A framework for modelling ecological communities and their interactions with the environment

A generalized framework for representing the stochastic interactions between an ecological community and the environment is introduced. By modelling an ecosystem by a coupled system of stochastic differential equations, one can capture many of the ecosystem's salient features. For example, the framework permits each species within the community to have multiple developmental phases as well as for the environment to consist of several, interdependent environmental factors. This enables statistical inferences to be made on many pertinent ecological issues such as biodiversity and the allometric relationship between species abundance and biomass.The framework gives accurate predictions of the population number cumulants over time whilst avoiding the negative transition rates that often beset traditional moment closure approaches. This framework is used to analyze a predator-prey model subject to random environmental fluctuations. The prey is assumed to have both an adult and larval phase with the adults vulnerable to predation. The results obtained for this model are in keeping with conventional ecological theory—the abundances of r-selected species are more erratic than those of K-selected species. In addition, the r-selected species abundances are more correlated to the environment than the K-selected species abundances.

Research highlights▶ A statistical framework for modelling ecological communities is developed. ▶ The model accurately predicts the ecosystem's means, variances and higher cumulants. ▶ The framework may be used to study the biodiversity of an ecosystem.