Visualizing environmental correlates of species geographical range limits

Although many studies have treated aspects of species geographical distributions and numerous approaches exist for understanding overall ecological correlates of distributions, software tools for exploring environmental correlates of distributional limits are relatively few. We focused on the challenge of understanding spatial correlates of distributional limits, and developed an extension to ARCVIEW that provides a simple, univariate test and visualization for such explorations. The 'Boundary U-test' seeks out environmental variables that show steep gradients associated with user-defined boundaries across geography. We illustrate the tool and its applications with an example of the likely historical distribution of Mexican wolf (Canis lupus).