The consequences of demographic stochasticity for communities: Linking theory and experiments

Benjamin Gilbert, Leila Z Forsyth and Denon Start, Ecology and Evolutionary Biology, University of Toronto, Toronto, ON, Canada

Background/Question/Methods
Ecological drift is a process that structures species diversity, yet ecologists have a limited understanding of how important drift is across habitats and communities. Drift arises from demographic stochasticity and the reduction or propagation of that stochasticity through species interactions. Because drift depends on both the purely random nature of demographic stochasticity and deterministic processes that govern the propagation of random fluctuations, its importance is likely to vary across ecosystems, among communities with differing types of organisms, and with the strength of species interactions. Ecological theory makes specific predictions about the importance and consequences of drift in some scenarios, but there is yet to be a synthesis across the diversity of ecological interactions seen in most communities.
We used a multi-year field experiment in which we manipulated the size of plant fragments to limit community size, creating a gradient in the expected importance of ecological drift. We use this experiment to test several questions that lack sufficient theory -- using a combination of subsampling techniques and dispersal barriers, we asked 1) how important is ecological drift for the distribution of diversity? 2) how does the importance of ecological drift change across trophic levels? and 3) does the effect of drift on species diversity also alter community-level trait distributions?

Results/Conclusions
Ecological drift had a strong influence on plant beta diversity, causing it to increase by 1.25 times in our smallest fragments, yet its influence on alpha diversity could not be detected. A similar effect of drift was seen in consumer communities (arthropod herbivores and their predators), and this effect was more pronounced when dispersal among fragments was prevented. Ecological drift increased with trophic level in consumer communities, causing a 4-fold increase in predator beta diversity relative to that of herbivores. Despite the effects of drift on species turnover and trophic diversity, their consequences for community-level trait distributions were muted for plants, suggesting that the influence of drift at lower trophic levels was counteracted by functional compensation.
Overall, our research documents the pervasive influence of ecological drift and highlights how this influence depends on the characteristics of the community studied and the interactions among its species. It also identifies shortcomings in our understanding of the consequences of demographic stochasticity that limit basic and applied ecology.