Growth and age composition of northern shrimp Pandalus eous estimated by multiple length frequency analysis

Abstract

We examined growth of northern shrimp Pandalus eous in the Sea of Japan, off western Hokkaido, to improve estimations of catch-at-age for stock assessment. Multiple length frequency analysis based on length frequency data collected by a scientific research vessel was conducted to examine length-at-age in the shrimp population. Multi-normal distributions estimated using maximum likelihood indicated a good fit to length distributions. AIC values and regression analyses revealed annual growth variation and a decreasing trend in the length at several age classes in the shrimp population. We revised the method for estimating catch-at-age from the age-conversion table (ACT), which is a simple method for age determination, to age–length keys (ALK) calculated from the results of multiple length frequency analysis. Abundant year classes caught successively year after year could be more easily identified from the catch-at-age data computed using ALK than by using ACT. Our results suggested not only that the mean size of commercial landings fluctuated based on changes in age composition but also that a decrease in the length-at-age in the population influenced the consistent size decrease of commercial landings.