Info

Average

Average

Giving-up

Species

First visit

Last visit

visit

patch

density

G. a. allenbyi

2.61 h

8.95 h

5.82 h

0.75 g

0.506 g

G. pyramidum

1.63 h

6.54 h

4.02 h

1.42 g

0.734 g

Source: After Kotler, Brown, and Subach 1993.

Note: The table lists the average times of the first visit by each species in resource patches, the last visit of the night, and average time of a visit (all in units of hours after sunset). The table also lists the average value of a patch (in units of grams of millet seeds in the patch) and the giving-up density (in units of grams of millet seeds left in the patch).

Source: After Kotler, Brown, and Subach 1993.

Note: The table lists the average times of the first visit by each species in resource patches, the last visit of the night, and average time of a visit (all in units of hours after sunset). The table also lists the average value of a patch (in units of grams of millet seeds in the patch) and the giving-up density (in units of grams of millet seeds left in the patch).

the night and collected one tray for analysis every 90 minutes. This technique provided snapshots of gerbil activity and patch depletion during the night. The faster forager, G. pyramidum, started foraging earlier in the night than G. a. allenbyi, but G. pyramidum also stopped foraging earlier (table 12.1). More importantly, G. pyramidum encountered richer resource patches, on average twice as rich as those G. a. allenbyi encountered. In contrast, the more efficient G. a. allenbyi extracted 0.25 g more seeds from each patch (table 12.1; Kotler, Brown, and Subach 1993). Thus, this pair of gerbils partitions nightly seed resources temporally. Each species biases its activity toward times of the night and resource densities in which it is the superior competitor.

In the second experiment, the researchers used fenced enclosures to test for interference (Ziv et al. 1993). They created experimental communities that differed in the presence or absence of G. pyramidum, and then recorded the intensity and timing of gerbil activity using sand tracking. When both species were present, the gerbil species showed dramatic temporal partitioning (fig. 12.5), with G. pyramidum dominating the early hours of the night and G. a. allenbyi the hours toward dawn. In the absence of the larger G. pyramidum, however, G. a. allenbyi expanded its activity to include all hours of the night. Indeed, the level ofactivity it achieved early in the night without a competitor was higher than previously observed in the presence of G. pyramidum. Thus, the small gerbil species compensated for all the "missing" G.pyramidum activity caused by the larger species' absence.

The following picture emerges. When the gerbils emerge from their burrows in the evening, they find rich patches of seeds created by afternoon winds, but as the night wears on, their foraging necessarily reduces the quality of these patches. G. pyramidum feeds quickly and aggressively outcompetes G. a. allenbyi in the rich patches of the early evening. The more energetically efficient G. a. allenbyi can extract more from each patch and outcompetes G. pyramidum in the depleted patches that occur late at night.

Figure 12.5. Timing of nightly activity for two species of gerbils, (A) Gerbillus andersoni allenbyi and (B) G. pyramidum, in experimental communities. G. pyramidum always forages early in the night, when seed resource patches are rich. When these species co-occur, they display temporal partitioning, with G. a. allenbyi foraging later in the night than G. pyramidum and thereby experiencing poorer resource patches ("control" communities and "GP present—enclosures" communities). When G. a. allenbyi lives without G. pyramidum ("GA alone" communities), it expands its time of activity to include the earlier hours ofthe night. (After Ziv etal. 1993.)

Figure 12.5. Timing of nightly activity for two species of gerbils, (A) Gerbillus andersoni allenbyi and (B) G. pyramidum, in experimental communities. G. pyramidum always forages early in the night, when seed resource patches are rich. When these species co-occur, they display temporal partitioning, with G. a. allenbyi foraging later in the night than G. pyramidum and thereby experiencing poorer resource patches ("control" communities and "GP present—enclosures" communities). When G. a. allenbyi lives without G. pyramidum ("GA alone" communities), it expands its time of activity to include the earlier hours ofthe night. (After Ziv etal. 1993.)

The study described above represents a mechanistic approach to communities, in which mechanisms for the coexistence of competitors are sought in the costs and benefits of adaptive behaviors. Foraging theory can be applied to reveal the mechanisms by which the species of a community coexist. Optimal foragers reveal their preferences, aptitudes, and handicaps through their foraging decisions. When faced with several options, an optimal forager should choose the one that yields the highest marginal value in terms of fitness (i.e., makes the largest contribution to per capita population growth rate). Thus, the abilities and the liabilities ofthe individuals determine the foods individuals exploit, the times, habitats, and microhabitats they utilize, their vigilance, and so on. The foragers' abilities, in the context of the environment, determine where and when they can forage profitably. Foraging theory allows us to quantify their behaviors, measure their costs and benefits of foraging, and test possible mechanisms of coexistence. This approach makes it possible to identify salient features ofthe environment and relevant characteristics ofthe organisms that allow species coexistence just by asking the animals.

