Evolution and maintenance of functional equivalence in simulated plant communities

Background/Question/Methods Understanding how functional traits of species may evolve under varying ecological conditions, as well as how those traits impact long-term coexistence directly addresses fundamental mechanisms driving community assembly. Traditional ecological theory has posited that species that are too ecologically similar cannot coexist, and that competition should lead to niche differentiation or to extinction of one of the competing species. However, Hurtt and Pacala (1995) suggested that dispersal and recruitment limitation can severely retard competitive exclusion. Hubbell (2006) has shown circumstances that can lead to the evolution of functionally equivalent species that subsequently coexist at relatively stable abundances for thousands of community turn-overs. However, a number of important questions remained, which we address here: 1. What is the impact of increasing numbers of species in a local community on trait evolution and species coexistence? 2. What is the impact of different strengths of dispersal limitation? 3. Do multiple uncorrelated environmental gradients and linked traits lead to a greater degree of specialization of species? Results/Conclusions Combining the unidimensional circle model implemented by Hurtt and Pacala (1995) with two evolving traits governed by multiple genes of small effect, we explore the interacting impacts of species richness, habitat heterogeneity, multiple, linked traits, and varying levels of dispersal limitation on trait evolution and community persistence. We find that fine-scale habitat heterogeneity always leads to broad ecological equivalence and long-term species persistence, while the outcome of simulations with coarse-scale heterogeneity depends on other factors. With weak dispersal limitation and no linkage, many species typically become extinct, and one or two species dominate most of trait and habitat space. With stronger dispersal limitation, and clumped initial species distribution, classical niche differentiation evolves, while random initial distribution of species leads to species displaying multiple ecotypes. So long as dispersal limitation is strong, most or all species persist for the duration of the simulations.