Large Carnivores and the Conservation of Biodiversity

About Island Press

Island Press is the only nonprofit organization in the United States whose principal purpose is the publication of books on environmental issues and natural resource management. We provide solutions-oriented information to professionals, public officials, business and community leaders, and concerned citizens who are shaping responses to environmental problems.

In 2005, Island Press celebrates its twenty-first anniversary as the leading provider of timely and practical books that take a multidisciplinary approach to critical environmental concerns. Our growing list of titles reflects our commitment to bringing the best of an expanding body of literature to the environmental community throughout North America and the world.

Support for Island Press is provided by the Agua Fund, Brainerd Foundation, Geraldine R. Dodge Foundation, Doris Duke Charitable Foundation, Educational Foundation of America, The Ford Foundation, The George Gund Foundation, The William and Flora Hewlett Foundation, Henry Luce Foundation, The John D. and Catherine T. MacArthur Foundation, The Andrew W. Mellon Foundation, The Curtis and Edith Munson Foundation, National Environmental Trust, The New-Land Foundation, Oak Foundation, The Overbrook Foundation, The David and Lucile Packard Foundation, The Pew Charitable Trusts, The Rockefeller Foundation, The Winslow Foundation, and other generous donors.

The opinions expressed in this book are those of the author(s) and do not necessarily reflect the views of these foundations.

Copyright © 2005 Island Press

All rights reserved under International and Pan-American Copyright Conventions. No part of this book may be reproduced in any form or by any means without permission in writing from the publisher: Island Press, 1718 Connecticut Ave., NW, Suite 300, Washington, DC 20009.

ISLAND PRESS is a trademark of The Center for Resource Economics.

No copyright claim is made in the work of James A. Estes, William J. McShea, Douglas W. Smith, and S. Joseph Wright, employees of the federal government.

Library of Congress Cataloging-in-Publication data.

Large carnivores and the conservation of biodiversity / edited by Justina C. Ray . . . [et al.].

p. cm.

Includes bibliographical references and index.

9781597266093

1. Carnivora—Conservation. 2. Carnivora—Ecology. 3. Predation (Biology) 4. Biological diversity conservation. I. Ray, Justina C.

QL737.C2C794 2005

591.716—dc22

2004021349

British Cataloguing-in-Publication data available.

Printed on recycled, acid-free paper e9781597266093_i0002.jpg

Design by Paul Hotvedt

Manufactured in the United States of America

10 9 8 7 6 5 4 3 2 1

Table of Contents

About Island Press

Title Page

Copyright Page

Acknowledgments

CHAPTER 1 - Introduction: How to Value Large Carnivorous Animals

PART I - Setting the Stage

CHAPTER 2 - An Ecological Context for the Role of Large Carnivores in Conserving Biodiversity

CHAPTER 3 - Large Carnivorous Animals as Tools for Conserving Biodiversity: Assumptions and Uncertainties

PART II - The Scientific Context for Understanding the Role of Predation

CHAPTER 4 - Carnivory and Trophic Connectivity in Kelp Forests

CHAPTER 5 - The Green World Hypothesis Revisited

CHAPTER 6 - Restoring Functionality in Yellowstone with Recovering Carnivores: Gains and Uncertainties

CHAPTER 7 - Large Marine Carnivores: Trophic Cascades and Top-Down Controls in Coastal Ecosystems Past and Present

CHAPTER 8 - Forest Ecosystems without Carnivores: When Ungulates Rule the World

CHAPTER 9 - King of the Beasts? Evidence for Guild Redundancy among Large Mammalian Carnivores

PART III - From Largely Intact to Human-Dominated Systems: Insight on the Role of Predation Derived from Long-Term Studies

CHAPTER 10 - Tigers and Wolves in the Russian Far East: Competitive Exclusion, Functional Redundancy, and Conservation Implications

CHAPTER 11 - Large Carnivores and Biodiversity in African Savanna Ecosystems

CHAPTER 12 - Large Carnivores and Ungulates in European Temperate Forest Ecosystems: Bottom-Up and Top-Down Control

CHAPTER 13 - Recovery of Carnivores, Trophic Cascades, and Diversity in Coral Reef Marine Parks

CHAPTER 14 - Human-Induced Changes in the Effect of Top Carnivores on Biodiversity in the Patagonian Steppe

PART IV - Achieving Conservation and Management Goals through Focus on Large Carnivorous Animals

CHAPTER 15 - Large Carnivores, Herbivores, and Omnivores in South Florida: An Evolutionary Approach to Conserving Landscapes and Biodiversity

CHAPTER 16 - Hunting by Carnivores and Humans: Does Functional Redundancy Occur and Does It Matter?

CHAPTER 17 - Detecting Top-Down versus Bottom-Up Regulation of Ungulates by Large Carnivores: Implications for Conservation of Biodiversity

CHAPTER 18 - Top Carnivores and Biodiversity Conservation in the Boreal Forest

CHAPTER 19 - The Linkage between Conservation Strategies for Large Carnivores and Biodiversity: The View From the Half-Full Forests of Europe

CHAPTER 20 - Conclusion: Is Large Carnivore Conservation Equivalent to Biodiversity Conservation and How Can We Achieve Both?

References

List of Contributors

Index

Island Press Board of Directors

Acknowledgments

We are grateful to the attendants of the workshop held at the White Oak Plantation in May 2003, of which this volume is a direct result. They generously contributed their time and thoughts to addressing the complicated set of questions we have posed here, and then spent considerable energy formulating these beautifully written essays in a relatively rapid timeframe. Thanks also to nonattend-ing coauthors who enthusiastically shared their expertise, often with very short notice, to fill critical gaps in our coverage. It was our sincere pleasure and honor to work with such high-caliber scientists.

