Disentangling multi-species aggregate versus overlapping distributions

Aim: Single-species distributions are well known to present random, aggregate, or regular patterns. However, at the community level, multi-species aggregate distribution patterns are not simply the lumped aggregate distribution of each species in the community. Actually, the superimposition of specimens of different species on a distribution map can present an interacting distributional pattern: different species may overlap with each other (or conversely, avoid each other). To this end, other than aggregation, it is necessary to further quantify distributionally overlapping patterns of different species at the community level. Methods and results: In this study, we combined the conspecific-encounter index v and Moran's I index to disentangle community-level distributional aggregation versus superimposition. Numerical simulations demonstrated that the two indices, calculated from biodiversity data collected using the sequential sampling of specimens along line transects, were effective and complementary in predicting multi-species distribution patterns. The conspecific-encounter index was sensitive to the change in the degree of distributional aggregation while being insensitive to the change in the distributionally overlapping pattern of different species at the community level. In contrast, Moran's I index can detect the change in the degree of distributional overlap while being insensitive to the change in the level of distributional aggregation at the community level. Moreover, through additional simulations, we showed that the conspecific-encounter index was robust despite many specimens from dominant species being omitted in the sampling. Finally, an empirical test on the Barro Colorado Island (BCI) forest plot in Panama revealed that tree assemblages in that region presented both random and non-overlapping distributions, the latter being reported herein for the first time. Conclusions: The performance of Moran's I index and the conspecific-encounter index are complementary when disentangling multi-species aggregate versus overlapping distributions at the community level.