Speciation and the evolution of male breeding coloration in darters

The means by which sexual selection can promote speciation is a topic of much debate and study. Darters, a species-rich clade of freshwater fishes that exhibit widespread sexual dichromatism, may represent a prime example of speciation by sexual selection. I investigated mechanisms of sexual selection and reproductive isolation in two closely related, ecologically and behaviorally similar species: the rainbow darter (Etheostoma caeruleum) and the orangethroat darter (E. spectabile).;First, I quantified variation in male breeding coloration across multiple populations of E. caeruleum and E. spectabile using digital photography and image processing software. Male color traits in both species can be categorized as "blue" and "red" components, which are based on chromoprotein and carotenoid pigments, respectively. Blue colors tended to be more conserved and better correlated with body size, whereas red colors tended to be more diverse among species and populations and less correlated with body size. These disparate patterns suggest that the blue and red components of male coloration serve different functions and/or are subject to different selective forces.;I then investigated the specifics of behavioral isolation between E. caeruleum and E. spectabile, and whether species discrimination is mediated by male coloration. I observed interactions between male E. spectabile, female E. spectabile, and male E. caeruleum under different lighting conditions that permitted or impaired the fish's perception of red coloration. Species recognition was controlled by males, who preferentially pursued conspecific females and directly aggression toward conspecific males. Females did not exhibit overt preference for any type of male, instead responding proportionately to the amount of male pursuit. Disrupting color perception inhibited male-male aggression, suggesting that male coloration allows males to identify conspecific rivals in the context of male-male competition.;As previous studies have found no effect of female choice on male fitness in E. caeruleum or E. spectabile, I tested whether male-male competition underlies sexual selection. I allowed multiple male E. spectabile to compete for a single female and looked for phenotypic correlates of reproductive success. Relatively larger males were competitively superior and initiated more spawning events with females. After correcting for size, I also found that the relative quality and quantity of male coloration was correlated with success in guarding the female against rivals and in spawning with the female. These results indicate that male coloration is under sexual selection through male-male competition.;Apart from sexual selection, behavioral isolation may also evolve through natural selection against heterospecific mating, i.e. reinforcement. I performed conspecific and heterospecific crosses using E. caeruleum and E. spectabile and compared reproductive isolation between these species in a sympatric population versus allopatric populations. Reproductive isolation was high between female E. caeruleum and male E. spectabile in both sympatry and allopatry. In contrast, reproduction isolation was high between female E. spectabile and male E. caeruleum in sympatry but low in allopatry. This pattern is consistent with reinforcement acting asymmetrically on behavioral isolation between these species.;Finally, I conducted a long-term investigation of postzygotic isolation between E. caeruleum and E. spectabile. I created purebred and hybrid F1 fish via artificial fertilization and raised them under common garden conditions. Fertilization success, hatching success, and larval survival to 10 months did not differ between purebred and hybrid families, indicating little hybrid inviability. However, hybrid families from both reciprocal heterospecific crosses were heavily biased toward males, while the purebred families did not differ from the 1:1 sex ratio found in nature. This sex ratio distortion suggests that heterospecific spawning between E. caeruleum and E. spectabile is detrimental, and provides a possible mechanism for driving reinforcement between these species.