Hence, a mechanism of species coexistence has two necessary ingredients: an axis of environmental heterogeneity or niche axis along which the species can segregate, and an evolutionary trade-off such that each species has a part of the axis at which it profits more than any of its competitors (Brown 1989a, 1989b). G. pyramidum and G. a. allenbyi can coexist because patch quality in their environment varies during the night. This nightly variation forms the necessary "axis of environmental heterogeneity" for our gerbils, but axes of environmental heterogeneity come in many forms: differences in food size, resource density, temperature, cover, or even predator type or density. In addition, our gerbils specialize on different parts of the patch quality axis because they have responded differently to the evolutionary trade-offbetween foraging speed and efficiency, primarily through body size. More generally, coexistence requires that each organism profit more than its competitors along some part of the niche axis. This usually happens via specialization: ajack-of-all-trades is a master of none (MacArthur 1972).

When both conditions hold, even when one competitor is at its carrying capacity and exploiting all ofits profitable opportunities, the other species can still find profitable opportunities. The heterogeneity must be great enough, and the trade-off severe enough, that a competitor at low density can obtain more than the resources it needs for its maintenance and replacement. If so, then the species will coexist because each can increase when rare and invade a community that its competitors dominate. This mutual invasibility criterion provides a sufficient condition for coexistence (Chesson 2000). We can find more precise conditions for coexistence using game theory (e.g., Chase et al. 2001), but simpler mutual invasibility criteria provide valuable assays for empirical testing.

Researchers have identified several potential mechanisms of species coexistence, and future work will probably uncover many more. Table 12.2 lists six important coexistence mechanisms, along with the axis ofenvironmental heterogeneity and the corresponding trade-offthat promotes coexistence in each case. The mechanisms range from resource partitioning to habitat selection. Habitat selection can include partitioning of time or space, or partitioning of spatial or temporal variation in resources or hazards.

Table 12.2 Axes of environmental heterogeneity and evolutionary trade-offs that permit niche partitioning for six major mechanisms of species coexistence

Axis

Trade-off

1. Food resource partitioning Foraging efficiencies on different food types that may vary in encounter rates, handing times, energetic content, nutrients, toxins, gut passage rates, etc. Each species must have a food type on which it profits more than its competitor.

2. Bush/open microhabitat selection Foraging efficiencies in bush vs. open microhabitats based on differences in energetic cost of foraging, harvest rates ofresources, or risk ofpredation. Each species must have a microhabitat in which it has the lowest giving-up density.

3. Habitat selection in a mosaic Foraging efficiencies in different habitats based on differences in energetic cost of foraging, harvest rates ofresources, or risk ofpredation. Each species must have a habitat in which it has the lowest giving-up density.

4. Spatial variation in resource abundance Foraging versus traveling efficiencies.

5. Temporal variation in resource abundance (daily or annual)

Foraging versus maintenance efficiency or foraging efficiency at high versus low resource abundance.

6. Temporal variation in foraging costs (daily or annual)

Foraging costs and efficiencies during different time periods. Each species must have a time period in which it has the highest foraging efficiency.

Source: After Brown, Kotler, and Mitchell 1994.

This mechanistic approach has been applied to other communities. One example involves seed-eating rodents of the Sonoran Desert (Brown 1989b). Here, akangaroo rat (Dipodomysmerriami), apocket mouse (Perognathusamplus), an antelope squirrel (Ammospermophilus harrisii), and a ground squirrel (Sper-mophilus teretecaudus) all coexist. The kangaroo rat, the pocket mouse, and the ground squirrel coexist via a seasonal rotation of foraging efficiency wherein each species has a time of year during which it is superior to its competitors. Differential susceptibilities of the three species to a seasonally changing array of predators drive the rotation. At the same time, the kangaroo rat coexists with the antelope squirrel via a mechanism involving spatial variation in resource abundance and a trade-off between foraging costs within patches versus the costs of traveling among patches. Effectively, the larger antelope squirrel uses its superior speed to move among rich patches and skim off the

"cream," while the smaller kangaroo rat uses its relatively low metabolic costs to forage more efficiently within patches on the remaining "crumbs."

