Experimental Tests of Speciation Mechanisms in Drosophila melanogaster

Since Darwin published the Origin of Species, biologists have contended that divergent natural selection is a common force in creating reproductive isolation. Others have disagreed, arguing for other evolutionary mechanisms instead. We used replicate outbred populations of Drosophila melanogaster that have been experimentally evolved for hundreds of generations under contrasting as well as parallel selection regimes, in order to test the importance of divergent selection compared to other evolutionary mechanisms in initiating reproductive isolation between allopatric populations. There is extensive genome-wide differentiation both between and within these groups of populations, according to analysis of their single nucleotide polymorphisms (SNP). Parental populations and their crosses were phenotypically assayed for the following characters: (1) mate-choice, (2) mortality, (3) fecundity, and (4) developmental success. In Chapter 1 we provide a literature review of the theories of speciation. We make a distinction between Darwin’s theory of ecological speciation and the “null” theories of speciation in which ecological differentiation plays no role in the evolution of reproductive isolation. In Chapter 2 we tested the null theory of speciation by conducting within treatment crosses of three groups of Drosophila melanogaster populations. We found only four population crosses out of 15 that demonstrated at most modest evidence for the null theory of speciation. In Chapter 3 we tested the ecological speciation hypothesis by conducting between treatment crosses of two groups D. melanogaster populations that have adapted to different selection regimes. We found strong evidence for the role of ecologically divergent selection among all population crosses in producing incipient reproductive isolation. In Chapter 4 we tested for the interaction between ecological selection and other evolutionary mechanisms in speciation. We did not find a statistically greater signal of reproductive isolation when the populations involved in any of these crosses had a greater time since their last shared common ancestor. This result suggests that there is no detectable interaction of time and selective differentiation in initiating reproductive isolation. Overall, our conclusion is that differences in selection regime have greater relative importance than evolutionary time in fostering reproductive isolation between allopatric populations