or months of life before dispersing across the lagoon (Bonnet et al. 2014). Another species of sea krait, Laticauda semifasciata, oviposits eggs inside tidal caves that have an opening to the sea (Tu et al. 1990; see Chapter 4, this volume). Mating can be observed in remote, offshore and coastal colonies. However, oviposition appears to be more restricted, and coastal igneous sites attract gravid females and accommodate nurseries. Overall, the specific requirements of reproduction (i.e., acquisition of resources, mating, oviposition, and early survival of juveniles) are associated with a variety of terrestrial refuges with sharp differences among the species of sea kraits. Broadly, complex shorelines are more likely to offer the density and diversity of refuges required for the reproduction of sea kraits. Therefore, cleaning sandy beaches, removing logs and leaves, and clearing thick vegetation to accommodate tourists are highly detrimental.

Costs and Benefits of an Amphibious Lifestyle Performing major activities on land is advantageous for sea kraits, providing that suitable and easily accessible shelters are available. Digestion, ecdysis, healing of injuries, resting, and mating are likely facilitated and optimized on land. The expenditure of energy is minimized in motionless individuals, hydromineral balance is easier to maintain out of the hyperosmotic seawater, precise thermoregulation is possible, and the risks of predation are likely to be minimal on land compared with water. These advantages may explain why the largest colonies of snakes are usually observed in amphibious species, such as natricines or homalospids (Brooks et al. 2007; Ajtić et al. 2013). A piscivorous diet, however, imposes periodic shifts between oceanic foraging grounds and terrestrial sites, and living in two worlds likely involves costs or limitations (Shine and Shetty 2001; Brischoux et al. 2008). Species that divide their life cycle across two highly contrasted habitats are limited by the availability of microhabitats that often need to be geographically close, or even contiguous. Clearly, to be very selective for microhabitats over two different environments reduces the potential geographic range, and thus distribution, of these species. Such a high level of specialization may also affect the persistence of populations insofar as disturbances affecting only one of the two environments are likely to be detrimental to survival.

86

Figure  3.7  Juvenile Yellow Sea Kraits (Laticauda saintgironsi) ≤6 months old aggregate in the thick herbaceous vegetation just above the berm of the shore. All the pictured snakes were found in a coastal nursery, deeply sheltered in the vegetation; they were measured, marked, and released at the place of capture. Source: Photographs by Xavier Bonnet.

87

Sea Kraits: Implication for Conservation  • 

87

Another set of costs induced by amphibious lifestyle relates to trade-​offs that can occur for a variety of physiological and morphofunctional aspects of the life of amphibious species. Indeed, physical and chemical properties of these environments are so contrasted that they often involve antagonistic selective pressures on major phenotypic traits; thus, being optimized for life in one environment (e.g., land) inevitably compromises performances in the other (e.g., water). Intensive studies of amphibious snakes have identified that such trade-​offs affect such key life history traits as thermoregulation (Aubret and Michniewicz 2010), locomotion (Shine et al. 2003; Wang et al. 2013), or osmoregulation (Lillywhite et al. 2008, 2009). Finally, living in two worlds includes being in contact with two different sets of predators and thus may involve increased risk of predation. For example, amphibious snakes might be susceptible to predation by carnivorous mammals and birds on land but also by fish in the water (Ineich and Laboute 2002). Interestingly, the same issue may apply to susceptibility to parasites and epibionts that can colonize amphibious snakes (Figure 3.8; Pfaller et al. 2012). However, in the case of sea kraits, survival of parasites from one habitat may be jeopardized in the other. Ticks (Amblyomma nitidum) can be found dead on snakes returning from foraging trips at sea (François Brischoux and Xavier Bonnet, personal observations), whereas barnacles do not survive for long periods on land. The

Figure  3.8 Ticks (Amblyomma nitidum) attached to a Blue Sea Krait. The ticks selected a damaged part of the skin. The gray ticks (three large and one small) are female; the small orange tick is a male. Source: Photograph by Xavier Bonnet.

8

8 8   • 

I s l a n d s a n d  S n a k e s

“sanitary” consequences of amphibious lifestyle, and especially the susceptibility to both airborne and waterborne diseases, have yet to be investigated.

The Spatial Distribution of Terrestrial Habitats and Sea Kraits In a previous study based on the description of the shores of 10 islets, we highlighted that the proportion of peculiar structures, namely beach rocks with crevices or not attached to the substrate situated in the intertidal zone, was strongly linked to the proportion of L. laticaudata captured at a given site (Figures 3.9 and 3.10; Bonnet et al. 2009). Interestingly, the study also showed that L. saintgironsi was also dependent, but to a lesser extent, on such specific microhabitats (Bonnet et al. 2009; see also Chapter 4, this volume). For the current chapter, we have conducted a similar analysis by extending our initial sample to a total of 32 islets. It is noteworthy that adding new sites allowed us to increase the breadth of the lagoon’s characteristics surrounding each site. Most notably, we have now added sites that are situated in wider and more open parts of the lagoon, as well as sites that are in very narrow parts (for a map, see Bonnet et al. 2015). These sites encompass a set of variables

Figure  3.9  Blue Sea Kraits sheltering under beach rocks. Several beach rocks were lifted to expose the snakes. Source: Photograph by Xavier Bonnet.

89

Sea Kraits: Implication for Conservation  • 

89

Figure 3.10  An adult male Blue Sea Krait taking a breath. Individuals shelter under the well-​buffered beach rocks situated in the intertidal zone. At high tide, they must partly leave their refuge to reach the water’s surface. Two air bubbles (one per nostril) expelled just before surfacing can be seen at the right of the snake. Source: Photograph by Xavier Bonnet.

including climate (e.g., rainfall), distance to the mainland or to the open ocean, and width of lagoon. Examination of the data (Figure 3.11) compared with Figure  5 from Bonnet et al. (2009) suggests that the strong trend we previously detected (i.e., L. laticaudata being dependent on the availability of beach rocks) was challenged by increased sample size (i.e., many sites without beach rocks actually sheltered a significant proportion of L.  laticaudata; Figure 3.12). However, a precise examination of the peculiarities of the shores at each site reveals that small-​scale habitat heterogeneity almost always robustly explains the presence of datum “outliers” (for explanations and details, see legend to Figure 3.11). Interestingly, addition of new sites highlighted the complexity of the selection of terrestrial microhabitats by these two species of sea kraits and emphasized its importance. Large colonies of sea kraits were systematically observed at sites provided with abundant refuges near the shore—​beach rocks, crevices, boulders, large trees, or thick vegetation. The most prosperous colonies were generally found in protected areas, either natural reserves or remote islets. However, in some of the larger colonies, population structure suggests that they could not function independently from other colonies. For example, one islet was only populated by thousands of adults with no local productions of recruits,

90

1.0 0.9 Proportion of L. laticaudata

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

100

200

300

400

1000

Beach rock shoreline (m)

Figure  3.11  Relationship between the dimension of the beach rock formation along the shoreline measured in 32 sea krait colonies (X axis) and the proportion of Laticauda laticaudata relative to Laticauda saintgironsi (Y axis, recaptures excluded). The former species cannot penetrate far inland and thus must shelter in the intertidal zone or very near the shore, whereas L. saintgironsi can use a greater diversity of terrestrial habitats. A  positive relationship is observed among typical coralline islets (black circles, n  =  22 sites); the lack of beach rocks precludes the occupancy of the site by L.  laticaudata. The gray hexagons (n  =  5 sites) also indicate coralline islets spread out in the lagoon, but sandy beaches represent the only type of shore. The presence of L. laticaudata (up to 50% of the snakes) is thus counterintuitive at these sites. However, almost all individuals (L. laticaudata + L. saintgironsi) were found curled and well sheltered among the roots of large trees very close to the tidal limit (see Figure 3.12). Crossed diamonds (n = 4 sites) indicate mainland coastal colonies where partly submerged boulders are abundant (artificial jetties and walls were considered as beach rock). The two symbols connected by a gray arrow represent a single site—​a coastal nursery for L. saintgironsi; when neonates (all L. saintgironsi) are excluded from calculations, the proportion of L. laticaudata increases. The white diamond indicates a wreck on the barrier reef (>30 km offshore). Although many artificial shelters are available, the barrier reef is mostly visited by L.  saintgironsi (Brischoux et al. 2007). Finally, the white triangle indicates a very peculiar microhabitat in an igneous islet without beach rock; an accumulation of dead coral at the base of a cliff offers a wet, hot, and very well-​protected refuge for a colony of L. laticaudata. Many snakes were found deeply sheltered there. Overall, roots of large trees, artificial buildings made of large rocks and/​or with cavities (walls, jetties, wharfs, and terraces), dead coral, and likely various other materials that offer wet and well-​buffered terrestrial conditions are used as alternative refuges to beach rocks.

