Oikos 116: 1431
1437, 2007 doi: 10.1111/j.2007.0030-1299.15833.x, Copyright # Oikos 2007, ISSN 0030-1299 Subject Editor: Kenneth Schmidt, Accepted 24 April 2007

Individual foraging specialization: niche width versus niche overlap Brooke L. Sargeant B. L. Sargeant ([email protected]), Dept of Biology, Georgetown Univ., Washington, DC 20057, USA. Present address: Dept of Biological Sciences, Florida International Univ., 11200 SW 8th St., Miami, FL 33199, USA.

A recent review of individual specialization suggests it is widespread and has important implications for conservation, ecology and evolution (Bolnick et al. 2003). Dozens of studies have identified ‘‘individual foraging specialization’’ in a variety of species, yet upon closer examination these studies differ markedly in how they define and quantify specialization. Given such diverse approaches, characteristics of niches that provide information about individual differences and quantify individual foraging specialization need to be clearly identified. Individual foraging specialization has been defined in many ways, although all definitions involve the use of different dietary items or foraging tactics by individuals within a single population. A direct corollary is that at least some individuals must have narrower niches than the population niche overall. Following Bolnick and colleagues’ (2002) interpretation of Roughgarden’s (1972) niche width concept, population-level niche width (‘‘total niche width’’, TNW) can be subdivided into a ‘‘within-individual component’’ (WIC) and a ‘‘between-individual component’’ (BIC). Thus, WIC/ TNW provides a measure of how much variation in TNW can be explained by variation within or between individuals. As WIC/TNW increases, most of the variation in TNW results from variation within individuals, but as it decreases, variation in TNW results from variation between individuals. Thus, unless all individuals have identical niches, at least some niches will be narrower than that of the population and individuals will differ from one another. Individual differences clearly differentiate individual foraging specialization from ecological specialization in species or populations (Levins 1968, Fox and Morrow 1981, Futuyma and Moreno 1988) and predict

individual differences in niches. But further distinctions about how specialization is defined and therefore who is most specialized vary substantially between studies. While it may initially seem that a strict definition could be applied (e.g. a specialist uses 50% of resources used by the population), biologically meaningful patterns of niche use are likely to depend on what resource and species are being considered. For example, Futuyma and Moreno (1988) recommended against a strict, rigid definition of resource specialization, viewing specialization and generalization as extremes of a continuum based on Hutchinson’s (1957) n-dimensional hypervolume of niche space. Therefore, similar to traditional approaches to generalization and specialization, individual specialization should describe individual niche use only in terms relative to the population’s niche use. So what, exactly, is individual foraging specialization? Generally, two different conceptual approaches have emerged, although both compare an individual’s niche to the population’s niche. In the first, which I will refer to as the niche width specialization concept, individual foraging specialists are individuals with narrower foraging niches than the population’s niche (Roughgarden 1972, West 1986, Werner and Sherry 1987, Holbrook and Schmitt 1992, reviewed by Bolnick et al. 2003). This should include definitions that describe ‘‘the consistent use’’ of resources or tactics in which alternative resources or tactics are used by the population (Werner et al. 1981, Partridge and Green 1985, Schindler et al. 1997), as consistency should result in a narrower niche by reducing richness or evenness (below). In the second approach, which I will refer to as the niche overlap concept, specialists have niches that exhibit little overlap with the population

1431

niche (Bolnick et al. 2002, Estes et al. 2003, Svanba¨ck and Persson 2004). A variant of this concept is that specialists are those who use rare dietary items or tactics. While the use of a rare resource is suggestive of low overlap with population dietary values, the relative use of the rare resource by the individual and by the population will influence such overlap measures. For the purposes of this discussion, the use of a ‘‘rare’’ resource will refer to the rarity of this resource in the population’s realized niche, but does not imply the actual availability of this resource. Some have referred to such behaviors themselves as ‘‘foraging specializations’’, which does not reflect the process of ecological specialization and may be more similar to terminology regarding morphological or evolutionary ‘‘specializations’’ as opposed to ecological specialization (Robinson and Wilson 1998, Ferry-Graham et al. 2002). These behaviors are better termed ‘‘rare’’ or ‘‘novel’’, rather than ‘‘specialized’’ or ‘‘specializations’’. The use of multiple definitions in studies of individual foraging specialization presents three main problems. First, without unified approaches, broadscale comparative studies are hindered, especially because different measures of individual foraging specialization do not always correspond. Second, some definitions conflict with the traditional view of ecological specialization as having reduced niche width. For example, the use of a rare dietary resource or tactic or having little dietary overlap with the population may not reflect niche width. Third, this ambiguity has blurred analysis of the functions and evolution of individual foraging specialization. The inherently flexible nature of the generalization
specialization continuum, however, does not preclude biologically

