ICES Interbenchmark protocol on eastern English Channel sole (IBPSOL7d)

The Interbenchmark protocol on eastern English Channel sole (IBPSOL7d) was convened because the Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak (WGNSSK) rejected the 2022 stock assessment due to a likely trend in survey catchability of the UK beam trawl survey (UK BTS) index and poor fit of the State-space Assessment Model (SAM) to observed catches. IBPSOL7d was primarily tasked with adapting the method of the calculation of the Belgian commercial beam trawl tuning fleet (BEL CBT), adapting the configuration of SAM, and updating the reference points. During this IBP, the calculation of the BEL CBT index was slightly modified to allow for a more robust index over time. More specifically, the threshold on the rectangles included in the calculation was removed. The possibility of allowing trends in the catchability of the UK BTS survey index in the SAM model was explored but was eventually rejected because similar model outcomes could be achieved with alternative model parameterisation and concerns about the appropriateness of this feature. In the final SAM model agreed at the IBP, the revised BEL CBT index was included, the auto-regressive correlation structure for catches at age (“obsCorStruct”) was removed to allow for more flexibility of the modelled catches to better match observations, and UK BTS data for ages 1–3 in 2010–2021 were removed because the inclusion of these data resulted in a deterioration of cohort tracking, possibly caused by ageing problems or changes in the survey catchability. These changes improved the model fit compared to the previous model and model diagnostics were appropriate. The calculation of the reference points was conducted with Eqsim and following the ICES technical guidelines. The stock-recruitment relationship was evaluated as type 5, showing a stock with no evidence that recruitment has been impaired or with no clear relation between stock and recruitment. The short-term forecast methodology was modified to provide a more consistent estimation of recruitment. In particular, forecast recruitment is now the median value from resampling over the period 2012–2020. This shorter time period was chosen because of the lower recruitment regime in the most recent years. The last data year is not considered as uncertainty is larger.