Thursday, August 6, 2009 - 11:10 AM

COS 92-10: Modeling food-web interactions in a metacommunity

Pradeep Pillai, Andrew Gonzalez, and Michel Loreau. McGill

Background/Question/Methods

Food webs are often investigated without regard to the spatial scale at which the various trophic interactions they encompass occur. Considering food webs from a spatial perspective may be one way to explain the noted discrepancy between theoretical predictions that food webs should be species poor with few interactions, and observations from field studies demonstrating actual food webs as being species rich with numerous, complex interactions. Most empirically compiled food webs are actually aggregate representations of all possible feeding relationships, and do not consider the variation in trophic links that can occur over time and space.  Studying food webs from a metacommunity perspective may allow us to incorporate such variation, as well as to consider how food web properties, like species richness and network complexity, can arise due to the spatial subdivision of trophic interactions across a landscape. We introduce a metacommunity framework based on an extension of the classic metapopulation (patch-dynamic) models of Levins and Culver (1971) in order to model the effects of spatial variation in food web interactions. In our approach food webs are an aggregation at the regional scale of interactions that occur at the local community scale. Dispersal between local communities allows species, and hence their accompanying trophic interactions, to persist at the regional scale despite the interactions being locally unstable.  

Results/Conclusions

We first demonstrate the application of our approach to the study of simple food web modules, showing how selective pressures on the dispersal ability of both consumers and resources vary depending on the configuration of trophic interactions that both are embedded within.  Strong and locally unstable consumer-resource interactions that drive both predator and prey locally extinct at high frequencies will lead only to strong positive selective pressures on predator dispersal rates, but little or no selective pressure on prey dispersal rates; the selective pressures on  both predator and prey are reversed when an alternative prey resource for the predator is present , as in the case of intra-guild predation (IGP). We then demonstrate, using a spatially-explicit lattice-based version of the model how the richness and network properties of more complex food webs, aggregated at the metacommunity scale, vary with the scale and rate of dispersal.