Measuring a mammal's metabolic rate is technically fairly straightforward these days. But as Roberto Nespolo explains, an animal that has been reared in clement conditions might have a very different metabolic rate then an animal that has been shivering in the cold. How much of an effect acclimation has on an animal's metabolic rate wasn't clear, until Nespolo measured the metabolic rates of tiny leaf-eared mice in the lab that had been acclimatised to either 30°C or 12°C. The animals' metabolic rates varied enormously,depending on the conditions the rodents had experienced. Which could mean that metabolic measurements made on animals in the wild could be unreliable, and strongly affected by the environmental conditions the animal has experienced(p. 2145).

Knowing that metabolic rate is highly correlated with body size, Nespolo also tested how well several physical characteristics that are often taken as good indicators of body size, correlate with the rodent's metabolic rates. In each case he found a correlation, but each characteristic was related to metabolic rate in a different way, so they couldn't be taken as a reliable indicator of body size. Nespolo decided to develop a more accurate way of defining the rodent's body size. He used a statistical analysis known as`structural equation modelling' on three easily measured physical parameters,and found that of the statistical analysis was a much truer way of defining the animal's body size.

Nespolo, R. F., Arim, M. and Bozinovic, F.(
). Body size as a latent variable in a structural equation model: thermal acclimation and energetics of the leaf-eared mouse.
J. Exp. Biol.