meaningful and consistent definitions and measurements. Perhaps a useful approach is to first consider what characterizes a niche, and then determine how these characteristics relate to the two concepts of individual foraging specialization described above. Additionally, hypothesized ultimate explanations for individual foraging specialization will be discussed in terms of their predicted relationships with niche characteristics.

Measuring individual niches At least four critical characteristics of a niche can be identified with regard to different resource categories (e.g. prey species): niche width, richness, evenness, and niche overlap. For each characteristic, indices such as those in Table 1 can be calculated for each individual and index means can be used to describe the characteristic for the population. Niche width (sometimes referred to as breadth) summarizes the diversity of an individual’s niche, often in terms of the proportion of their resources in various categories (Roughgarden 1979, Schindler et al. 1997, Bolnick et al. 2002, Svanba¨ck and Persson 2004). Individual niche width (WICsi, where s designates Shannon index and i designates individual) can be quantified using Roughgarden’s (1979) adaptation of the Shannon biodiversity index. Similarly, WICs/ TNWs is a categorical equivalent of WIC/TNW described above (Bolnick et al. 2002). When measured using the Shannon index, individual niche width is influenced simultaneously by richness and evenness (Magurran 2004), richness (Ri) being the number of resource categories used and evenness (Ei) referring to

Table 1. Characteristics of individual niches for categorical resource data. Niche characteristic

Index for individual

Incorporates population niche?

Width

/

WICsi aj pij lnpij

No

Richness

/

Ri ji

No

Evenness

/

Overlap

Ei 

WICsi lnji

1=ni

Wi(li)

/

; liPj

!nij qj pij

Notes Roughgarden’s (1979:528) measure of individual niche width for categorical niches, from Bolnick et al. (2002)

No

Adapted from Pielou’s (1975:15) evenness measure

Yes

Adapted from Petraitis’ W (1979) for interspecific diet overlap by Bolnick et al. (2002)

Individuals (i) have consumed a number of items (n) in several resource categories (j), such that: 1) ji is the number of resource categories used by the individual i; 2) nij is the number of items in the jth resource category for individual i; 3) ni is the number of items for individual i; 4) pij is the proportion of items in the jth resource category for individual i; and 5) qj is the proportion of items in the jth resource category for the population (see Bolnick et al. 2002 for additional details).

1432

the relative use of each resource. Evenness can be calculated using Pielou’s (1975) evenness index adapted for individual niche width. Because niche width (as measured using WICsi) is maximized (for a given number of resource categories) when all categories are used equally, a comparison of the observed width to the maximized width provides an index of evenness (Pielou 1975). Richness and evenness correlate positively with niche width, such that width increases with increasing numbers of resources used and is maximized where all resources are used equally. An individual with higher richness and lower evenness than another individual may have higher or lower niche width, depending on the strength of the differences, although the Shannon index is generally influenced more heavily by richness (Magurran 2004). Individual niches may be evaluated also in terms of deviation from population resource values (Bolnick et al. 2002), which I refer to as niche overlap (Wi). Bolnick and colleagues (2002) provide an adaptation of Petraitis’ W, originally calculated for interspecific diet overlap (Petraitis 1979), to calculate intraspecific overlap. Unlike the other niche characteristics, the definition and calculation of niche overlap incorporate the resource use of the population. The lower the Wi value, the less an individual overlaps with the population’s resource use. Several further points detailing the definition of individual specialization are worth considering. While biologists typically measure realized niches when studying individual specialization in nature, consideration of both realized and fundamental niches may be required to understand the causes of individual specialization (Bolnick et al. 2003). For example, individuals may have different fundamental niches (resulting in different realized niches) due to phenotypic differences. Alternatively, individuals may have identical fundamental niches, but differences in ability to exploit that niche, as in interference competition, could generate individual differences. Long-term consistency is another important component of niche use (Schindler et al. 1997, Bolnick et al. 2002), and can be described in terms of niche characteristics by comparing their values over time. Additionally, depending on the niche that is being measured, individual foraging specializations may be defined in terms of dietary species or sizes, behavioral

