Search

Like other fisheries models, multispecies models are subject to various sources of error. However, with regard to their use for ecosystem-based fisheries management (EBFM) between model errors are likely to be most important. As multispecies models are by definition many-dimensional, comparing them is potentially a complex task. The paper uses a simple approach. This is to calculate the Jacobian matrix of long term steady state catch by species with respect to the fishing mortality relative to status quo levels on all species. This enables the comparison of the relative strength of species interactions among models both within and between regions. This Jacobian matrix approach to comparing multispecies models is applied to available models for the North Sea, the Baltic Sea and from Icelandic waters. Moreover, this information is used to provide the basis for estimating a multidimensional quadratic yield surface for each model in the near field. Used this way it is possible to compare different model estimations of fishing mortality rate changes needed to approach yield-related management goals. The results suggest considerable variation between models in their detailed results but more coherence in suggesting directions for changing fishing mortality rate. Thus the approach is of considerable importance in specifying the confidence with which it is possible to make multispecies predictions for EBFM.

The pan-regional Working Group on Multispecies Assessment Methods (WGSAM) met in San Sebastian, Spain, 16–20 October 2017. In this eleventh report of the group, work focused on three of the multi-annual ToRs (B, C, D). Based on their knowledge, participants provided an updated inventory of progress of multispecies models in ICES Ecoregions (ToR A), noting those regions where no information was available. A Key Run (ToR B) of the North Sea Stochastic Multispecies Model (SMS) was presented and reviewed in detail by 4 WGSAM experts, and approved by the group following implementation of changes agreed in plenary at the meeting and verified by a subset of experts post-meeting. The Key Run is documented in detail in Annex for ToR B, with key outputs summarised in Section 5 and data files made available on the WGSAM webpage and the ICES expert group Github (https://github.com/iceseg/wg_WGSAM). Since the M2 values are used for the assessment of important North Sea stocks, it is recommended to publish the annex also on the official stock annex website. In addition, WGSAM does not recommend updating existing data series of natural mortality by simply adding the latest three new years. The timeseries as a whole shows patterns which are not retained by this procedure. Multispecies model skill assessment (ToR C) and multi-model ensemble methods (ToR D) were emphasized this year. Considerable progress has been made towards advancing both aspects of multispecies modelling. Investigation of skill assessment and ensemble methods and case studies is critical to ensure that outputs of multispecies assessment models are reliable for use in operational assessment and to inform management decisions. Progress was also made on investigations of top predator impacts on managed fish across several regions (ToR F), including the North Sea where new information was included in the SMS key run. Further progress was also made on multispecies and ecosystem level reference points and harvest contr