Stock assessment of yellowfin tuna in the western and central Pacific Ocean. M. Vincent, N, Ducharme-Barth, P. Hamer, J. Hampton, P. Williams, G. Pilling. Oceanic Fisheries Programme, The Pacific Community

This paper describes the 2020 stock assessment of yellowfin tuna (Thunnus albacares) in the western and central Pacific Ocean. A further three years of data were available since the last full stock assessment conducted in 2017, and the model extends through to the end of 2018. New developments to the stock assessment include addressing recommendations of the 2017 stock assessment report (Tremblay-Boyer et al., 2017), new information on age and growth from otoliths and the integration of growth information from tag recaptures, as well as the implementation of the Richards growth model; updates to the definition of reproductive potential; implementation of a composite ‘index’ longline fishery for each model region which received that region’s standardized CPUE index, and concurrently a ‘pseudo catch conditioned’ approach taken for the assessment. Based upon recommendations of SPC’s 2020 pre-assessment workshop, only the 10? N spatial structure was considered within this assessment. The assessment is supported by the analysis of longline catch-per-unit-ofeffort (CPUE) data (Ducharme-Barth et al., 2020b), background analyses of biological parameters, compilation of the tagging data and definition of the fisheries structures (Vincent and DucharmeBarth, 2020), tagging data treatment (Peatman, 2020; Scutt Phillips et al., 2020), preparation of the length and weight composition data (Peatman et al., 2020), new otolith aging and growth work (Farley et al., 2020) and estimation of growth using a tag-integrated model (Eveson et al., 2020).

Key changes made in the progression from the 2017 to 2020 diagnostic models include:

• Updating all data up to the end of 2018.
• Implementation of updated models for tag data, purse seine catch estimates and size composition data.
• Implementation of the ‘index fishery’ approach, which used a geo-statistically standardized CPUE index.
• Utilizing updated biological parameters for the length-weight relationship and reproductive potential, and extension of the number of quarterly age classes in the model to 40.
• Changes to gear selectivity settings.
• Implementation of growth using the conditional age-at-length otolith data.

In addition to the diagnostic model, we report the results of one-off sensitivity models to explore the relative impacts of key data and model assumptions for the diagnostic model on the stock assessment results and conclusions. We also undertook a structural uncertainty analysis (model grid) for consideration in developing management advice where all possible combinations of those key areas of uncertainty were included.