Modelling animal populations in changing landscapes

Abstract

Models of Mobile Animal Populations (MAP models) simulate long‐term land use changes, population trends and patterns of biological diversity on landscapes of 103‐105 ha. MAP models can incorporate information about past land‐use patterns and management practices and can project future patterns based on management plans. We illustrate this approach with an example of how implementation of a U.S. Forest Service management plan at the Savannah River Site in South Carolina, U.S.A., might influence population trends of Bachman's Sparrow Aimophila aestivalis, a relatively rare and declining species in southeastern pine forests. In this case, a management plan, largely designed to improve conditions for an endangered species, Red‐cockaded Woodpecker Picoides borealis, may have a negative impact, at least in the short term, on another species of management concern, Bachman's Sparrow.In a parallel processing version of the MAP models, a single landscape that would ordinarily be too large or detailed to be simulated on a single computer is subdivided into a number of smaller landscapes, and each landscape is simulated in parallel, either on a single multi‐tasking machine or on a group of networked machines. With this approach we are attempting to determine just how large a landscape must be before the dynamics of a population within it are more or less independent of factors beyond the landscape boundaries.