All chapters of this volume benefited enormously from rigorous reviews provided by the following individuals, who generously donated their time and creative energy to improve clarity and content of each contribution: Liz Bennett, Richard Bodmer, Luigi Boitani, Stan Boutin, Terry Bowyer, Rodrigo Bustamante, Carlos Carroll, Emmett Duffy, Sarah Durant, Jim Estes, Graham Forbes, Todd Fuller, Josh Ginsberg, Jodi Hilty, Luke Hunter, John Linnell, David Maehr, Tim McClanahan, Bill McShea, Brian Miller, Reed Noss, François Messier, Andrés Novaro, Tim O’Brien, John Robinson, Mel Sunquist, Rick Sweitzer, and Adrian Treves. To all we are grateful.

Barbara Dean at Island Press was a bottomless source of enthusiasm from the moment she heard the first seeds of our proposal. She and Barbara Youngblood responded to each and every query, guiding us with steady hands through the small details and the complexities alike. The preparation of this volume in the relatively short timeframe from conception to printing would simply not have been possible without the cheerful and able assistance of Joanna Zigouris, whose careful attention to detail and understanding of the process helped at every turn of the way.

The workshop was made possible through the generous support of the White Oak Conservation Center of Gilman International Conservation. In particular we thank John Lukas. We gratefully acknowledge the Wildlife Conservation Society for its support through the development of the volume. Finally, we extend our deep appreciation for the patience and encouragement of our families.

CHAPTER 1

Introduction: How to Value Large Carnivorous Animals

Kent H. Redford

According to a quote attributed to Marjory Stoneham Douglas, The Everglades is a test, if we pass, we get to keep the planet. This evocative challenge can be applied equally to the conservation of large carnivorous animals. Over the entire surface of the globe, these animals, wolves (Canis spp.), bears (Ursus spp.), large cats, sharks, and orcas (Orcinus orca) are fighting a rearguard action for survival. As the world increasingly becomes a handmaiden to the human race, saving these species has become one of the most difficult tests we face in biological conservation. The urgent need to develop and implement strategies to conserve such creatures has led to two approaches, one based on the ecological roles or services played by these species in maintaining biodiversity, and the other on their intrinsic value as a component of biodiversity.

In this volume we probe the relationship between these two approaches and the science underlying them, seeking to understand the relationship between the presence of large carnivorous animals and the conservation of all attributes and components of biodiversity. We specifically address the conservation challenges of these species, whose diets consist mainly or exclusively of large animal prey. As such, they are often in direct and indirect conflict with humans (c.f. Treves and Karanth 2003). We describe the species of interest as large and carnivorous because we are interested in the conservation problems and opportunities posed by species with these characteristics, and not in their taxonomic position per se. We have worked to include studies from the marine as well as the terrestrial realm. We have been only partially successful in this effort; although the loss of large carnivores in the marine realm has been well documented (e.g., Myers and Worm 2003), the effects of such loss have only recently begun to be examined. We were unable to find cases that include reptiles and avian species and welcome further analysis that extends the scope of this volume and its conclusions. The book is intended as a representative review of what is currently known about the relationship between carnivores and biodiversity and how it relates to the conservation of both. It is aimed at practitioners and academics, with a hope that the work of the former can more effectively inform the work of the latter.

This volume is based on a conference convened by the Wildlife Conservation Society (WCS) at the White Oak Conservation Center of the Gilman Foundation in 2003. We invited professionals who had worked on the relationship between large carnivorous animals and biodiversity or had long-term data sets that might be used to examine this issue. WCS is a conservation organization with a mission of conserving wildlife and wildlands using science-based, field-grounded work. As well as operating the world’s largest set of urban zoos in New York City, WCS is engaged in field conservation at over 300 sites in over 50 countries. In many places WCS works with large carnivorous animals and everywhere finds this work complicated.

Conservation organizations and individual conservation biologists have been very effective in drawing attention to the plight of these animals by arguing that they play critical roles in the conservation of their ecosystems. As such, conservation of these species is said to be a prerequisite for achieving larger-scale conservation.

This is a very important claim. As practitioners of science-based conservation, and strong supporters of the value of large carnivorous animals, we thought it a critical time to bring together experts from the scientific community to assess this link. Therefore, we convened the White Oak meeting to address the question: Is the conservation of large carnivorous animals equivalent to the conservation of biological diversity? Aware of the complicated history of ecological thought that addresses versions of this question, we have worked to place our volume in the perspective of larger ecological theory. Also aware of the efforts of others addressing similar questions, we have placed our work in the context of the work of others. We have also found that as we worked on this book the question we had posed to the workshop has proven more complicated than expected to answer, and our efforts to answer it have led us into unexpected quarters (see the concluding chapter). We have found that our search for a simple answer has been frustrated by, but also informed by, the paucity of scientific investigation addressing the role of large carnivores; the conclusion that even when there is sufficient science, the answer will depend on context; and the rich, complicated mix of ethics, values, and science that envelops and obscures virtually everything having to do with the interactions between humans and large carnivores. But despite this ambiguity, we have worked throughout this book to bring to the surface the management implications of and actions connected with conserving both large carnivorous animals and the biodiversity that enrobes them. This constant eye on conservation action makes this book different from many others. We hope this book will be of use to those charged with the conservation and management of both wildlife and wildlands.

The book is organized into four parts. The first part, Setting the Stage, lays out the theoretical and practical issues underlying the question of whether conservation of carnivores is equivalent to the conservation of biodiversity. The two chapters review both the ecological foundations that are at the core of this question and the assumptions and uncertainties underlying the ways in which large carnivorous animals have been used as tools for conserving biodiversity. Part Two, The Scientific Context for Understanding the Role of Predation, consists of six chapters. The first set presents several of the best-known research projects that have examined the ecosystem-structuring role of large carnivorous animals including sea otter (Enhydra lutris)–kelp systems, Lago Guri, Venezuela, and wolves (Canis lupus) of the Greater Yellowstone Ecosystem. The remaining chapters contribute through examining research results from a set of systems less frequently appreciated as central to the topic of this book. These include examining the general phenomenon of trophic cascades and top-down controls in large marine carnivores, the forests of the northeastern United States where large carnivores are gone and ungulates rule the world, and what is known about redundant roles in groups of large carnivores, focusing particularly on the African guilds.

