The stability of real ecosystems

Most current theoretical work on whole ecological communities is based on a conception of community dynamics (see, for example, refs 1, 2) in which the community resides in a neighbourhood of equilibrium. However, there is very little evidence that this is a realistic viewpoint. To test this fundamental assumption ‘empirically’, I have constructed here plausible community matrices1,2 corresponding to 40 real food webs3 and studied their local stability properties. I find that stability (in the sense of tending to return to the presumed equilibrium after a small perturbation) is far more likely if the interaction strengths are chosen in accord with the nature of the particular organisms in each food web, rather than strictly at random. Either this is a monumental coincidence, or the equilibrium viewpoint really is appropriate for quite a few real communities. Additional results of this analysis are that stability generally requires a degree of intraspecific interference on the part of some consumer species, whereas interspecific interference tends to exert a destabilizing influence.