91

Sea Kraits: Implication for Conservation  • 

91

Figure  3.12  Several Yellow Sea Kraits (Laticauda saintgironsi) sheltering in hot and wet interstices in roots in the sandy berm of the shore. Source: Photograph by Xavier Bonnet.

and thus there was obligatory reliance on recruits coming from other sites (Bonnet et al. 2014). Overall, our large sample size (numerous sites spread over a large distance within the lagoon and characterized by a specific physiognomy) highlights the fact that despite their marked philopatry (Shetty and Shine 2002; Brischoux et  al. 2009a), sea kraits are generally unlikely to perform their entire life cycle at a single site because their terrestrial requirements are so variable between different activities. Indeed, there are only four sites for which we have unequivocal evidence that at least one of the two neo-​Caledonian species of sea krait can perform its entire life cycle and where populations are stable without immigration (i.e., 1.5 m and body mass > 2 kg; Schwaner and Sarre 1988; Shine 1991; Aubret et al. 2004a), whereas “dwarf ” populations occur where only small lizards are available as prey (i.e., Roxby Island and South Australia; snout–​vent length < 0.8 m and body mass < 700 g; Boback 2003; Keogh et al. 2005). Furthermore, the size of snakes at birth also varies tremendously across the tiger snake’s range, to a degree never reported before within a vertebrate species: 2.5-​fold difference for mean body mass and 1.3-​fold difference for mean snout–​vent length (see Table 6.1). Comparative data suggest that these striking variations in body size are linked to prey availability by adaptive rigid genetic expressions (i.e., canalized

Figure 6.5  Adult female tiger snake feeding on a Silver Gull chick on Carnac Island. Source: Photograph by Xavier Bonnet, with permission.

415

Table 6.1  Body Size in Adult and Neonate Tiger Snakes Among Insular and Mainland Populationsa Location (No. of Adult Snakes/​No. of Litters)

Adult Size

Carnac Island (124/​22)

Adults

Neonates

Isolation/​Introduction Time (Years Ago)

Body Mass (g)

Snout–​Vent Length (cm)

Body Mass (g) Snout–​Vent Length (cm)

Normal

425.75 ± 122.12

87.41 ± 10.24

5.46 ± 1.09

18.64 ± 1.56

100

Herdsman Lake (159/​25)

Dwarf

242.62 ± 73.77

77.17 ± 8.03

4.74 ± 0.75

17.47 ± 1.15

Mainland

Joondalup Lake (28/​4)

Dwarf

255.28 ± 55.29

76.54 ± 8.18

4.35 ± 0.66

17.87 ± 0.88

Mainland

Williams Island (40/​4)

Giant

800.47 ± 292.98

111.46 ± 16.96

8.55 ± 1.23

22.24 ± 1.82

9,100

Reevesby Island (44/​5)

Giant

682.14 ± 238.22

100.13 ± 14.03

3.49 ± 0.19

16.90 ± 0.83

7,700

Hopkins Island (62 adults)

Giant

649.04 ± 430.74

103.63 ± 24.59

–​

–​

8,400

Christmas Island (58/​6)

Giant

566.33 ± 256.35

102.56 ± 17.35

6.05 ± 0.70

20.95 ± 0.84

6,000

New-​year Island (33/​4)

Giant

621.47 ± 248.88

104.73 ± 15.86

5.31 ± 0.62

19.48 ± 1.04

6,000

Swan Island (15 adults)

Giant

728.13 ± 412.98

100.67 ± 16.02

–​

–​

6,250

Trefoil Island (9/​3)

Normal

430.11 ± 238.88

95.06 ± 18.00

4.11 ± 1.17

19.75 ± 1.80

50

Tasmania (14/​4)

Normal

–​

91.86 ± 16.00

4.44 ± 0.48

20.25 ± 0.71

Mainland

Chappell Island (10/​10)b

Giant

821.00 ± 27.6

125.10 ± 9.80

7.95

23.19

9,100

Means ± SD are given.

a

Data from Schwaner and Sarre (1988).

b

416

1 4 6   • 

I s l a n d s a n d  S n a k e s

shifts in life history traits such as body size at birth or growth rates) and/​ or by developmentally plastic responses (i.e., adaptive reaction norm to food availability; Aubret et al. 2004b; Keogh et al. 2005; Aubret and Shine 2009; Aubret 2012, 2015). Field observations highlighted the fact that both neonate and adult snakes were larger in places where food resources consisted of larger prey (notably, chicks of seabirds) and smaller where the snakes were restricted to smaller prey (skinks; Aubret 2012; see Table 6.1). These findings collectively suggest that the size of prey was the main driver for the evolution of body size in gape-​limited predators and that variation of adult size reflected both selective forces acting on earlier life stages and the availability of resources during ontogeny (notably abundance and diversity of prey; Shine 1987; Schwaner and Sarre 1990; Aubret 2012). Perhaps the best studied of all populations of tiger snakes is the one at Carnac Island in Western Australia. Although several species of bird (including Pied Cormorants, Little Penguins [Eudyptula minor], and Wedge-​ tailed Shearwaters [Puffinus pacificus]) nest on Carnac Island, the Silver Gull is by far the most abundant, with estimates of 3,000–​4,000 breeding pairs in 1981 (Lane 1979; Dunlop and Storr 1981). Historically, the number of gulls increased around Perth in response to the availability of food at domestic rubbish disposal sites (Dunlop and Storr 1981), thereby increasing the nesting populations of these birds on nearby offshore islands, notably Carnac Island, and thus comprising a significant portion of the island’s biomass. Terrestrial vertebrate fauna on Carnac Island includes the King Skink (Egernia kingii), the Shrubland Skink (Morethia obscura), a gecko (Christinus marmoratus), and the introduced House Mouse (Mus domesticus). Skinks and mice constitute the major dietary components of neonatal and juvenile snakes, whereas the adults feed mostly on chicks of Silver Gulls, which comprise 83% of prey items, with mice (15%) and lizards (2%) constituting the remainder of the adult diet (Bonnet et  al. 2002). The population of tiger snakes on Carnac Island is estimated to be approximately 300–​4 00 adults (Bonnet et al. 2002). Although very small (i.e., 19 ha), Carnac Island supports a stable population of snakes, and the density recorded is one of the highest reported for sedentary vertebrates (19–​25 adult snakes per hectare; Bonnet et al. 2002). In comparison, Schwaner and Sarre (1988) estimated densities of 4–​13 snakes per hectare in insular populations of tiger snakes in South Australia. Even higher densities are known in other species—​for example, 55 snakes per hectare in South American Bothrops insularis (Martins et al. 2008) and approximately 200 snakes per hectare in the Shedao Island pit viper Gloydius

417

Insular Tiger Snakes and Seabirds in Australia  • 

147

shedaoensis from China (Shine et  al. 2002). Undoubtedly, the large supply of Silver Gull chicks must have played a key role in allowing tiger snakes on Carnac Island to thrive to the point that Carnac Island was coined a “heaven” for snakes by X. Bonnet et al. (2002; see Chapters 5, 8, and 9, this volume).

