Abstract
This study used 3D landmark-based geometric morphometrics under an object-symmetry framework to quantify pelvic asymmetry in domestic cats and dogs while explicitly accounting for measurement error through replicate digitizations. Procrustes ANOVA revealed significant components of both directional asymmetry (DA) and fluctuating asymmetry (FA) in the overall dataset, and DA was further visualized as a structured, non-random pattern across the landmark configuration. Measurement error remained smaller than the FA component, yielding high repeatability, indicating that the detected asymmetry patterns were not driven by landmarking imprecision. Group-wise summaries are presented as descriptive patterns of the sample (rather than direct between-group inference). In these descriptive summaries, cats tended to show a more coherent DA pattern, whereas dogs showed greater individual variation consistent with a relatively stronger FA component; males also tended to exhibit greater FA-related dispersion than females. In regression models of FA, age showed no association, body mass exhibited only a weak trend, and sex emerged as a significant predictor, while species showed no detectable effect when covariates were included. Collectively, these results support the hypothesis that pelvic shape contains both systematic (DA) and individual-specific (FA) asymmetry components, with sex-related differences in FA magnitude, but limited evidence for age- or weight-related effects within the sampled range. The study provides a repeatable framework and baseline reference for pelvic asymmetry in cats and dogs.