Initial colonization constraints on food web assembly

Background/Question/Methods

Island biogeography theory predicts area and isolation to drive assembly irrespective of traits, a view supported by aspects of metacommunity theory where initial colonization is mostly constrained by stochasticity while species sorting only occurs later in assembly. However, These assumptions are unlikely to hold true for food webs where interactions between trophic levels constitute an important deterministic and immediate constraint in the early stages of assembly. Here, we describe these interactions in a 25 ha large-scale meta-community island experiment in tallgrass prairie, testing for biogegraphic versus diet-based deterministic limitations on plant and arthropod assembly. The design includes three different island sizes (25,100 and 400 m²), and random isolation distance from the mainland (source of propagules), allowing us to decouple the effects of size from distance.

Results/Conclusions

We showed that food web structures derived from mechanisms operating regionally in a way such that each local community structure could not be predicted solely by diet-based traits or by the spatial characteristics of the islands but rather by a complex interaction between both. Food web colonization was the result of biogeographic and trophic factors co-occurring to produce three distinct food webs: (i) grass dominated (ii) forb dominated, and (iii) communities of the above two with or without certain dispersal limited herbivores. This work clarifies how complexity shapes assembly through simultaneous and sequential interactions of spatial and trophic processes, consistent with recent theoretical predictions on trophic island biogeography