Paradox of otolith shape indices: routine but overestimated use

The identification of fish species using otolith shape has been common in many fields the marine science. Different analytical processes can be applied for morphological discrimination, but reviewing literature we have found conceptual and statistical limitations use indices wavelets (contour analysis), being specially worrying first case due to their widespread routine use. In present study, 42 were classified applying traditional machine learning classifiers performance measures (accuracy, Cohen’s kappa statistic, sensitivity, precision). Our results conclusive; a more adequate option classification than indices, independently considered. artificial neural network support vector provided highest values all wavelets. cases, sensitivity precision pointed out higher confusion between some patterns indices. Therefore, strongly discourage species.