More than one mechanism can operate in a community, and a single species can be involved in more than one mechanism within and across communities. D. merriami provides a good example. In the example above, its predator avoidance abilities help it coexist with pocket mice and ground squirrels through a mechanism involving seasonal rotation of foraging efficiencies, but its small body size and low metabolic costs help it coexist with antelope squirrels through a mechanism involving spatial variation in seed densities. In different communities, its predator avoidance abilities again come into play, but this time in promoting bush versus open microhabitat partitioning with still more energetically efficient pocket mice (Kotler 1984). D. merriami occurs in communities containing at least 88 different combinations of coexisting species (Brown and Kurzius 1987) that vary in their numbers of species and their characteristics. The mechanisms by which D. merriami coexists in all of these situations must vary. As the environmental conditions change from location to location, so too will the axes of heterogeneity and the relevant trade-offs among the species that allow for their coexistence.

We can combine the heterogeneities outlined here to generate still further, unique mechanisms. One such example involves larger, more arboreal red squirrels (Tamiasciurus hudsonicus) and smaller, more efficient eastern chipmunks (Tamias striatus) in Quebec, Canada. Quebec experiences strong seasonality and offers a range of forest types ranging from coniferous to mixed deciduous forests (Guerra and Vickery 1998). Red squirrels have exclusive access to resources during the winter, when chipmunks hibernate. Measurements of giving-up densities reveal that red squirrels forage more efficiently in spring in coniferous forest, while chipmunks forage more efficiently in all other forest types. Squirrels and chipmunks coexist via a combination of habitat selection in time and in space. Different forest types and different seasons provide the necessary environmental heterogeneity, and differences in body size, torpor strategies, and arboreal abilities provide the necessary trade-offs.

Studying the foraging behaviors of two or more coexisting species often suggests the mechanisms of coexistence underlying the community's biodiversity. Two examples include the interactions of tropical nectar-feeding hummingbirds (Feinsinger and Colwell 1978) and of Darwin's finches (Grant 1986). Hummingbirds may partition flower species according to dispersion and nectar reward. Tropical hummingbird species can be categorized by the length of their bills. Among short-billed hummingbirds, some species have higher wing disc loading than others (wing disc loading is the ratio of body mass to the area swept out by a wing beat and indicates the power needed for hovering). Species with high wing disc loading have short, broad wings that provide greater maneuverability and good interference ability, but high wing disc loading also makes flight more expensive (Feinsinger 1976). Hummingbirds with high wing disc loading use their fighting ability to defend territories with large clumps of moderately rewarding to rich flowers. Hummingbirds with low wing disc loading have longer wings, lower flight costs, and reduced interference ability. They cannot defend territories, but can forage profitably on dispersed or poor flowers. Finally, long-billed hummingbirds with low wing disc loading and large body sizes need very rewarding flowers to forage profitably. These hummingbirds are particularly apt at harvesting nectar from flowers with long corolla tubes (which exclude the short-billed hummingbird species) and with wide dispersions (precluding territorial hummingbird species). Hummingbird species arrange themselves across communities along axes of flower density and corolla length, based on trade-offs of body size and wing size that influence flight costs, flight speed, maneuverability, and interference ability.

Darwin's finches (Geospiza) partition seeds according to seed size based on their beak depth (Grant 1986). Birds with larger beaks can open larger and harder seeds than those with smaller beaks. Larger beaks also permit faster handling of larger seeds. Birds with smaller beaks can handle smaller seeds more quickly, but cannot open many large seed species. So, large-beaked finches profit most from the largest seeds, while small-beaked finches can exploit smaller seeds most efficiently. In the field, finches specialize on the seeds they can harvest most efficiently and coexist by resource partitioning according to seed size.

Mechanistic approaches to the study of ecological communities based on foraging theory hold much promise. So far, advocates of this approach have examined only a handful ofcommunities, identifying only a tiny subset ofco-existence mechanisms. We look forward to a much larger sample before we can answer even simple questions such as "Do coexistence mechanisms vary more within or between continents?" or "How do mechanisms of species coexistence change along clines of species diversity?" Our ability to answer such questions may help us conserve biodiversity, manage natural and artificial ecosystems, and meet the challenge of global climate change.

