The ecology and geography of speciation: A case study using an adaptive radiation of threespine stickleback in Alaska (Gasterosteus aculeatus)

The objective of my research is to understand the relative importance of geographical isolation and adaptive phenotypic evolution for the formation of new species. Geographical speciation is a model in which species are formed primarily by geographical isolation and overall population genetic divergence, whereas ecological speciation models the formation of new species as a consequence of adaptive phenotypic evolution. The relative frequency of these two modes of speciation in the wild is still unknown. Populations undergoing divergence leading to speciation provide the most promising avenue of research. Thus, I used a combination of field and laboratory experiments to study a model system for evolutionary biology, the threespine stickleback Gasterosteus aculeatus. The repeated independent colonization of freshwater habitats by anadromous stickleback has established a large scale replicated evolutionary experiment that could never be done in a laboratory setting. I utilized this natural experiment to determine whether speciation in stickleback could best be explained by the geographical or the ecological model.; I first tested a prediction of ecological speciation that phenotypic divergence leading to speciation can occur in sympatry. I describe the existence of two types of stickleback, benthic and limnetic, in Benka Lake. Divergence into these two forms has previously been shown important for speciation in stickleback and other fishes, and I show how the phenotypic distribution in Benka Lake parallels the divergence seen between sympatric stickleback species pairs in British Columbia. I then present data on stable isotopes of carbon and nitrogen that I used to examine whether resource partitioning is occurring among phenotypes in Benka Lake. The fitness distribution of stickleback phenotypes when they are confined to alternative habitats is then presented, along with a description of the neutral genetic background of the stickleback population in Benka Lake.; Next, I study whether selection is necessary to cause speciation in stickleback, or whether geographic isolation is enough by itself. A prediction from geographical speciation is that populations diverging in allopatry along phenotypic lines leading to speciation should also show significant divergence in neutral population genetic structure. To test this prediction, I examine the distribution and frequency of six microsatellite loci among allopatric Alaskan stickleback populations. Finally, I reverse the frame of reference in order to test another prediction of geographical speciation that isolation in the absence of divergent selection is sufficient to cause speciation. I examine two ancient clades of stickleback that have been geographically isolated over extended periods of time, but have experienced similar selective regimes during isolation. These clades have recently extended their ranges and are now sympatric in some Alaskan stickleback populations. I examine linkage disequilibrium between mtDNA and microsatellite markers to determine whether these two lineages have become reproductively isolated from one another during their extended period of geographical isolation, or whether they still interbreed during secondary contact.; The data I present in my dissertation reveal no role for long term isolation, or genetic revolutions over short periods of time, in stickleback speciation. Analysis of a polytypic population of stickleback showed that geographic isolation may not be necessary at all. These data indicate that speciation in stickleback is largely a result of adaptive phenotypic evolution, and that speciation in stickleback is primarily ecological, with geographic isolation playing an auxiliary, not causative, role.