Testing generalities in the shape of patch occupancy frequency distributions

Bimodal patch occupancy frequency distributions regularly appear in studies of a variety of taxa at a variety of spatial scales. Metapopulation models associated with the core-satellite hypothesis and a strong rescue effect predict the occurrence of this bimodality. However, there have been numerous other explanations, both artifactual and biological, of bimodality in patch occupancy frequency distributions. One approach to a better understanding of the phenomenon is to conduct studies on a variety of taxa and spatial scales. Here we examine patterns of patch occupancy frequency distributions for plants, dung beetles, and birds at four spatial scales in sand forest and mixed woodland in Tembe Elephant Park and Sileza Nature Reserve in South Africa. We then examine the predictions of each alternative mechanism proposed to explain bimodality with respect to the patterns found in this study. The incidence of bimodality was high. Little evidence was found for any of the artifactual or nonmetapopulation-based explanations of bimodality. Although there were no between-taxon differences in the shape of occupancy distributions, there was a clear trend of decreasing frequency of bimodality with an increase in the size of the area examined. The general shape of range-size frequency distributions in combination with an area effect is suggested as a mechanism underlying this pattern. Metapopulation dynamics associated with a strong rescue effect could also not be excluded as a possible mechanism playing a role in generating the bimodality found at the patch scale.