In Part Three, From Largely Intact to Human-Dominated Systems: Insight on the Role of Predation Derived from Long-Term Studies, five case studies are presented by ecologists who have worked on a long-term basis in various systems and provide information essential for determining whether the functional importance of carnivores necessarily means that focusing conservation efforts on them will achieve conservation of biodiversity. Their contributions are arranged from those that examine relatively intact ecosystems to those heavily influenced by humans, in the Russian Far East, African savannas, European temperate forests, tropical coral reefs, and Patagonian Steppe. In Part Four, Achieving Conservation and Management Goals through Focus on Large Carnivorous Animals, five chapters address the practical applications that may be derived from the science of understanding carnivory. These include discussions of how long-term studies on carnivores designed to address management issues can play a role in conserving landscapes and biodiversity, an analysis of whether hunting by humans and hunting by other large carnivorous animals are functionally redundant, and a conceptual framework for assessing whether populations of large herbivores are regulated by top-down or bottom-up processes. The final two chapters in Part Four offer contrasting perspectives on how top carnivores are related to biodiversity conservation in boreal forest ecosystems and the half-full forests of Europe.

This book is not meant to be another book about carnivores. It is intended to be a book about the relationship between carnivores and conservation. All authors were asked specifically to address the conservation implications of their work. The book concludes with Chapter 20, an overall synthesis that draws the conservation implications from the rich mix of chapters, making the point that despite the lack of a simple answer to a complicated question, there are ways to improve our thinking and action to conserve both large carnivorous animals and biodiversity.

There has been a good deal of ecological work done on the impact of biodiversity loss on ecosystem structure and function (Scheffer and Carpenter 2003), with trophic interactions appearing to play important roles in these processes (Worm and Duffy 2003). But there continues to be debate about the relative role of consumer-driven (top-down) versus resource-driven (bottom-up) control, with both appearing to operate at some times, in some systems (Meserve et al. 2003; Sinclair et al. 2003). Yet little of this work has provided tools that would help conservation practitioners in their efforts to conserve biodiversity.

To us, the question, Is the conservation of large carnivorous animals equivalent to the conservation of biological diversity? is a vital one for the conservation community to address head-on. It is fashionable to argue in some quarters that large carnivorous animals are a tool whose presence is required in order to achieve conservation of all components and attributes of biodiversity. And further, this argument states that restoration of this full spectrum of biodiversity is not possible without reintroduction of large carnivorous animals. If this utilitarian approach to large carnivore conservation is correct, then we must be able to prove the vital role played by these species. If it is not correct, then we must proceed with caution (c.f. Warren et al. 1990), for these species may not be necessary (in this utilitarian sense) and, given the negative costs of their presence and the conservative nature of scientific proof, a limited version of conservation success might be easier to achieve in their absence. Difficult though this question is, it exists as a reality in the world of the Designer Ark (Weber in press).

A different, though perhaps complementary, argument for the conservation of large carnivorous animals is value based and draws on the long-intertwined history of humans and these species and the roles they played and play in the human psyche. As Quammen (2003: 13) has written: "For as long as Homo sapiens has been sapient . . . alpha predators have kept us acutely aware of our membership within the natural world. They’ve done it by reminding us that to them we’re just another flavor of meat." The power that large carnivorous animals had over humans is bred in the bone and has resulted in complex accounting of the relationship between the two (Redford and Robinson 2002). Origin myths place humans descended from jaguars (Panthera onca) (Benson 1997) or sharing the same mother as tigers (Panthera tigris) (Wessing 1986). And a common theme is the blurred boundary between the two with lycanthropy, or humans turning into wolves, found in Europe (Otten 1986), echoing beliefs from throughout the world that humans transform into jaguars, pumas (Puma concolor), leopards (Panthera pardus), lions (Panthera leo), tigers, and bears (Boomgaard 2001).

The power of the relationship between humans and large carnivorous animals lies in its ambiguity and blurring of boundaries (Wessing 1986; Benson 1997; Boomgaard 2001). For example, in some of the early European illustrations of the New World—such as a Dutch woodcut published in 1695—there is a conflating of human and jaguar, with the jaguar pictured standing in a human position (Saunders 1990). Large carnivorous animals are symbols of the nonhuman world both within and outside of the human body, as illustrated by the human–lion hybrid, the sphinx. The nature of this relationship between such animals and the nonhuman world is well illustrated in a Javanese tale related by Crawford (1967) (in Wessing 1986):

Make choice of an equal friend, and do not like the tiger and the forest. A tiger and a forest had united in close friendship, and they afforded each other mutual protection. When men wanted to take wood or leaves from the forest, they were dissuaded by their fear of the tiger, and when they would take the tiger, he was concealed by the forest. After a long time, the forest was rendered foul by the residence of the tiger and it began to be estranged from him. The tiger thereupon quitted the forest, and men having found out that it was no longer guarded came in numbers and cut down the wood, and robbed the leaves, so that, in a short time, the forest was destroyed, and became a bare place. The tiger, leaving the forest, was seen and although he attempted to hide himself in clefts and valleys, men attacked him and killed him, and thus, by their disagreement the forest was terminated, and the tiger lost his life.

Undoubtedly, there is no single unifying theory to tell us when the tiger and the forest are locked into this symbiotic relationship. We do not know enough to predict the role of large carnivorous animals in the ecology of every place, time, and circumstance. We must therefore be careful not to assume that we know when and where such species must be conserved in order to conserve other components and attributes of biodiversity. Their existence is worth more than just the role they play in ecosystems.

