ABSTRACT: The cochlea and semicircular canals (SCCs) of the inner ear are vital neurosensory devices. There are associations between the anatomy of these sensorineural structures, their function, and the function of related biological systems, for example, hearing ability, gaze stabilization, locomotor agility, and posture. The endosseous labyrinth is frequently used as a proxy to infer the performance of the hearing and vestibular systems, locomotor abilities, and ecology of extinct species. Such fossil inferences are often based on single specimens or even a single ear, representing an entire species. To address whether a single ear is representative of a population, we used geometric morphometrics to quantitatively assess the variation in shape and symmetry in a sample of endosseous labyrinths of wild turkeys Meleagris gallopavo of southern Ohio. We predicted that ears would be symmetrical both within individuals and across the sample; that labyrinth shape and size would covary; that labyrinth shape would vary with the size of the brain, measured as width of the endocranium at the cerebellum; and that labyrinths would be morphologically integrated. To test these predictions, we microCT-scanned the heads of 26 cadaveric turkeys, digitally segmented their endosseous labyrinths in Avizo, and assigned 15 manual landmarks and 20 sliding semilandmarks to each digital model. Following Procrustes alignment, we conducted an analysis of bilateral symmetry, a Procrustes regression analysis for allometry and other covariates including side and replicate, and analyses of global integration and modularity. Based on Procrustes distances, no individual's left and right ears were clearly different from each other. When comparing the ears of different specimens, statistically clear differences in shape were found in only 66 of more than 1,300 contrasts. Moreover, effects of both directional and fluctuating asymmetry were very small-generally, two orders of magnitude smaller than the variance explained by individual variation. Statistical tests disagreed on whether these asymmetric effects crossed the threshold of significance, possibly due to non-isotropic variation among landmarks. Regardless, labyrinths appeared to primarily vary in shape symmetrically. Neither labyrinth size nor endocranial width was correlated with labyrinth shape, contrary to our expectations. Finally, labyrinths were found to be moderately integrated in a global sense, but four weakly separated modules-the three SCCs and cochlea-were recovered using a maximum-likelihood analysis. The results show that both fluctuating and directional asymmetry play a larger role in shape variation than expected-but nonetheless, endosseous labyrinths are symmetrical within individuals and at the level of the population, and their shape varies symmetrically. Thus, inferences about populations, and very possibly species, may be confidently made when only a single specimen, or even a single ear, is available for study.
