Ecological Aspects of the Geographical Distribution and Diversity of Mammalian Species

Abstract

Using the comprehensive collection of range maps compiled for 679 species of North American mammals by Hall, we compute the distribution in the sizes of the geographical ranges of these mammalian species. The distribution is roughly lognormal. This corresponds to high skewing on a linear scale toward small geographical ranges, with a median geographical range of around 1% of the area of North America; only about 2% of the species have ranges exceeding 50% of this area. There are differences among taxa; skewing is high for bats, rodents, insectivores, and lagomorphs and less pronounced for carnivores and artiodactyls. In contrast to Rabinowitz et al.'s study of "seven forms of rarity," which found most of the British flora in the study to be widely distributed throughout Britain, we find that most mammalian species in North America have relatively restricted geographical ranges. This difference may derive from systematic differences between plants and animals, or from the small size and relative homogeneity of Britain, or from other causes. The fact that half or more of North American mammals have ranges smaller than the combined size of Nicaragua and Honduras has implications for conservation biology, especially because such restricted patterns are more pronounced in the tropics, where deforestation is currently most severe. As mentioned by earlier authors, there are both latitudinal and longitudinal gradients in range size, with average ranges increasing to the North and to the East. At any one latitude, however, there is wide variability in range size. Corresponding to these gradients in range size are latitudinal and longitudinal gradients in species density, with density increasing as average range size decreases. These latitudinal gradients have long been appreciated, but we think that the longitudinal gradients may ultimately prove more illuminating because they are less confounded by the kinds of systematic changes (in temperature, etc.) that accompany latitudinal change. Following Aldrich and Loftas, we identify 23 distinct habitat types in North America and show that patterns both in geographical range and in species density are correlated with the number of different habitats a species range includes: high density of species and small geographical range sizes occur in areas of high habitat diversity. Correlations between the number of habitats (and the average geographical range of habitats) versus latitude and longitude explain some, but not all, of the geographic variation in species' range patterns noted in this article. Finally, we show that species are more habitat specific (having ranges embracing fewer distinct habitat types) toward more southern latitudes and toward more western longitudes.