12.9 The Evolutionary Ecology of Communities

Two types of three-spined sticklebacks live in Paxton Lake, British Columbia. One form feeds on the lake bottom near the shore on an array of large aquatic invertebrates. The other feeds on small zooplankton in open water, although it feeds near the shore during the nesting season. The two types vary morphologically as well. The bottom feeder has a deep body, a big mouth (for its larger prey), and a small number of gill rakers. The open-water feeder has a slender body, a narrow upturned mouth, and many long gill rakers. Schluter and McPhail (1992) recognize these forms as separate species recently descended from a common ancestor via sympatric speciation. Schluter and McPhail have not formally described the two species, so following their practice, we call the bottom feeder the benthic species and the open-water feeder the limnetic species.

The two species choose habitats adaptively. The limnetic species captures more food per strike and has a higher energetic intake rate than the benthic species in open water. The benthic species captures more food per strike and has a higher energetic intake rate than the limnetic species in the benthic habitat (Schluter 1993). Within species, the benthic species has a higher food capture rate in the benthic habitat than in open water; the limnetic species has approximately equal feeding rates in both habitats. Coadaptations between morphology and behavior contribute to the species-specific performances in the two habitats. In open water, the fish lunge at prey using characteristic "S-start" strikes; in the benthic habitat, they take mouthfuls of sediment. The limnetic species' slender body makes it much better at "S-start" strikes, while the wide mouth of the benthic species allows it to take bigger mouthfuls of sediment. These differences in foraging performance translate into differences in individual growth rates. The benthic species grows about twice as fast as the limnetic species when both species feed in the benthic habitat; the limnetic species grows twice as fast as the benthic species when both feed in open water (Schluter 1995). Interestingly, hybrids have intermediate characteristics and thrive in the laboratory, yet they grow only 73% as fast as either parent species in the wild (Hatfield and Schluter 1999).

In addition to their behaviorally flexible food capture strategies for each habitat, the sticklebacks' morphology exhibits adaptive phenotypic plasticity (Day et al. 1994). When Day et al. fed each species its competitor's diet, it developed morphological features that more closely resembled those of the competitor, especially in the length of the gill rakers and in head depth. The limnetic species showed greater plasticity, consistent with its more opportunistic habitat use, and less skewed habitat-specific feeding rates. Yet, even when fed the competitor's diet, the two species remain morphologically distinct, suggesting that many of the differences between them are fixed and heritable. Reaction norm is a term often used to describe the interaction between genes and environment in determining an organism's phenotype. It formalizes the idea of phenotypic plasticity. While both sticklebacks exhibit appropriate and similarly directed phenotypic plasticity, they each exhibit this plasticity according to a distinct and species-specific reaction norm.

Overall, each species' heritable reaction norm, foraging ability, foraging tactics, and diet choice represent a coadapted syndrome in response to habitat variation (benthic versus limnetic), food type, and food availability. The recent divergence ofthese species, together with the fact that they have always lived together in Paxton Lake, suggests sympatric speciation driven by differences in optimal diet and optimal habitat use. The reaction norm of an animal feeding benthically produces a morphology that enhances aptitude within that habitat at the expense of aptitude in the limnetic habitat. Once a fish possesses this morphology, it is more likely to direct its foraging behavior toward the benthic habitat than a fish ofthe same species that has moved along the reaction norm toward a more limnetic morphology. Once fish exhibit directed foraging behavior based on their phenotypes, the possibility exists for natural selection to favor an exaggeration of these morphological differences by selecting for a divergence of reaction norms. Eventually, the coadaptation of morphology and feeding strategies produces a community with two species.

The resulting two species become defined by their foraging behaviors and the form of their reaction norms. Empirically, the superior performance of each species on its characteristic diet and in its characteristic habitat, along with the inferior fitness of intermediate types, attests to strong disruptive selection. It appears, then, that the distinct ecological opportunities offered by the two habitats to a phenotypically plastic species led to the speciation of an intermediate species into two daughter species with more extreme reaction norms. Foraging behavior is key to this process. Varied feeding strategies select for phe-notypic plasticity, the coadaptation of behavior and morphology selects for divergent reaction norms, and divergent reaction norms define the new species.

