Variation in sexual dimorphism of mouse os coxae shape, volume, and bone mineral density in response to selection for high voluntary wheel running

Pelvic (os coxae and sacrum) dimorphism is common in mammals. Studies (primarily on human and non‐human primates) of pelvic dimorphism suggest that dimorphism results from an evolutionary trade‐off between reproduction and locomotion. Previous work has shown that pelvic morphology varies in relation to locomotor modes among species and that, along with other girdle elements such as the scapula, high levels of voluntary activity cause training responses in mice. We analyzed three lines of laboratory mice, from an ongoing selection experiment, two of which have been bred for 59 generations for high levels of voluntary wheel running (HR) and one that is a non‐selected control (C) line. One selected line (HR MINI ) is fixed for a “mighty mini‐muscle” mutation that reduces skeletal muscle mass by up to 50%. The other selected line (HR NORMAL ) does not possess this mutation. We used two dimensional geometric morphometric techniques to produce quantitative descriptors of os coxae shape (Relative Warp scores) and geometric (centroid) size. In addition, we used X‐ray micro computed tomography (μCT) to measure elements of bone microstructure (bone mineral density (BMD) and bone volume (BV)). ANOVAs showed significant differences in body size (length and mass) among the lines and between the sexes. Both HR lines were smaller than the C line, and the HR MINI line was smaller than the HR NORMAL line. However, although females were smaller than males in all three lines, os coxae centroid size was greater in females. Finally, for all size metrics, no significant sex‐by‐line interaction was present, indicating no differential sexual dimorphism among lines. Analyses of μCT data, with body size as a covariate, showed significant differences in BMD and BV of the os coxae among lines and between sexes. Females had higher BMD and BV than males, and HR lines had lower values than the C line (the HR MINI line showed the largest decrease). Additionally, a significant sex‐by‐line interaction in the analysis of BMD suggests a stronger effect of selection on males than on females. Finally, MANCOVAs, with body size as a covariate, revealed significant shape differences among lines and between sexes. A significant sex‐by‐line interaction was also found, indicating different patterns of shape difference between males and females, depending on line. These results indicate that although there is no effect of selective breeding or the mini‐muscle mutation on patterns of sexual dimorphism in body and os coxae size, this is not the case for patterns of sexual dimorphism in other components of coxal morphology. Male and female differences in os coxae shape and bone microstructure have different trajectories and magnitudes depending on line, suggesting that in some cases, males and females respond to selection and the mini‐muscle mutation in different ways. Support or Funding Information MJ Murdock Charitable Trust to HS, KB & CH U.C. Riverside Chancellor's Postdoctoral Fellowship to HS NSF IOS‐1121273 to TG