Niche partitioning and the storage effect facilitate coexistence in an amphibian community.

Virtually all natural community assemblages are dominated by a handful of common species. Dominant species can exert negative impacts on biodiversity through competitive exclusion, and thus there is a strong incentive to understand imbalances in community composition, changes in dominance hierarchies through time, and mechanisms of coexistence. Pond-breeding amphibians that utilize ephemeral wetlands provide an excellent opportunity to evaluate theoretical predictions of community composition in stochastic environments. One of the most striking features of pond-breeding amphibians is the marked stochastic fluctuations in abundance across years. Given strong theoretical and empirical links between evenness and biomass, one would expect community evenness to change from year to year. Moreover, if different species exhibit different boom-and-bust reproductive cycles, then a storage effect may help to explain why one species does not outcompete all others. Here, we explore the interplay between biotic and abiotic conditions in shaping amphibian communities at two ephemeral wetlands on Eglin Air Force Base, Florida. We document consistent community composition over 6 years of monitoring, resulting from a lack of species turnover and similar responses of all community members to environmental conditions. The similar dynamics of species argues against a storage effect as the sole mechanism for coexistence and instead points to niche partitioning as a more important factor. In support of this conclusion, we show that the degree of synchrony in breeding migrations only correlates with environmental conditions within species, not between species. The lack of pattern seen between species implies that individuals are somewhat constrained in the timing of breeding migrations, perhaps owing in part to competition with other community members. We hope that our work reinvigorates interest in amphibian communities and highlights ephemeral wetlands as model systems to study community dynamics in stochastic environments.