Pain: The Costs of Avian Predation on Snakes Although nesting birds provide a reliable food supply to insular tiger snakes, these snakes may be preyed on by other avian species such as the Barn Owl (Tyto alba; nesting on Reevesby Island, South Australia) or the White-​bellied Sea Eagle (Haliaeetus leucogaster; observed on Williams Island and Hopkins Island, South Australia). Furthermore, adult Silver Gulls or Pacific Gulls (Larus pacificus) might prey upon neonate and small juvenile tiger snakes. No records exist, however. In July 2007, an adult female N. scutatus with a large (~2.5 × 2.5 cm; Figures 6.6a and 6.6b), healed scar on the left side of her head was captured on Hopkins Island, South Australia. A large area of tissue and bone as well as the left eye were missing. The left fang was embedded in scar tissue and was nonfunctional. The snake had a low body mass for its length (356 g and 1130 mm snout–​vent length) compared to the island average (almost 650 g for a comparable length; see Table 6.1). The injury appeared to be quite old, indicating that this snake had continued to survive despite its severe head mutilation. A  large adult male N.  scutatus with a large, healed scar at mid-​body was also captured on Hopkins Island. We presume that such injuries are due to attacks by large predatory birds. No other terrestrial snake predators occur on Hopkins Island. White-​bellied Sea Eagles and Wedge-​ tailed Eagles (Aquila audax) are both present on Hopkins Island, and both appear capable of attacking adult N.  scutatus and inflicting severe injuries (Aubret and Thomas 2009). Although the chicks of seabirds appear to provide easy prey for insular tiger snakes, a striking drawback to this foraging behavior became apparent after an unusual discovery on Carnac Island (Bonnet et al. 1999): A large number of otherwise apparently healthy adult snakes were observed with severe head injuries, to the point of skull exposure (see Figure 6.6d). Of the estimated 400 adult snakes that appear to thrive on Carnac Island, 7.5% were, in fact, totally blind, and 6.6% were half-​blinded. This was a truly astonishing finding because survival is often compromised in animals when vision is partially or totally lost as a result of injury or disease (Martin 1981; A. S. Brown et al. 1984; Gauthier 1991). It quickly became evident that snakes were sustaining these injuries while feeding on chicks of Silver Gulls, which elicited repeated and

418

1 4 8   • 

I s l a n d s a n d  S n a k e s

Figure  6.6  Predatory attacks by White-​bellied Sea Eagle or Wedge-​tailed Eagle may result in serious head injuries and partial blindness in tiger snakes (a and b; Hopkins Island, South Australia). Head injuries sustained from Silver Gulls defending their chicks (c; Christmas Island, Tasmania) may eventually lead to total blindness (d; Carnac Island, Western Australia). Sources: Photographs by Fabien Aubret (a–​c) and Xavier Bonnet (d), with permission.

aggressive attacks of nesting Silver Gulls protecting their chicks. A similar observation was made on Christmas Island (7 ha) off the coast of King Island, Tasmania, in the Bass Strait (Aubret and Thomas 2009; see Figure 6.6c) and is probably more common than first thought. Out of 33 adult snakes (17 males and 16 females) captured on Christmas Island, 3 large adult males (9.1%) had significant head injuries. These injuries were characteristic of Carnac Island tiger snakes and recognized from numerous observations of wounds inflicted by nesting birds defending their chicks. Many bird species nest on Christmas Island, including Short-​ tailed Shearwater, Little Penguins, and Silver Gulls. Chicks of these three species are preyed upon by N. scutatus. However, there is no evidence that shearwaters or penguins defend their chicks from the snakes, whereas silver gulls have been observed displaying this defensive behavior. Although there are many islands with tiger snakes in southern Australia where Silver Gulls do not nest, these unusual interactions are likely to occur where both species coexist.

419

Insular Tiger Snakes and Seabirds in Australia  • 

149

The most surprising aspect of the snake–​g ull interactions was that injured snakes seemingly survived. Their body condition was no different than that of non-​injured animals, suggesting that they fed normally. Injured animals were also clearly involved in reproductive events (Bonnet et al. 1999). How could adult snakes feed without the sense of vision? Both visual and chemical (olfactory and taste) senses play major roles in the feeding behavior of snakes (Herzog and Burghardt 1974; Drummond 1985; Teather 1991; Cooper et al. 2000). Laboratory studies have shown that blindfolded, adult, field-​caught tiger snakes had great difficulty in capturing live adult mice (Aubret et al. 2005). The time taken to initiate a first strike at the prey increased 20-​fold, and the time to successful envenomation increased 28-​fold, compared to sighted snakes. Blindfolded juvenile laboratory-​born snakes were also less successful at biting a young dead mouse offered with tweezers (Aubret 2016). When successful, young snakes required more time and more tongue flicks than sighted snakes to bite the prey. The frequency of tongue flicking, however, was not increased by blindness. That is, there was no immediate compensation of loss of vision with increased chemosensory sampling by means of tongue flicks (although this might occur over time in the case of permanent blindness). Overall, a combination of sight and olfaction triggered more successful and faster bites than did olfaction alone. Field data showed that mobile prey (mice and skinks) only constituted 5% of blind snakes’ diets on Carnac Island, in contrast to 20% for snakes with full vision, indicating reduced predation on mobile prey by blind snakes. It is plausible that some blind snakes may also forage by scavenging, feeding on dead prey, because several records indicated there were maggots found on regurgitated prey items (see Chapter 9, this volume). However, the ability of large adult tiger snakes to survive without vision may well be attributable to the availability of abundant immobile prey in these insular ecosystems having few or no predators on adult snakes (Aubret et  al. 2005). Vision does not appear to be critical in the capture of abundant and sessile silver gull chicks, which are the primary prey (Aubret et al. 2005).

Conclusion Survival and adaptation following isolation on islands comprise two of the more studied ecological challenges because of their relevance to speciation and, more recently, adaptation to the changing world climate. The combined research of various scientists spanning nearly three decades sheds light on the extraordinary survival of semiaquatic, frog-​eating snakes on arid, bird-​infested

510

1 5 0   • 

I s l a n d s a n d  S n a k e s

offshore islands (Figure 6.7). Tiger snakes have adapted to insular life and alien diets by altering many of their life-​history traits and, above all, body size. Moreover, these snakes are able to coexist with other species of snakes on some islands where, for example, another sizeable elapid is also present (Figure 6.8). The unique and periodic isolation of tiger snake populations due

Figure  6.7  Typical habitats of tiger snakes found on the mainland (a; Joondalup, Western Australia) versus the insular environments on Carnac Island (c; Western Australia) and Reevesby Island and Williams Island (b and d, respectively; South Australia). Sources:  Photographs by Fabien Aubret (a and b) and Radika Michniewicz (c and d), with permission.

Figure  6.8  Death Adders (Acanthophis pyrrhus) photographed on Reevesby Island, where tiger snakes and Death Adders co-​occur. The photograph on the right features a pair of Death Adders in courtship (the male is on the top). Source: Photographs by Fabien Aubret.

15

Insular Tiger Snakes and Seabirds in Australia  • 

151

to the geological history of island formation, as well as recent introductions, provides a window from which we may view rapid evolution and speciation in process.

Acknowledgments I thank the Department of Conservation and Land Management (Western Australia; permit Nos. SF004604 and CE000347; SF005274 and CE001216), the Department of Environment and Heritage (South Australia; permit No. M25082 1), and the Department of Primary Industries, Water and Environment (Tasmania; permit No. FA 07282) for the issuing of licenses and for their ongoing support of the study. The Animal Ethics Committees of the University of Western Australia (project No. 01/​100/​177) and the University of Sydney (project No. L04/​3-​2006/​4297) approved all procedures. Funding was provided by the University of Western Australia, the Région Poitou-​ Charentes, the Australian Research Council (ARC), and the Centre National de la Recherche Scientifique (CNRS). I  also thank W.  Gigg, D.  Roberts, D.  Bradshaw, X.  Bonnet, D.  Pearson, R.  Shine, J.  Thomas, M.  Elphick, R. Michniewicz, as well as numerous helpers throughout the years of collection of the data presented in this chapter. Léa Lapin and Maxim Kiki assisted in the maintenance of captive snakes.

References Aubret, F. 2012. Body size evolution on islands: Are adult size variations in tiger snakes a non-​adaptive consequence of selection on birth size? The American Naturalist 179:756–​767. Aubret, F.  2013. Role and evolution of adaptive plasticity during the colonization of novel environments. In J.  B. Valentino and P.  C. Harrelson (Eds.), Phenotypic Plasticity, Molecular Mechanisms, Evolutionary Significance and Impact on Speciation. New York: Novinka, pp. 1–​34. Aubret, F. 2015. Island colonisation and the evolutionary rates of body size in insular neonate snakes. Heredity 115:349–​56. Aubret, F. 2016. Effect of sudden loss of vision on foraging behavior in captive born tiger snakes Notechis scutatus (Serpentes:  Elapidae). Phyllomedusa:  Journal of Herpetology 15:75–​78. Aubret, F., X. Bonnet, S. Maumelat, D. Bradshaw, and T. Schwaner. 2004a. Diet divergence, jaw size and scale counts in two neighbouring populations of tiger snakes (Notechis scutatus). Amphibia–​Reptilia 25:9–​17. Aubret, F., X.  Bonnet, D.  Pearson, and R.  Shine. 2005. How can blind tiger snakes (Notechis scutatus) forage successfully? Australian Journal of Zoology 53:283–​288.

