An Agent-Based Model of Ecological Disturbance

Peter T. Hraber

Abstract

Gene flow, the flux of biological information across the landscape, plays a key role in the evolution of ecosystem structure. As humans continue to interact with ecosystems in innovative manners, the consequences of drastically altered rates of gene flow are of increasing concern. Introduction of barriers, removal of habitat, and introduction of exotic or chimeral genotypes to a stable ecosystem are all means by which humans restructure their living surroundings. While it may never be possible to assess a priori. the effects of such disturbance events, ecology has had numerous successes in elucidating general priciples of ecosystem organization via theoretical models and numerical simulation.

One such model, still in development, is the Echo-class of adaptive agent simulations, developed by J. Holland (1994). Using this model, as well as a neutral model to control for specific assumptions in the Echo model, I consider a scenario in which isolated habitats are subjected to perturbation via migration. Species diversity exhibits periods of punctuated equilibria in the original model but not in the neutral model. Fluctuations in species diversity demonstrate a scaling relation between magnitude of fluctuation and probability of occurence. Scaling exponents describing fluctuations differ significantly across model types and disturbance regimes. The fact that ecosystems are driven by processes which interact at many scales may explain scaling properties of speciation and extinction.