Benchmark workshop on Baltic cod stocks (WKBALTCOD2)

The ICES Benchmark Workshop on Baltic Cod Stocks (WKBALTCOD2) met at ICES Headquarters in Copenhagen, Denmark on 4–8 February 2019, following a data evaluation workshop (Chair Johan Lövgren) and several preparatory web conference meetings. The meeting, cochaired by Meaghan Bryan, USA (External Chair) and Michele Casini, Sweden (ICES Chair), was attended by two invited external experts, Vladlena Gertseva, USA, and Verena Trenkel, France, and 47 participants from 10 countries. Participants represented a diversity of groups including industry, NGOs, managers, and scientists. The objectives of WKBALTCOD2 were to evaluate the appropriateness of the data and the assessment methods to determine stock status for the Western Baltic cod (SD 22-24) and the Eastern Baltic cod (SD 24-32) stocks, evaluate the short-term forecasting methods, re-examine and update the reference points, and update the stock annex as appropriate to these stocks. The workshop started with a group discussion of data issues and decisions that were integral to both assessments. Subsequent, stock-specific discussions were held in smaller subgroups and the main results and conclusions were presented in plenary. Around 15–20 persons, together with at least one external reviewer/co-chair, participated in each of the subgroups. A single modelling approach, the state-space stock assessment model (SAM), was presented for Western Baltic cod. This model has been used previously to determine stock status of Western Baltic cod and the panel agreed that this model should be used for the current assessment. The main issues that emerged and were addressed during the workshop for this stock were stock mixing in SD 24, the inclusion of new recreational data from Sweden and Denmark, the inclusion of a German pound-net survey index of age-0 fish in the assessment model, the extension of the time-series of the assessment back in time, and the update of reference points. Future efforts should be made to update the mixing proportion of Eastern and Western Baltic cod in SD 24 by length and season, and continue improving the stock splitting methods and the geographical coverage of the samples. An analytical quantitative assessment had been lacking since 2014 for the Eastern Baltic cod stock. The key issues addressed during the benchmark included how to best account for changes in productivity (e.g. growth, mortality, maturity) in the assessment model, stock splitting in SD 24, the use of age–length keys in the assessment, and ageing error and bias. Stock assessment models using Stock Synthesis and the stochastic surplus production model (SPiCT) were put forward for the benchmark as two possible model candidates. The panel agreed that the Stock Synthesis model assuming time-varying growth, natural mortality, and maturity to account for changes in productivity was acceptable to provide scientific advice, while the SPiCT model should be maintained as an alternative approach. The accepted model exhibited some residual patterns that were likely due to assuming that ages were precisely known. Stock Synthesis can accommodate an ageing error matrix to account for precision and bias. Ageing error and bias statistics should be developed and thoroughly reviewed for the next benchmark. Additionally, future work should focus on improving the growth estimates, which will allow a more precise separation between growth and natural mortality. The Stock Synthesis model would also benefit from information on sample size associated with length distributions of commercial catches. Estimates of fishing mortality compatible with a precautionary FMSY was not attainable for this stock, therefore probabilistic forecasts with Markov chain Monte Carlo methods were proposed to be used instead.