NONMETABOLIC EXPLANATIONS FOR THE RELATIONSHIP BETWEEN BODY-SIZE AND ANIMAL ABUNDANCE

1. Metabolic constraints are the usual explanation for the relationshi p between body size and species abundance in natural assemblages of an imals. In some assemblages, abundance scales with body weight to the - 0.75 power. Metabolic rate scales as weight raised to the (plus) 0-75 power, therefore, on average equal amounts of energy are available to each species in a community. This equality has been taken as evidence that a species' abundance is limited by its energetic requirements. 2. Criticisms of these arguments notice that most species in samples of complete assemblages cannot be energy limited (although the most abund ant species, at the 'upper bound' may be). These arguments also ignore the frequency distributions of species' body size and species' abunda nce which underlie abundance versus size plots. Both frequency distrib utions have theoretical explanations, which combined could explain abu ndance versus size patterns independently of metabolic arguments. 3. W e test the hypothesis that concatenating the underlying frequency dist ributions of species' body size and species' abundance can directly ac count for observed patterns in plots of size versus abundance in assem blages of animals. We compare the negative slope of the upper boundary of plots of size against abundance from real assemblages, with the sa me slopes derived from models with no energy constraints. It is specie s at this upper boundary that are most likely to be energy limited. 4. Our models give estimates for the upper bound slopes that are very si milar to slopes calculated from real assemblages. We produce realistic patterns in plots of abundance versus size without recourse to metabo lic arguments. 5. We conclude that patterns of abundance versus body s ize in natural assemblages may not be constrained by species' energy r equirements, and do not require explanations independent of those for the constituent frequency distributions of size and abundance.