tactics, and potentially any other foraging niche. As such, they are not restricted solely to differences in prey species. Further, ecological specialization is niche specific; an individual could be a generalist in terms of prey species but be more specialized in prey size, for example. Similarly, the degree to which an individual is an ecological specialist in terms of niche width may not address morphological or evolutionary ‘‘specializations’’, which are not measured in terms of ecological niches (Ferry-Graham et al. 2002). Finally, the calculations presented here refer to cases in which the niche axis is categorical and would not be applicable to a continuous niche axis (e.g. prey size) (but see Bolnick et al. 2002 for an example of an index for continuous niches). Additionally, several other indices may be used to measure niche components, each with different strengths and weaknesses (Levins 1968, Bolnick et al. 2002, Magurran 2004). The indices presented here were selected due to their ease of calculation, common usage, and ability to estimate desired niche properties, but have some weaknesses. For example, WICsi is zero for monophagous individuals and can artificially depress population WICs values. Often the niche characteristics presented above are positively correlated, but this does not have to be the case, particularly when niche overlap is compared to niche width or richness. For example, consider a hypothetical population (Table 2) in which most individuals (A
D) utilize only one resource, thereby having a niche width (WICsi) of 0. However, individual E has begun to use two additional resources; this individual has a calculated niche width of 1.10. Thus, individual E uses two rare resources and has a wider niche than the other members of the population. When niche overlap (Wi) is calculated in terms of each individual’s overlap with population diet proportions, individual E exhibits the least niche overlap as a result of its divergence from others. According to the niche width concept of individual foraging specialization, E is the least specialized in the population. Individual E is the most specialized, on the other hand, when viewed in terms of the niche overlap concept. Clearly, the niche width concept and the niche overlap concept can provide conflicting results regarding who is most specialized.

Table 2. Niche width compared to niche overlap. Individual A B C D E a

Prey type 1

Prey type 2

Prey type 3

Ri

Eai

WICsi

Wi

90 90 90 90 30

0 0 0 0 30

0 0 0 0 30

1 1 1 1 3

. . . . 1

0 0 0 0 1.10

0.87 0.87 0.87 0.87 0.47

Evenness cannot be calculated for monophagous individuals.

1433

As the previous example illustrates, use of a rare resource does not necessarily reflect niche width. A similar exercise reveals that defining individual foraging specialization in terms of the use of a rare resource may not match the outcome of the niche overlap concept (Table 3). Although individual A is the only individual using prey type 1, according to the niche width concept, B has the narrowest niche and would be the most individually specialized forager. But because all individuals except B use prey type 3 and B has an unusually high diet proportion of prey type 2, B shows the least overlap with the population’s niche. In this case, individual B is most specialized in terms of both the niche width concept and the niche overlap concept, even though only A uses prey type 1. Admittedly, these examples have been constructed to illustrate the potential conflicts among these concepts of individual foraging specialization, and the frequency of such conflicts in biological populations is unclear. Despite numerous studies documenting individual differences, relatively few explicitly measure niche use in terms of the characteristics described here. Yet it is not hard to imagine a scenario in which such conflicts would occur. For example, consider a case of niche expansion in response to high intraspecific competition (Bolnick 2001 for an experimental example), where (at least at the start of expansion) a few individuals could use a wide range of poorer quality resources while the majority of individuals use a small range of high quality resources. Such niche width expansion incorporating lower quality resources has been predicted to occur in response to competition (Roughgarden 1972, Wilson and Turelli 1986). Additionally, niche width has been shown to be uncorrelated with niche overlap in wild Indian Ocean bottlenose dolphins (Tursiops sp.) (Sargeant 2005). Clearly, more studies are needed to identify the frequency with which these patterns occur. Historically, ecological specialization has generally referred to niche width, with specialists having narrower niches than either the available niche or the population niche (Levins 1968, Fox and Morrow 1981, Futuyma and Moreno 1988, Jaenike 1990, Bolnick et al. 2003). Importantly, many studies do not clearly define individual specialization or distinguish findings based on niche width and niche overlap. In many cases, even when individual specialization is explicitly defined using

