Get the results of TWRI-funded research through technical and special reports. Find abstracts and full-text online for our reports.


Shared and unique features of predator-associated morphological divergence in three livebearing fishes

R. Brian Langerhans, Craig A. Layman, A. Mona Shokrollahi, Thomas J. DeWitt

Divergent natural selection (fitness trade-offs) between environments often drives intraspecific diversification. When multiple species face a common environmental gradient their patterns of divergence might exhibit both shared and unique elements. We investigated shared and unique elements of diversification for three species of livebearing fishes (Poeciliidae) across a predator gradient (piscivorous fish vs. no piscivorous fish). All species (Gambuia affinis from the United States, G. hubbsi from the Bahamas and P. reticulata from Trinidad) exhibited a posterior shift in body allocation and more elongate bodies in predator populations. This morphological shift applied not only across species, but across age/gender classes, particular predator species and gross habitat types. Each species also had unique features of divergence, which may reflect peculiarities unique to their natal environments. We hypothesize much of the predator-associated morphology enhances fitness in the presence of predators but decreases fitness in the absence of predators. Preliminary evidence supports this hypothesis. We additionally found gonopodium size divergence in both Gambusia species that was opposite a trend previously found in P. reticulata. Partitioning phenotypic variation for multiple species into shared and unique responses across environmental gradients is a useful tool for elucidating general and localized mechanisms of diversification.

Divergent natural selection (DNS) is a powerful mechanism that generates and maintains phenotypic diversity (Rice and Hostert 1993; Orr and Smith 1998; Schluter 2000b). DNS is created by functional trade-offs between alternative environments (Robinson and Wilson 1994; Robinson et al. 1996; Schluter 2000b). Depending on how environments vary (e.g. fine- or coarse-grained variation; sensu Levins 1968), DNS can lead to evolution of either canalized genetic differences between populations or phenotypic plasticity (West-Eberhard 1989; Robinson and Wilson 1994; Orr and Smith 1998; Schluter 2000b). DNS results in population differentiation through either of these mechanisms. That is, populations may differ from each other because of fixed genetic differences or because each is similarly plastic but their current environments differ. Environmental differences that typically produce population divergence include predation, competition, or abiotic stresses (Magurran and May 1999; Schluter 2000b).

Consider the case where the sign of selection on focal traits differs across environments. Consider also that several species are evolving adaptive solutions in these environments. Some elements of evolutionary response will be similar across species. For example, species A and B may largely respond in a similar manner. Yet other aspects of response to DNS will be speciesspecific (A responds one way; B responds differently). In theory, after accounting for gross differences between species, the unique and shared elements of response to DNS can be identified. This approach allows one to integrate the study of phenotypic variation across selective environments, populations and species.

Empirical studies of diversification in vertebrates has focused on resource competition and niche partitioning (Robinson and Wilson 1994; Van Valkenburgh and Wayne 1994; Smith and Skulason 1996; Schluter 2000a). However, predation is among the most important factors structuring natural populations (Sih et al. 1985; Kerfoot and Sih 1987; Jackson et al. 2001). Predators should not only change the distribution and abundance of species (via lethal and behavioral impacts on prey), but should also produce diversification—changes in the distribution and abundance of phenotypes within species (Endler 1995; DeWitt et al. 2000; Trussell 2000; Van Buskirk and Schmidt 2000).

Although predation is often thought responsible for the evolution of inducible defenses among invertebrates (Harvell 1986; Havel 1987; Tollrian and Harvell 1999), studies investigating the importance of predation in vertebrate diversification are less common. Most studies of vertebrate evolutionary responses to predation center on life history changes (e.g. Skelly and Werner 1990; Rodd and Reznick 1997; Belk 1998; Johnson and Belk 2001; Reznick et al. 2001). Investigation of morphological divergence between alternative predator environments in vertebrates is less common (Brönmark and Miner 1992; Van Buskirk et al. 1997; Walker 1997; Lardner 2000; Relyea 2001).

In this study, we examined body shape divergence in three species of livebearing fishes across alternative predator regimes. Furthermore we evaluate trends within and between species to identify both unique and shared aspects of diversification among prey species.

Back to Top