In an evolutionarily abrupt turning of the tables, humans are now responsible for the survival of large carnivorous animals. Will Quammen (2003) be correct in his prediction of the year 2150 as a probable end point to the special relationship between humans and alpha predators? We certainly hope not. Boomgaard (2001) recounts early Dutch reports from Indonesia documenting the existence of a kind of tiger called the volgtiger, literally a following or attendant tiger or a familiar. The concept of a familiar (meaning a spirit, usually taking the form of an animal but also a close friend or companion) that helps someone (often a witch) is apt in this context. Large carnivorous animals are a part of humans and of our past. But they are also a test of our humanity and of our ability to save the earth. Perhaps it is true, as told in a Colombian indigenous myth, that the jaguar was sent to the world as a test of the will and integrity of the first humans (Davis 1996). If we are to save ourselves, we must save all the parts of our humanity. As go these wild animals, so goes the human soul.

PART I

Setting the Stage

There are both theoretical and practical aspects to the question regarding the relationship between large carnivorous animals and biodiversity posed by this book. Although the link between the two is often acknowledged, conservation scientists and practitioners have generally remained in two separate camps. Scientific inquiries examining the role of large carnivorous animals in structuring biological communities do not generally delve into how the science translates into practical terms. By the same token, practitioners utilizing large carnivores as tools to increase the efficiency of attaining conservation goals do not often probe deeper into the labyrinth of exceptions and uncertainties that form the scientific basis of the work.

This introductory section is composed of two chapters that lay out the theoretical and practical foundations of this topic. The first, by Robert Steneck (Chapter 2), provides a theoretical framework for exploring the ecosystem role of large carnivorous animals. Although most research on this topic has focused on small-bodied predatory animals in relatively closed systems, there is a strong theoretical basis for extending many conclusions from this work to large-bodied predators. Large carnivores can affect local and regional biodiversity, but it is important to consider the conditions that might be necessary for their influence on ecosystem properties to be strong. Such questions are central to the scientific basis for conserving biodiversity. Justina Ray (Chapter 3) takes a first step in considering the conservation applications stemming from the growing body of research on the relationship between large carnivorous animals and biodiversity. Although a substantial shift in attitude toward top predators from obstacles to instruments for achieving conservation goals has enabled their increasing use as centerpieces of conservation strategy, the assumptions behind this use have undergone little scrutiny. Dr. Ray examines the rationale and underlying assumptions that characterize conservation tools that have been developed using large carnivores in both terrestrial and marine settings.

Together, these two chapters set the stage for the remainder of the volume in which the scientific context and practical implications of the role of large carnivores in conserving biodiversity are explored in finer detail.

CHAPTER 2

An Ecological Context for the Role of Large Carnivores in Conserving Biodiversity

Robert S. Steneck

How important are large carnivorous animals for conserving biodiversity? Today they are rare or absent from most terrestrial, aquatic, and marine ecosystems. Should we invest heavily in political and real capital to restore them? These questions require that we understand their ecological roles in ecosystems. However, most studies are too limited in scope to provide answers to such broad questions. One way around this lack of data that would allow a more holistic perspective is to apply ecological theory to help sort out which concepts are most appropriate, most compelling, and most robust.

Most general ecological concepts begin with observations made in nature. Fortunately, people have always been keen observers of predators. Cave paintings in France made 35,000 years ago depict large carnivores stalking prey; some of which were humans. Obviously, our preoccupation with carnivores is primal, and it has resulted in a wealth of knowledge about them and their effects. However, over time perceptions of their roles in ecosystems have changed and, accordingly, observations could have been colored by prevailing dogma and existing social and scientific paradigms. Therefore, we can better understand contemporary concepts by knowing their conceptual history.

Considerable empirical and theoretical ecological research supports the thesis that large predators can affect community structure and biodiversity. It is less well known under which conditions predators do exert, or could exert, major influences on the structure and functioning of ecosystems. Today, relatively few ecosystems have large predators that play important roles, often due to extirpations induced by hunting and habitat change. But it is also possible that some habitats and ecosystems never had ecologically significant large predators. Under what conditions should we expect predators to be important regulators of biodiversity? Obviously there is no point in trying to restore large carnivores if they were never major players in the system.

The present is often not the key to the past because baselines for most structuring processes have slid so far that they may now be unrecognizable from their former selves and, worse, they may be unrestorable. This is not wholly a question for science because it depends on how we weigh human values relative to other conservation values. To help sort out what we can do from what we should do, this chapter considers the effects of large carnivores on ecosystem biodiversity within the context of contemporary ecological theory. Specifically I will review the origin and evolution of ecological theories that led to our current understanding of effects of large carnivores on biodiversity. I will consider under what conditions and in which ecosystems predator impacts are greatest, and where and when those impacts translate to lower trophic levels. That is, where predator impacts affect the biodiversity of the entire ecosystem. Finally, I will discuss the seductive nature of predator baselines that have been sliding for centuries and causing generations of people to redefine what we see as natural.

The World Is Green Revolution

In 1960 Hairston, Smith, and Slobodkin (HSS) wrote a deceptively simple paper entitled Community structure, population control and competition (Hairston et al. 1960). This may be the ecologist’s only parallel to Albert Einstein’s thought experiments in physics, since both were based entirely on a logical interpretation of the world as understood by the authors. HSS argued that, since the world looks green, it is not overgrazed by herbivores. They pointed to coal deposits as an indication of accumulating plant matter over geological time. If herbivores are seldom food-limited, they are most likely to be predator-limited. Thus they concluded that density-dependent processes regulate carnivores at the top of food webs and producers at the base, but density-independent processes (i.e., carnivory) regulate the herbivores in the middle.

