Frontiers in Marine Science (Dec 2015)
Otolith morphometrics distinguish difficult species of rockfishes in the Gulf of Alaska
Abstract
The oceans around Alaska are home to at least 40 species of rockfishes (Sebastidae). A number of these species are commercially important and are subject to strict catch limitations. However, regulators are hindered by the fact that some species are difficult to distinguish based on external characteristics and can only be conclusively identified using genetics. The rougheye (S. aleutianus) and blackspotted (S. melanostictus) rockfishes are one such example. Survey evidence suggests that these species segregate by depth, and that the shallower blackspotted is more prevalent in the fishery, raising concern that managing the species as a mixed stock might mask detrimental effects to the blackspotted population. This study examines otolith shape and morphometric analysis as tools for distinguishing these closely related species. Seven hundred and thirty-eight rockfish specimens were collected, their identity verified via genetic analysis, and their ages read. We then analyzed the data using several procedures including simple morphometrics and non-dimensional shape indices. We found that simple logistic regression across eight parameters was sufficient to correctly identify the species for more than 90% of the specimens. The use of shape indices did not improve prediction accuracy, but did eliminate allometric correlations among the predictor variables, which may be useful in some cases. A crucial aspect of the analysis is the inclusion of interactions with age in the model, which improved discrimination success. This may suggest an unrecognized difference in the growth rates of these species. Overall, we are capable of predicting species identity with fewer than 5% errors, which is sufficient to allow for separate management plans for the species. As such, we recommend that otolith stock discrimination techniques like these could provide an easy and effective way to discriminate many difficult-to-identify stocks, and further recommend that age data may be a useful addition to standard morphometric analyses.
Keywords
