Food web dynamics in correlated and autocorrelated environments

The densities of populations in a community or food web vary as a consequence of both population interactions and environmental (e.g. weather) fluctuations. Populations often respond to the same kinds of environmental fluctuations, and therefore experience correlated environments. Furthermore, some environmental factors change slowly over time, thereby producing positive environmental autocorrelation. We show that the effects of environmental correlation and autocorrelation on the dynamics of the populations in a food web can be large and unintuitive, but can be understood by analyzing the eigenvectors of the community (system) matrix of interactions among populations. For example, environmental correlation and autocorrelation may either obscure or enhance the cyclic dynamics that generally characterize predator-prey interactions even when there is no direct effect of the environment on how species interact. Thus, understanding the population dynamics of species in a food web requires explicit attention to the correlation structure of environmental factors affecting all species. (C) 2003 Elsevier Inc. All rights reserved.