Ecological drift is a potentially important but poorly understood process that may alter the diversity and composition of natural communities. Despite a rich history in evolutionary biology, the study of drift as an ecological process has largely been limited to neutral theory. The lack of integration of ecological drift with other processes that structure species composition, such as competition and dispersal, has resulted in poorly defined conceptual models about the effects of drift on community composition. While we know that demographic stochasticity, which causes drift, can have large impacts on individual populations, it is not known whether these impacts increase, decrease, or remain constant as a result of interactions with other species. This uncertainty led us to design a study that couples a field experiment with modeling to ask: 1) how important is drift in isolated communities, and how does it scale with community size? and, 2) are species interactions expected to increase or decrease drift, and through what mechanisms? Our field study is based in an annual grassland. We designed 135 replicate plant communities that differed in size (total number of individuals), but contained identical initial densities of viable seed from six plant species.
Results/Conclusions
The first two years of our study indicate a much more important role for drift in the smallest communities: the variance in relative abundances in small communities was from 3 to over 20 times greater than in large communities. This variation was largely driven by germination dynamics in the first year. In the second-year, drift was driven by germination dynamics and individual-level variation in seed production, with the joint effects of these processes overwhelming those of germination alone. Our modeling suggests that interactions amongst competing species will most often lead to greater levels of ecological drift than would be expected in the absence of competitors, due to depressed population sizes. However, for a given equilibrium population size, the degree of ecological drift tends to be lower than expected when competitive dynamics confer a rare-species advantage. Overall, our results indicate that, much like genetic drift, the effects of ecological drift are strongly tied to the size of a community (i.e. total number of individuals). However, the diverse effects of species interactions on ecological drift suggest that drift may vary among communities with different constituent species or environmental conditions due to the nature of species interactions within those communities.