Interactive Effects of Dragonfly Predation in Expermental Pond Communities

Abstract

One goal of ecology is to generate dynamic models that predict the species composition of complex natural communities. The general applicability of such models will depend on how often unpredictable higher order interactions occur between competitors and predators. I experimentally tested for interactions in a two predator—multiple prey community by manipulating the presence or absence of two species of predatory dragonfly larvae (Anax junius and Tramea carolina) in replicated outdoor artificial ponds containing identical number of four species of anuran tadpoles and a diverse zooplankton assemblage. The impact of predation on prey communities was measured by the relative abundances and larval performances (size at metamorphosis, length of larval period, and growth rate) of the anurans, and by the densities of zooplankton sampled at 10—d intervals. Zooplankton reached high densities in the experimental ponds (up to 2700 animals/L). Dragonfly predation had a marginally significant impact on zooplankton species composition, and the effects on zooplankton size structure were not significant. Predation by Anax and Tramea profoundly modified the species composition of anurans: the relative abundance of toads (Bufo americanus) increased in the presence of predators, as two hylid species (Hyla crucifer and Pseudacris triseriata) decreased. The interaction between Anax and Tramea on anuran species composition was additive. The performance of surviving larvae of two anurans (Bufo and Pseudacris) was significantly improved by predation (growth rate increased and larval period shortened), whereas the performance of two others (Hyla and Rana utricularia) showed no response to dragonflies. The absence of strong interactions between the predatory odonates implies that in some cases it may be valid to build models of complex ecological communities by additively combining results from more simple subsets of species.