The context and the consequences of the HSS paper are underappreciated today. At the time the paper was written, density-dependent processes (specifically competition) were thought to be the primary processes structuring natural populations. This idea was championed by the luminaries of the day, including Andrewartha, Birch, Hutchinson, and MacArthur. HSS elegantly sensitized ecologists to interactions among trophic levels as well as to the fact that different processes may act at different trophic levels. They also pointed out that predators at upper trophic levels might control the distribution and abundance of consumers at lower trophic levels.

The HSS paper evoked numerous responses and stimulated several avenues of research that are still actively pursued (e.g., Terborgh, this volume). Arguably, their paper defined how we now consider biodiversity (Box 2.1). It also spawned contemporary concepts such as the intermediate disturbance hypothesis, trophic cascades, top-down forces in food web structure and facilitation, and positive and indirect interactions. To understand contemporary concepts on the role of carnivory in preserving biodiversity, it is useful to appreciate where these ideas originated and how they have been shaped over the past several decades.

Paradigm Gained—Predation as a Structuring Process

A year after HSS’s publication, Connell (1961b) wrote a paper entitled: "The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. This classic paper demonstrated the importance of competition in determining dominance among intertidal barnacles. There was no reference to HSS, although Nelson Hairston was thanked for critiquing it. Nevertheless, one of the other factors" in the title turned out to be predation by snails. Connell showed that only in the absence of their predator did carpets of barnacles (i.e., prey) cover the rocks and intensely compete with each other for space. From this, he and others generalized that the intensity of competition varies inversely with the intensity of predation. That conclusion may have initiated a paradigm shift that focused on the role of consumers in structuring natural communities.

A few years later Paine (1966) (Fig. 2.1) observed that intertidal mussels and a few other herbivorous, suspension-feeding animals form monopolies that will outcompete other organisms unless they get eaten by predators. He advanced the hypothesis that: Local species diversity is directly related to the efficiency with which predators prevent the monopolization of the major environmental requisites by one species (Paine 1966: 65).

Box 2.1

Defining Biodiversity (a brief history)

A simple question posed by G. Evenly Hutchinson (1959), [W]hy are there so many kinds of animals? is difficult to answer because it can be interpreted in so many ways. How do we count the many kinds of animals? Today, biodiversity has ecological and evolutionary connotations. The term itself has evolved from Hutchinson’s focus on simply the number of species found at a site (now called species richness) to what is now almost synonymous with some definitions of ecology.

Today’s term biodiversity has its roots in the phrase species diversity. Beginning in the 1940s, information indices were developed to go beyond simple lists of the number of species (i.e., species richness or species density for a fixed area) to integrate taxa abundance. These indices were useful to demonstrate that a biota was dominated by few species (e.g., Simpson’s dominance index, Simpson 1949), or if they were more evenly distributed (e.g., the Shannon diversity index or index of evenness, Shannon and Weaver 1949). Information indices, however, had their critics too. They were thought by some to represent hyperreductionism that was generating index numbers to complex natural communities. Statistics could not be performed on them; they could not be arranged in linear order along a diversity scale (see Hurlbert 1971). What did it mean if a peat bog in New England had the same diversity index as a mudflat in New Zealand?

Ecologists focusing on the structure of communities and ecosystems (e.g., Whittaker 1975) saw species diversity (defined as the abundance-weighted distribution of species) as being useful if it was grouped by habitat type. Whittaker identified within-habitat diversity (alpha diversity), as different from between-habitat diversity (beta diversity) and different still from the entire pool of species within a region (gamma diversity). Although the latter results from evolutionary processes over geological time, the former results from habitat and process-driven ecological differences (Huston 1994). Thus patterns in species diversity may reflect regional or local pools of species but not necessarily processes driving those patterns. Some ecologists saw this as a serious shortcoming. They were particularly frustrated that interspecific interactions were not part of the species diversity indices (Hurlbert 1971).

The term biodiversity is a contraction of the phrase biological diversity (Wilson and Peter 1988). It was intended to encompass all scales of diversity from genomic to species, populations, communities, ecosystems, and landscapes. Significantly, it also includes ecological interactions among the species (Huston 1994). Thus, Biodiversity refers to the natural variety and variability among living organisms, the ecological complexes in which they naturally occur, and the ways in which they interact with each other and with the physical environment (Redford and Richter 1999: 1247). Although this is close to many definitions of ecology, it is commonly used by conservation biologists and it allows us to consider holistically the relative role of large carnivorous animals as they interact with species at lower trophic levels all in the context of the ecosystem’s other physical and biological processes.

Monopoly busting was only one part of the species diversity story. As predation pressure or other forms of disturbance increase to high levels, fewer species can persist. Therefore, at very low or very high levels of disturbance, diversity is low; so it follows that the highest diversity will be between those two extremes. This became known as the Intermediate Disturbance Hypothesis (its origins have been attributed to several papers; Paine and Vadas 1969; Connell 1978; Lubchenco 1978; Huston 1979), and it established a strong link between the process of predation and the local species diversity (see Fig. 2.1).

