The comparative ecology of a killifish (Rivulus hartii) across aquatic communities differing in predation intensity

Background: Life-history theory predicts that populations experiencing different patterns of ageor size-specific mortality will evolve divergent life histories. Higher mortality can also cause indirect effects by reducing population density and increasing resources for survivors. How life histories evolve can ultimately be shaped by the interactions between the direct impact of predators on mortality, their indirect effects on resource availability, and the age specificity of these combined effects. Prior research on the killifish, Rivulus hartii, suggests differences among aquatic communities in both predation risk and resource availability but has failed to characterize the age specificity of these effects. Study organism and site: We studied Rivulus hartii in Ramdeen stream, a second-order tributary of the Arima River, on the south slope of the Northern Range Mountains in Trinidad. We used four sites near each other: in the two Rivulus-only sites, killifish were found alone; in the Rivulus/guppy site, guppies (a potential predator of juvenile Rivulus) also occurred; and in the high-predation site, Rivulus occurred with predators. Hypotheses: (1) If guppies shape the evolution of Rivulus life histories by increasing juvenile mortality rates and indirectly increasing food availability to the survivors, then juvenile mortality and growth rates should be higher when guppies are present than when they are absent. (2) If larger predators shape Rivulus life histories by selectively preying on adult Rivulus, then the added mortality associated with predators should be greater in the larger size classes. Methods: We performed mark–recapture studies on juvenile and adult Rivulus in the field (>12 mm total length), which allowed us to estimate size-specific mortality, growth rate, and density. Results: Rivulus experienced the highest mortality in the high predation site, but predation was not selectively focused on adults. Furthermore, the higher mortality was coupled with reduced population density and increased adult growth rates. In Rivulus/guppy and Rivulusonly sites, all size classes had the same survival rate. Laboratory study confirmed that adult guppies can prey upon hatchling Rivulus (<7 mm) and are restricted to preying on this size class. Lastly, juvenile Rivulus from Rivulus/guppy localities had higher growth rates than those from Rivulus-only localities, as predicted, but such an effect disappeared in the adult stage.