This paper reports on the update of the Bayesian sex- and age-structured integrated stock assessment model for Antarctic toothfish (Dissostichus mawsoni) in the Ross Sea region (RSR; Subareas 88.1 and Small-Scale Research Units (SSRUs) 88.2A-B) using the most recent available data for the RSR Antarctic toothfish fishery. The assessment model used catch data for 1998–2023, tag-release data for 2004–2022 and associated tag-recapture observations for 2005–2023, fishery catch-at-age data for 1998–2022, abundance observations from the Ross Sea Shelf Survey (RSSS) for 2012–2023, and age observations from the RSSS for 2012–2022. Data (releases and recaptures of tagged fish, and length data) from vessel trips that had been quarantined since the 2019 assessment were excluded. The model structure was the same as that used in 2021, updated to the software Casal2 and including the recommendations of the 2023 Working Group on Statistics, Assessments and Modelling.
The 2023 base case model Markov chain Monte Carlo estimated MCMC estimated B0 as 77 855 t (95% CIs 71 954–85 115 t) and the current stock status (B2023) as 64.3% B0 (95% CIs 61.3–67.3% B0). The MPD estimated status in 2021 from the 2023 assessment (66.8%) was almost the same as that estimated by the 2021 assessment (66.0%). The maximum posterior density estimate of spawning stock biomass trajectory was almost identical in the equivalent model run in CASAL.
Multiple sensitivity runs were carried out, following recommendations from the 2023 Working Group on Statistics, Assessments and Modelling. Models fitting to all recaptured tagged fish rather than only those at liberty for six years or less resulted in a higher initial biomass, potentially due to the potential bias of tag mixing assumptions and tagging loss rate estimate. Models with time-varying selectivities resulted in much more complex models with little improvement in the catch-at-age residuals and had similar biomass outcomes to the base case. Models estimating year class strength further back in time provided no improvement in the performance of the models and unexpected estimates of early year class strengths. When combined with time-varying selectivities, the age composition residuals looked more promising although not markedly so. Models fitting to tagging data by age rather than by length resulted in higher initial biomass yet similar estimates of other parameters and fits and require further investigation.
The precautionary yield calculated using the Commission for the Conservation of Antarctic Marine Living Resources toothfish decision rules and applying a catch split of 19% for the area north of 70° S, 66% for south of 70° S, and 15% in the Special Research Zone was 3499 t with the base case model. We recommend a catch limit of 3499 t for RSR Antarctic toothfish in the 2023/24 and 2024/25 fishing seasons. We also recommend further work on model selectivity assumptions and the use of age-based release and recapture data of tagged fish in particular.