MEPS 534:163-177 (2015)  -  DOI: https://doi.org/10.3354/meps11376

ABSTRACT: Otolith shape analysis provides a practical basis for stock separation, useful in fisheries management. A wide range of factors affect otolith morphology and a fundamental challenge in morphometric stock identification is to develop a consensus on the biological interpretation of otolith shape variation. To date, there are few comprehensive reports that have examined various factors simultaneously at multiple scales in natural conditions. In this study, I identified sources of otolith shape variation across spatial and taxonomic scales using a new hierarchical partitioning method embedded in a geometric morphometric framework. Various environmental, taxonomic and endogenic factors which affect otolith shape were quantitatively investigated in 2077 coral reef fishes. Using the new partitioning method, allometry accounted for a considerable degree of otolith shape variation at all scales and contributed more variation to regional differences than did habitats or islands. While large-scale variations are expected to be associated with significant shape variation, the study provides quantitative evidence that both local environmental variables and large-scale patterns contribute equally to total otolith shape variation. Models that aim at discriminating stocks and forecasting stock boundaries implicitly assume that the within-stock morphological variation is negligible in regard to the variation that exists among stocks. The importance of local environmental variables may therefore act as an important confounding effect into those predictive models. More generally, the degree of contribution of some variables differed substantially among the taxonomic scales, as did their relationship among spatial scales. A scale-dependent understanding of factors affecting otolith morphology is critical in understanding the integrity of fish populations and is an effective tool in management of fisheries resources.

KEY WORDS: Morphological disparity · Morphospace · Regression tree · Coral reef fish · Determinants · Taxonomic scales · Geographic scales