Conclusions Strong seasonal trends in G and M interact to determine the change in biomass of cohorts. The fastest growing cohorts, those hatched in May, are rapidly lost to predators in most years. In years with abundant prey early in the year, cohorts hatching in February and March experience lower mortality and can increase rapidly in biomass. These early cohorts may ultimately make up the bulk of the survivors. In agreement with the conclusions above, otolith microstructure analysis of haddock surviving to the fall in the Gulf of Maine (Lapolla and Buckley 2005) and the North Sea (Wright and Gibb 2005) indicate strong negative selection on hatch date. Our data support Pope et al’s (1995) “optimal fish strategy” in a seasonally-perturbed, size spectrum.  Successful cod and haddock hatch ahead of the peak abundance of prey, reaching a large size before being overtaken by the wave of abundant predators.