Environmental influences on geographic variation in body size of western bobcats

Hypotheses that explain geographic variation in body size were examined usi ng cranial measurements of 950 bobcats (Lynx rufus) from western North Amer ica. Bobcats were divided into 25 geographic localities of similar habitats and landform (based on ecoregions). Principal component analyses were used to derive a single estimate of size from scores on the first principal com ponent. Males and females were examined separately, because they were signi ficantly dimorphic in body size and because sex and locality exhibited a si gnificant interaction. We expected that female body size would best reflect environmental influences, because male size may be influenced by sexual se lection. We found significant geographic variation in bobcat body size, wit h about 44% of the variation in males and 47% of the variation in females a ccounted for by comparison among the localities. We also found that variati on in body size was associated with Bergmann's rule, as indicated by signif icant multiple regression of body size of males (R-2 = 0.426) and females ( R-2 = 0.480) on latitude and elevation. Using correlation and regression an alyses, we examined the association of body size with selected environmenta l variables that represent the classical physiological explanation of Bergm ann's rule, James' moisture-humidity modification of Bergmann's rule, Rosen zweig's productivity hypothesis, and Boyce's seasonality hypothesis. Only t he productivity hypothesis was not supported. The relative strengths of ass ociations suggested, however, that James' modification was better supported than the classical explanation for Bergmann's rule. Path analyses permitte d further discrimination of hypotheses, and only the seasonality hypothesis received significant support. As expected, this support was only evident f or females. Path analysis may provide a tool for evaluating relative streng ths of competing but correlated explanations of geographic variation.