Long-Term Studies of Vertebrate Communities

By Martin L. Cody

This unique book synthesizes the ongoing long-term community ecology studies of fish, amphibians, reptiles, birds, and mammals. The studies have been conducted from deserts to rainforests as well as in terrestrial, freshwater, and marine habitats and provide valuable insight that can be obtained only through persistent, diligent, and year-after-year investigation.
Long-Term Studies of Vertebrate Communities is ideal for faculty, researchers, graduate students, and undergraduates in vertebrate biology, ecology, and evolutionary biology, including ecology, natural history, and systematics.

• Provides unique perspectives of community stability and variation
• Details the influence of natural and other perturbations on community structure
• Includes synopses by well-known authors
• Presents results from a broad range of vertebrate taxa
• Studies were conducted at different latitudes and in different habitats

• Language: English
• Publisher: Academic Press
• Release date: Oct 24, 1996
• ISBN: 9780080535623

Book preview

School

CHAPTER 1

Introduction to Long-Term Community Ecological Studies

MARTIN L. CODY,     Department of Biology, University of California, Los Angeles, California 90095

I. Introduction

I. Contents of the Book

References

I. INTRODUCTION

A. Background

In view of the wide disparity of approaches among ecologists to their science, it is notable that considerable concordance exists in at least one respect: long-term studies are widely regarded as indispensable, and thus no elaborate case need be made for their justification. The potential for change in the populations that make up a local biota and its various assemblages and communities is an accepted fact and is compelling in its many sources. Some of the potential comes directly from variability in abiotic factors, such as climatic trends over the longer term, or in weather variations from year to year in the mid to shorter terms. Some will come indirectly from such abiotic factors, as they influence species’ geographic ranges and lead to expansion or contraction; thus, at the local level some populations are favored at the expense of others as conditions for their persistence or increase are enhanced. Yet other factors for long-term change are strictly intrinsic to organisms, as they adapt and evolve and thereby occupy ecological roles in a specific area or at a given site that change or shift over time. And overall, inexorable and perhaps inevitable, is the increasing pressure on environments and habitats from human activities, providing a potential for biotic change that can be ignored only in very few circumstances. Thus, the nature, causes, and effects of variability in ecological systems can be understood only through extended study over far longer time periods than those typical of many sciences and of most other biological sciences.

Clearly, with a longer research commitment the capacity to resolve shortterm phenomena is enhanced, and the possibilities for discovering and addressing those of longer term are revealed. As Gilbert White, the astute and dedicated natural historian of his southern English parish, remarked in 1779, It is now more that forty years that I have paid some attention to the ornithology of this district, without being able to exhaust the subject; new occurrences still arise as long as any inquiries are kept alive (White, 1877). Calls for the investment of more time, effort, and resources in long-term ecological studies have become much more persistent over the last decade or so. They appear attributable to a recognition of two rather distinct aspects of ecology, the dynamical nature of ecological systems themselves and the uses to which our understanding of the systems might be put.

First, there is the recognition that ecological processes are, innately, of longterm resolution and that change is as justifiably a quality of ecological systems as is permanence or stability. Many sorts of natural perturbation or disturbance are rare and/or aperiodic, and their effects may persist for centuries or longer. And many ecological processes, such as natural succession, the outcome of competition for limited resources, or the cycling of predator-prey systems, may take inordinately long times for documentation let alone resolution, even without persistent disturbance. Thus with an overlay of variability from abiotic factors, from long-term climatic change to shorter term variations in weather, and from inevitably pervasive anthropogenic influences, it is apparent that many ecological questions require an unusual commitment of research effort.

Second is the need for and dependence upon long-term data bases in the management of critical species, habitats, and resources. This need is often promoted by the legal aspects of species conservation and resource exploitation and becomes apparent only when the letter of such legislation is followed. Resource managers, whether of large animals in small national parks, of habitatspecialized kangaroo rats or gnatcatchers facing urban development, or of desert landscapes faced with large-scale mining concerns, inevitably conclude that their data bases are deficient and that their capacity for prediction or extrapolation is limited.

