Report of the Workshop on Blue Whiting (Micromesistius poutassou) Long Term Management Strategy Evaluation (WKBWMS)

The Workshop on Blue Whiting Long Term Management Strategy Evaluation (WKBWMSE) was hosted at ICES headquarters on August 30, 2016. The aim of the workshop was to re-evaluate biological reference points and evaluate a proposed long-term management strategy (LTMS) for the Northeast Atlantic blue whiting (BW, Micromesistius poutassou) stock (whb-comb; Subareas 1–9, 12 and 14). This was done in a response to a recommendation from the Inter-Benchmark Protocol of Blue Whiting, where a new assessment of the stock was produced, and a NEAFC request to ICES. There were 20 participants from 10 countries: 18 scientists and 2 industry representa-tives. Two reviewers reviewed the analytical work and the conclusions of the workshop. See full participant list in Appendix 1.
A stochastic simulation model, R-package called “EQSIM”, was used to evaluate the biological reference points and the LTMS. Two characteristics of the input data were identified that could influence the simulations: highly variable recruitment, with peri-ods of high and low productivity, and high uncertainty in the stock assessment. Re-cruitment is a keystone variable in LTMS as it has a large impact on the potential yield that can be taken from a fish stock. There is no relationship between BW spawning stock biomass (SSB) and recruitment, but there is relationship between recruitment and SSB two years later. BW recruitment, from 1981 to 2016, is highly variable (the highest value is 17x the lowest value) and has one period of high recruitment, from 1996 to 2005, causing high auto-correlation in estimated recruitment. WKBWMSE agreed that in the LMTS evaluation the whole recruitment time series should be used as there is no scientific reason to reject some years and accept others. The stock assessment depends primarily on a single acoustic index of abundance that in the past has had obvious year effects. This has led to notable retrospective revisions in estimates of stock size and fishing mortality (F).  Past advice uncertainty, rather than parametric model uncertainty, was estimated and used to simulate error in the perceived state of the stock and the fishery. 

EQSIM was amended to conduct a long term simulation of the blue whiting stock, ra-ther than simply estimating equilibrium conditions. The amended model is called ‘SimpSIM’ and is R package. WKBWMSE concluded that a Blim of 1.5 million tonnes and a Bpa of 2.25 million tonnes was still appropriate and should remain unchanged but re-evaluated Fmsy (= 0.32), Fpa (= 0.58) and Flim (= 0.88). The proposed LTMS included a harvest control rule (HCR) with two breakpoints (at Bpa and Blim), a 10% interannual quota flexibility (“banking and borrowing”), and a 20% TAC change limits. This HCR was evaluated to ensure that the probability of SSB going below Blim in any given year should be less than 5%. WKBWMSE concluded that the proposed HCR was precau-tionary for the re-evaluated estimates of Blim (1.5 million tonnes), Bpa (2.25 million tonnes) and Fmsy (= 0.32). Including a 10% interannual quota flexibility in the LTMS had an insignificant effect on the performance of the HCR. The HCR was precautionary both with and without the 20% TAC change limits above Bpa. However, not applying TAC change limits can lead to advised TAC being lowered considerablyif the stock is estimated to be below Bpa, and it limits how quickly advised TAC can increase once the stock is estimated to have recovered above Bpa.
WKBWMSE argued that it was wise to explore sensitivity of the conclusions to recruit-ment by evaluating the HCR for a shorter recruitment series (excluding years of high and very low recruitment). The proposed HCR was also considered precautionary us-ing the shorter recruitment series. Finally, the workshop decided to evaluate the HCR using another stochastic simulation model software (HCS) which has been used to evaluate blue whiting HCRs in the past. The HCS model supported the SimpSIM re-sults as the HCR was found to be precautionary, both when using the whole recruit-ment time series and when using the shorter recruitment series on this platform as well.
Two external reviewers checked the simulation frameworks used and evaluated con-clusions from the work done by WKBWMSE (the full reviews are annexed to this re-port). The first reviewer, H. Björnsson, raised concerns about some of the ICES guidelines in place for reference point (FMSY and PA) definitions and calculation. In par-ticular the requirement for MSY Btrigger to be larger than Bpa is considered unnecessary, since equally appropriate, but more stable, management could be obtained with lower F targets and a lower biomass trigger. In addition, it was felt that the simplistic recruit-ment generation procedure used in the simulations could be inappropriate. As such this reviewer felt FMSY=0.32 may be too high for this stock. Simulations conducted on an alternative platform to those presented in this report indicated that while the NEAFC rule could be considered precautionary, it was not necessarily recommended due to management getting ‘stuck’ at low Fs (as highlighted in this report).
The second reviewer, A. Whitten, concluded that the methodology was well docu-mented and transparent. The evaluation was considered credible and understandable, and the assumptions and methods appropriate. The difficulty in trying to estimate that reference points given the lack of a relationship between stock and recruitment was acknowledged. The conclusions on the proposed LTMS were supported.