512

1 5 2   • 

I s l a n d s a n d  S n a k e s

Aubret, F., X. Bonnet, and R. Shine. 2007. A role for adaptive plasticity in a major evolutionary transition:  Early aquatic experience affects locomotor performance of terrestrial snakes. Functional Ecology 21:1154–​1161. Aubret, F., G. M. Burghardt, S. Maumelat, X. Bonnet, and D. Bradshaw. 2006. Feeding preferences in two disjunct populations of tiger snakes, Notechis scutatus. Austral Ecology 17:716–​725. Aubret, F., R. J. Michniewicz, and R. Shine. 2011. Correlated geographic variation in predation risk and antipredator behaviour within a wide-​ranging snake species (Notechis scutatus, Elapidae). Austral Ecology 36:446–​452. Aubret, F., and R. Shine. 2009. Genetic assimilation and the post-​colonization erosion of phenotypic plasticity in island tiger snakes. Current Biology 19:1932–​1936. Aubret, F., and R. Shine. 2010. Fitness costs may explain the post-​colonisation erosion of phenotypic plasticity. Journal of Experimental Biology 213:735–​739. Aubret, F., R.  Shine, X., and Bonnet. 2004b. Adaptive developmental plasticity in snakes. Nature 431:261–​262. Aubret, F., and J.  Thomas. 2009. Notechis scutatus (Australian Tiger snake):  Injuries. Herpetological Review 40:100–​101. Blackburn, T. M., and K. J. Gaston. 1994. The distribution of body sizes of the world’s bird species. Oikos 70:127–​130. Boback, S. M. 2003. Body size evolution in snakes: Evidence from island populations. Copeia 2003:81–​94. Bonnet, X., F.  Aubret, O.  Lourdais, M.  Ladyman, D.  Bradshaw, and S.  Maumelat. 2005. Do “quiet” places make animals placid? Island versus mainland tiger snakes. Ethology 111:573–​592. Bonnet, X., D. Bradshaw, R. Shine, and D. Pearson. 1999. Why do snakes have eyes? The (non-​)effect of blindness in island tiger snakes (Notechis scutatus). Behavioral Ecology and Sociobiology 46:267–​272. Bonnet, X., S. Lorioux, D. Pearson, F. Aubret, D. Bradshaw, V. Delmas, and T. Fauvel. 2011. Which proximate factor determines sexual size dimorphism in tiger snakes? Biological Journal of the Linnean Society 103:668–​680. Bonnet, X., D. Pearson, M. Ladyman, O. Lourdais, and S. D. Bradshaw. 2002. “Heaven” for serpents? A mark–​recapture study of tiger snakes (Notechis scutatus) on Carnac Island, Western Australia. Austral Ecology 27:442–​450. Brown, A.  S., F.  N. Carrick, and G.  Gordon, 1984. Infertility and other chlamydial diseases and their effects on Koala populations. Australian Mammalian Society Bulletin 8:97. Brown, J. H. 1995. Macroecology. Chicago: University of Chicago Press. Brown, J. H., and B. A. Maurer. 1989. Macroecology: The division of food and space among species on continents. Science 243:1145–​1150. Brown, J. H., P. A. Marquet, and M. L. Taper. 1993. Evolution of body size: Consequences of an energetic definition of fitness. The American Naturalist 142:573–​584.

513

Insular Tiger Snakes and Seabirds in Australia  • 

153

Case, T.  J. 1978. A  general explanation for insular body size trends in terrestrial vertebrates. Ecology 59:1–​18. Case, T.  J., and T.  D. Schwaner. 1993. Island/​mainland body size differences in Australian varanid lizards. Oecologia 94:102–​109. Cogger, H.  G., and H.  Heatwole. 1981. The Australian reptiles:  Origins, biogeography, distribution patterns and island evolution. In A.  J. Keast (Ed.), Ecological Biogeography of Australia. The Hague, the Netherlands: Junk, pp. 1331–​1373. Cooper, W. E., Jr., G. M. Burghardt, and W. S. Brown. 2000. Behavioural responses by hatchling racers (Coluber constrictor) from two geographically distinct populations to chemical stimuli from potential prey and predators. Amphibia–​Reptilia 21:103–​116. Drummond, H. 1985. The role of vision in the predatory behaviour of natricine snakes. Animal Behaviour 33:206–​215. Dunlop, J. N., and G. M. Storr. 1981. Seabird islands: Carnac Island, Western Australia. Corella 5:71–​74. Forsman, A.  1991. Adaptive variation in head size in Vipera berus L.  populations. Biological Journal of the Linnean Society 43:281–​296. Foster, J. B. 1964. Evolution of mammals on islands. Nature 202:234–​235. Frisch, J. E., and J. E. Vercoe, 1977. Food intake, eating rate, weight gains, metabolic rate and efficiency of feed utilization in Bos taurus and Bos indicus crossbred cattle. Animal Production 25:343–​358. Gauthier, D.  1991. La kérato-​ conjonctivite infectieuse du chamois:  étude épidémiologique dans le département de la Savoie 1983–​1990. Thesis, ENVL, Lyon, France. Gorman, G.  C. 1968. The relationships of Anolis of the roquet species group (Sauria: Iguanidae): III. Comparative study of display. Breviora 284:1–​31. Gotelli, N. J., and G. R. Graves. 1990. Body size and the occurrence of avian species on land-​bridge islands. Journal of Biogeography 17:315–​325. Grayton, B. D., and F. W. H. Beamish. 1977. Effects of feeding frequency on food intake, growth and body composition of rainbow trout (Salmo gairdneri). Aquaculture 11:159–​172. Herzog, H. A., and G. M. Burghardt. 1974. Prey movement and predatory behavior of juvenile western yellow-​bellied racers, Coluber constrictor mormon. Herpetologica 30:285–​289. Keogh, J. S., I. A. Scottand, and C. Hayes. 2005. Rapid and repeated origin of insular gigantism and dwarfism in Australian tiger snakes. Evolution 59:226–​233. Lahti, D. C., N. A. Johnson, B. C. Ajie, S. P. Otto, A. P. Hendry, D. T. Blumstein, R. G. Coss, K. Donohue, and S. A. Foster, 2009. Relaxed selection in the wild. Trends in Ecology and Evolution 24:487–​496. Lande, R. 1976. Natural selection and random genetic drift in phenotypic evolution. Evolution 30:314–​334.

514

1 5 4   • 

I s l a n d s a n d  S n a k e s

Lane, S. G. 1979. Breeding seabirds on Carnac Island, Western Australia. The Western Australian Naturalist 14:134–​135. Lomolino, M. V. 1985. Body size of mammals on islands: The island rule re-​examined. The American Naturalist 125:310–​316. Lomolino, M. V. 2005. Body size evolution in insular vertebrates: Generality of the island rule. Journal of Biogeography 32:1683–​1699. Losos, J. B. 1995. Community evolution in greater antillean Anolis lizards: Phylogenetic patterns and experimental tests. Philosophical Transactions of the Royal Society B: Biological Sciences 349:69–​75. Madsen, T., and R. Shine. 2000. Silver spoons and snake sizes: Prey availability early in life influences long-​term growth rates of free-​ranging pythons. Journal of Animal Ecology 69:952–​958. Marquet, P. A., and M. L. Taper. 1998. On size and area: Patterns of mammalian body size extremes across landmasses. Evolutionary Ecology 12:127–​139. Martin, R. W. 1981. Age-​specific fertility in three populations of the koala, Phascolarctos cinereus Goldfuss, in Victoria. Australian Wildlife Research 8:275–​283. Martins, M., R. J. Sawaya, and Marques, O. A. 2008. A first estimate of the population size of the critically endangered lancehead, Bothrops insularis. South American Journal of Herpetology 3:168–​174. Maurer, B. A. 1998. The evolution of body size in birds: II. The role of reproductive power. Evolutionary Ecology 12:935–​944. Mirtschin, P.  J., and R.  Davis. 1992. Snakes of Australia:  Dangerous and Harmless. Melbourne: Hill of Content. Mori, A. 1994. Ecological and morphological characteristics of the Japanese rat snake, Elaphe climacophora, on Kammuri-​jimais land: A possible case of insular gigantism. Snake 26:11–​18. Morse, D. R., N. E. Stork, and J. H. Lawton. 1988. Species number, species abundance and body length relationships of arboreal beetles in Bornean lowland rain forest trees. Ecological Entomology 13:25–​37. Rawlinson, P. A. 1974. Biogeography and ecology of the reptiles of Tasmania and the Bass Strait area. In E. D. Williams (Ed.), Biogeography and Ecology in Tasmania. The Hague, the Netherlands: Springer, pp. 291–​338. Robinson, R., P. Canty, T. Mooney, and P. Rudduck. 1996. South Australia’s Offshore Islands. Canberra, Australia: Australian Heritage Commission. Roy, K., D. Jablonski, and K. K. Martien. 2000. Invariant size frequency distributions along a latitudinal gradient in marine bivalves. Proceedings of the National Academy of Sciences of the USA 97:13150–​13155. Schwaner, T. D. 1985. Population structure of black tiger snakes, Notechis ater niger, on off-​shore islands of South Australia. In G.  Grigg, R.  Shine, and H.  Ehmann (Eds.), Biology of Australasian Frogs and Reptiles. Sydney, Australia: Surrey Beatty, pp. 35–​4 6.