Several indicators of the importance and perceived value of long-term studies are visible. An overview of the strategic or policy aspects of conducting and funding long-term studies is given in Long-Term Ecological Research (SCOPE—Scientific Committee on Problems of the Environment—Volume 47, Risser, 1991), and implicit in the imperatives for monitoring our dwindling resources is the realization that such monitoring efforts will be ongoing, likely to become an essential part of environmental management in perpetuity (e.g., Spellerberg, 1991). Long-Term Studies in Ecology (Likens, 1989) resulted from the 2nd Cary Conference, which identified a critical need for commitment to a longer-term approach and adopted a statement reading, in part:

Sustained ecological research is one of the essential approaches for developing [an] understanding [of] processes that vary over long periods of time. However, to fulfill its promise, sustained ecological research requires a new commitment on the part of both management agencies and research institutions [and] should include longer funding cycles, new sources of funding, and increased emphasis and support from academic and research institutions.

The US/LTER (Long-Term Ecological Research) program, instigated in 1977, addresses these needs (Callahan, 1991), but its aims for representative sites in major North American biomes have yet to be met.

B. Ecological Studies over Spatial versus Temporal Scales

There has been a good deal of attention in the ecological literature paid to the notion of scale (e.g., Wiens et al., 1986, for an overview), which has dual spatial and temporal aspects. However, much of the development of the concept has addressed the spatial rather than the temporal concerns, leading, among other important aspects, to treatment of populations in fragmented habitats as meta-populations (e.g., Hastings and Wolin, 1989) and their associations in the patches as metacommunities (Wilson, 1992). Models that emphasize spatial structure already have a considerable history, beginning with the work of Levins (1969, 1970), Levins and Culver (1971), Horn and MacArthur (1972), and Yodzis (1978); their implications for genetics, predation (Kareiva, 1986), competition and coexistence (e.g., Chesson, 1986), diversity (e.g., Hanski et al., 1993), and conservation (e.g., the MVP—Minimum Viable Population—problem, Gilpin and Soulé, 1986) are profound. Levin’s (1992) perspective of these developments and their Significance is particularly valuable.

A corresponding theoretical development of temporal analogs of models with spatial heterogeneity seems lacking in the literature. In the same way that a greater understanding of communities in patchy and spatially heterogeneous habitats is gained from studies with a broader, more regional perspective than from single, site-specific snapshots, so it is with communities that experience environmental heterogeneity over time. The broader perspective that comes from long-term research contributes to the understanding of how communities cope with year-to-year, and decade-to-decade, variations in conditions at a particular site.

A diagrammatic comparison of temporal versus spatial scale broadening is given in Fig. 1, but the duality of the two scales is exacerbated by their obvious qualitative differences. Logistically, a broadening of perspective in spatial scales can be achieved relatively easily, with adjustments that might take advantage of, e.g., further site replication, wider habitat gradients, a better microscope, or better satellite imagery, but there is not such a ready correction for restricted temporal scales: time alone can serve to broaden a data base in the temporal dimension. Specifically with respect to questions of community structure, stability, and replicability, there are advantages and disadvantages to broadening scales in the spatial versus temporal dimensions, the far-flung versus the long-term approach. By replicating sites within a limited time over the geographical range of a habitat (Fig. 1, left–right axis), more habitat variability is likely to be encountered (front–back axis), and the precision of intersite comparisons reduced. However, questions of convergent and parallel evolution can be addressed, as geographic counterparts may substitute in distant sites, and the effects and influences of species from adjacent (different) habitats (spillover effects, mass effects; Shmida and Ellner, 1984, Cody, 1993, 1994) can be evaluated, because throughout a habitat’s range the extent and type of other habitats will likely vary. The long-term approach will lack these potential advantages, but also lack the disadvantage of establishing replicability (e.g., in terms of habitat structure) over multiple sites.

FIGURE 1 Diagrammatical representation of two different views of extended community-level studies, contrasting the far-flung approach, in which sites are replicated over space, with the longterm approach, in which a Single site is studied over an extended time period. See text for discussion.

However, at a single site studied over a longer term, site stability is by no means assured; indeed, site-specific variation over time can be dramatic and, if calibrated, can allow for the investigation of covarying community structure. Such attributes as density compensation, shifts in habitat use or territory preference, and changes in interspecific interactions can thus be studied as the site undergoes year-to-year variation in resource availability and community composition.