An intriguing question in community ecology concerns whether communities evolutionarily take species or make species (Kotler and Brown 1988; Wilson and Richards 2000; Chase et al. 2001). To what extent does a given community represent that selection regime that shaped the characteristics of the species coexisting within it, or to what extent did the species currently coexisting within a community evolve characteristics in response to other circumstances that exist elsewhere in the species' ranges? If a community is primarily a species taker, then invasions from a regional species pool filter through mechanisms ofcoexistence to assemble communities. Ifa community is a species maker, then interactions within the community act through natural selection to shape the diversity and characteristics ofthe species within the community. The Paxton Lake sticklebacks appear to conform to a "species-maker" scenario. The current bird community ofthe Hawaiian Islands, with its preponderance of introduced exotic species, clearly conforms to a "species-taker" scenario. The formation and composition ofspecies in most communities probably result from the joint action ofspecies-taking processes (in which selective forces occur elsewhere) and species-making processes (in which selection acts in the community on the community). With time, both processes work together to shape ecological communities. Mechanisms of coexistence provide the species, and evolution helps shape morphologies and behaviors of individuals to fit the community context and may further promote speci-ation. As a result, coexistence often represents an ESS (evolutionarily stable strategy) (Wilson and Richards 2000).

At an ESS, a species' heritable phenotype maximizes fitness given the circumstances. In this case, it is the fitness of the evolutionary strategy that is optimized, and it is optimized over all ofthe circumstances in which individuals possessing the strategy find themselves. Generally, the scale at which the phenotype ofa species represents an ESS is probably larger than the scale over which mechanisms of coexistence operate. It is this disjunction of the evolutionary scale of optimal phenotypes and the scale of ecological contingencies relating to coexistence that gives a central role to flexible feeding behaviors (and reaction norms) in revealing and influencing community organization (whether the community represents a coevolved ESS or not). The next subsection examines the tools used for modeling the coevolution ofcommunities within the context of evolutionarily stable strategies.

Models of Evolution in Communities

Models of evolution in communities show that the coevolution of interacting populations can lead to speciation and place limits on species diversity. Mitchell (2000) modeled a community in which individuals move around a landscape, randomly encountering habitat patches. These patches represent a continuum of habitat properties in which habitat type varies continuously from stressful to benign. Based on the habitat properties of a patch, a forager can choose to exploit the habitat or move on to another patch (this model can be modified into a diet model in which patches are food items instead of habitat patches). The foragers possess an evolutionary strategy (heritable phenotype) that determines their ability to exploit habitats according to stress. Often, ecological models with a habitat continuum permit the coexistence of an unlimited number of species (Abrams 1988; Tilman and Pacala 1994), each one specialized at a point along the continuum. Mitchell's model, in which the species can evolve, results in a discrete set ofspecies at the ESS despite the habitat continuum.

In Mitchell's model, foragers pay travel costs when moving between habitat patches and foraging costs when exploiting a habitat patch. Regardless of evolutionary strategy, all forager species have their lowest costs in the least stressful habitat type. Foraging costs increase with habitat stress. Foragers with a stress-intolerant strategy have very low foraging costs in benign habitats, but their foraging costs increase rapidly with habitat stress. Foragers with a stress-tolerant strategy have relatively high foraging costs in benign habitats (relative to the stress-intolerant strategy), but their foraging costs increase much more slowly with habitat stress. Depending on their evolutionary strategies, foragers will have ranges of habitat stresses at which they are relatively superior to their competitors and make larger foraging profits. For simplicity, Mitchell defined a species as the population of individuals possessing the same value for the evolutionary strategy. Given that a species enjoys an absolute advantage when it is in its best habitat type, should it always select that habitat type?

Travel costs affect the cost of habitat selection. The more selective a forager is, the farther it must travel to reach the next suitable patch, and this increases the cost of habitat selection. So, optimal habitat selection often predicts something less than strict selectivity. The species' stress tolerance strategy will affect its optimal habitat selection behavior. At the same time, an animal's habitat selection behavior and that of others will affect the optimal value of its stress tolerance strategy. As with the hummingbirds and the sticklebacks, we see a coadaptation between optimal feeding behaviors and heritable phenotypes. And the strategies and behaviors ofothers influence the optimal combinations of behaviors and morphology. We need game theory to analyze the evolution of stress tolerance because the profitability of a patch depends on the condition in which other foragers have left it. Mitchell's model combines this logic by finding the behavioral ESS for habitat selection given the interacting individuals' stress tolerance, while finding ESS values for stress tolerance for individuals that choose habitats optimally.