51

Insular Tiger Snakes and Seabirds in Australia  • 

155

Schwaner, T. D. 1991. Spatial patterns in Tiger snakes (Notechis ater) on offshore islands of southern Australia. Journal of Herpetology 25:278–​283. Schwaner, T. D., and S. D. Sarre. 1988. Body size of tiger snakes in southern Australia, with particular reference to Notechis ater serventyi (Elapidae) on Chappell Island. Journal of Herpetology 22:24–​33. Schwaner, T. D., and S. D. Sarre. 1990. Body size and sexual dimorphism in mainland and island tiger snakes. Journal of Herpetology 24:320–​322. Scott, I.  A. W., C.  Hayes, J.  S. Keogh, and J.  K. Webb. 2001. Isolation and characterization of novel microsatellite markers from the Australian tiger snakes (Elapidae: Notechis) and amplification in the closely related genus Hoplocephalus. Molecular Ecology Notes 1:117–​119. Shine, R.  1977. Habitats, diets and sympatry in snakes:  A study from Australia. Canadian Journal of Zoology 55:1118–​1128. Shine, R.  1987. Ecological comparisons of island and mainland populations of Australian tiger snakes (Notechis: Elapidae). Herpetologica 43:233–​240. Shine, R.  1991. Why do larger snakes eat larger prey items? Functional Ecology 5:493–​502. Shine, R., L.  X. Sun, M.  Kearney, and M.  Fitzgerald. 2002. Thermal correlates of foraging-​site selection by Chinese pit-​vipers (Gloydius shedaoensis, Viperidae). Journal of Thermal Biology 27:405–​412. Teather, K. L. 1991. The relative importance of visual and chemical cues for foraging in newborn blue-​striped garter snakes (Thamnophis sirtalis sirtalis). Behavior 117:255–​261. Van Valen, L. 1973. A new evolutionary law. Evolutionary Theory 1:1–​33. Wilson, E. O., and W. H. Bossert. 1971. A primer of population biology. Sunderland, MA: Sinauer.

516

7

T H E E Y E S H AV E   I T

Watching Treeboas on the Grenada Bank Robert W. Henderson

I like everything about the treeboas known as Corallus grenadensis. I like where they live, I like looking for them, I like watching them, and I  like marking and then releasing them. I  even like that my hands and arms are smeared with my blood after multiple bites and I smell like treeboa cloacal voidings after a successful night in the field. When I am not in the field, I wonder what the boas are doing and what dramas I am missing. For some time, I have been trying to put into words how much studying C.  grenadensis has meant to me. On a few occasions, I  thought the right words were coming but, if they were, they quickly disappeared; they always seem just out of reach. Of course, the snakes themselves are not the only appeal of the C. grenadensis experience. The islands on which they occur are as important as the snakes. Besides their appealing ambiance, the islands are, in essence, responsible for the species being as intriguing as it is; they in no small measure contribute to the overall experience of studying these remarkable snakes.

Treeboas I became enamored of treeboas (Corallus) the first time I  saw pictures of them in books as a school kid. Their slender necks supported heads that were rather chunky, and they had conspicuous labial pits that gave them a menacing appearance. I liked their laterally compressed bodies and the descriptions of their irascible behavior and arboreal lifestyle. I  was especially taken with those exhibiting the yellow color morph. Photographs of these wonderful

517

Treeboas on the Grenada Bank  • 

157

yellow boas appeared in books by Ditmars (1942: Plate 11) and Schmidt and Inger (1957: Plate 69), and they fueled my daydreams of one day seeing them where they lived. I saw my first Corallus in the field in 1974. It was at Mishana, a site on the Río Nanay, approximately 6 hours west of Iquitos, Peru, by water taxi. It was an Amazon Treeboa (Corallus hortulanus), and although not a yellow morph, it was an exciting encounter. Thirteen years passed before I again encountered Corallus in the wild (1987), this time Corallus cookii on St. Vincent (although still no yellow morphs). From then on, however, I would never go more than a few months without being in the field with one or another species of Corallus. My experiences with treeboas have been concentrated on the two species in the Lesser Antilles (C. cookii and C. grenadensis), but especially C. grenadensis on the Grenada Bank. Currently, nine species are assigned to the genus (Henderson 2015). Geographically, they occur from southeastern Guatemala in Central America to southeastern Brazil in South America, as well as islands at the southern end of the Windward group in the Lesser Antilles. Representatives rarely exceed 2.0 m in total length. Dorsal ground color can be extremely variable (gray, many shades of brown, taupe, red, orange, and yellow) in C. hortulanus and C. grenadensis, and somewhat less so in Corallus ruschenbergerii. Two species, Corallus batesii and Corallus caninus, undergo a dramatic ontogenetic change in dorsal coloration, from yellowish or brick red to beautiful emerald green. Species occur in a wide range of habitats, from xeric scrub to lush rainforest, and several species can be found living in proximity to humans. Treeboas take a wide range of prey (Henderson and Pauers 2012) that includes frogs, lizards, birds (including parrots), and a variety of mammals (but mostly rodents, bats, and marsupials). The phylogeny of Colston et  al. (2013) has C.  caninus as sister to all other species of Corallus and C. cropanii as sister to all species exclusive of C. caninus. Reynolds et al. (2014), as well as the earlier work of Vidal et al. (2005), support the distinction between C. caninus and C. batesii. Corallus cookii and C.  grenadensis were recovered as sister taxa but nested within C.  hortulanus. Corallus ruschenbergerii is sister to the clade composed of C. cookii, C. grenadensis, and C. hortulanus; Corallus annulatus and Corallus blombergii (T. J. Colston, unpublished data) are sister to that clade. Although the molecular data of Colston et al. (2013) and Reynolds et al. (2014) cast doubt as to the species-​level validity of C. cookii and C. grenadensis, morphological data (color and pattern and several scale characters) and geographic isolation remain in support of their taxonomic legitimacy (Henderson 1997).

518

1 5 8   • 

I s l a n d s a n d  S n a k e s

Again, based on the molecular work of Colston et  al. (2013), Corallus most likely originated in South America during the middle Eocene approximately 49 mya. Corallus caninus and C. cropanii originated in South America approximately 35–​49 mya in the Eocene. Corallus annulatus dispersed from South America to Central America during the late Miocene approximately 10 mya. A second dispersal to Central America occurred during the Pleistocene (~1.5 mya) when C. ruschenbergerii expanded its range northward from its inferred ancestral area of South America. The most geographically widespread member of the genus, C. hortulanus, arose in South America during the middle Miocene (~11.9 mya). The common ancestor of C.  cookii and C. grenadensis was inferred to have been distributed in both South America and the Lesser Antilles, thus requiring overwater dispersal to the Lesser Antilles by a northern Guianan Shield ancestor (as previously suggested by Henderson and Hedges 1995; see also Hedges 2006). Divergence estimates (Colston et al. 2013) place the age of the most recent common ancestor of C. cookii, C. grenadensis, and C. hortulanus from Guyana at approximately 2 mya in the early Pleistocene.

