Influence of the direct and indirect effects of interspecific interactions on life history evolution in a Trinidadian killifish (Rivulus hartii)

Predators influence evolutionary change via direct consumption of prey. However, indirect effects of predators and competitors often alter growth rates and food availability, which provide an additional link to evolution. The killifish, Rivulus hartii, are located across a gradient in predation in Trinidad. Rivulus are located in: (1) 'high predation' sites with large piscivores, (2) Rivuluslguppy' sites with guppies, and (3) 'Rivulus-only' sites with just Rivulus. Rivulus suffer higher mortality with predators, and guppies likely prey upon young Rivulu,s in Rivuluslguppy environments. In turn, population densities decline while growth increases in both localities compared to Rivulus-only sites. To explore how these interactions influence evolution in Rivulus, I utilize a multifaceted approach consisting of field and laboratory components.;Rivulus phenotypes from high predation sites exhibited a smaller size at maturation and a higher reproductive allotment than Rivulus-only sites. A common garden experiment demonstrated these differences are heritable and such divergence is consistent with theoretical predictions consider predation on adults. To determine if the indirect effects of predators influence evolutionary change, I reared 2nd generation born fish under two food levels that approximated natural differences in resources. This approach demonstrated that the magnitude of trait divergence between high predation and Rivulus-only sites were either reduced or disappeared when Rivulus were fed a ration that approximated growth in a Rivulus-only environment, which indicate that the indirect effects of predators have evolutionary consequences.;Rivulus from sites with guppies have evolved earlier maturation and increased reproductive investment compared with Rivulus -only localities. Such divergence is consistent with the predictions from theory that incorporate changes in growth, population density, and resource availability. Moreover, common garden experiments show that these differences were only apparent on a food level typical of Rivulus/guppy environments; the opposite pattern was observed when food was reduced to levels approximating Rivulus-only environments. Such interactions further argue that differences in resource availability dictate evolution in Rivulus. To prove that guppies cause evolution in Rivulus, I evaluated the response to selection after guppies were introduced into two Rivulus-only tributaries. This perturbation caused a similar trajectory of evolution, including analogous food-dependent patterns of trait variation within 25 years.