overlap measures, ‘‘niche width’’, ‘‘niche breadth’’, ‘‘variation’’, ‘‘overlap’’, and ‘‘specialization’’ appear interchangeably throughout the text. For example, in some cases both individual calculations of niche width and niche overlap have been considered to ‘‘estimate individual specialization for each individual’’ (Bolnick et al. 2002). The measures of population level incidence of individual specialization used by Bolnick and colleagues (2002) (including WICs/TNWs and W (mean of the Wi values)), however, may have greater agreement than similar measures for individuals. Indeed, they found high correlation between WICs/ TNWs and W using simulated data sets. Given the historical framework of ecological specialization, ‘‘individual foraging specialization’’ should be defined as the use of a narrower niche than the population niche and should be measured on a continuum of width. WICs/TNWs can be calculated to assess how much of the population’s niche width is a reflection of variation within or between individuals, and WICsi/TNWs can be used to calculate a measure of specialization for each individual. Because TNWs is equal for all individuals in the same population, this is the same as comparing WICs values directly when examining a single population. Therefore, individuals with lower niche widths are relatively more specialized than individuals with higher niche widths within the same population. While niche overlap should not be equated with measures of specialization, it does provide useful information regarding individual patterns of niche use and has important implications when individual foraging specializations result from density- and frequency-dependent fitness effects. Therefore, to understand the nature and evolution of individual foraging specialization, niche overlap should also be measured. For the rest of this paper, ‘‘individual foraging specialists’’ will refer to individuals with relatively narrow niche widths compared to the population.

Functions of individual foraging specialization Discriminating between niche overlap and niche width is more than just a semantic argument. Such clarification

Table 3. Use of rare resources compared to niche width and niche overlap. Individual A B C D E a

Prey type 1

Prey type 2

Prey type 3

Ri

Eai

WICsi

Wi

30 0 0 0 0

30 90 45 45 45

30 0 45 45 45

3 1 2 2 2

1 . 1 1 1

1.10 0 0.69 0.69 0.69

0.72 0.57 0.91 0.91 0.91

Evenness cannot be calculated for monophagous individuals.

1434

should illuminate discussion of the functions of individual foraging specialization. Proximate explanations for individual foraging specialization include differences in genotype, phenotype, environment, individual learning, or social learning, or combinations of these factors (Partridge and Green 1985). These address how behaviors develop through genetic, environmental, sensory, and/or hormonal pathways at a mechanistic level. At the ultimate level, which addresses how a behavior has evolved, tradeoffs related to phenotypic differences, habitat heterogeneity, or other characteristics, and intraspecific competition have been proposed to explain the function and evolution of individual foraging specialization (Partridge and Green 1985). Individuals may differ in phenotype, such that diets or tactics provide different energy intake rates depending on these phenotypes (Roughgarden 1974, Price 1987). A second evolutionary explanation is that individual resource use is limited by access to resources, but individuals have different local environments (Holbrook and Schmitt 1992). When individuals are restricted in their habitat use (by costs of travel, territoriality, etc.), they experience tradeoffs between using one habitat and another. In this case, patchy habitats could drive population variation in resource use by providing different prey types or physical substrates, such that use of particular resources or tactics provides higher foraging success. Phenotypic and habitat tradeoffs have been commonly discussed, but use of multiple resources can result in tradeoffs with foraging efficiency through additional mechanisms (reviewed by Bolnick et al. 2003). Individuals in this scenario benefit by having a low niche width that affords decreased handling time, increased digestive efficiency, or other benefits, but individuals differ from conspecifics due to differences in ecology, phenotype, or learning (Werner et al. 1981). Tradeoffs may be critical to individual specialization, as models predict individuals should expand their niches to match the population niche unless prevented by some cost(s) (Taper and Case 1985, Ackermann and Doebeli 2004). Finally, frequency-dependent selection could also be occurring, with the best strategy depending on the strategies used by others when there is intense intraspecific competition (McLaughlin et al. 1999, Bolnick 2001, McLaughlin 2001, Estes et al. 2003, Svanba¨ck and Persson 2004). While all of these hypotheses predict individual differences, they do not always predict who will be most specialized in terms of niche width. For an individual, hypotheses regarding tradeoffs generally predict decreasing niche width with increasing constraints or in some cases increasing fitness (if reducing niche width increases energy intake). Because niche overlap is influenced by what the rest of the population is doing, however, a theoretical prediction of the relationship between tradeoffs experienced by an individual and