Why Islands? My eventual focus on island snakes came about gradually. A stint as a volunteer in the US Peace Corps in Belize (then British Honduras) allowed me to conduct a mark–​recapture study of the vinesnake Oxybelis aeneus (Henderson 1974). Post-​Peace Corps, an intriguing paper by Henry Horn (1969) and a suggestion by Roy McDiarmid pointed me in the direction of Hispaniola and snakes in the endemic vinesnake genus Uromacer. On Isla Saona (a satellite of the Dominican Republic), I frequently encountered 10–​15 Uromacer catesbyi and Uromacer oxyrhynchus during a single night as they slept on Vachellia branches. The abundance of certain species of snakes on Hispaniola compared to the neotropical mainland was amazing; this was life-​changing! Over a period of 6 years, I made multiple trips to the Dominican Republic and Haiti in pursuit of natural history data pertaining to species of Uromacer (Henderson et al. 1981). That was followed by substantial island hopping in order to do surveys for racers (genus Alsophis) in the Lesser Antilles. Racers were abundant on Saba, St. Eustatius, and Dominica, reinforcing my estimation of snake densities on West Indian islands. In the meantime, my boyhood interest in treeboas had been rekindled and, in the midst of the racer surveys, it was an easy jump to fly from Dominica to St. Vincent and, eventually, Grenada (Figure 7.1). I have previously described my initial encounters

519

Figure  7.1  Map of the St. Vincent and Grenada banks. Corallus cookii is endemic to St. Vincent, and Corallus grenadensis is endemic to the Grenada Bank (Grenada and the Grenadines).

610

1 6 0   • 

I s l a n d s a n d  S n a k e s

with C. cookii (Henderson 2002) and C. grenadensis (Henderson 2015). On both of those islands, treeboas were seemingly everywhere, and conducting multiyear projects with one or both species seemed eminently feasible. Thus, islands had a twofold attraction for me. First, and most important, they harbored snake species that occurred in population densities that were exceedingly rare or unheard of on the neotropical mainland. They therefore had the potential to yield information about certain species in a reasonable span of time. By opting to work on Lesser Antillean islands, I  knew I  was foregoing the amazing diversity of snakes and other reptiles that occurs on the mainland, but the prospect of seeing 20‒30 snakes of one species over the span of a couple of hours was too appealing to ignore. The second reason was the ambiance of islands. I liked knowing that just about every day was going to be sun-​filled, that coconut palms were ubiquitous, that refreshing breezes were almost continuous, and that the Caribbean Sea would most likely be in view several times a day.

A Little History Corallus had a long history in the West Indies (Colston et al. 2013) prior to the arrival of humans. While working at sites along Grenada’s windward coast, I have tried to imagine the first treeboas to successfully reach a West Indian landfall. After many days or weeks adrift on a raft of flotsam, the unwilling immigrants likely were hungry and dehydrated (although rainfall could have provided fresh drinking water) when they finally washed ashore. Those first successful colonizers, however, had arrived at a land of opportunity: a plentiful food supply and habitat devoid of other arboreal snakes. Typically associated with forest edges, the distribution of ancestral C.  grenadensis on the Grenada Bank prior to the arrival of humans probably was restricted to margins of lakes, rivers, coastlines, and other natural breaks in the distribution of forests, including those resulting from hurricanes and other disturbances. The boas fed largely on native lizards and now-​extinct rice rats of the genera Oligoryzomys or Oryzomys, and they were preyed upon by raptors (e.g., Buteo platypterus). Treeboas on Grenada have had a shared history with humans for at least 4,000 years, and possibly longer. Caribbean Archaic sites (ca. 3,000–​4 00 bc) are rare and relatively small. Treeboa distribution on Grenada likely changed little between 3,000 and 400 bc when human occupants were primarily fisher–​foragers with an emphasis on marine foods (Newsom and Wing 2004). If small forest clearings were made, suitable edge habitat may have

16

Treeboas on the Grenada Bank  • 

161

increased, thereby providing additional treeboa habitat. Treeboas did, however, gain a new enemy, as they almost certainly were killed when encountered by human invaders. Pearls was the major Ceramic-​age (400 bc–​1,500 ad) site on Grenada. Although early human activity at Pearls probably altered habitats, large-​ scale deforestation was unlikely. A  large area would have been cleared for the village and surrounding gardens. It would be planted for 4 or 5 years and then allowed to return to secondary forest before being cleared again (perhaps 20  years later) using a slash-​and-​burn technique (W. Keegan, in litt., 6.xi.2008). Thus, a waxing and waning of potential treeboa habitat would have been associated with early settlements, and the impact of humans on pre-​ Columbian environments in the West Indies should not be underestimated (Newsom and Wing 2004; Fitzpatrick and Keegan 2007). The first successful European colonization of Grenada was in 1650 by the French. Ultimately, however, the island was ceded to England in 1763 (Brizan 1984), and C.  grenadensis distribution at this time may not have been significantly different than during prediscovery or late Ceramic periods. The 18th century, however, saw the onset of the plantation system and large-​scale deforestation, wherein forests were largely eliminated on all but the steepest slopes (Beard 1949). By 1772, 125 of 334 estates were devoted to sugar cultivation (Brizan 1984), and those 125 estates extended over nearly 13,000 ha, accounting for 42% of Grenada’s surface area (311 km2). Not all of Grenada is suitable C.  grenadensis habitat. Treeboas are uncommon or absent at elevations much above 530 m. Because roughly 9% of Grenada lies above 500 m, approximately 50% of potential C.  grenadensis habitat was lost to sugarcane cultivation by 1772. By the 1850s, however, the economy was shifting from sugar to cocoa, and by 1878, the land area devoted to cacao trees surpassed that for sugarcane. By abandoning sugar for cocoa, more land was devoted to habitat that could be exploited by arboreal snakes. Today, orchard trees (e.g., mango, citrus, breadfruit, cacao, and nutmeg) are often among the most productive habitats for encountering treeboas (Powell et al. 2007; Henderson 2015). The Grenadines likewise have had a long history of human occupation. Fitzpatrick et al. (2010) suggested that the time of the most intensive occupation on Carriacou was ad 500–​1,000 (but that may be an artifact of 14C data sampling). Bullen and Bullen (1972) found evidence of early human occupation on Bequia, Petit Nevis, Isle à Quatre, Mustique, Canouan, Mayreau, Union, and Frigate Island (satellite to Union). A Western presence (French and British) in the Grenadines has probably been ongoing since the

612

1 6 2   • 

I s l a n d s a n d  S n a k e s

mid-​1600s. Based on accounts from the 19th century (Shephard 1831; Nichols 1891), forests on the major islands were largely eliminated to accommodate agriculture (primarily cotton) and lumbering. That any primary forest occurs on the islands is unlikely, and Union has what is considered the best and last stand of intact secondary forest in the island chain.

To Grenada . . . Grenada is the southernmost of the Lesser Antilles and is situated approximately 125 km from the nearest point on the South American mainland and 115 km from Tobago; the highest peak on Grenada is Morne St. Catherine at 839 m. Habitats on the island range from xeric scrub to lush rainforest, but the likelihood of any original forest remaining is slim. The origin of the herpetofauna on the Grenada Bank is largely South American. The St. Vincent passage, lying between St. Lucia to the north and St. Vincent to the south, is considered a faunal division of “unrivaled importance” in the southern Lesser Antilles (Lescure 1987; Censky and Kaiser 1999). Genera of frogs (e.g., Pristimantis), lizards (e.g., Copeoglossum, Marisora, Gonatodes, and Bachia), and snakes (e.g., Clelia, Mastigodryas, and Amerotyphlops) are represented usually by a single species in the West Indies, and generally not north of St. Vincent (Iguana and Thecadactylus are exceptions), and Pristimantis, Gonatodes, Corallus, Chironius, and Amerotyphlops have endemic representatives on the St. Vincent and/​or Grenada banks. Accompanied by my wife Rose and beginning with our first full day on Grenada in 1988, I  quickly realized this place was, from the perspective of treeboa biology, very special. Aside from finding my first yellow treeboa at Pearls, we found them in spectacular numbers at almost every site we visited. I eventually had the epiphany to hunt at sites that did not look suitable for C. grenadensis. With one exception (a monoculture of a tree I did not recognize), we still found boas. We were in treeboa paradise.

