Dissimilarity-overlap analysis of community dynamics: Opportunities and pitfalls

Abstract

Do multiple different communities have similar dynamics? This novel question was recently addressed with the development of the dissimilarity–overlap curve (DOC) method, designed to identify sets of communities with similar (‘universal’) dynamics using pairwise comparison of communities’ composition. However, the theoretical properties of the method were tested using very restrictive and, in some cases, unrealistic assumptions. We test the ability of the DOC method to discriminate universal from non-universal dynamics in more realistic scenarios—under various levels and regimes of stochasticity and sampling errors, in cases where species respond to environmental gradients or geographical axes and in cases when only a fraction of the communities have universal dynamics. We also examine the ability of the method to test the Neutral Theory of Biodiversity. We found that multiple factors and processes that produce variability among observed communities lead in general to a spurious detection of universal dynamics. In particular, sets of communities subject to environmental gradients, differences in stochasticity and in sampling quality among communities, and cases in which only a small fraction of the communities are similar, may be detected as having universal dynamics. We suggest the DOC method as a potential for testing the symmetric interactions assumption of neutral or symmetric theories, superior to the common techniques which are based on species abundance statistics. Despite the promise of the DOC method, interpreting its results may be challenging and further developments are needed to address the question of similarity in dynamics.