In general, the changes that occur within a site over the course of a long-term study are among the more formidable challenges faced by this type of research. The site may change year-to-year as a consequence of, e.g., natural variability in weather patterns or may show long-term trends in response to broader scale climatic patterns or factors related to such variation, as discussed above. Figure 2 shows examples of four of the more obvious patterns in within-site variation over time, in which variation is random y–y with a modest range (Fig. 2a) or with a more conspicuous range (Fig. 2b), in which a unidirectional change occurs over time throughout the study (Fig. 2c) or in which conspicuous perturbations occur periodically and from which the site gradually regains its initial (quasi-equilibrial) value (Fig. 2d). Few long-term studies at a fixed site are expected to be free from one or another of such variations or changes, and the monitoring of the site ideally, even necessarily, would encompass both the consumers (in the community of interest) and their site-based resources. Furthermore, the distinction between data sets that have basically chaotic versus nonchaotic or deterministic generators, whether in time series data of resource variations or of consumer populations that are purportedly tracking them, may be extremely subtle, as simulations have shown (Ellner and Turchin, 1995).

FIGURE 2 Synthetic patterns of long-term environmental variation (ordinate) over time (ab-scissa) at a hypothetical site of interest. Random variation (a) with a modest range and (b) with a more pronounced range among years. (c) An obvious trend for increase over time in the environmental parameter measured; (d) periodic perturbations of differing severity (at arrows) followed by asymptotic return, over time, to the original status quo.

The extent to which the community is an open rather than a closed system will potentially affect studies at both broad spatial and temporal scales, but is more likely a handicap in long-term studies. If the community disbands and reassembles with a seasonal tempo, then factors operating beyond the study site may be important but largely unknown. For example, factors affecting overwintering survival may affect the densities of migrant birds in their breeding communities, or those influencing adult salamander survival in woodlands could constrain populations at the breeding ponds. These and other aspects of longterm studies are recurrent themes in the chapters assembled in this book.

C. Studies of Vertebrate Communities

The chapters of this book represent a diversity of approaches to longterm ecological studies, but they have in common a community rather than a populational perspective, and they address in particular communities of vertebrates. Few of the authors become embroiled in the notion of what constitutes a community; rather, each author adopts de facto his or her own working version of Charles Elton’s (1927) view of community, as distinct from population, studies as those addressing a larger fragment of nature… of at any rate a group of ecologically inter-related species. Some authors, notably those who perforce sample sight unseen from below the waterline of rivers and lakes or who trap at population confluences near breeding ponds, choose to talk of assemblages rather than communities, but all avoid extending a debate on definition that has generally proved of no value to the field.

Our restriction of contributions to studies of vertebrates does have implications for community ecology because it is traditionally among vertebrates that we expect to find several of the properties prerequisite for a meaningful community ecology: the operation of density-dependent factors in population regulation; the involvement of larger, often longer-lived and wider-ranging animals and of generalized foraging habitats (whether herbivores or carnivores) and therefore a potential for interactions among species of a competitive nature; and the sort of proactive ecological life styles that come with the higher-order behavior typical of vertebrates, including perhaps acoustical, olfactory, or visual recognition. Such behavior in turn leads to habitats being evaluated and choices made, community membership or exclusion being more than a passive aggregation or a random sample, and a more discrete set of questions for ecological determination than would characterize, e.g., the plankton caught in a towed net, the insects gassed from trees, or the weedy plants found in a vacant lot.

II. CONTENTS OF THE BOOK

A. Origins

This book originated after a symposium on long-term studies of vertebrate communities was convened for the Ecological Society of America’s 1993 annual meeting, held in Madison, Wisconsin. Several of the chapters in this book were first presented as symposium papers, and most symposium contributors agreed that a synthesis volume on the topic was timely and valuable. Additional topics were added as other researchers with long-term data sets were identified and recruited to the task. In some cases the chapters present a current synthesis of an ongoing study with a considerable history of publication (e.g., on Lake Opinicon fishes, Chap. 3; Galapagos finches, Chap. 12; and experimental desert rodent communities, Chap. 17). In others, the chapter might represent the first overview of a long-term research project (e.g., on prairie ducks, Chap. 13) or an attempt to relate a body of work that has been treated thus far as separate contributions (e.g., on successional small mammals in Australia, Chap. 15). All authors have addressed similar questions in as far as the specifics of their work allow: What changes are apparent over longer time periods in vertebrate communities in terms of species identities and densities? How important are such changes relative to aspects of constancy in the communities? Is community change related directly or indirectly to identifiable factors operating locally (within-site) or more distantly, with general or specific impact, or is it rather the result of the vagaries of various indeterminate or random processes? And in segregating the determinate from the indeterminate, how important is the longer time span in allowing for, first, the quantification and qualification of the changes and, second, attribution, or at least suggestion, of their causes?

