Rarity and Body Size: Importance of Generality

Abstract

The identification of those species at greatest risk of extinction has been a major concern of conservation biologists (Burton 1984; Munton 1987; Mace & Lande 1991; Groombridge 1993; Collar et al. 1994). A starting point in many such assessments is an understanding of the relative abundances and range sizes of species, on the assumption that scarce and restricted species on average will have a greater likelihood of extinction than will those that are abundant and widespread, or will at least constitute a greater proportion of the species at highest risk. Because abundance and range size are insufficiently well correlated with likelihood of extinction to act as adequate predictors of that likelihood, assessments must often be refined further by the consideration of additional variables. The question thus arises as to which variables should be used and how. Building on work by Arita et al. (1990), Dobson and Yu (1993) have argued that comparisons of the abundances and range sizes of animal species would be most meaningful they controlled for differences in body mass; they assert that if a mouse and an elephant had the same local density, the mouse would be at relatively greater risk of extinction. They illustrated the consequences of controlling for body size with reference to data for 100 species of Neotropical forest mammals. Gaston and Blackburn (1995a) urged caution in the application of such an approach on the grounds that it was unclear why body size in particular should be controlled for, that there were potentially serious methodological difficulties to be overcome in attempting to control for the effects of body size in many assemblages, and that it was unlikely to be applicable in many-perhaps the vast majority of-cases. Dobson et al. (1995) have responded to a number