| Summary: | The benchmark on Pandalus borealis in Skagerrak and Norwegian Deep Sea (WKPAND), chaired by Anders Nielsen, Denmark, and Carmen Fernández, ICES, met in Bergen (Norway) from 20–22 January 2016 to perform a benchmark assess-ment of the Pandalus stock. The benchmark successfully addressed all the terms of reference: it agreed on a new stock assessment method for this stock, calculated refer-ence points and developed a Stock Annex. Part of the work (finalization of reference point calculations, and most of the writing of the report and Stock Annex) had to be done by correspondence after the meeting, with a final web-conference for bench-mark approval of results on February 19. Most of the benchmark work focused on exploring two alternative length-based models: one of them had already been presented at the previous inter-benchmark process for this stock (ICES, 2013), whereas the other one, implemented in Stock Syn-thesis (SS3), was developed for this benchmark. The fits to the data were better for the model implemented in SS3, particularly for the survey length–frequency distribu-tions, which are a very important source of information to determine the strength of the incoming age-1 group. The model developed in SS3 has internally a quarterly time-step and the selection pattern of the fishery is modelled as length-based. This allows the shrimp to be increasingly selected by the fishery as they grow through the year, which is particularly relevant to age-1 shrimp and appears to be a determining factor in achieving good model performance. By contrast, in the already-existing length-based model, the selection pattern of the fishery is assumed to be age-based and unchanged throughout the year and the fits to the survey data were problematic; an ad hoc modification incorporating linear changes in selection-at-age through the year was tried during the workshop and resulted in improved fits, but reformulating the model assuming length-based selection is considered by the benchmark to be a more appropriate solution.
|
|---|