SIZE-DEPENDENT VULNERABILITY OF MARINE FISH LARVAE TO PREDATION - AN INDIVIDUAL-BASED NUMERICAL EXPERIMENT

Twenty-day simulations of responses by a larval fish cohort were inves tigated with an individual-based model of predation by ctenophore, med usa, and planktivorous fish predators. Results indicate that the relat ionship between larval size and vulnerability to predation was general ly dome-shaped for invertebrate predators and could be dome-shaped for fish predators if they foraged optimally by size, and depended upon a ttributes of both predators and larval fish prey. For the predators th at did not forage optimally, cohort-specific mortality generally decre ased as the mean length of larvae in a cohort increased, but bigger or faster-growing larvae within a cohort were not always most likely to survive. Until larvae grew through a ''window'' of vulnerability and r eached a threshold length when susceptibility to the predators decreas ed more rapidly with larval length than encounter rate increased, mean length or growth rate of surviving larvae on each day was slightly lo wer, or not different from those that died in most of the simulations. After the threshold length was reached, predators began to catch smal ler larvae, which resulted in larger survivors. The time necessary to grow through the window and reach the threshold length depended on gro wth rate of the larvae, size of the predators, and the variance struct ure of these parameters. These results indicate that size and growth r ate of fish larvae are partially decoupled by the predation process an d, ultimately, act differentially to determine cohort survival rate, a lthough both may be most important after larvae have reached the thres hold length. In these simulations, the threshold length was reached af ter a significant portion (56-99%) of total larval mortality had occur red; time to reach the threshold was generally shorter for the faster growing cohorts. Initially, both fast- and slow-growing individuals wi thin a larval cohort differed little in size and, therefore, were near ly equally vulnerable to predation. However, reduced risk of predation occurred when all members of a cohort had reached the threshold lengt h, which suggests that mean growth rate of individuals within a cohort , not their size, is probably the more important parameter affecting c umulative mortality, especially when the rate is high. We propose that characteristics of larval survivors may be more influenced by attribu tes of the predators to which they were exposed in early life, rather than by their initial status within a cohort with respect to length at hatching and potential growth rate. (C) 1996 International Council fo r the Exploration of the Sea