Print PageMost models studying the emergence of network complexity in food webs have ignored the spatial and temporal scales of food web interactions. Investigating food webs as spatially extended networks of trophic interactions can provide important insights into the assembly and evolution of complex trophic communities. Here we use a patch-dynamic metacommunity model to study the evolutionary assembly of spatial food web networks. Our metacommunity model assumes local patches contain relatively simple communities of trophically interacting species linked to each other through species movements and colonisation. The complexity of the full food web arises as a spatial aggregate of all local interactions at the metacommunity scale. Using this modelling approach we consider how a food web can be evolutionarily assembled at large spatial scales when we allow species to be defined along a size gradient where all species feeding relationships are structured both by size – with larger species feeding on smaller prey some distance below them along the gradient, and by niche width, where those species feeding on a wider array of prey resources are competitively inferior to those with a more specialised or narrower diet breadth.
Results/Conclusions
We demonstrate how food web complexity arising from network branching can evolve due to the structural support of omnivore and generalist links, and further demonstrate how omnivore feeding modules, far from being destabilising as predicted by previous theory, can actually provide critical support for network complexity in spatial food webs. We also show how in evolutionary assembled spatial food webs, omnivore and generalist feeding appears to be only selectively advantageous at higher trophic levels, whereas lower trophic species evolve towards resource specialisation.
