Density-versus temperature-dependent growth of Atlantic cod (Gadus morhua) in the Gulf of St. Lawrence and on the Scotian Shelf

Dramatic changes in length-at-age have been observed in many populations of Atlantic cod (Gadus morhua) in the Northwest Atlantic over the past 20-30 years. We tested for density- and temperature-dependent effects on length-at-age over this period for cod in four populations, the southern Gulf of St. Lawrence stock (referred to here as 4T cod), the northern Gulf of St. Lawrence stock (3Pn4RS), the northeastern Scotian Shelf stock (4VsW) and the southwestern Scotian Shelf stock (4X). Effects were tested using a modified von Bertalanffy growth model, in which L-infinity was density- and/or temperature-dependent. For the 4T stock, effects of both density and temperature were significant, with faster growth at lower densities and higher temperatures. Results were similar for the 4VsW stock, except that the effect of density was slight compared to the effect of temperature. Only the effect of temperature was significant for the 3Pn4RS stock, with faster growth at wanner temperatures. Age alone accounted for most of the interannual variation in mean length of 4X cod, and neither density nor temperature effects were significant for this stock. Measures of both cod ambient temperature (bottom temperature weighted by cod distribution) and general temperature conditions in the environment (i.e., temperatures unweighted by cod distribution) were used for the 4T stock, but only the latter were available for the other stocks. For the 4T stock, ambient temperature was the most important explanatory variable in 4-parameter models (i.e., three von Bertalanffy parameters plus one additional explanatory variable). However, more of the variation in mean length was explained by the 5-parameter model with cod abundance and general environmental temperature than by either of the 5-parameter models that included ambient temperature. Strong patterns were evident in the residuals from the growth models, especially for the 4T, 4VsW and 3Pn4RS stocks. These patterns were similar between stocks, and suggested the action of some factor not accounted for in the models, possibly variation in size-selective mortality. Crown Copyright (C) 2002 Published by Elsevier Science B.V. All rights reserved.