Report on the Long-term Management Strategy Evaluation for Northern Shrimp (Pandalus borealis) in Division 4.a East and Subdivision 20 (PandLTMS)

This report provides the technical basis to support the ICES response to the EU-Norway request to ICES on a long-term management strategy for Pandalus in Skagerrak and the Norwegian Deep. The report was prepared by correspondence by a group of scientists who are members of the Joint NAFO/ICES Pandalus Assessment Working Group (NIPAG), led by Carmen Fernández (not a NIPAG member). Two external sci-entists have reviewed the work.  
A Management Strategy Evaluation (MSE) was developed to investigate via simulation the performance of different harvest control rules (HCR), which would be applied to the results of a stock assessment conducted in March of a given year. The catch given by the HCR would apply to the same year (January–December), and a preliminary TAC for the following year should be provided at the same time. The request indicates that, alternatively, the TAC year could go from May 1 of the assessment year to April 30 of the following year. Both TAC year options were examined in the evaluation, without detecting any significant difference in performance metrics (risk of low stock bio-mass, yield, interannual yield variability). For providing catch advice, a January–December TAC year seems more natural, because ICES calculates the spawning–stock biomass (SSB) for this stock as corresponding to January 1, and the trigger point of the HCR is defined as the SSB at the start of the TAC year. The preliminary TAC for the following year could then be based on a catch forecast, using the same HCR, for the first two quarters of that year.
The Operating Model for the MSE was based on the latest available assessment from March (March 2017). Given the reduction in recruitment observed in the last decade, Historic and Low (based on the last ten years) recruitment scenarios were considered in the evaluation. The consequences of selecting an HCR based on a particular scenario when future recruitment follows the other scenario were examined, and risk-yield trade-offs were found. Uncertainty / error in the assessment and short-term forecast was analysed retrospectively and error parameters obtained from that analysis were included in the MSE.
The HCRs evaluated consist of (Ftarget, Btrigger) combinations. The catch corresponds to the Ftarget fishing mortality; if the SSB at the start of the TAC year is estimated to be below Btrigger, then the Ftarget is linearly reduced to zero. A clause in the HCR (paragraph 3) sets the overriding constraint that there should be no more than 5% probability of the SSB falling below Blim at the beginning of the following TAC year. Paragraph 3 posed considerable implementation difficulties and could not be implemented exactly. Therefore, the HCRs evaluated either did not include paragraph 3, or included para-graph 3 implemented in an approximate form by requiring that the SSB from a deter-ministic forecast should be at or above Bpa at the beginning of the following TAC year. The technical aspects of this implementation were not straightforward and results for HCRs including paragraph 3 should be regarded as approximate. The report presents detailed results for HCRs with Btrigger=MSY Btrigger=9900 t, as well as for HCRs that consider multiple (Ftarget, Btrigger) combinations.
As requested, the (Ftarget, Btrigger) combination that maximises long-term yield (median value) while resulting in no more than 5% probability of the SSB falling below Blim in any 20-year period was identified. Assuming Historic recruitment, this combination is (0.48, 13 000 t) or (0.48, 12 000 t), depending on whether the HCR excludes or includes paragraph 3. If, on the other hand, Low recruitment is assumed in the evaluation, this combination is (0.36, 12 000 t) or (0.37, 11 000 t), depending on whether the HCR ex-cludes or includes paragraph 3. The peformance of the HCRs excluding or including paragraph 3 is similar.
The requirement in the request that there should be no more than 5% probability of the SSB falling below Blim in any 20-year period is more stringent than the ICES standard criterion to consider an HCR as precautionary, which is that there should be no more than 5% probability of the SSB falling below Blim in any year. If HCRs are selected based on maximising the median long-term yield while achieving a risk at or below 5%, then applying the 20-year-based risk criterion leads to HCRs that, in the long term, result in a higher median SSB (roughly of the order of 25%–35% higher) and lower median yield (roughly of the order of 5%–15% lower) than if the ICES standard risk criterion was applied.
Some scenarios of interannual quota flexibility of +/-10% were evaluated, and resulted in a small increase of the probability of SSB falling below Blim. In this case, some reduc-tion (of the order of 5% reduction) in the Ftarget is necessary in order not to exceed the 5% risk level.
As requested separately to the group, FMSY ranges were calculated following the stand-ard ICES procedure and are included in the report. It should be noted that the standard ICES risk definition was used in this calculation.