Can current food web models explain the structural stability of observed food webs?
Fox, Jeremy*,1, 1 University of Calgary, Calgary, AB, Canada
ABSTRACT- Qualitiative food webs describe ′who eats who,′ thereby summarizing community tophic structure. Qualitative food webs from different systems exhibit similar, non-random structural features, suggesting that general rules govern food web structure. Current food web models successfully predict many aspects of food web structure (e.g., number of omnivorous species) from knowledge of species richness and connectance (fraction of possible predator-prey links that are actually observed). Here I examine the ability of three food web models (cascade, niche, and nested hierarchy models) plus a random null model to predict a new measure of food web structure, structural stability. Structural stability summarizes ′overall′ qualitative food web structure in a single number, and is sensitive to small changes in that structure. Structural stabilities of observed webs differ significantly from those of random webs, and are poorly predicted by food web models. Poor model performance partially reflects bias (e.g., systematic over- or under-prediction of structural stability), which might be corrected by modifying the models. However, poor performance also reflects the fact that observed webs of similar species richness and connectance can vary greatly in structural stability, while model webs cannot. This variability cannot be accounted for by a model that takes species richness and connectance as the only free parameters, except by adding random noise that would degrade predictive precision without providing ecological insight. Further improvement in our understanding of the determinants of food web structure likely will require the development of new models that do not take species richness and connectance as the only free parameters.
Key words: food webs, structural stability, connectance, trophic structure
All materials copyright The Ecological Society of America (ESA), and may not be used without written permission.