Abstract

With many species interacting in nature, determining which describe community dynamics is nontrivial. By applying a new Bayesian-sparse modelling approach to an extensive field survey, we assessed the importance of interactions from con- and hetero-specific plants, pollinators, and insect herbivores on plant performance. We compared the inclusion of the interaction effects as aggregate "generic" terms versus specific terms. We found that a continuum of positive to negative interactions, containing mostly generic but a few strong specific interactions, was sufficient to describe variation in plant performance. While interactions with herbivores and conspecifics varied from weakly negative to weakly positive, heterospecific plants mainly promoted competition and pollinators facilitated plants. The consistency of these empirical findings over three years suggests that a broad resolution, including the generic effects of guilds and a few specific groups rather than all pairwise and high-order interactions, can accurately describe species variation in plant performance across natural communities.