their niche overlap is unclear. Unlike ultimate hypotheses for the adaptive function of individual foraging specialization based on tradeoffs, which offer no a priori predictions about niche overlap, competition could result in a negative relationship between fitness and overlap because initially uncommon patterns of niche use should provide decreased competition. Applying the theory of ideal free distributions (Fretwell and Lucas 1970) to inter-individual niche overlap, individuals may eventually have equal fitness at equilibrium if they are flexible in their behavior (Partridge and Green 1985, Estes et al. 2003). At equilibrium, no correlation would be expected between niche overlap and fitness. Thus, density- and frequency-dependent fitness benefits associated with use of different resources may favor individual foraging specialization secondarily and generate predicted correlations between foraging success and niche overlap, but will not generate predictions about how specialization in terms of niche width will relate to foraging success. Svanba¨ck and Bolnick (2005) used optimal diet theory to provide a framework for how, once individuals experience phenotypic tradeoffs, high population density can promote inter-individual diet variation. In short, ultimate explanations for the evolution of the individual differences underlying individual specialization may be divided into those that reduce intraindividual variation (reduce niche width) and those that increase inter-individual variation (reduce niche overlap). Tradeoffs or interactions with conspecifics (e.g. social hierarchies) may cause individuals to specialize by generating different optima for different individuals and favor reduced niche width, while high competition resulting in frequency dependence can favor individuals with low niche overlap. Both mechanisms may occur, and in particular the effects of competition may depend heavily on the strength of competition and the intensity of tradeoffs and differences in prey rankings (Svanba¨ck and Bolnick 2005). The above examples are simplified cases, and multiple factors may actually drive niche diversification. Nonetheless, discussions regarding the function and evolution of individual foraging specialization should be explicit about how the factors driving niche diversification and population niche expansion are predicted to influence particular niche characteristics. Clearly, hypotheses regarding tradeoffs are relevant to niche width, whereas those regarding competition are more closely tied to niche overlap.

Conclusions It is now evident that individual foraging specialization is frequent and has important ecological, evolutionary, and conservation consequences (Bolnick et al. 2003).

1435

Particularly, individual foraging specialization is relevant to recent attention to disruptive selection and speciation (Bolnick 2004) and social learning and animal culture (Werner and Sherry 1987, Rendell and Whitehead 2001, Estes et al. 2003, Tinker 2004). Given the wide applicability of individual foraging specialization, it is vital that studies measure not only differences among individuals, but also the degree of specialization in populations and individuals in a quantitative way (Bolnick et al. 2002). The niche characteristics described here provide useful ways of describing an individual’s niche, and niche width in particular can be used to quantify the degree to which an individual is specialized relative to the population and can provide information about an individual’s prey preferences and tradeoffs that may constrain their resource use. Describing the use of a rare resource or deviation from the population niche provides useful information about individual differences and the origin of individual foraging specialization because measures of niche overlap can reveal how much the individual is competing with the rest of the population for resources. However, because niche overlap and rarity do not necessarily correspond to individual niche width, they should not be used to describe whether an individual is itself a specialist. Use of multiple standardized and quantitative methods within and among studies will advance our understanding of the function and evolution of individual foraging specialization, providing direction for future research.

Acknowledgements
Rachel Barr, Dan Bolnick, Sasha Dall, Larry Douglass, Janet Mann, Richard Svanba¨ck, Martha Weiss and Aaron Wirsing provided helpful discussions and comments on drafts of this manuscript.

References Ackermann, M. and Doebeli, M. 2004. Evolution of niche width and adaptive diversification.
Evolution 58: 2599
2612. Bolnick, D. I. 2001. Intraspecific competition favours niche width expansion in Drosophila melanogaster .
Nature 410: 463
466. Bolnick, D. I. 2004. Can intraspecific competition drive disruptive selection? An experimental test in natural populations of sticklebacks.
Evolution 58: 608
618. Bolnick, D. I. et al. 2002. Measuring individual-level resource specialization.
Ecology 83: 2936
2941. Bolnick, D. I. et al. 2003. The ecology of individuals: incidence and implications of individual specialization.
Am. Nat. 161: 1
28. Estes, J. A. et al. 2003. Individual variation in prey selection by sea otters: patterns, causes, and implications.
J. Anim. Ecol. 72: 144
155.