B. Representation by Taxonomic, Regional, and Habitat Diversity

Although the sixteen chapters that follow are similar in their overall goals, particularly in the hope of reaching new or deeper insights into communities through long-term study, they are of course unlike in many respects. Some of the most obvious differences are listed in Table I, and foremost among these is the wide range of subject taxa. The chapters are evenly distributed over the four major vertebrate groups (fish, herpetiles, birds, and mammals), a pleasing equanimity from the taxonomic standpoint, but one implying a diversity in approaches, from sampling or census methodology to the details of foraging ecology or behavioral interactions within and among species. The relative constraints and advantages to research on different taxa have obvious implications for the type of community studies that are feasible.

TABLE I

Synopsis of Chapter Characteristics

Note. Entries in the table are chapter numbers.

Although marine fishes can be directly counted along transects or within patch reefs by divers (Chaps. 2 and 4), freshwater fish communities are studied largely by netting samples taken from murky waters (Chaps. 3 and 5); foraging habits are derived from stomach contents and deductions based on the structure of fishes’ trophic equipment. Whereas most lizards are diurnal and countable, and their foraging behaviors directly measurable (Chap. 8), most salamanders are cryptic and/or nocturnal (Chaps. 6 and 7), and, along with turtles (Chap. 9), their numbers are best measured when they congregate at aquatic habitats for breeding purposes. Their ecology in the terrestrial environments, where most of their growth and survival (or otherwise) occurs, is less tractable and remains less well known. Birds are not only diurnal and active but often colorful, noisy, and obvious, and they are thus easily counted; their activities can be plotted, usually sex-specifically, and a good deal of behavior relevant to resource acquisition and partitioning is then built into an evaluation of the community (Chaps. 10–13). Like lizards, birds can be, and often are, caught and recaught, tagged, weighed, and measured, and must rival lizards (Pianka, 1986; Vitt and Pianka, 1995) as the ideal vertebrate organism for ecological studies. The case is also made, with some justification, that coral reef fishes share many of the same advantages prized by field ecologists (Sale, 1991). On the other hand, most large mammals are scarce, elusive, and/or dangerous. Small mammals have none of these qualities but are nearly invariably nocturnal, and their communities must be assessed by trapping, as in terrestrial rodents and their marsupial equivalents (Chaps. 14, 15, and 17), or netting, as in the volant bats (Chap. 16). Food and foraging habits may be studied occasionally by direct observation, e.g., in the former on moonlit nights, with attached reflectors, or via the contents of convenient cheek pouches, but in bats such information is less directly obtained, by location and location-specific auditory spectra and receiver characteristics or by examination of gut or fecal samples.

The chapters are also diverse in terms of geographic region and collectively span four continents (Table I); around one-half of the studies were conducted in North America, with the remainder reporting on research from the Neotropics, Australia, and Africa. Several chapters include a strong comparative component in which results from different regions are compared. In Chap. 4 communities of coral reef fishes from the Caribbean and Indo-West Pacific regions are contrasted, and Chap. 5 relates the structure of freshwater fish assemblages from similarly seasonally inundated habitats around Venezuelan and Zambian river systems. References are made in Chap. 15 to results on small mammal succession in post-fire vegetation in southeastern Australia and in other fire-prone habitats .worldwide, and in Chap. 10 to other comprehensive work on tropical rainforest birds besides that reported from Gabon, West Africa.

A conspicuous aspect of the diversity of chapter topics is the extremely wide range of habitats in which the long-term research has been conducted. By latitude, the coverage extends from the north temperate (the lake fishes of Chap. 3) to the tropics (rainforest birds and bats, Chaps. 10 and 16; island finch communities, Chap. 12). Both aquatic and terrestrial habitats are represented. Research from aquatic habitats is exemplified by marine fish studies of both temperate (Chap. 2) and tropical (Chap. 4) reefs and by studies of freshwater habitats in temperate lake (Chap. 3) and tropical river (Chap. 5) systems. Several chapters (amphibians, Chaps. 7 and 9; turtles, Chap. 6; prairie ducks, Chap. 13) concern animals that depend on both aquatic and terrestrial habitats; for the amphibians and turtles, the aquatic habitat is prerequisite for reproduction, whereas in one component of the duck communities, it is the terrestrial habitat that provides the nesting opportunities. Desert communities are represented by Australian lizards (Chap. 8) and North American rodents (Chap. 17); grasslands to shrubland, successional or otherwise, by Rocky Mountain birds (Chap. 11), Galapagos finches (Chap. 12), Kansas rodents (Chap. 14), and small Australian mammals (Chap. 15); and forests by tropical bird and bat communities (Chaps. 10 and 16).

