Understanding community responses to altered disturbance regimes will be crucial for managing the effects of global change. Models of community assembly across environmental gradients predict biotic interactions should be stronger in physically benign compared to stressful habitats. However there have been surprisingly few empirical investigations, especially of temporal variance in interaction strength. The strength of biotic interactions in physically stressful habitats is likely to be more dynamic than in stable habitats, due to succession after disturbances, variable habitat morphology, and the onset of seasonal time constraints. We investigated temporal shifts in interaction strength by examining predator-prey interactions in ponds varying in habitat permanence due to drying. In our study sites, permanent ponds have more diverse guilds of larger predators than temporary ponds, which may result in stronger predator-prey interactions. However, the strength of predator-prey interactions are likely to increase in temporary habitats over time as predator size and biomass increases after colonisation, habitats contract, and individuals become more active as ponds dry. To test these hypotheses, we manipulated predator biomass in replicated cages within multiple temporary and permanent ponds, three times over the season of filling and drying.

Results/Conclusions

The biomass of predators in temporary ponds soon after filling was two orders of magnitude lower than in permanent ponds; however our assays indicated no difference in predator impact on prey between the two habitat types. Over time, predator biomass converged between permanent and temporary ponds due to a 10 fold increase in predators in temporary ponds, combined with a 50% decrease in biomass in permanent ponds. Similarly, predator size increased over time in temporary ponds, but remained static in permanent ponds. Changes in the predator fauna were reflected in predator impact, which increased dramatically during the drying phase of temporary ponds, ending up an order of magnitude higher than in permanent ponds. These results indicate controls of community assembly in ephemeral habitats are likely to shift from opportunistic colonisation processes to deterministic biotic interactions as time constraints approach. Furthermore, temporal shifts in the strength of biotic interactions over an environmental gradient may result in more intense interactions in physically stressful than benign habitats. These conclusions indicate contemporary models of community assembly will have to consider the temporal, as well as spatial dynamics of biotic interaction strength, in order to predict the consequence of alterations in disturbance regimes.