Interbenchmark workshop on the assessment of northeast Atlantic mackerel (IBPNEAMAC)

The NEA mackerel inter-benchmark group met at ICES HQ for a data workshop 11-13 December 2018 and a final second meeting 4-7 March 2019 at Wageningen Marine Research, Netherlands. The group consisted of scientists (national labs, universities and industry scientists), stakeholders, managers and re-viewers from in total 15 different countries. The ToRs of the group show a specific focus on the influence of the tagging data on the assessment, since at the 2018 working group meeting (WGWIDE) sensitivity analyses showed a substantial change in perception of the mackerel stock when tagging data was excluded from the assessment. This indicates a high, and potentially incorrectly specified, weight of this data source. The group furthermore focused on a revised recruitment index and at the quality of the catch sampling. After thorough review of the data and analyses the group made a number of decisions with regard to the data that should be included in the assessment model and the statistical approach to model these data. The updated recruitment index was considered to be of good quality and was implemented in the assessment. There were concerns related to the age reading quality, directly applicable to the quality of the catch-at-age data and potentially survey data. Analyses showed that experienced age readers are in agreement in ~41% on the age of older fish (age 7+) and that national catch-at-age frequency distributions originating from similar types of fisheries diverged substantially. The impact of ageing error on the perception of the stock turned out to be minimal and was therefore not considered an issue for this inter-benchmark, although recommendations are made to improve in this field. Several analyses were undertaken to investigate the quality of the RFID tagging data, including spatial coverage of the recapture stations, detection rate of RFID tags in factories, data exploration to infer exploitation rate from the tagging data itself, changes in perception of cohort size as inferred from the tagging data itself and a variety of alternative assessment model configuration in which the RFID dataset was truncated in time, by age, by number of years tagged fish were al-lowed to have been out at sea (years-out) and models in which the original tagging data was replaced with alternative methods to aggregate the tagging data into e.g. indices of abundance...