Predatory cues have repeatedly been shown to alter behavioral, morphological, and life history traits in prey. These trait changes are especially prevalent in aquatic ecosystems, as prey can often detect chemical signals from predators released into the water. These individual-level changes in prey have been shown to adjust consumption rates by predators, and can modify population trajectories of predators and prey. However, in spite of this body of research, little is known about how individual-level changes in different prey species affect the overall structure of natural communities, and how this may affect ecosystem processes. We present the results of a field experiment using the aquatic, detritus-based communities housed in the leaves of tank bromeliads from the neo-tropics. Bromeliads are natural microcosms, and allow highly replicated, community level experiments to be carried out in the field. Within these communities, larval damselflies are voracious predators consuming larval mosquitoes, chironomids, tipulids and scirtid beetles. In order to tease apart the trait-mediated and direct effects of the damselfly predators, we divided our plants into 3 treatments – caged predators (trait-mediated effects only), free predators (direct and trait-mediated changes), and controls (no predator effects).
Results/Conclusions
We show using non-metric distance scaling that the presence of predators significantly altered the structure of the detritivore community. Both caged predator and free predator treatments differed significantly from controls, however, were similar to each other. This demonstrates the high contribution trait-mediated effects make to the total effect of predators. We detected no significant differences in the total density of detritivores between all treatments. However, the relative abundance of various species differed significantly. In the presence of caged or free predators, detritivore species richness was greatly reduced, the mosquito community was dominated by Culex jenningsii while all species of Anopheles were absent, and numbers of chironimids and tipulids were reduced. We also found that the trait-mediated changes to community structure cascaded down the food chain and affected ecosystem processes, specifically CO2 production. Both caged and free predator treatments produced significantly higher levels of CO2 than controls, however they were not significantly different from each other. We believe this is because the predator-related reduction in detritivore species richness reduced grazing rates on CO2-producing micro-organisms, increasing their densities.