Reliability of stock size estimates from adaptive framework virtual population analysis

Abstract

The stock size estimates from adaptive framework virtual population analysis are sometimes very sensitive to the model's assumptions. The uniqueness and precision of stock size estimates from adaptive framework virtual population analysis are examined analytically and numerically. The results indicate that (i) when the fishing mortalities remain constant for all years, the stock size estimates are not determined uniquely without additional assumptions, and (ii) when the cumulative fishing mortalities are similar among cohorts, the precision of stock size estimates will be poor, even if the abundance indices are precise and informative. For the precise estimation of stock size, it is necessary that fishing mortalities vary among years, that cumulative fishing mortalities vary among cohorts and that auxiliary information such as selectivity is available in addition to precise abundance indices.