Dispersal allows species to immigrate and emigrate to and from habitat patches and is an important factor in determining community structure. The influence of species dispersal in a metacommunity is poorly understood, particularly its effect at the local community level. We aimed to address this deficiency by evaluating the potential influence of dispersal on local community structure in a rock pool metacommunity. Short term dispersal was quantified over an 11 day period by intercepting propagules dispersing via overflowing pool water and via wind. The composition of dispersing organisms was compared to natural local communities in the rock pools surveyed annually on 11 occasions. On average, the composition and abundance of dispersing species was approximately 54.1 ± 9.3% (mean ± SD) similar to the established pool community. Some species were more abundant among dispersers than they were in the pool community. This may be attributed to several factors including a variation in tolerance to environmental conditions, dispersal capacity, and local scale species interactions (predation and competition). In general, we found considerable similarity between short term dispersal and long term local community structure across a metacommunity. Differences in abundance patterns between the resident rock pool community and the dispersal assemblage emphasize that dispersal, a regional process, must interact with local factors in structuring communities.
The contribution of local (e.g., competition) and regional (e.g., dispersal) processes in the structure of communities remains an unresolved issue. In general, a tendency to assume local processes to be deterministic and regional to be stochastic dominates, although it is challenged. Fortunately, it can be cast as a testable proposition: if correct, the degree of determinism in the final community structure might indicate which process is more prominent in the control of community structure. However, recent findings have also suggested that stochastic patterns can arise from local processes and that dispersal can homogenize communities, which would make them appear deterministic irrespective of the mechanism involved. To evaluate these competing expectations we conducted an experiment where the initial communities had the same composition and species abundances. We hypothesized that if local processes dominate, then arrays of communities will show divergence of community structures whether connected by dispersal or not (i.e., being fully isolated). Alternatively, if regional processes dominate, the dispersal connected communities should converge while isolated ones should not. We found, however, that both groups of experimental communities showed similar patterns of change - a decline in similarity and a tendency to diverge. This suggests that biological interactions, demographic stochasticity, or both, exert noticeable control over community structure such that they reduce similarity among replicate communities and diversify their final states. We speculate that these mechanisms enhance potential for species additions, particularly in conjunction with factors such as dispersal and the size of the regional species pool.