By the 1970s, community ecologists were becoming increasingly convinced that predators could control community structure. Some viewed this paradigm shift from competition-based to predator-based control of community structure as revolutionary. After all, niche theory was founded on the notion that animal communities appear qualitatively to be constructed as if competition were regulating their structure (Hutchinson 1957: 419). However, Hutchinson himself in his Homage to Santa Rosalia (Hutchinson 1959) suggested diversity relates at least partially to the trophic organization of the community. He cited Odum’s (1953) textbook treatment of trophic structure (called a predator chain) and remarked the lowest link is a green plant, the next a herbivorous animal, the next a primary carnivore and the next a secondary carnivore, etc. Hutchinson (1959: 147). Hutchinson further pointed out the assertion of both Wallace (1858) and Elton (1927) that food webs were constructed such that the predator at each level is larger and rarer than its prey Hutchinson (1959: 147). This pattern became known as the Eltonian food pyramid and it developed into the field of trophic-dynamics (sensu Lindeman 1942). In this view, each trophic level is successively dependent upon the preceding level as a source of energy (Lindeman 1942: 415). In other words, the primary interactions resulted from lower tropic levels fueling those at the top. Today, this is called bottom-up control of community structure (sensu Power 1992). What Hairston et al. (1960) proposed to the world was decidedly different. Rather than resources at lower trophic levels fueling higher trophic levels (bottom up), consumers at higher trophic levels limit the abundance of lower trophic levels (top-down) (sensu Power 1992). This paradigm shift is much more than changing terminology. While predators had long been considered part of natural communities, they had been thought of as passengers carried by the resources available in the ecosystem. What had been underappreciated was that predators could be drivers of the system by limiting the abundance of their prey. This new way of thinking opened new avenues of ecological theory focusing on the communitywide impacts of higher-order predators on organisms at lower trophic levels.

Figure 2.1

The intermediate disturbance hypothesis (after Lubchenco 1978). Maximum species diversity falls between the extremes of no predation where one or a few competitively dominant species thrive and high predation pressure where only a few predator-resistant species persist.

Top-Down Forces in Food Webs: Keystones to Trophic Cascades

The top-down manner by which predators drive the structure of ecosystems was illustrated in several compelling studies published in rapid succession, beginning in the mid-1960s. These early studies from widely divergent ecosystems all found that a single predator can control the distribution, abundance, body size, and species diversity of all other species in the system. One classic example came from Robert Paine who had been a student of Frederick Smith (the first S of HSS). Paine observed that the intertidal sea star, Pisaster ochraceus, controlled the abundance of the competitively dominant large mussel, Mytilus californianus in the Pacific Northwest (Paine 1966). Without the carpet of mussels, a variety of algae and other organisms flourished. About the same time Brooks and Dodson (1965) observed in freshwater lakes that a planktivorous predatory fish, the alewife (Alosa pseudoharengus), consumed most large herbivorous zooplanktons, thereby allowing small, nonpreferred, competitively inferior species to thrive. Finally, in the Aleutian Islands of Alaska, the sea otter (Enhydra lutris) was shown to control the distribution and abundance of herbivorous sea urchins, which in turn control kelp forest development (Estes and Palmisano 1974; reviewed in Estes, this volume). In all of these examples, a single predator affected the entire community by removing either a dominant spatial competitor or a dominant herbivore. Thus the larger impacts resulted from a release of ecological control by those competitors or herbivores.

Single species that greatly affect communities but constitute only a low proportion of the community biomass are called keystone species (sensu Paine 1966; Power et al. 1996; Fig. 2.2). Most widely recognized keystone species are apex predators, such as the sea stars and sea otters (already described), large predator snails (Concholepas concholepas) (Castilla and Paine 1987), and freshwater bass (Power et al. 1996). Curiously, these and most examples of keystone predators are from either marine or freshwater aquatic ecosystems. It could be that terrestrial predator–prey interactions are more difficult to observe because they play out over much larger areas and over a much longer period of time. Nevertheless, the effects of keystone species can become evident when they are reintroduced to isolated terrestrial ecosystems such National Parks (Ripple and Larsen 2000; Berger et al. 2001a; Berger and Smith, this volume) or islands. For example, fluctuating population densities of wolves (Canis lupus) on an island in Lake Superior control the abundance of moose that in turn control the abundance of the island’s balsam fir (Abes balsamea) trees (McLaren and Peterson 1994).

Keystone species need not be carnivores, but most are, because of the stipulation that they have a great impact at low abundance. Other species that have a large impact but are abundant in the system are called foundation (Dayton 1975; Soulé et al. 2003) or just dominant species (Power et al. 1996; see Fig. 2.2). Often, herbivores rather than carnivores were the dominant species in ecosystems. Although large herds of wildebeest and other ungulates in Africa (Sinclair and Norton-Griffths 1979) or sea urchins in numerous shallow marine habitats (Steneck et al. 2002) control the structure of lower trophic levels (i.e., plant communities), they do so by the brute force of numbers and as such they do not qualify as keystone species (see Fig. 2.2).

Usually there are relatively few carnivorous species at the highest trophic levels. These apex predators are so named because no predator controls their abundance (i.e., they are resource limited according to HSS). Thus all keystone predators are apex predators, but the reverse is not true. There are relatively few keystone predators in the world. They are rare or absent from most highly diverse ecosystems. Arguably, it is immaterial whether a single or several predators are controlling prey densities. What matters most is that carnivores at or near the top level control consumers at lower trophic levels, thus creating ripple effects throughout the food web.

Robert Paine pointed out that strong interactions by consumers cascade through the community, transmitted by a chain of strongly interacting links (Paine 1980: 674). Such trophic cascades(Paine 1980: 676) result from the top-down control of consumers on their prey. If prey are themselves strong interactors, then their prey, at yet lower trophic levels, are also affected. In this way top-down impacts cascade from apex predators to primary producers. Typically, trophic cascades must show inverse patterns of abundance between a consumer and its prey across more than one trophic level in a food web.

Figure 2.2

Keystone and dominant species. Their functional importance of species relative to their abundance (after Power et al. 1996, with permission—Copyright, American Institute of Biological Sciences). Important species include Pisaster (P), sea otter (O), the Chilean predatory whelk Concholepas (C), sea urchins (U), trees (T), kelp (K), grass (G), and reef-building corals (Cr). Note keystone species are only those with a high impact relative to their abundance.