When foragers experience high travel costs, the ESS results in a single species with a stress tolerance strategy that utilizes a wide range of habitats (fig. 12.6A). The species cannot afford to restrict itself to its best habitat because travel costs greatly reduce the value of being picky. Furthermore, the ESS population of foragers reduces the profitability of the "preferred" habitat. When the species has a non-ESS value for its stress tolerance strategy, two processes produce Darwinian evolution toward the ESS value: natural selection can favor variants that more closely resemble the ESS, and immigrants with strategies closer to the ESS can invade and displace resident values farther from the ESS. As the community slowly approaches evolutionary equilibrium, it can support two species at ecological equilibrium, one on either side of the ESS. Eventually, the community will contain a single strategy, the ESS. If we equate strategies with species, then travel costs set limits on the numbers and characteristics ofthe species that the community can contain, even before reaching the ESS. In Mitchell's model, the community away from the ESS

Figure 12.6. The frequency-dependent adaptive landscape plotting fitness for strategy u for a range of values for u in an environment containing a continuum of habitat types that vary in stress. (A) Foragers experience high travel costs. The result is a single ESS. The high travel costs make specialization forthe best habitats too costly. (B) Foragers experience lower travel costs, and greater habitat selectivity is now possible. The ESS community now contains a greater number of species. (After Mitchell 2000.)

Figure 12.6. The frequency-dependent adaptive landscape plotting fitness for strategy u for a range of values for u in an environment containing a continuum of habitat types that vary in stress. (A) Foragers experience high travel costs. The result is a single ESS. The high travel costs make specialization forthe best habitats too costly. (B) Foragers experience lower travel costs, and greater habitat selectivity is now possible. The ESS community now contains a greater number of species. (After Mitchell 2000.)

can produce up to twice as many species as the community at the ESS (see Cohen et al. 1999; Vincent and Brown 2004).

At lower travel costs, the number of species at the ESS grows from 1 to 2, and then from 2 to 3 (fig. 12.6B). So long as there is a finite travel cost, the number of forager species at the ESS will always be finite. All of these model communities represent the interplay between foraging behavior, phenotypic evolution, and community structure.

This model considers only a single mechanism of coexistence: habitat selection. Many more types of heterogeneity influence real-world communities, so we can expect actual ESS limits to be greater. Nonetheless, this model demonstrates that competition among locally adapted organisms can promote a fixed and finite number of species within a community. While regional processes and the size of the regional species pool set the rate at which new species arrive in a community, local ecological and evolutionary processes determine the characteristics and numbers of the coexisting species.

12.10 Summary

Foraging theory gives us unique insights into the coexistence mechanisms and the forces that structure and shape assemblages ofspecies. For a species to coexist with its competitors, members ofthe species must experience positive fitness at some point; that is, the strategy of an individual in the population must lead to a positive per capita growth rate. The fitness of an individual depends on its foraging profit. Thus, the characteristics that really matter for the community are the characteristics that really matter for foragers. There are many such characteristics, including properties of the forager such as encounter rates and energetic costs; properties ofprey such as handling times, energetic value, nutrient content, the bulk of various food items (digestion time), and search time; and properties ofpredators such as mortality risk. This chapter shows how parameters like these determine the intensity of species interactions, conditions for and mechanisms of species coexistence, and even the characteristics ofcoevolved species in an ESS community.

12.11 Suggested Readings

Chesson (2000) provides an excellent review and synthesis of the theory of species coexistence. Other important theoretical treatments of species coexistence include the consumer-resource models of Holt et al. (1994) for a pair of competitors that share a common predator and of Vincent et al. (1996) for optimally foraging competitors exploiting resources that may differ in quality or in spatial distribution. Mitchell (2000) shows how coevolution among optimal foragers can lead to communities whose species are shaped and determined by the ESS conditions and in which species interactions set local limits on species diversity. Morris (1988, 1996) provides theoretical explanations and empirical examples of the application of isodars, and Rosenzweig and Abramsky (1997) do likewise for isolegs.

Foraging theory provides the tools for understanding a community in depth, as demonstrated in gerbils. The behavior of individuals of the constituent species (Kotler et al. 1991), the salient features of the environment and the species that promote coexistence (Kotler, Brown, and Subach 1993; Ziv et al. 1993; Brown et al. 1994), and even the resolution of the foraging game played among competitors and their predators (Kotler et al. 2002; Kotler, Brown, and Bouskila 2004; Kotler, Brown et al. 2004) can be understood by applying foraging theory. The article by Rosenzweig and Abramsky noted above, which concerns the gerbils, provides an excellent summary of the application ofisolegs, isoclines, and isodars to better understand this community.

Was this article helpful?

0 0

Post a comment