Phylogenetic tools elucidate effects of trophic interactions on community assembly

Background/Question/Methods

Ecologists are increasingly interested in both food webs and explaining species co-occurrence patterns based on their traits. Food webs describe which species each species eats and is eaten by.  These interactions are fundamental traits of species. Here, we apply tools from trait-based community phylogenetics to the study of food webs to investigate how co-occurrence patterns may illuminate fundamental aspects of trophic niche space, community structure, and species' interactions. We do this by exploring how species’ trophic similarity within local communities differs from what would be expected under a random assembly model. We ask, "Are species in local communities more or less similar in their diets than expected by chance?" If more, species' co-occurrence may be restricted by the preys that are available. If less, species' co-occurrence may be driven by competition for food. We also ask, "Are co-occurring species more or less than randomly similar in what eats them?” If more, species co-occurrence may be enhanced by diluting predation pressure among species eaten by the same predators. If less, species co-occurrence may be driven by apparent competition by which prey more tolerant of predation out compete other prey less tolerant of predation by the same predators.    

Results/Conclusions

We find that our approach allows us to effectively test for “trophic underdispersion” (species in local communities are more trophically similar than expected by chance), and “trophic overdispersion” (species in local communities are more trophically dissimilar than expected by chance) within 50 food webs from 50 lakes from the Adirondack mountains of New York State (USA). Species were neither more nor less similar in what eats them than expected under random assembly, suggesting that the indirect effects of facilitation or competition mediated by consumers was not a major driving force in the assembly of the lakes. However, species' diets were much less similar than the random expectation. This relatively low trophic overlap suggests that competition for food reduces the likelihood of species co-occurring if they eat the same resource species and that competitive exclusion for food is a major driving force of assembly in the lakes. Our results shed light on the processes driving community composition and demonstrate the utility of integrating approaches from community phylogenetics into ecological network research.