Environments are changing rapidly, which renders many local populations susceptible to extinction unless they can adapt to these changes. Studies of rapid adaptation commonly document the evolution of individual traits. Overall adaptation however, is a function of fitness itself, rather than the individual traits that contribute to fitness. Although numerous studies provide evidence for the evolution of specific traits on contemporary time scales, no published studies of wild animal populations have examined the evolution of a major fitness component following environmental change. My research demonstrates that an introduced population of guppies (Poecilia reticulata) has adapted to its new environment in less than ten years (13-26 generations). This adaptation consists of several phenotypic traits that have changed in the expected direction. Most critically, the introduced population now has higher survival than its ancestral source population when both are tested together in the introduction site. These results show that important components of fitness can evolve rapidly in populations, and that this evolution might influence the persistence of populations in the face of environmental change.

No description available.

Two processes may lead to genetic divergence among populations. One is mediated by geography, whereby physical barriers and geographic distance limit gene flow among populations, resulting in divergence due to drift or mutation. Another is ecological speciation, whereby populations adapt to their local environments via natural selection, and gene flow is impeded by selection against dispersers in favor of adapted residents. I used natural populations of guppies (Poecilia reticulata) to determine the relative influence of these two processes in the structuring of populations. If geography is playing a strong role, I predicted that gene flow would be greatly impeded by physical barriers and geographic distance. If ecology is playing a strong role, I predicted that gene flow would decrease with increasing strength of divergent selection among populations. Specifically, I examined the relative roles of physical barriers, geographic distance, predation, and various other habitat features (e.g. canopy cover, water velocity) on the amount of gene flow among populations. I was thus able to determine whether natural selection or decreased dispersal plays a greater role in the reduction of gene flow. I found that physical barriers and geographic distance played a large role in the regulation of gene flow among populations. Predation and physical habitat features did not play a role in the reduction of gene flow. My research clarifies the mechanisms involved in speciation and the production and maintenance of biodiversity, important issues in conservation and evolutionary biology.

The colour of guppies (Poecilia reticulata) evolves as a compromise between sexual selection (favouring conspicuousness) and natural selection (favouring crypsis). However, guppies live in a variety of habitats and with a variety of predators and consequently in a variety of selective environments. I investigated how habitat and predator's visual systems affect the evolution of colour. I used regressions to assess the importance of habitat features on the evolution of colour for 29 guppy populations. I then quantified the colour of guppies living in the presence and absence of two predators. The prawn predator is insensitive to orange light while the fish predator is insensitive to ultraviolet light. Habitat explained some variation in colour, but not in a consistent manner. Guppies living with the prawn were more orange and guppies living with the fish had more ultraviolet reflectance, providing evidence for the use of these aspects of colour as private signals.

No description available.

No description available.

Ecological speciation involves the evolution of reproductive isolation (RI) as a by-product of adaptation to different selective environments. Parallel patterns of non-random mating by environment type provide strong evidence that ecological speciation has occurred. The processes involved in the origins of RI are more difficult to detect however. One mechanism involves the correlated evolution of mating preferences and sexually selected traits. I developed a conceptual model for detecting RI under various scenarios of mate preference evolution. The model predicts that RI will not evolve if preferences are evolutionarily constrained relative to the preferred traits, but is detectable as long as preferences evolve in parallel. I then applied this framework to an empirical system with populations of guppies (Poecilia reticulata) adapted to low- and high-predation environments. I measured female mate preferences for male colour and size; traits which are divergent between the two environment types. Preference functions for colour also diverged in the predicted direction. The parallel pattern of preference divergence suggests that divergent natural selection from predators may be contributing to RI between guppy populations.

No description available.