Integrating edge detection and dynamic modeling in quantitative analyses of ecological boundaries

Abstract

Habitat boundaries profoundly influence the structure and function of landscapes, influencing ecological processes both locally and over larger scales. In addition, boundaries themselves are dynamic entities whose changes can influence diverse populations, communities, and ecosystems by way of feedback effects. These two issues, scale dependence and spatiotemporal dynamics, underlie much of the now considerable attention that modelers and statisticians have devoted to the quantitative study of ecological edges and boundaries. We present the linkages between methods of delineating boundaries, monitoring boundary changes, and modeling edge-related dynamics. In the process, we clarify statistical and mathe matical approaches to the study of ecological edges and boundaries, and we discuss important remaining issues in the area of quantitative edge research. In particular, we address conceptual and methodological problems faced by statisticians and modelers, while highlighting topics that would benefit from a collaborative approach.