Component models of assembly in a metacommunity of tropical arboreal ants

Background/Question/Methods: Empirical community studies testing for the occurrence of all four metacommunity paradigms (species sorting, patch dynamics, mass effects, and neutral) have found patterns not clearly fitting the predictions of any one paradigm. This may be because all four may be acting simultaneously to structure metacommunities. One way this may occur is if component subsets of species within the metacommunity each follow a different assembly model. Here, we utilize the manageable complexity of a shaded tropical coffee agroecosystem to asses the extent to which component species in a community of arboreal twig-nesting ants in coffee can be described by each of these four paradigms. Results/Conclusions: Data from a survey of the community indicate that most species utilize equivalent nest sites, show a uniform response to a gradient in the density of nest sites and to disturbance and that there is close correspondence between local species colonization, nesting, and co-occurrence patterns and metacommunity abundance. These data, coupled with laboratory experiments and simulations, further indicate random overall patterns of species co-occurrence, nonequilibrium priority effects for nest sites overriding a dominance hierarchy, and an absence of community-wide competition-fecundity or competition colonization trade-offs. However, a finer scale species level analysis suggests metacommunity dynamics are dominated by the commonest four species, whereas rare species may persist via mass effects from ant communities in the shade tree canopy. Additional experimental and survey data suggest a role for patch and/or neutral dynamics occurring between some species pairs and for species sorting occurring at a landscape scale along a gradient in the management intensity of coffee farms (shade to sun coffee). In sum, these data support a component species approach to synthesizing metacommunity paradigms.