Non-zero-sum neutrality test for the tropical rain forest community using long-term between-census data

For community ecologists, "neutral or not?" is a fundamental question, and thus, rejecting neutrality is an important first step before investigating the deterministic processes underlying community dynamics. Hubbell's neutral model is an important contribution to the exploration of community dynamics, both technically and philosophically. However, the neutrality tests for this model are limited by a lack of statistical power, partly because the zero-sum assumption of the model is unrealistic. In this study, we developed a neutrality test for local communities that implements non-zero-sum community dynamics and determines the number of new species (N _sp) between observations. For the non-zero-sum neutrality test, the model distributed the expected N _sp, as calculated by extensive simulations, which allowed us to investigate the neutrality of the observed community by comparing the observed N _sp with distributions of the expected N _sp derived from the simulations. For this comparison, we developed a new "non-zero-sum N _sp test," which we validated by running multiple neutral simulations using different parameter settings. We found that the non-zero-sum N _sp test rejected neutrality at a near-significance level, which justified the validity of our approach. For an empirical test, the non-zero-sum N _sp test was applied to real tropical tree communities in Panama and Malaysia. The non-zero-sum N _sp test rejected neutrality in both communities when the observation interval was long and N _sp was large. Hence, the non-zero-sum N _sp test is an effective way to examine neutrality and has reasonable statistical power to reject the neutral model, especially when the observed N _sp is large. This unique and simple approach is statistically powerful, even though it only employs two temporal sequences of community data. Thus, this test can be easily applied to existing datasets. In addition, application of the test will provide significant benefits for detecting changing biodiversity under climate change and anthropogenic disturbance.