Variability of Trophic Cascades

HSS described a hypothetical trophic cascade that became the classical standard: predators regulate herbivores allowing edible plants to be limited only by resources available to them. This was a food web of three trophic levels (i.e., an odd number). However, food webs can have fewer or more than three levels (Morin and Lawler 1995). If top-down forces dominate the system, then a higher-order apex predator representing a fourth trophic level will effectively control the predators of herbivores (Fig. 2.3). In even-numbered trophic-level food webs with four or more levels, herbivore populations can expand and overgraze plant communities (Fretwell 1977, 1987; Oksanen et al. 1981). An excellent example is the tri-trophic sea otter/sea urchin/kelp system described by Estes (this volume; Estes and Palmisano 1974; Estes and Duggins 1995). The sea otter in Alaska is a reintroduced apex predator that controlled sea urchin population densities until otter-eating killer whales (Orcinus orca) entered this coastal ecosystem in the 1990s. The addition of this fourth trophic level eliminated sea otters, causing herbivorous sea urchin populations to explode and denude kelp forests over vast areas (Estes et al. 1998). This also illustrates the context-dependent nature of top-down controls (Pace et al. 1999). With the unprecedented attacks on sea otters by killer whales, beginning in the 1990s, sea otters lost their status as the system’s apex predator.

Figure 2.3

Number of trophic levels and effects. Large arrows indicate large effect from strong interactors. Small arrows indicate small effect. Relative abundance of organisms in any given trophic level is indicated by the circle diameter. A indicates apex predators.

Not all predation from upper trophic levels cascades to lower levels. There are several reasons for this attenuation (sensu Schmitz et al. 2000). Edibility (that something can be consumed) and palatability (that something edible is chosen to be consumed) control what is eaten. For example, many woody mature plants are inedible, so changes in herbivores will have little immediate impact on them. In contrast, saplings are usually edible and thus are more likely to show the effects of herbivory.

Megaherbivores can grow to a size at which they are relatively inedible and thus immune to apex predators (Owen-Smith 1988; Sinclair et al. 2003). Herbivory can weed out the most edible and palatable plants from a community, leaving plants that are avoided or impossible to eat. Thus the world can be green and herbivores could be trophically limited (reviewed in Terborgh, this volume).

Even among undefended prey in highly diverse ecosystems, the effect of the predator and herbivore guilds can become so diffuse that their per capita impacts become very low (Duffy 2002). In that case, loss of a species may have modest or undetectable communitywide implications. Attenuation of top-down effects in highly diverse ecosystems caused some to question whether trophic cascades are important or even possible there (Strong 1992; Polis and Strong 1996). Most examples of cascading effects are at the community level (e.g., those already described here), but there are also cases where top predators strongly affect one or two species but because those species are either not strong interactors or constitute a small fraction of the community, there is little or no communitywide impact (Polis 1999). It is possible that greater prey diversity can reduce the penetration of trophic cascades beyond one trophic level (Duffy 2002). However, explicit tests showing biodiversity per se can attenuate trophic cascades are generally lacking (Duffy 2002). Classic marine examples of trophic cascades are primarily confined to ecosystems with naturally low biodiversity (e.g., kelp forests of Maine and Alaska; Steneck et al. 2002) or more diverse systems that have lost functional diversity (e.g., Caribbean coral reefs due to overfishing and disease; Hughes 1994). Nevertheless, even some highly diverse ecosystems have been shown to have trophic cascades (e.g., Pace et al. 1999; Terborgh et al. 2001).

Perhaps the more important question about variability of trophic cascades is why they are so evident in some ecosystems but not in others. There are at least four factors that can diminish the expression of strong trophic cascades in which predator effects conspicuously translate to change among plants or other basal trophic level organisms within the community. They include, in increasing order of importance: (1) compensatory community changes initiated by top-down forces that result in an environment hostile to herbivores, (2) poorly defined trophic structure resulting from widespread omnivory that blurs trophic-level distinctions and functions, (3) reduced interaction strengths of predators or herbivores due to reduced consumer body size from biogeographic or anthropogenic effects, and (4) environmental regulation of interaction strengths via physiological stress. Following here, I will describe each of these factors.

Compensatory Community Changes

Several compensatory mechanisms can dampen or eliminate trophic cascades (Pace et al. 1999). These include changes to ecosystems that reduce the effectiveness of consumers. Some good examples come from shallow marine ecosystems where the vegetation responds quickly to changes in herbivory. Predator-induced declines in herbivores can result in rapid increases of macroalgae that change habitat architecture, creating a predator-free refuge for small herbivores and other mesopredators in which to hide from visual predators (Hacker and Steneck 1990; McNaught 1999). Similarly, some distasteful, toxic or heavily armored plants that are avoided by consumers create an effective defense for organisms closely associated with them (Hay 1986; Pfister and Hay 1988; Bruno et al. 2003). On coral reefs, reductions in herbivory resulting in marked increases in vegetation reduce the susceptibility of the plant community to subsequent herbivory (McClanahan et al. 2001b). Thus consumer-driven changes to ecosystems can reduce the effectiveness and impact of other consumers that drive local trophic cascades.

Poorly Defined Trophic Structures

Consumers’ structuring effects on food webs can be difficult to characterize because trophic levels can be hard to define. Omnivores and detritivores are ubiquitous and can switch facultatively among trophic levels, resulting in trophic interactions that are more reticulate than those classically structured into distinct trophic levels (Polis 1999). This is well known and was specifically addressed by Paine (1980) when he first described trophic cascades. He was clear to dispel the notion that the study of food webs and community structure possesses the crisp determinism of physics. He considered food webs as idealized local abstractions or nontrivial determinism (sensu Pascual and Levin 1999) of dynamic, complex trophic interactions. Thus these and many other food web studies do show communitywide effects on lower trophic levels from functionally distinct higher trophic levels, even if their exact placement in the food web remains unclear. Further, Menge and Sutherland (1976) suggested that it matters less in the abstraction of food webs if consumers eat meat or plants because, in most systems, larger consumers eat smaller species. Because apex predators are often the largest consumers in the community preying on smaller carnivores, omnivores, and herbivores, the cascading effects resulting from them will vary primarily as a function of their interaction strength.

