Common principal components analysis (CPCA) of covariance matrices among stonechat subspecies.

The step-up approach (where a low p-value indicates a low probability that the higher model is better than the lower model) suggests that covariance matrices among subspecies share all PCs, but have differing eigenvalues (CPC), and the model building approach (where the lowest AIC value indicates the best model) suggests that the matrices are equal (see Table 1 for a complete description of the step-up and model building approaches).