The patterns of omnivory
When discussing food webs it is convenient to define an omnivore as a species that feeds on more than one trophic level (Pimm & Lawton, 1978). In some food webs, there are few omnivores ( Fig. 7.1(a)), while in others omnivores are common ( Fig. 7.1(b)). This chapter examines patterns of omnivory using the method outlined earlier, that is, real webs are predicted to have those structures that enhance stability in models. This method gives several predictions that can be tested against real food web structures. These predictions are supported by the data, and this constitutes the most certain evidence that dynamics impose constraints on food web structures. Moreover, the success in predicting specific patterns of omnivory provides the least equivocal evidence that increasing complexity decreases the likelihood of stability in the real world. In Chapter 5, I argued that the relationship of complexity to species number, which is suggested by dynamical considerations, could as easily be explained by simple biological assumptions. Now, other things being equal, increased omnivory means increased complexity. If omnivory is constrained by dynamics, then so is complexity. Unlike the patterns of complexity, the more specific patterns of omnivory cannot be so easily related to simple biological assumptions. Only dynamical constraints predict the detailed patterns of omnivory observed in nature.
KeywordsTrophic Level Stable Model Return Time Capita Effect Food Chain Length
Unable to display preview. Download preview PDF.