While competition is ubiquitous in natural ecosystems, its consequence in the context of community assembly remains an issue of much debate. Existing theories predict that communities will have lower than expected functional or phylogenetic diversity (clustering) when interspecific difference in competitive ability is configured in a hierarchy, while communities will have high diversity (overdispersion) when similar species competes for similar resources. Such theories, however, have implicitly assumed competitive equilibrium. Here, we offer a new perspective on how competitive processes translate into community patterns by introducing the non-equilibrium paradigm.

To simulate community patterns during plant succession, we used a lattice-structured model in which interspecific trade-off between competitive ability and growth rate was incorporated. Our model predicted that the competitive hierarchy could drive both community overdispersion and clustering. These patterns were driven by the accumulation of the past competition history among neighbor individuals. Also, the observed patterns depended on spatial scales, tree sizes, and the definition of metacommunity. Our results highlight the importance of having a non-equilibrium perspective in inferring community assembly.