AN APPRAISAL OF THE SIZE-DEPENDENT MORTALITY HYPOTHESIS FOR LARVAL FISH - COMPARISON OF A MULTISPECIES STUDY WITH AN EMPIRICAL REVIEW

This study presents an assessment of the size-dependent mortality hypothesis for larval fish from a multispecies survey of Conception Bay, Newfoundland, Canada. Mortality rates are estimated using a length-based method (per millimetre). The results from this survey are consistent with previous studies which indicate that losses decrease with increasing size of fish. However, for each species within this survey, mortality rates are constant. Comparison of mortality rates within species among surveys indicates that as the range of size categories sampled increases, the estimated mortality rates decrease, despite evidence of adequate fit to the length-based model. The findings indicate that previous relationships between size or stage and mortality of larval fish should be reevaluated. Length-based methodology used to estimate mortality rates of larval fish appears to provide biased estimates of this vital characteristic. It is suggested that using size as a proxy for biological age (i.e., assuming a constant growth rate) may be an invalid assumption. Future surveys will need to provide accurate information about the age structure of larvae sampled in order to properly estimate mortality rates.