Otolith analysis of juvenile walleye pollock Theragra chalcogramma from the western Gulf of Alaska

Growth rate and hatch date distributions were estimated for juvenile walleye pollock Theragra chalcogramma (Pallas, 1814) collected in autumn 1987 from the western Gulf of Alaska. Mean juvenile growth rates varied geographically by as much as 45%. A trend for slower growth around Unimak Pass and the Shumagin Islands and faster growth upstream in the Alaska Coastal Current towards Kodiak Island was noted; the fastest growth did not occur in the main habitat region. Juvenile hatch date distributions were compared to identify regional differences that might reflect stock structure. Juvenile hatch dates were compared with hatch dates of the same cohort sampled as larvae to test for selective mortality. Regional differences in hatch dates, along with other information, indicated several minor spawning populations located around Kodiak Island and near Unimak Pass. For the main aggregation of pollock in the Shumagin Island region, hatch date distributions were not significantly different among the early larval cohort sampled in late May, the late larvae sampled in mid-June to early July, and the juveniles sampled in autumn. Neither growth-rate nor size-dependent mortality of pollock between the larval and juvenile stages appears to be a dominant factor in determining survival patterns. An alternative test was attempted, by which lengths-at-age during larval life were back-calculated from juvenile otoliths and compared with lengths-at-age of the population sampled as larvae in May and June. Pollock surviving as juveniles in autumn were not larger as larvae than the general larval population. This result is an example of the observation that back-calculated lengths are almost always smaller than the lengths of fish sampled at age (Lee's phenomenon). Problems in determining survival patterns based on otolith back-calculations and comparison of hatch date distributions are discussed.