Using otolith shape analysis to distinguish eastern Gulf of Mexico and Atlantic Ocean stocks of king mackerel

Abstract In winter, king mackerel from eastern Gulf of Mexico and Atlantic stocks mix off southeast Florida, where they support a large fishery. Neither tagging nor genetics has yielded a way to estimate mixing rates accurately. For management purposes, and based on tagging data from the mid-1970s, all of these fish have been considered to be from the Gulf stock. Our objectives were to examine the feasibility of using otolith shape data to distinguish the two stocks, and if the method proved feasible, to use it to estimate stock composition in the winter mixed-stock fishery. In the feasibility phase of the study, we collected shape data from sagittae of 355 female king mackerel taken during summer spawning seasons, 1986–1993, outside the winter mixing area. Gulf fish were taken off the Florida panhandle and Atlantic fish from waters north of Cape Canaveral, FL. Shape data, obtained from the posterior half of the sagitta, included area, perimeter, and standardized Fourier amplitudes. Using a training set of 250 fish and a stepwise discriminant procedure, we selected a set of variables that correctly classified 80.4% of Atlantic and 85.7% of eastern Gulf fish of the remaining 105 individuals. In the application phase of the study, we estimated mixed-stock composition in the 1996–1997 winter fishery. Shape data were extracted from sagittae of 363 females taken from distinct areas during the 1996 spawning season. A training subset of the data and a stepwise discriminant procedure were used to select a set of variables that correctly classified 71.1% of Atlantic and 77.5% of eastern Gulf fish in the remaining data (test set). We used that set of variables and a maximum likelihood method to estimate composition of the ensuing mixed-stock fishery from a sample of 463 females taken in that fishery. The resulting estimate was that 99.8% of individuals in the winter landings were from the Atlantic stock and only 0.2% were from the eastern Gulf stock, with a standard error of 3.4%. Using otolith shape data, it is possible to distinguish individuals from eastern Gulf and Atlantic stocks of king mackerel and to estimate stock composition in the mixed-stock fishery. Our estimate of that stock composition is consistent with other recent studies, but is markedly different from the composition presently assumed in management. Because management can be ineffective if based on inaccurate estimates of stock composition, we recommend that such information be used regularly in management and that corresponding estimates be made in future years. Such continued analysis would serve both to confirm our results and to estimate year-to-year variability in stock composition of the mixed fishery.