. . . and the Grenadines As much as I enjoy working on Grenada, I like fieldwork on the Grenadines even more. Just getting there is part of the adventure. One can fly to several of the islands, but that is way too easy. The Osprey ferry makes a round trip from St. George’s (Grenada) to Carriacou (the largest of the Grenadines)

613

Treeboas on the Grenada Bank  • 

163

every day. Once in the Grenadines, travel between islands is by water taxi or local fishermen. The water taxis are very efficient, especially if one enjoys terrifying, white-​knuckle, ass-​slamming travel (but then, who doesn’t?). To me, the Grenadines are the quintessential Caribbean islands: quiet, laid back, usually friendly locals, not many tourists, unpaved roads, rainbow splashes of bougainvillea, and palm-​lined white sand beaches, all surrounded by translucent water in exquisite shades of blue and green that words cannot adequately describe. Oh, and they harbor wonderful treeboas. The Grenadines represent the exposed peaks on a single bank of submerged mountains. According to Howard (1952, p. 7), “If the current sea level were lowered 125 feet [38 m] all of the Grenadines between Bequia and Carriacou would be united.” No geological evidence indicates that the Grenada Bank ever had a mainland connection. I cannot with any certainty state how many emergent land masses comprise the Grenadines. Based on nautical charts and aerial photographs, Kingsbury (1960) put the number at 125, and he noted estimates of up to 600. Frogs and reptiles have been documented on 33 islands in the Grenadines archipelago. Those islands range in area from 0.01 km2 (Jamesby, one of the Tobago Cays) to 32 km2 (Carriacou). Three lizard species dominate in terms of the number of islands on which they occur: Anolis aeneus (documented on 32 of 33 islands, but likely on all 33), Iguana iguana (21), and Ameiva aquilina (18). Next in number are the native snake Mastigodryas bruesi on 13 islands and the introduced human commensal Hemidactylus mabouia on 12 islands. Corallus grenadensis is the only other species that occurs on at least 10 islands. Among frogs, the introduced Eleutherodactylus johnstonei has the widest distribution, occurring on 5 islands, but so far only on those that receive commercial boat traffic. According to Daudin and de Silva (2011), today only 9 of the islands have human inhabitants; more important, at least 10 support populations of C. grenadensis (but I suspect it occurs on additional islands). I have visited 7 of the 10 Grenadines known to be inhabited by C.  grenadensis. On most of those islands, treeboa numbers were low or occurred in localized pockets. On Petite Martinique (0.7 km2), we found only two boas over two nights of searching, and those were the first records for the island in 133 years (Henderson and Berg 2012). When we questioned people about treeboas, most were unaware of their presence on the island. On Mayreau (2.6 km2), although we did not find boas in numbers, they appeared to occur throughout the island. Efforts on Canouan (7.4 km2) were fruitless until I encountered a small enclave in a patch of roadside scrub. Encounter rates (1.32 ± 0.86 boas/​hour) with C. grenadensis on Union Island (8.1 km2)

614

1 6 4   • 

I s l a n d s a n d  S n a k e s

were comparable to those for xeric scrub on Grenada (Quinn et  al. 2011). Treeboas have been encountered in a variety of habitats on Carriacou, the largest of the Grenadines. Between 1989 and 2019, many more trips to the West Indian range of Corallus followed, and hundreds of encounters with treeboas provided insights into various aspects of their natural history.

Working at Night In the field, our workday usually begins somewhere between late afternoon and dusk. Bottles of water, a loaf of bread, and a jar of peanut butter get tossed into our jeep (ideally, some cookies as well), along with whatever equipment we need for our work. I can think of few things more enjoyable or exciting than searching for reptiles, and snakes in particular, at night (Figure 7.2). One’s world is reduced to the beam of light produced by a headlamp or a handheld flashlight or spotlight. On Grenada, night work is invariably accompanied by the incessant serenading of the introduced frog E. johnstonei. Going anywhere on the island at night and being out of earshot of these horny amphibians is virtually impossible. In general, searching for snakes at night requires great concentration, and one must have developed an appropriate search image for seeing snakes, often in a complex, three-​dimensional matrix. They can be on the ground in the middle of a trail and fairly conspicuous, off to the edge of a trail and partially hidden by trailside litter and other vegetation, or they might be in a tree or bush. If in an arboreal situation, the snake could be below, at, or above eye level and, if motionless, could easily blend with surrounding vegetation. Species of Corallus are arboreal and closely associated with trees and bushes covered in leafy, vision-​obstructing vegetation. However, when a beam of light meets the eye of a treeboa, it produces an amazingly bright reflection (Figure 7.3). Therefore, when searching specifically for treeboas, one’s search image is not that of a snake’s body but, rather, the brilliant red-​orange reflection from a boa’s eye. In an unobstructed line of vision, a large (1.5 m) C. grenadensis can be seen from a distance of 40 m. Smaller boas, with proportionately smaller eyes, are less conspicuous, but pencil-​thin neonates can certainly be spotted from a distance of 10 m. My co-​workers and I have been in habitats in which one could stand in one spot, do a 360-​degree turn, and see boa reflections in multiple directions. My colleague, William Lamar, recalls seeing so many eye reflections of C. ruschenbergerii in mangrove habitat on Isla de Salamanca (off the coast of Colombia) that they reminded him of a nocturnal cityscape. Corallus ruschenbergerii can be very large (~2.4 m

615

Figure 7.2  Using an extensible pole, a treeboa is being “fished’ out of a tree at a site in central Grenada. Source: Photograph by Richard Sajdak, with permission.

61

1 6 6   • 

I s l a n d s a n d  S n a k e s

Figure 7.3 A Corallus hortulanus from Amazonian Ecuador exhibiting a conspicuous reflection from its eye. Source: Photograph by Zach Marchetti, with permission.

total length), and I have seen eye reflections in Trinidad that I was sure were from a decent-​sized mammal. In French Guiana, mammalogist Rob Voss mistakenly shot a C. hortulanus believing it to be an opossum.

Making a Living in Today’s World Habitat After our first few days of treeboa hunting on Grenada, I  felt reasonably confident that I  would be able to study aspects of C.  grenadensis natural history. After 30 years, hundreds of hours searching for treeboas, and more than 1,000 encounters, I  have a fairly good idea of where C.  grenadensis lives (and does not live). Snakes can be encountered in virtually any habitat on Grenada below approximately 530 m, including cactus–​Vachellia scrub, mangroves (including sleeping or foraging over water), seasonal evergreen and evergreen forest, fruit orchards, suburban or rural landscapes (including residential yards), hotel grounds, roadside trees and shrubs, banana plantations, deciduous evergreen forest, semi-​deciduous forest, stands of bamboo, evergreen coastal forest, and botanical gardens (Figure 7.4). One has the potential to walk out of an upscale hotel and encounter C. grenadensis within minutes. Some habitats, of course, are more favorable than others. I have not

617

Treeboas on the Grenada Bank  • 

167

Figure 7.4  Excellent habitat for Corallus grenadensis at a site in central Grenada. Source: Photograph by Robert W. Henderson.

encountered treeboas in elfin or Sierra Palm cloud forest or in any treeless expanse. However, if that treeless expanse is surrounded by even a single line of trees with contiguous crowns, C. grenadensis might inhabit it. Being a habitat generalist is only part of the ecological plasticity exhibited by C. grenadensis. Since much of Grenada and virtually all of the Grenadines lost their original forests, many of the habitats in which the species is found include (or are dominated by) species of trees that did not occur on the island prior to the arrival of Europeans. Treeboas are frequently found in orchard trees (mango, breadfruit, cacao, nutmeg, and citrus) and occasionally in Musa (banana) plantations, coconut palms, and papaya trees. I consider ideal habitat a mixture of native vegetation and orchard trees. Treeboas will occasionally enter homes and outbuildings, cross roads in urban areas by way of power lines, and forage near streetlights. Different age/​ size classes of C.  grenadensis partition arboreal habitat by using perches of different heights and diameters and sometimes by utilizing different tree species. At Mt. Hartman Bay, a xeric site at sea level in southeastern Grenada with sharply demarcated Vachellia scrub and mangrove habitats, I encountered juvenile and subadult boas most frequently in Vachellia, whereas larger boas were more often associated with mangroves.

