RECRUITMENT IN MARINE FISHES - IS IT REGULATED BY STARVATION AND PREDATION IN THE EGG AND LARVAL STAGES

We used literature reports to evaluate the hypotheses that: 1. year-cl ass strength in marine fishes is determined by mortality operating dur ing the pre-juvenile stage of the life history, and 2. recruitment in marine fishes can be regulated by starvation and predation in the egg and larval stages. The available evidence is largely consistent with t he first hypothesis, although mortality operating during the juvenile and post-juvenile stage may moderate the variation induced at the pre- juvenile stage. The hypothesis that recruitment can be regulated by st arvation during the larval stage was assessed in relation to Hjort's'c ritical period' and Cushing's 'match-mismatch' hypotheses. The availab le evidence does not support a major link between food abundance at th e time of first feeding, and recruitment (Hjort's 'critical period'). The hypothesized relationship between recruitment and the coincidence between the seasonal timing of plankton production and the seasonal ab undance of larvae (Cushing's 'match-mismatch') is generally supported. However, the relationships are weak and the importance of the strengt h of the coupling between seasonal cycles in plankton and larval abund ance appears to be weaker than had previously been surmised. Recent ev idence, which suggests that failure to distinguish between food abunda nce in the environment and the availability of food to individual larv ae may have compromised the evaluation of these hypotheses, is reviewe d. The relationship between mortality due to predation and its potenti al effects on recruitment was evaluated with reference to two emerging paradigms, the 'bigger is better' and the 'stage duration' hypotheses . We conclude that failure to fully evaluate the assumptions underlyin g these hypotheses may have led to erroneous generalizations regarding the importance of size at age/stage and growth rate on the probabilit y of death due to predation. Neither the 'bigger is better', nor the ' stage duration' hypotheses is unequivocally supported.