Results/Conclusions We found ample evidence for non-random spatial pattering, and the type and frequency of spatial association patterns changed with scale. Trees of most species pairs co-occurred at all spatial scales investigated less than expected by chance and positive associations were rare at small neighborhoods. When selectively testing for small-scale effects we found that one third of all species pairs maintained significant and mostly negative small-scale associations. Negative associations occurred generally between species which shared one or more attributes such as light guild, fruit type, or dispersal agent. Selectively testing for positive larger-scale association revealed that only 8% of all species pairs co-occurred at large scales.

       The finding that aligned species showed frequently significant and negative small scale associations provides strong evidence for the occurrence of competitive interactions. This is not in accordance with neutral theory. Analogous analyses of data from tropical forests, however, have shown that only roughly 5% of all species pairs maintained significant small-scale associations. This suggests that the degree of neutrality in a forest may increase with species richness. The techniques presented here allow for a detailed analysis of spatial association that can be widely applied to gain a better understanding of species coexistence.