Influence of the quality and quantity of data from a multi-year tagging program on bias and precision of biomass estimates from an integrated stock assessment – update

By way of simulation, this study investigated how the bias and precision of biomass estimates from an integrated tag-based assessment are influenced by various aspects of a multi-year tagging program, in particular the effects of the size (tag size-overlap) and numbers of tagged fish, the duration of the tagging program, the type of auxiliary data used in the assessment, and the catch history. Biomass estimates generally improved with more and better quality tagging data, however important nuances emerged within this overall trend. The results show that in the early stages of the tagging program, a high tag size-overlap is imperative to maximise the likelihood of a robust assessment. Tagging fish with a low tag size-overlap, even with large numbers, is likely to result in overestimates of biomass and the resulting data should not be used in stock assessments with only few years of data available. In contrast, with a 100% tag size-overlap even low numbers of releases and subsequent recaptures collected in short tagging programs were sufficient for relatively robust assessments. Estimation errors of SSB0 and SSB status in the assessment year stabilized or were relatively small with a tag size-overlap over 60-70%. Biomass estimates were largely unaffected by the type of catch history. Therefore, creating ‘contrast’ in the data by strongly reducing the fish biomass is not needed for tag-based assessments, since data obtained from a tagging program in recent years only can be sufficiently informative as long as estimates of the catch history are available.