618

1 6 8   • 

I s l a n d s a n d  S n a k e s

Anolis lizards, the primary prey of young treeboas, also were more abundant in Vachellia. Corallus grenadensis exhibits an extraordinary array of colors and patterns (Figure 7.5). Thomas Barbour (1914) described this species (as Boa grenadensis) based on specimens that were yellow and patternless or nearly so and concluded that it was a species distinct from Grenadian specimens that were conspicuously patterned (Boa hortulana). By 1930, Barbour had doubts as to its status as a distinct species, and in 1935, he relegated it to a subspecies of Boa cookii (but still sharing the island with B. hortulana). That same year, Stull (1935) synonymized the two species as Boa enydris cookii, thereby recognizing only one species of treeboa on Grenada. Eventually, I (Henderson 1997) resurrected grenadensis and elevated it back to species level (as the lone species of Corallus on the Grenada Bank).

Figure  7.5 Color and pattern variation in Corallus grenadensis. Clockwise from the upper left:  Les Avocats, St. David Parish; Grand Bras, St. Andrew Parish; Petite Martinique, Grenada Grenadines; and Grand Bras, St. Andrew Parish. Sources: Photographs by Robert W. Henderson.

619

Treeboas on the Grenada Bank  • 

169

In xeric, sun-​drenched locales (e.g., Mt. Hartman, sea level), one can be fairly certain the snake will be pale (71.1% yellowish), whereas in cool, overcast, wet habitats (e.g., Grand Etang forest, 530 m), many boas (50%; Henderson 2015) are dark brown with fine, lace-​like, white markings. In less extreme situations, taupe (gray-​brown) boas are common, but individuals might be patternless yellow or orange, red, yellow with a rose wash, yellow with a brown pattern, black and dirty white, shades of gray, shades of brown, orange and cream, red and off-​white, or something else. Talk about childhood daydreams coming true, my colleague Marie Rush introduced me to a site at which more than 90% of boas are yellow. We see little color pattern variation on the Grenadines. On Carriacou, only approximately 1% of boas are not shades of gray with gray to brown markings. The same holds true for Union Island: one yellow among dozens of gray snakes. On Mayreau, virtually all are orange with darker markings. More often than not, iris color will match the predominant dorsal coloration, and the tongue, although not matching in color, will have markings that match the degree of darkness of the boa (i.e., yellow snakes with few or no markings and dark snakes with heavily marked or solid black tongues; Henderson 2004). The color and pattern variation seen in C. grenadensis (and C. hortulanus on the South American mainland) may be unmatched by any other species of snake. Foraging and Diet Among species of Corallus, only the West Indian taxa routinely include ectotherms in their diet. Of 93 prey records for C. cookii and C. grenadensis combined, 63 (67.7%) were either lizards (61) or frogs (2). Among mainland species, ectotherms accounted for only 6.6% of 196 records. Neonatal and juvenile C. grenadensis prey almost exclusively (82.5%) on native Anolis lizards (A. aeneus and Anolis richardii), but with increasing age and size, the diet shifts to endothermic prey (73.7% of boas ≥800 mm snout–​vent length [SVL]; Henderson 2015). Rice rats likely were the most common prey species for adult C.  grenadensis prior to the arrival of Amerindian colonizers from South America, and that probably continued until Amerindian activities drove the rats to extinction. The arrival of Europeans settlers, however, brought new prey in the form of colonizing mice (Mus musculus) and rats (Rattus rattus) that are now the boas’ primary mammalian prey; birds are rarely taken. Prey mass ratios (prey mass: boa mass) are generally under 0.30, but I have recorded one at 1.22; that is, the prey item (likely a rat) had a mass that was 122% of the boa’s mass.

710

1 7 0   • 

I s l a n d s a n d  S n a k e s

Insofar as treeboas are nocturnal, sleeping anoles are located by active foraging. Young C. grenadensis move slowly through the vegetation, usually at or near the distal ends of branches, tongue-​flicking leaf and branch surfaces, and frequently on branches other than the one supporting their body (Yorks et al. 2003). I assume they are searching for scent trails laid down by anoles as they go to roost at dusk; anoles often sleep at or near the distal ends of branches or on associated leaf surfaces. When a sleeping lizard is encountered, the boa assumes a strategy of incredible stealth. Based on detailed video analysis, the boa’s approach to the lizard occurs at a rate of 0.8–​1.4 cm/​minute over the last 20 cm, and the boa is virtually touching the anole before it opens its jaws for the capture (Yorks et al. 2003). Similar foraging behavior has been observed in the closely related species C. cookii hunting anoles on St. Vincent (Henderson et  al. 2007)  and in C.  hortulanus stalking a sleeping iguana (I. iguana) in Brazil (da Costa Silva et al. 2012). In contrast to lizard-​eating juveniles, adults often assume a sit-​and-​wait (= ambush) foraging mode. That is, they feed on nocturnally active prey, so an ambush foraging boa will often perch low (0.5–​1.5 m) in a tree or bush with its head angled toward the ground and wait for a rat or mouse to pass within striking distance (Figure 7.6; see also Chapters 5 and 8, this volume). Treeboas likely choose ambush sites based on scent, seeking trails on the ground that are routinely used by active rodents and that retain olfactory information. Adult C. grenadensis also ambush prey that is active above the ground. Rats and mice are active in trees and use branches as aerial pathways. By doing so, they again lay down odor trails that provide olfactory cues for the boas. We have been privy to only one episode of predation on a rat. At a site on Carriacou, a boa’s head was visible in profile while perched in a tree at approximately 5 m. Seconds later, it suddenly turned its head toward the interior of the tree’s foliage, and my colleague Billie Harrison heard a shrill squeal. For a few minutes, the boa was lost in the crown of the tree; when Harrison was able to relocate it, the rat was held in coils toward the posterior of the boa’s body and was obviously dead. This was our first observation of rodent predation in the field, and this particular episode was especially interesting for two reasons. First, we assume the boa was aware of a rodent being in proximity due at least to olfactory information but also to infrared (thermal) and, possibly, visual cues. The boa’s head and the forepart of its body altered their direction in an instant to capture the rat. This is in sharp contrast to observations of active foraging treeboas stalking quiescent Anolis lizards. The capture of the rat required rapid information processing and a quick strike (from an unknown distance). The second observation of note was that after the capture

17

Treeboas on the Grenada Bank  • 

171

Figure 7.6  Corallus grenadensis in a sit-​and-​wait (ambush) posture at a site in central Grenada. Source: Photograph by Robert W. Henderson.

and killing, the rat was observed in coils of the posterior portion of the snake’s body and its tail. Approximately 15 minutes later, swallowing was nearly complete. Sometime between capture and the start of deglutition, the boa had moved to a higher and, presumably, safer location in the tree. We have observed transporting prey from the site of capture to a different nearby position within the habitat by C. grenadensis with a tropical mockingbird (Mimus gilvus; Sajdak and Henderson 2017; Figure 7.7) and also by C. hortulanus with a green iguana (I. iguana; da Costa Silva et al. 2012) and a bare-​tailed woolly opossum (Caluromys philander; da Costa Silva and Henderson 2014).

712

1 7 2   • 

I s l a n d s a n d  S n a k e s

Figure 7.7  Corallus grenadensis beginning to swallow a fledgling Tropical Mockingbird (Mimus gilvus) on Carriacou, Grenada Grenadines. Source: Photograph by Richard Sajdak, with permission; See Sajdak and Henderson (2017).

Abundance Corallus grenadensis occurs on an island archipelago where the entire range of the species is a small fraction of the range of mainland species of Corallus. In addition, Grenada has the highest human population density (306/​km2; Food and Agricultural Organization of the United Nations 2010)  of any country in which species of Corallus are found, and it has the smallest area of all forested land and also of primary forest (Henderson 2015). Despite this, C. grenadensis likely occurs at population densities unmatched by any other species in the genus. Three likely reasons account for the abundance of Corallus on West Indian islands: (1) Foremost is density compensation—​that is, the presence of fewer species on islands facilitates niche expansion (greater access to food and space), which can result in unusually high population densities (e.g., Rodda and Dean-​Bradley 2002; see also Chapters 6 and 9, this volume, and references

713

Treeboas on the Grenada Bank  • 

173

therein); (2) the superabundance of anoles (which also is a consequence of density compensation); and (3)  the covert lifestyle of the boas (inconspicuous by day, active at night). On the mainland, C. hortulanus has the widest distribution of any member in the genus, and it preys on more species than any of its congeners. Among boas