Geographic variation in body size: the effects of ambient temperature and precipitation

Latitudinal trends in body size have been explained as a response to temperature- or water-related factors, which are predictors of primary production. We used the first principal component calculated from three body parameters (weight, body length and the greatest length of the skull) of a sample of mammals from Israel and Sinai to determine those species that vary in size geographically, and whether such variation is related to annual rainfall, average minimum January temperature and average maximum August temperature. We used a conservative approach to discern the effects of precipitation and temperature by applying sequential regression. Variable priorities were assigned according to their bivariate correlation with body size, except for rainfall and its interactions that entered into the model last. Eleven species (Acomys cahirinus, Apodemus mystacinus, Canis lupus, Crocidura suaveolens, Gerbillus dasyurus, Hyaena hyaena, Lepus capensis, Meles meles, Meriones tristrami, Rousettus aegyptius and Vulpes vulpes) of the 17 species examined varied in size geographically. In five of them, rainfall was positively related to body size, while in one species it was negatively related to it. Contrary to the prediction of Bergmann's rule, mean minimum January temperature was positively related to body size in five species and negatively related to body size in two species (C. suaveolens and G. dasyurus). As predicted by Bergmann's rule, maximum June temperature was negatively related to body size in three species, and positively so in one (L. capensis). Primary production, particularly in desert and semi-desert areas, is determined mainly by precipitation. The above results indicate that, in our sample, primary production has an important effect on body size of several species of mammals. This is evident from the considerable proportion of