Report of the Benchmark Workshop on Roundfish Stocks (WKROUND)
The Benchmark Workshop on Roundfish Stocks (WKROUND) chaired by Mike Armstrong (UK) and Jim Berkson (USA), with invited external experts Martin Dorn (USA) and Steve Ralston (USA ) took place from 8-10 February 2013 to develop benchmark assessments for whiting in Subarea IV (North Sea) and Division VIId (Eastern Chan-nel); haddock in Division VIIa (Irish Sea); anglerfish (Lophius piscatorius and L. budegassa) in Division IIIa (Skagerrak and Kattegat) and Subareas IV (North Sea) and VI (west of Scotland); and haddock in Division Va (Iceland). The objectives of the meet-ing were to evaluate the appropriateness of data and methods to determine stock status in relation to existing or proposed new biological reference points, and to investigate methods for short term outlook taking agreed or proposed management plans into account. As part of the benchmark process, a data compilation workshop was carried out from 8 to 10 January at the Marine Institute in Galway, to review and agree the data, parameters and other information to be used in the benchmark assessments, and to consider the quality of data including discards estimates and misreporting of landings.
North Sea whiting: WKROUND carried out comparative XSA, SAM and SURBAR assessments on North Sea whiting. The previously noted discrepancy during the first half of the time series between survey trends and population numbers estimated from the catch at age data was again noted. The pre-1990 catch data would need to be reduced by at least 75% for the FLXSA SSB estimates to resemble those from SUR-BAR. This discrepancy is very difficult (if not impossible) to resolve. WKROUND explored the hypothesis that one of the key problems with the assessment of North Sea whiting was an incorrect assumption about stock structure and identity, and recommended the compilation of data sets to explore the performance of assessments of a joint VIa-IVa stock and a stock south of the Dogger Bank. The results would be analysed to test whether the new stock units lead to an improvement in assessment consistency, and/or any change in management advice. Only if the new stock units lead to a different perception about stock dynamics would consideration be given to adopting them for future assessments. WKROUND recommends that WGNSSK ex-plores further the potential adoption of SAM/FLSAM as a potential baseline model following sensitivity analysis of SAM settings. Future update assessments should (until the next benchmark) continue to apply the FLXSA stock assessment model to data from the whole of Subarea IV and Division VIId, using the time-series from 1990 onwards. Supporting exploratory assessments should be produced using the SAM and SURBAR models, with settings as given by WKROUND. Further work to be considered in the future should include:
- Further exploration of the potential adoption of SAM/FLSAM as a potential baseline model following sensitivity analysis of SAM settings.
- Development of candidate fishing mortality and biomass reference points us-ing meta-analytic (or other) approaches, based on inter alia FLXSA results, and testing of their utility as part of a future management strategy evaluation process.
- Continued development of methods to generate long-term time-series of catch data for more biologically-justifiable stock units of whiting.
- Analysis of the area- and component-specific time-series of catch-at-age to ensure that the total catch data used in the assessment are consistent.
Irish Sea haddock: WKROUND explored the use of the SAM model to fit fishery land-ings/ discards data and trawl survey indices of abundance, but a number of serious issues with the model configuration and results for this stock precluded reaching any agreement on an analytical assessment method and associated biological reference points. As it was not possible, during or following the meeting, to understand or re-solve these issues, the work has not been written up and will be deferred to a future benchmark when sufficient knowledge has been gained to progress the assessment. In the meantime, WKROUND proposes continuation of the survey-based trends assessment adopted in previous WGCSE meetings. A detailed compilation and evalua-tion of fishery landings and discards data were completed at the data compilation workshop and presented at the assessment workshop. This is provided in Annex 11 of WKROUND.
Northern anglerfish: WKROUND evaluated an updated survey-at-age-based version of the assessment model developed at WKFLAT in 2012 and examined the sensitivity of the model to alternative assumptions about age given length. The model used data from the SCO-IV-VI-AMISS-Q2 surveys and reported catches from 2005-2011. The assessment model exploited the assumption that the SCO-IV-VI-AMISS-Q2 survey is able to provide absolute estimates of abundance beyond age 5 following from consid-erable work carried out to directly estimate catchability of the survey. Survey age compositions from otoliths appeared to give coherent year class signals and catch curves with slope (~total mortality) of 0.44 at ages 6-8 for years 2009-11. Due to con-cerns in the accuracy of age readings the sensitivity of the model to alternative assumptions about age were investigated. As requested by ICES WKFLAT, the alternative assumptions were based on the parameters of the von Bertalanffy growth models of Landa et al (2008) and Quincoces et al (1998). A transition matrix was built which converted numbers ate age from the existing index at age dataset to numbers at age if those ages had been determined by the aforementioned growth models. These modified indices were then input to the original version of the assessment model. The relative trends in abundance from the model were similar, but more marked differences in fishing mortality and selectivity were obtained. WKROUND proposed three options for developing future assessments of anglerfish:
- The existing procedure using the SCO-IV-VI-AMISS-Q2 survey to provide advice on catches by applying method 3.2.0 of ICES’ Implementation of RGLIFE advice on Data Limited Stocks (ICES 2012).
- A simple harvest ratio approach as used in e.g. Nephrops.
- The age-structured survey model explored in WKFLAT and WKROUND. This option may be used to determine certain parameters for the former two simpler approaches, if it is deemed unacceptable for use per se. However, further work on all three of these options needs to be made available and presented prior to any firm conclusions as to the way forward with anglerfish assessment.
Icelandic haddock: Since 2007 the assessment of this stock has been conduced with an Adapt type model assuming M=0.2 and tuned with two groundfish surveys from March and October. The assessment has been consistent without major problems. Haddock are heavily discarded because large and small haddock are more mixed together in the fishing areas than most other species. Discards have not been previ-ously included in the assessment. WKROUND evaluated available data on discarding and examined the sensitivity of the Adapt model to including discards. A range of different assessment models was applied (XSA, Adapt, TSA, a separable model and Surba). WKROUND concluded that the Adapt model should continue to be used as basis for advice, as it gives similar results to other models based on the same data. Tuning with both the surveys is recommended as there is some discrepancy between the surveys and nothing is known about which survey gives the more correct picture of the stock. Retrospective pattern and comparison with survey indices show strong consistency. WKROUND proposes the following work to improve the assessment and advice:
- Studies to improve the understanding of the form and magnitude of unaccounted mortality caused by discards, mesh penetration and predation.
- Improve the modelling of changes in overall fishery selectivity over time through use of fleet disaggregated models incorporating separate catch at age data for trawlers and longliners. Additional fleet disaggregation could in-clude large and small longliners, and gillnets that used to be more important.
- Studies to understand the difference in perception of the stock from the two surveys.
- Improved procedures for forecasting weights at age, including spatial distribution effects.
Published under the auspices of the following ICES Steering Group or Committee