Femoral strength and posture in terrestrial birds and non-avian theropods.

Osteological and experimental evidence suggest a change in femoral posture between non-avian dinosaurs (in which the femur presumably was carried in a subvertical position) and birds (in which the femur is held nearly horizontal during most phases of terrestrial locomotion). In this study, we used a broad comparative sample to test the hypothesis that cross-sectional properties of the femur records evidence of this presumed change in posture. I(max) and I(min) (second moment of area, related to resistance to bending) and cross-sectional area (indicating resistance to compression) were measured from computed tomography scans of the femora of 30 species of flightless or primarily terrestrial birds, one probable non-dinosaur dinosauromorph, and at least four species of non-avian theropods. It was predicted that birds should have more eccentrically shaped femoral midshafts as measured by I(max)/II(min) (reflecting greater bending) and comparatively smaller cross-sectional areas than non-avians. Results show that no significant differences occur between non-avian dinosaurs and birds for any parameter, and the samples overlapped broadly in many cases. Thus, cross-sectional properties cannot be used to infer differences in femoral posture between the two groups. This surprising finding might be explained by the fact that femoral postures were not drastically different or that a gradation of postures occurred in each sample. It is also possible that bone loading during life was not closely correlated with cross-sectional morphology. We conclude that cross-sectional properties should be used with caution in determining the posture and behaviors of extinct animals, and only in conjunction with other morphological information. (c) 2009 Wiley-Liss, Inc.