An Individual Based Model of Arctic cod ( Boreogadus saida) early life in Arctic polynyas: II. Length-dependent and growth-dependent mortality

Abstract

A bioenergetics individual based model (IBM) of early growth is used to investigate the relative importance of length-dependent and growth-dependent mortality during the early life (0–45 d) of Arctic cod in the Northeast Water (NEW) in 1993 and the North Water (NOW) in 1998. In the model, individual growth is forced by the observed temperature and prey concentration histories as prescribed by the hatch date of a larva. The IBM reproduced well the observed length-at-age and revealed large ontogenetic and interregional fluctuations in instantaneous growth. Four mortality scenarios were compared for each population: (1) constant mortality (estimated from catch-at-age data); (2) length-dependent mortality; (3) growth-dependent mortality; and (4) combined length- and growth-dependent mortality. Scenarios 2, 3, and 4 were parameterized to achieve the final survival produced by the constant mortality rates estimated from observations (scenario 1). Scenario 2 accounted well for declining mortality with size but not for the large variations in growth-dependent mortality. Scenario 3 failed to capture the decreasing vulnerability of surviving larvae to predation. Only scenario 4 accounted for both the large fluctuations in growth-dependent mortality and the progressive shift in dominance from length-dependent to growth-dependent mortality as the survivors increased in size. Sub-sampling the model output to reproduce the limited temporal resolution of sampling at sea improved the fit between observed and modeled frequencies-at-age, and pointed to the under-sampling of the smallest larvae as a major sampling bias.