Effects of Environment on Phenotypic Variation and Sexual Dimorphism in Dipodomys simulans (Rodentia: Heteromyidae)

Abstract

The magnitude of sexual dimorphism in size among populations of some small mammals living in different habitats may be a reflection of the degree of habitat partitioning between sexes or nutritional quality of available food. We tested these hypotheses by quantifying between-sex differences in size in populations of the Dulzura kangaroo rat ( Dipodomys simulons ) from several macrohabitat types at regional and local geographic scales. On a regional scale, our results indicated a significant shift in morphology of males and females along a north-south latitudinal gradient. Large kangaroo rats inhabited warm and arid southernmost latitudes, whereas small animals inhabited relatively cool and moist regions to the north. This morphologic gradient represented an increase in size with less seasonal variation in precipitation and temperature. This pattern was clearly counter to Bergmann's rule, but consistent with the hypothesis that temporal variation in availability of food may be an important factor leading to geographic variation in size. External and cranial measures of size dimorphism were not correlated significantly with any major geographic factor, and only cranial measures were associated significantly with variation in climate. At a local level, kangaroo rats living in different vegetation communities within the same latitudinal zone also exhibited significant sexual dimorphism in size; however, macrohabitat heterogeneity appeared to affect males and females equally. Both sexes were larger in Sierran Montane Conifer Forest and Californian Chaparral macrohabitats. Smaller animals occupied Californian Grassland and Coastal-scrub, and Vizcaino vegetation communities. This pattern of variation was observed in all three-way comparisons and suggests the possibility of habitat-mediated phenotypic responses. Finally, only one of eight comparisons involving kangaroo rats living in adjacent plant communities showed a significant two-way interaction between dimorphism and macrohabitat type. Our results, therefore, did not provide strong evidence to substantiate the hypotheses that the magnitude of dimorphism among populations of D. simulans living in different macrohabitats is a reflection of the degree of partitioning between sexes or nutritional quality of available food.