Landscape pattern, spatial behavior, and a dynamic state variable model

Abstract

Evaluation of landscape pattern for wildlife is a difficult problem, one most often addressed by spatial indexes. Models of animal movement may be an alternative way to assess the impacts of landscape pattern. A dynamic programming model is formulated in this paper to address such factors as edge effects and travel cost, as well as conditions of patches such as food resources and predation. The approach works well for species that prefer a mix of habitat types, such as the northern bobwhite ( Colinus virginianus). The model, with parameters estimated for quail in the Great Plains of the United States, shows promise for evaluation of agricultural landscapes for suitability for the animals. The model also suggests hypotheses about the effects of spatial arrangement of habitat elements on wildlife.