C. Other Variables in Long-Term Community Studies

1. Spatial Scales

Apart from variation in geographical locale, the spatial scale of the studies summarized in this volume also varies widely. Marine fish studies have used transect counts in harbors and on natural reefs and counts at isolated patch reefs (Chaps. 2 and 4). Freshwater fish studies were confined to sampling in a single lake and associated streams (Chap. 3) or broad sampling over a river system and its adjacent floodplains and marshes (Chap. 5). A local pond system and neighboring marshes were used in Michigan turtle studies (Chap. 6), broader elevational transects were employed to census Appalachian salamanders (Chap. 7), and a single pond was the focus of the amphibian community discussed in Chap. 9.

The lizard communities of Australian desert habitats (Chap. 8) have been monitored over several study plots of some hectares each within a sandplains region, and in bird community studies, data collection was focused in areally discrete plots that varied in size from 2 (tropical forest birds, Chap. 10) to 1 km² (prairie ducks, Chap. 13) to study plots of a few hectares (shrubland birds in the Rockies, Chap. 11). However, in Galapagos finches, coexisting populations are delineated by the islands on which they occur (Chap. 12). The studies of small terrestrial mammals were conducted on fixed plots from ¼ to 2 ha in size (Chaps. 14, 15, 17), but for the vagile and probably wide-ranging tropical bats (Chap. 16), trapping stations were located at fixed sites, often fruiting trees, on Barro Colorado Island in Gatun lake, Panama.

2. Study Duration

What constitutes a long-term community study is of course open to question and ideally would be related to population turnover rates or the longevity of the member species. Most of the chapters report data collection periods of around two decades (mean: 20.3 years± 5.0 SD), although some are notably longer (e.g., 30 years for Keast’s lake fish studies, 25 years for prairie ducks, desert lizards, and Australian successional mammals). In temperate passerine birds and small mammals, with average life expectancies of at most a few years, a quarter-century study does seem to qualify as long-term, although for some lower vertebrates with reduced metabolic rates and long life expectancies, such as fish and turtles, the same duration is relatively modest. The 20-year turtle study of Chap. 6, for example, is about equivalent to age at first reproduction of these species, but individuals can live at least three times this time period. The details of life expectations and replacement rates would be required to assess the adequacy of the study periods relative to community changes, and in general these details are not available. Further, important but rare events that can drastically alter community structure, such as periodic droughts and fires, might occur with a frequency much lower than every couple of decades, and recovery from such events might be slow. Thus, there is little objectivity in our classification of studies as long-term, and in general we simply mean that lots of data were collected using extended efforts over many years.

3. Community and Species’ Sizes

The communities that are the subjects of these studies are likewise variable in size and in the number of species included in the researchers’ operational definitions of community or assemblage. Some encompass all of the organisms within a broad taxonomic group that can be collected or sampled using a given technique (e.g., netted fishes or bats), whereas others are more narrowly focused on a suite, or guild, of species that are closer in size, ecology, resource use, and behavior (e.g., insectivorous birds or seed-eating rodents). Some indication of the variations in both community size and the size range of constituent organisms is shown in Fig. 3. Note that collectively the studies span two orders of magnitude in community sizes (±3–320) and an even wider range in body size. Even single community studies, such as those on fish, Australian lizards, or tropical bats, can include a range of species that differ in mass by at least 1.5–2 orders of magnitude.

FIGURE 3 Community studies reported in this volume vary widely in the number of component species (abscissa), in the sizes of component species (ordinate), and in the size range of component species both within and among studies. Symbols within the figure refer to chapter numbers.

4. Site Variation over the Duration of the Study

As pointed out above, variations are expected to occur at a study site within the duration of long-term studies, such that the conditions under which the community functions change over time. Weather shifts, newly introduced or extinct species, anthropogenic effects, etc., are some possibilities whose incidence will be experienced with increasing likelihood over extended study periods. These natural or unnatural perturbations can be regarded either as a hazard or as an opportunity to learn more about community organization and resilience.

Not surprisingly, the authors of our chapters report many such incidences of changing conditions during their periods of study. Often cited are unusual weather phenomena, sometimes operating at broad geographical scales, such as El Niño Southern Oscillation events that affect Californian reef fishes (Chap. 2) as well as the food supplies of finches on the Galapagos Islands (Chap. 12). Precipitation patterns are shown to be particularly important to communities of species as varied as sagebrush and willows passerine birds (Chap. 11) and Chihuahuan desert rodents (Chap. 17), and they are more obviously a factor in the breeding or feeding habitats of aquatic species, such as turtles and salamanders (Chaps. 6, 7, and 9), prairie ducks (Chap. 13), and floodplain fishes (Chap. 5). Changes to habitats are not always due to abiotic factors, however; the colonization of willow shrublands by beavers in the Rocky Mountains causes a considerable change in water levels and affects foraging opportunities for several bird species, eliminating some while facilitating others (Chap. 11).

