USEFULNESS OF FOURIER-ANALYSIS OF OTOLITH SHAPE FOR ATLANTIC MACKEREL (SCOMBER-SCOMBRUS) STOCK DISCRIMINATION

We compared shapes of Atlantic mackerel (Scomber scombrus) sagittae between the two contingents (i.e. spawning groups) from the Northwest Atlantic and between the stocks from the Northwest Atlantic and the North Sea to evaluate whether otolith shape could differentiate between the two contingents in a mixed fishery. We quantified shapes with the Fourier series, an objective and rapid method which decomposes a shape's outline into a series of sinusoids. To determine a correct way to compare contingents/stocks, we first assessed four intracontingent effects on otolith shapes. Age and year-class effects were significant, while sex and bilateral position effects were not. This temporal instability in shapes indicates that confounding effects of age and year-class on otolith shapes need to be assessed carefully before drawing conclusions on stock structure. It also shows that comparative studies of otolith shapes with Fourier descriptors are not useful for mackerel contingent discrimination. The power of discriminant functions to correctly classify test mackerel samples separated by age and sampling year, on the basis of otolith shape, was better for comparisons between the stocks than for those between the contingents.