Reduced Consumer Body Size

The functional role of consumers often scales with their body size. Apex predators that are strong interactors (sensu MacArthur 1972) often initiate the top-down control leading to trophic cascades. Such predators, by definition, have high per capita interaction strength (Sala and Graham 2002). The strength of carnivore effects often relates to their body size (Sinclair et al. 2003). Body size scaling dictates both predatory and competitive dominance (e.g., Connell 1983). Larger predators can eat larger prey. For example, there is a strong linear relationship between terrestrial predators ranging from 10–4 to 10³ kg and their prey ranging from 10–6 and 10³ kg (Peters 1983). Thus large predatory mammals and birds scale to the mass of their prey in the same manner as small predatory lizards, amphibians, and birds. Large predators also consume the widest range of prey sizes (Peters 1983), which magnifies their per capita impact to the structure of the food web. However, body sizes and predation capacity change ontogenetically, resulting in ecological niche shifts (Werner and Gilliam 1984). Slow-growing predators that are hunted or fished may not attain the size necessary to be strong interactors in the community (to be discussed further).

Cascades also vary geographically. Although it is the strong interactors that often define the web’s structure and function (called interaction webs by Menge and Sutherland 1987), those same species may be weak interactors in other parts of their geographic range (Paine 1980; Menge and Sutherland 1987). For example, on the coast of Washington State the sea star (Pisaster sp.) is the keystone predator (e.g., see Fig. 2.2) that limits the abundance of the dominant space competitor, the mussel (Mytilus californianus). In Alaska, those same species are present, but their interaction strength is low and thus neither species is a major player in that system (Paine 1980). In this example, the physical environment regulates the interaction strength of a keystone predator and thus regulates its structuring role in the community.

Interaction Strengths via Physiological Stress

Regional differences in predator effects can vary due to differences in productivity in the system. As was the case in coastal Alaska versus Washington State, predation potential can be controlled by environmental stress (Menge et al. 1994; Fig. 2.4a,b). Such consumer control of prey falls along a continuum of primary productivity of the system (Oksanen 1990). Under environmentally harsh physical conditions, predation becomes less important (Lubchenco 1978). This is obvious in deserts where evapotransporation is more important than top-down trophic cascades in controlling community structure. Where resources are more limited, they become more limiting to species. Under such conditions, competition drives the ecosystem, overriding predation effects (Menge et al. 1994) (see Fig. 2.4). In this view, physical factors and competition drive interactions at the highest trophic levels, but predation becomes increasingly important at lower levels. In systems with high environmental stress, the role of predation in community regulation is low (see Fig. 2.4b). However, some sessile organisms, such as terrestrial and intertidal vegetation or intertidal barnacles and mussels, create positive feedbacks in which abiotic stress is reduced, which increases competition strength (Bruno et al. 2003). Other organisms can reduce predation potential by providing associational defenses or refugia (Fig. 2.4c). Thus many aspects of trophic cascades are context dependent. Although this should give pause to some sweeping generalizations, some of the variability just described is more the exception than the rule. Other factors, such as low environmental stress (i.e., high productivity potential) driving consumption (e.g., see Fig. 2.4), may be much more important to the ubiquity and strength of trophic cascades in benthic marine and aquatic ecosystems.

Figure 2.4

Ecological processes structuring apex carnivores (a), mesopredators, herbivores, and producers in isolation (modified from Menge and Sutherland 1987) (b), and with positive feedback (facilitations) (c) such as associational refugia (defenses) and stress amelioration (from Bruno and Bertness 2001, with permission).

Do Marine Systems Have the Strongest Trophic Cascades?

Trophic cascades described from benthic aquatic ecosystems (both marine and freshwater), often have a higher impact on lower trophic levels (i.e., lower attenuation) than those from terrestrial ecosystems (Fig. 2.5) (Shurin et al. 2002). In fact, the only examples of predator change cascading to the complete denuding of all canopy-forming vegetation are from marine ecosystems (Polis 1999). The sea otter/sea urchin/kelp example of predator control of vegetation-denuding herbivores has many other marine parallels in tropical (McClanahan and Muthiga 1989; Sala et al. 1998) and temperate to arctic (Steneck et al. 2002) systems, whereas the best tri-trophic terrestrial cases have relatively modest vegetational impacts.

Figure 2.5

Attenuation of trophic cascade effects among different biomes (from Shurin et al. 2002 with permission). Changes in predators that result in changes in herbivore abundance are given in the x-axis. Changes in plant abundance that result from subsequent changes in herbivore abundance are given in the y-axis. If predators have relatively little impact in the system, they will be plotted near the origin (bottom right). If there is no attenuation of predator impacts, herbivores will decline and plants will increase in proportion with changes in carnivore abundance. This line of zero attenuation is represented by the dotted line. The solid line represents the average line of attenuation.

Predator effects on herbivores that cascade to plants exist in both marine and terrestrial systems, but the changes in higher-order terrestrial predators translate to relatively modest or undetectable cascading changes to plants (Fig. 2.5). In a review of 60 terrestrial studies, Schmitz et al. (2000) found evidence for trophic cascades in 45 of them. However, the evidence was strongest when measured as injuries to the plants rather than as changes in biomass. They suggested attenuation could have resulted from induced antiherbivore deterrents in grazed plants, which could slow biomass loss.

It is also possible that trophic cascades simply take longer to show themselves in terrestrial ecosystems. Could complete deforestation be occurring in areas where large predators have been extirpated, but it will take centuries to observe it? Several studies suggested historical declines in wolf populations in the Rocky Mountains of North America resulted in increased moose (Alces alces) and