No b to Rule Them All

+ See all authors and affiliations

Science  13 Nov 2009:
Vol. 326, Issue 5955, pp. 917
DOI: 10.1126/science.326.5955.917-a

The dependence of basal metabolic rate (BMR) on body mass (M) in mammals is generally expressed as BMR = aMb, but the value of the exponent b has long been disputed. Some researchers argue for geometric scaling (b = 0.67), as predicted by surface area–to–volume ratios. Others claim quarter-power scaling (b = 0.75), which has been supported by theoretical analyses of nutrient supply networks. However, most empirical “mouse-to-elephant” studies have not accounted for the shared evolutionary history of the sampled species. To do so, White et al. use two methods: phylogenetic generalized least squares and independent contrasts. They consider high-quality BMR data from 585 species (which include only measurements on inactive, fasted, nonreproductive adults in a thermoneutral environment) and a subset of 537 species that excludes lineages for which BMR may not be measurable. Their phylogenetically informed estimates of b fall between 0.67 and 0.75, and all of their statistically valid analyses reject both values. In addition, they find that b differs among lineages. Their results reinforce doubts as to the existence of a universal allometric relationship between mammalian BMR and body mass.

Evolution 63, 2658 (2009).

Navigate This Article