NESTED SPECIES SUBSETS AND GEOGRAPHIC ISOLATION - A CASE-STUDY

Insular systems tend to exhibit nonrandom patterns of species composition in which smaller species assemblages contain successive subsets of those found on richer ones. The fact that such ''nested'' patterns of species composition have usually been observed in systems where species number is correlated with island size has led to the hypothesis that differences among species in the effect of island size on the probability of extinction represent the main mechanism producing nested patterns of species composition. This study tested the hypothesis that differences among species in dispersal ability may interact with geographic isolation to produce nested subsets of species composition, This hypothesis was tested in a system of woody plants inhabiting artificial islands that were created by the filling of the Clarks Hill Reservoir, Georgia, USA. All of the studied islands had been logged prior to their separation from the mainland and their species number 34 yr later was found to be negatively correlated with distance to the closest mainland. The results of the present analysis indicated that (1) species occupancy of the whole island system was highly nested, (2) nestedness occurred with respect to geographic isolation but not with respect to island size, (3) only a small portion of the overall species included in the study contributed to the observed nestedness, and (4) the degree to which individual species contributed to the observed nestedness was correlated with their dispersal properties, with wind-dispersed species showing no evidence for nestedness and species lacking adaptations for long-range dispersal showing the strongest patterns of nestedness. These results indicate that isolation effects may interact with species-specific dispersal properties to produce nested subsets of species composition.