Landscape connectivity: A conservation application of graph theory

We use focal-species analysis to apply a graph-theoretic approach to landsc ape connectivity in the Coastal Plain of North Carolina. In doing so we dem onstrate the utility of a mathematical graph as an ecological construct wit h respect to habitat connectivity. Graph theory is a well established mains tay of information technology and is concerned with highly efficient networ k flow. It employs fast algorithms and compact data structures that are eas ily adapted to landscape-level focal species analysis. American mink (Muste la vison) and prothonotary warblers (Protonotaria citrea) share the same ha bitat but have different dispersal capabilities, and therefore provide inte resting comparisons on connections in the landscape. We built graphs using GIS coverages to define habitat patches and determined the functional dista nce between the patches with least-cost path modeling. Using graph operatio ns concerned with edge and node removal we found that the landscape is fund amentally connected for mink and fundamentally unconnected for prothonotary warblers. The advantage of a graph-theoretic approach over other modeling techniques is that it is a heuristic framework which can be applied with ve ry little data and improved from the initial results. We demonstrate the us e of graph theory in a metapopulation context and suggest that graph theory as applied to conservation biology can provide leverage on applications co ncerned with landscape connectivity. (C) 2000 Academic Press.