Some of the studies are conducted in habitats undergoing succession, where an assessment of vegetational change and its consequences to the vertebrate inhabitants is very much a part of the research design. Examples in this book are the post-fire successions in the Australian desert Triodia grassland and associated lizard community (Chap. 8) and in the Eucalyptus woodlands of southeastern Australia relative to small mammal succession (Chap. 15). Rodents occupying old fields in Kansas (Chap. 14) are further examples.

Anthropogenic influences are hard to avoid over the longer time periods. Human influence had much to do with introducing an alien fish, alewife (Alosa pseudoharengus), and weedy Eurasian milfoil (Myriophyllum spicatum) to Lake Opinicon (Chap. 3); surprisingly, the former had little effect on the fish community, whereas the latter was a significant contributor to habitat alteration. In the tropical forest bird study (Chap. 10), habitat clearing and road building were major factors in the incursion of second-growth plants and birds, and on Barro Colorado Island, forest recovery from partial cutting 70 years ago has lead to the reduction of bat species associated with secondary vegetation and its fruit and nectar resources.

5. Are the Communities Closed or Open Systems?

Community ecology studies are almost never conducted in (or treated as) completely closed systems, but the extent to which they are not varies dramatically with the site and with the life-history attributes of different subject taxa. At one extreme, reef fish may have widely pelagic larvae whose ambit necessarily is considered outside of the reef-centered community study. The terrestrial life of salamanders and turtles is similarly a black box from the viewpoint of pond-based community studies and one in which much of relevance presumably takes place, as it surely does on the migration routes and wintering grounds of birds studied on their temperate breeding grounds. Communities of the more vagile species, such as birds and bats, will presumably be more open, and community characteristics will be less completely attributable to onsite effects and influences than will be the case for more sedentary species such as lake fish, rodents, and herpetiles. Using desert rodents, Brown and associates (Chap. 17) have proven that it is feasible to constrain the animals in realistically large enclosures and thereby study essentially closed, or selectively closed, communities; but naturally the opportunities for such protocols will be very limited in volant or more vagile taxa.

6. Sampling, Observation, and Experiment

As a final dimension to the research presented in this volume, mention can be made of the nature or methodology of data collection. There has been a long tradition in vertebrate community studies, particularly of terrestrial species, of observation-based study; that is, data are collected by simple observation only, without the necessity for hands-on work with the organisms. This is the case particularly with bird research, although birds can easily be caught in many habitats, banded, and weighed or otherwise manipulated. With taxa that are more easily captured, such as most lizards and small rodents, observational data are supplemented or supplanted with information derived from trapped individuals. In other taxa (e.g., salamanders and turtles), discovery is nearly tantamount to capture, and in yet others practically the only source of information is the trapping data (e.g., in fish and bat communities).

When it is logistically feasible, legal, and morally acceptable, field ecologists like to enhance observational data with those derived from field manipulations and experiments. In general, the possibilities are narrower with vertebrates (in contrast, e.g., with plants and insects) and narrower still with higher vertebrates, and in open communities of higher vertebrates the potential for conducting controlled experiments is severely limited. A useful summary of the topic, aimed at community ecologists, is well presented in Diamond’s (1986) discussion.

Not surprisingly, then, the chapters here rely on field manipulation and experimentation to only a small degree, the more so as emphasis is on the long-term aspects of community constancy and change rather than on the mechanics of community structure and function. The most notable exceptions are the study on desert rodents (Chap. 17), which has a very strong component of controlled experimentation; the old-field rodent study (Chap. 14), where habitat fragmentation is experimentally manipulated; and the study in Chap. 4, where patch reefs are manipulated for settlement patterns in coral reef fishes. Further, there is a tradition of experimentation in studying the community ecology of pond vertebrates, such as fish and amphibians. Clearly, variation in data collection techniques is heavily taxon-dependent and is well reflected in the following chapters.

References

ICSU, Chapter 2 Callahan, J.T., Long-term ecological research in the United States: A federal perspective. Risser P.G., ed. Long-term Ecological Research: An International Perspective. SCOPE. Wiley; 1991;vol. 47:9–21.

