Community assembly is commonly viewed as a hierarchical process where species are first ‘filtered’ by the environmental conditions of a site based on their physiological tolerances. Those species that can tolerate the site then interact with one another as they compete for the limited resources within the site. However, predictive models of environmental filtering and species interactions have been developed in isolation even though they jointly influence community assembly. Given the known importance of interactions for influencing species distributions, the integration of these processes within a synthetic model would improve our predictions of ecological responses to global change. We propose that carrying capacities provide the conceptual and mathematical link for integrating interactions into models of environmental filtering because carrying capacities must vary along environmental gradients due to species physiological tolerances. Our new framework broadens the concept of carrying capacity to be the likelihood that a species can occur at a site based on the matching of its phenotype to the environment and is therefore analogous to a fundamental niche. Using Lotka-Volterra interaction matrices, these probabilistic carrying capacities can then be used to predict species relative abundances while accounting for competition.
Results/Conclusions
The model was tested in an ephemeral wetland in New Zealand, where rapid compositional turnover of a turf community occurs along a steep hydrological gradient. Environmental filtering along the gradient was driven strongly by variation in root aerenchyma among species, where species with high root aerenchyma were more likely to occur in frequently flooded zones. Within each local quadrat, we modeled the interactions by assuming a competitive hierarchy, where taller species would outcompete shorter species. The model predictions of species relative abundances were positively correlated with observed relative abundances, and therefore performed well in predicting species distributions along this hydrologic gradient. Our new quantitative synthesis offers a new way forward for predicting species and community-level responses to changing environmental conditions while simultaneously accounting for competitive interactions.