PRIMATES, 29(1): 135-137, January 1988 SHORT COMMUNICATION 135 Body Size Variation in Baboons: A Reconsideration of Ecological Determinism V. J. VITZTHUM University of Michigan, Ann Arbor ABSTRACT. Pope (1983) presented an intriguing argument regarding the covariation of body size in baboons and rainfall. However, a reanalysis of the data indicates that "Principle 2" of the model is not supported. Key Words: Papio; Body size; Ecology; Climate. INTRODUCTION Body size is a fundamental component of an animal's adaptation, at once reflecting and determining a vast array of morphological and life history parameters (HuTcHINSON & MACARTHUR, 1959; SCHMIDT-NIELSEN, 1984). POPV (1983) found a significant positive relationship between mean annual rainfall and body size in baboons (sexes analyzed separately). He suggested that the greater energy avail- ability in habitats with more rainfall shifts the focus of inter-male competition from food resources to females. As rainfall increases, increasingly aggressive interactions select for an increasingly greater body size in males. A prediction of this model, as stated by Povv (1983), is Principle 2: The degree of sexual dimorphism in body mass increases with increasing rainfall. Male body mass increases geometrically as a function of adult female body mass. Although Povv's sexual selection hypothesis represents an interesting approach to integrating selective factors affecting behavior and morphology, his data and analyses require re-examination. ANALYSES AND RESULTS Table 1 of Povv (1983) lists the data for body weight and rainfall in the 13 baboon samples used for his analyses. The single sample representing mandrill males has a mean of 50.5 kg, based on two captive animals. This observation is as much as twice the mean weight reported by any other study (cf., STAMMBACH, 1986; DECHOW, 1980 for several sources) and, as it is based on captive animals, appears likely to be a substantial overestimate of the average body weight of wild mandrill males. Therefore, in the reanalysis conducted here, this sample is omitted. Figure 2 (Povv, 1983), redrawn here as Figure 1, shows the regression of male against female mean body mass and depicts a power curve (dashed line) which "best described" the data and represents increasing sexual dimorphism with increasing body weight. Although not stated, calculations reveal that this curve can be determined by performing least squares