Turning maneuvers in sharks: Predicting body curvature from axial morphology.

Given the diversity of vertebral morphologies among fishes, it is tempting to propose causal links between axial morphology and body curvature. We propose that shape and size of the vertebrae, intervertebral joints, and the body will more accurately predict differences in body curvature during swimming rather than a single meristic such as total vertebral number alone. We examined the correlation between morphological features and maximum body curvature seen during routine turns in five species of shark: Triakis semifasciata, Heterodontus francisci, Chiloscyllium plagiosum, Chiloscyllium punctatum, and Hemiscyllium ocellatum. We quantified overall body curvature using three different metrics. From a separate group of size-matched individuals, we measured 16 morphological features from precaudal vertebrae and the body. As predicted, a larger pool of morphological features yielded a more robust prediction of maximal body curvature than vertebral number alone. Stepwise linear regression showed that up to 11 features were significant predictors of the three measures of body curvature, yielding highly significant multiple regressions with r(2) values of 0.523, 0.537, and 0.584. The second moment of area of the centrum was always the best predictor, followed by either centrum length or transverse height. Ranking as the fifth most important variable in three different models, the body's total length, fineness ratio, and width were the most important non-vertebral morphologies. Without considering the effects of muscle activity, these correlations suggest a dominant role for the vertebral column in providing the passive mechanical properties of the body that control, in part, body curvature during swimming.