1436

Ferry-Graham, L. A. et al. 2002. Using functional morphology to examine the ecology and evolution of speciation.
Integr. Comp. Biol. 42: 265
277. Fox, L. R. and Morrow, P. A. 1981. Specialization: species property or local phenomenon?
Science 211: 887
893. Fretwell, S. D. and Lucas, H. L., Jr. 1970. On territorial behavior and other factors influencing habitat distribution in birds I. Theoretical development.
Acta Biotheoret. 19: 16
36. Futuyma, D. J. and Moreno, G. 1988. The evolution of ecological specialization.
Annu. Rev. Ecol. Syst. 19: 207
233. Holbrook, S. J. and Schmitt, R. J. 1992. Causes and consequences of dietary specialization in surfperches: patch choice and intraspecific competition.
Ecology 73: 402
412. Hutchinson, G. E. 1957. Concluding remarks.
Cold Spring Harbor Symp. Quant. Biol 22: 415
427. Jaenike, J. 1990. Host specialization in phytophagous insects.
Annu. Rev. Ecol. Syst. 21: 243
273. Levins, R. 1968. Evolution in changing environments: some theoretical explorations.
Princeton Univ. Press. Magurran, A. E. 2004. Measuring biological diversity.
Blackwell. McLaughlin, R. L. 2001. Behavioral diversification in brook charr: adaptive responses to local conditions.
J. Anim. Ecol. 70: 325
337. McLaughlin, R. L. et al. 1999. Adaptive peaks and alternative foraging tactics in brook charr: evidence of short-term divergent selection for sitting-and-waiting and actively searching.
Behav. Ecol. Sociobiol. 45: 386
395. Partridge, L. and Green, P. 1985. Intraspecific feeding specializations and population dynamics.
In: Sibly R. M. and Smith, R. H. (eds), Behavioural ecology: ecological consequences of adaptive behaviour. Blackwell, pp. 207
226. Petraitis, P. S. 1979. Likelihood measures of niche breadth and overlap.
Ecology 60: 703
710. Pielou, E. C. 1975. Ecological diversity.
Wiley. Price, T. 1987. Diet variation in a population of Darwin’s finches.
Ecology 68: 1015
1028. Rendell, L. and Whitehead, H. 2001. Culture in whales and dolphins.
Behav. Brain Sci. 24: 309
382. Robinson, B. W. and Wilson, D. S. 1998. Optimal foraging, specialization, and a solution to Liem’s paradox.
Am. Nat. 151: 223
235. Roughgarden, J. 1972. Evolution of niche width.
Am. Nat. 106: 683
718. Roughgarden, J. 1974. Niche width: biogeographic patterns among Anolis lizard populations.
Am. Nat. 108: 429
442. Roughgarden, J. 1979. Theory of population genetics and evolutionary ecology: an introduction.
Macmillan. Sargeant, B. L. 2005. Foraging development and individual specialization in wild bottlenose dolphins (Tursiops sp.). PhD thesis.
Georgetown Univ. Schindler, D. E. et al. 1997. Density-dependent changes in individual foraging specialization of largemouth bass.
Oecologia 110: 592
600.

Svanba¨ck, R. and Persson, L. 2004. Individual diet specialization, niche width and population dynamics: implications for trophic polymorphisms.
J. Anim. Ecol. 73: 973
982. Svanba¨ck, R. and Bolnick, D. I. 2005. Intraspecific competition affects the strength of individual specialization: an optimal diet theory model.
Evol. Ecol. Res. 7: 993
1012. Taper, M. L. and Case, T. J. 1985. Quantitative genetic models for the coevolution of character displacement.
Ecology 66: 355
371. Tinker, M. T. 2004. Sources of variation in the foraging behavior and demography of the sea otter, Enhydra lutris . PhD thesis.
Univ. of California Santa Cruz.

Werner, E. E. et al. 1981. The role of foraging profitability and experience in habitat use by the bluegill sunfish.
Ecology 62: 116
125. Werner, T. K. and Sherry, T. W. 1987. Behavioral feeding specialization in Pinaroloxias inornata , the ‘‘Darwin’s finch’’ of Cocos Island, Costa Rica.
Proc. Natl Acad. Sci. USA 84: 5506
5510. West, L. 1986. Interindividual variation in prey selection by the snail Nucella ( Thais ) emarginata .
Ecology 67: 798
809. Wilson, D. S. and Turelli, M. 1986. Stable underdominance and the evolutionary invasion of empty niches.
Am. Nat. 127: 835
850.

1437