Chapter 14 Chesson, P.L. Environmental variation and the coexistence of species. In: Diamond J.M., Case T.J., eds. Community Ecology. New York: Harper & Row; 1986:240–256.

Chapter 13 Cody, M.L. Bird diversity components within and between habitats in Australia. In: Ricklefs R.E., Schluter D., eds. Species Diversity in Ecological Communities. New York: Univ. of Chicago Press; 1993:147–158.

Cody, M.L. Mulga bird communities. I. Species composition and predictability across Australia. Aust. J. Ecol.. 1994;19:206–219.

Chapter 1 Diamond, J.M. Overview: Laboratory experiments, field experiments, and natural experiments. In: Diamond J.M., Case T.J., eds. Community Ecology. Chicago: Harper & Row; 1986:3–22.

Ellner, S., Turchin, P. Chaos in a noisy world: New methods and evidence from time-series analysis. Am. Nat.. 1995;145:343–375.

Elton, C. Animal Ecology. New York: Sidgwick and Jackson, London. Ltd.; 1927.

Chapter 2 Gilpin, M.A., Soulé, M. Minimum viable populations: Processes of species extinction. In: Soulé M., ed. Conservtion Biology. Sinauer Assoc.; 1986:19–34.

Chapter 10 Hanski, I., Kouki, J., Halkka, A. Three explanations of the positive relationship between distribution and abundance of species. In: Ricklefs R.E., Schluter D., eds. Species Diversity in Ecological Communities. Sunderland, MA: Univ. of Chicago Press; 1993:108–116.

Hastings, A., Wolin, C.L. Within-patch dynamics in a metapopulation. Ecology. 1989;70:1261–1266.

Horn, H.S., MacArthur, R.H. Competition among fugitive species in a harlequin environment. Ecology. 1972;53:749–752.

Chapter 11 Kareiva, P. Patchiness, dispersal, and species interactions: Consequences for communities of herbivorous insects. In: Diamond J.M., Case T.J., eds. Community Ecology. Chicago: Harper & Row; 1986:192–206.

Levin, S.A. The problem of pattern and scale in ecology. Ecology. 1992;73:1943–1967.

Levins, R. Some demographic and genetic consequences of environmental heterogeneity for biological control. Bull. Entomol. Soc. Am.. 1969;15:237–240.

Levins, R. Extinction. In: Gerstenhaber M., ed. Some Mathematical Problems in Biology. New York: Am. Math. Soc.; 1970:77–107.

Levins, R., Culver, D., Regional coexistence of species and competition between rare species. Proc. Natl. Acad. Sci. U.S.A., 1971;68:1246–1248.

Likens G., ed. Long-term Studies in Ecology. Providence, RI: Springer-Verlag, 1989.

Pianka, E.R. Ecology and Natural History of Desert Lizards. New York: Princeton Univ. Press; 1986.

Risser P.G., ed. Long-term Ecological Research: An International Perspective. SCOPE. Princeton, NJ: ICSU. Wiley; 1991;Vol. 47.

Sale P.F., ed. The Ecology of Fishes on Coral Reefs. New York: Academic Press, 1991.

Shmida, A., Ellner, S.P. Coexistence of plant species with similar niches. Vegetatio. 1984;58:29–55.

Spellerberg, I.F. Monitoring Ecological Change. San Diego, CA: Cambridge Univ. Press; 1991.

Vitt L.J., Pianka E.R., eds. Lizard Ecology: Historical and Experimental Perspectives. Cambridge, UK and New York: Princeton Univ. Press, 1994.

London (first publ. 1779). White, G., Natural History and the Antiquities of Selborne. Thomas Bell; FRS, 1877.

Chapter 8 Wiens, J.A., Addicott, J.F., Case, T.J., Diamond, J. Overview: The importance of spatial and temporal scale in ecological investigations. In: Diamond J.M., Case T.J., eds. Community Ecology. Princeton, NJ: Harper & Row; 1986:145–153.

Wilson, D.S. Complex interactions in metacommunities, with implications for biodiversity and higher levels of selection. Ecology. 1992;73:1984–2000.

Yodzis, P. Competition for Space and the Structure of Ecological Communities. New York: Springer-Verlag; 1978.

SECTION I

Fish Communities

CHAPTER 2

On the Structure and Dynamics of Temperate Reef Fish Assemblages

Are Resources Tracked?

