A Metric for Analyzing Taxonomic Patterns of Extinction Risk

Spatial autocorrelation statistics can provide an efficient tool for exploring taxonomic patterns in extinction risk. Using Moran's I, we found that U.S. vertebrates exhibit little or no taxonomic clustering of extinction threat within orders, but much greater clustering within families and genera. Among amphibians, clustering was unusually high within families. Across classes, most groups had a similar degree of clustering at all taxonomic levels, despite their different evolutionary histories. Birds were the exception, with higher clustering at the genus and family level. Our results suggest that intrinsic traits play a large role in species endangerment. The taxonomic pattern produced can guide policymakers in their allocation of scarce conservation dollars and in their efforts to develop pre-emptive conservation programs.