The performance of a size-structured stock assessment method in the face of spatial heterogeneity in growth

Abstract

Monte Carlo simulation is used to assess the performance of a size-structured stock assessment method of the type commonly employed to assess rock lobster populations in Australia and New Zealand. The simulations consider the impact of measurement error and process error in catchability and the length at 50% selectivity, as well as the implications of pooling data across populations that differ in terms of growth rate. The ability to estimate the virgin biomass depends critically on having catch-rate or size-composition data for earliest years of exploitation; in the absence of such data the estimates can be highly biased and imprecise. Several of the reference points commonly reported for assessment purposes (e.g. the biomass at which maximum sustainable yield is achieved) are, however, based on the estimate of the virgin biomass. Estimation performance (bias and precision of estimated quantities) deteriorates with increasing process error. For most of the scenarios examined, the expected benefits of increased precision arising from pooling data across spatial zones are not realized and better performance can be achieved by conducting assessments at the level of population and subsequently aggregating results spatially.