SALLY J. HOLBROOK and RUSSELL J. SCHMITT,     Department of Biological Sciences and Coastal Research Center, Marine Science Institute, University of California, Santa Barbara, California 93106

I. Introduction

II. Methods and Study Sites

III. Results and Discussion

IV. Concluding Remarks

V. Summary

References

I. INTRODUCTION

Understanding processes that influence community structure and population dynamics of marine fishes associated with reefs has long been a cardinal goal of marine ecological research (Sale, 1991b). Historically, competition for reef resources was viewed as a dominant process shaping local patterns of species richness, species composition, and population abundances. This view largely has been replaced over the past decade in recognition of the fact that local assemblages and most constituent populations of reef fishes are open systems whose nature and dynamics are coupled to processes that occur in other locations (Sale, 1991a; Doherty, 1991). This arises from the bipartite life histories of most marine reef organisms: typically the propagule stage (larvae or spores) disperses in the plankton away from the point of birth before settling to the reef environment occupied by older life stages (Kingsford, 1988; Kingsford et al., 1989). The linkage among local populations via exchange of reproductive output is widely believed to decouple the amount of recruitment (here defined as the appearance of newly settled young fishes) to a given local population from the production of young by adults that reside there (Sale, 1991a). This effectively removes one source of local regulation. Current models that describe dynamics of populations and structures of reef fish communities emphasize nonequilibrium processes associated with larval recruitment (e.g., Doherty, 1991; Sale, 1991a).

With respect to abundance, the sizes of local populations are viewed as a balance between external inputs (larval settlement) and subsequent mortality (Keough, 1988; Mapstone and Fowler, 1988; Roughgarden et al., 1987, 1988; Warner and Hughes, 1988; Underwood and Fairweather, 1989; Hughes, 1990; Raimondi, 1990; Doherty, 1991; Doherty and Fowler, 1994; Sale, 1991a). Temporal variation in recruitment can be great and often is attributed to oceanographic processes that transport and deliver larvae (Doherty, 1991). One widely held notion, called the recruitment limitation model, is that reef resources often may not be limiting due to an undersupply of propagules (Sale, 1977, 1980, 1991a; Victor, 1983, 1986; Wellington and Victor, 1985, 1988; Lewin, 1986; Young, 1987; Doherty and Williams, 1988; Doherty, 1991; Doherty and Fowler, 1994; Underwood and Fairweather, 1989; Sutherland, 1990; Stoner, 1990). As a result, mortality after settlement may be largely density independent, and fluctuations in density of older life stages can mirror the pattern of fluctuation in recruitment (Doherty, 1991). Such direct correlations can be obscured by variation in the density-independent mortality rates on reef-associated life stages (Warner and Hughes, 1988) and obviously will be altered by density-dependent mortality (Jones, 1991). Thus far compensatory mortality has not been detected for reef fishes with open populations, although density-dependent growth has been found relatively frequently (e.g., Jones, 1991; For-rester, 1990). Because the duration of field experiments that explore density dependence in reef fishes typically is short, it has not been possible to determine whether compensation in growth rates results in density dependence in an individual’s probability of reaching adulthood. Nonetheless, the general failure to identify strong regulatory mechanisms that operate at the local reef scale reinforces the notion that the dynamics of local reef fish populations are shaped mostly by external forces and less so by local processes. Consequently, local populations of reef fishes are thought to be highly variable and relatively unaffected by local resources and events that occur on the reef.

Given the prevailing view for constituent populations, it is not surprising that current models describing attributes of local reef fish assemblages also emphasize nonequilibrial processes. Indeed, the first model that seriously challenged the resource competition perspective for reef fishes—the lottery hypothesis (Sale, 1977)—addressed aspects of community structure by positing how chance in securing a settlement location by larvae of different species could govern composition of the assemblage. Subsequently, it was recognized that strong recruitment periods can buffer local populations of perennial species (which most reef fishes are) from extinction—the storage effect (Chesson and Warner, 1981; Warner and Chesson, 1985)—and that fluctuations in recruitment could promote local coexistence of competing species with open populations (Warner and Chesson, 1985; Chesson, 1986). Of course, the existence of external sources of new young also means that local extinction of any given species of reef fish is likely to be temporary (Sale, 1991a; Holbrook et al., 1994). Further, widely fluctuating recruitment that is asynchronous among species will result in an assemblage whose structure is highly transient (Sale, 1991a). Thus, such community-level aspects as species richness, composition, and relative abundances of local reef fish assemblages are thought to be shaped primarily by recruitment history and are therefore unlikely to remain constant through