Högskolan i Skövde

The ICES Benchmark Workshop on Baltic Cod Stocks (WKBALTCOD), chaired by External Chair Jean-Jacques Maguire, Canada and ICES Chair Marie Storr-Paulsen, Denmark, and attended by two invited external experts Verena Trenkel, France and Meaghan Bryan, USA met in Rostock, Germany, 2–6 March 2015 with 39 participants and six countries represented. The objective of WKBALTCOD was to evaluate the appropriateness of data and methods to determine stock status and investigate meth-ods appropriate to use in the single-stock assessment for the cod stock in SD 22–24 and cod in SD 25–32 in the Baltic. Participants in the workshop were a large group with diverse backgrounds representing the industry, fisheries, NGOs, managers and scientists.The single-stock analytic assessment of the eastern Baltic stock was not accepted by the assessment working group (WGBFAS) in 2014 due to severe problems with the input data. The advice for the eastern Baltic cod was, therefore, based on the ICES approach for data-limited stocks. As an outcome ICES decided to establish a bench-mark for both cod stocks and to scope an integrated assessment for the Baltic cod stocks. The first meeting (WKSIBCA) was therefore meant to introduce the interces-sional work conducted since the assessment working group in April 2014, and to reach some conclusions on how to proceed both in the short term (Benchmark in March 2015) and longer term (2–3 years) and was seen as a data compilation work-shop, there is produced a separate report from this workshop. The WKBALTCOD was the 2nd meeting in the benchmark process and was intended to come up with a final stock assessment method, stock annex and input data for both stocks. As it was not possible to reach conclusive decision on the final model to be used for the east Baltic cod stock during the benchmark meeting and as more work on the preferable models was needed, it was decided by the ACOM leadership to prolong the bench-mark process until the assessment working group meeting in April 2015. This deci-sion has led to a relatively long process partly mixed with the assessment working group WGBFAS.It became clear during the benchmark process that although large effort has been put into explaining the underlying processes leading to the changes in the Baltic ecosys-tem, there is still some lack of understanding of the present situation in the eastern Baltic cod stock. Therefore, it was not possible to reach firm conclusions on the final model to be used and therefore not possible to set reference points. It was decided to continue to explore the most promising models and to continue to improve the input data until the assessment working group started in April.The main challenges still to be solved for the Eastern Baltic cod stock is the quantifi-cation of increased natural mortality and decrease in growth. Through several presentations during the workshop (both WKSIBCA and WKBALTCOD) it became clear that natural mortality very likely has increased in later years, due to decreased condition and increased parasite infection. A decrease in growth also seems plausible duo to a decrease in condition and/or selectivity-induced mortality of the largest in-dividuals. However, as none of these parameters are easily estimated, especially with the severe ageing problems, different model assumptions made the output very shaky.For the western Baltic cod, stock identification issues were examined in area SD 24, the intermediate area: based on otolith characteristics and genetics. Due to the results showing a large proportion of east cod in this area, it was decided to split the catch2 | ICES WKBALTCOD REPORT 2015and survey from SD 24 into either the western or eastern Baltic cod stock. It was pos-sible to derive proportions of eastern and western cod in SD 24 back to the mid-1990s.For the western Baltic cod stock a modelled survey indices was included in the as-sessment covering the western part of SD 24 and Area 22+23 and based on a smoothed ALK.Both cod stocks have in the past used commercial tuning fleet to have a better cov-ered of older age groups. It was decided to abound this time-series duo quality issues such as a limited coverage and problems with technical creeping.WKBALTCOD was not able to explore and define reference points for the Western Baltic cod stock during the meeting due to time constraints, but these were calculated and decided by correspondence after the meeting. The recent protocols on estimation procedures developed by WKMSYREF3 for stocks with a full analytical assessment and for data-limited stocks served as objective guidelines to obtain reference point estimates.

Uncertainty is an endemic condition of the Baltic Sea herring (Clupea harengus membras, L) fishery management. It is a condition exacerbated by the fishing fleet overcapacity and consequent exploitation of the herring stock at a level believed to be unsustainable. Some sources of uncertainty are mainly related to biology and fishing technique: the unsolved problem of herring assessment and management units, the recruitment–environment relationship and the reduction in mean weights-at-age, uncertain ageing of fish, the problem of unaccounted fishing mortality caused by the fish selection through the trawl net. Fishing fleet overcapacity is believed to be behind of the regulatory overfishing when setting the Total Allowable Catches (TACs) higher than the scientific advice (decision overfishing) and tolerating the extensive underreporting of catches (implementation overfishing). Two scenarios for the Central Baltic Sea herring fishery management options are constructed and the Bayesian networks are used to represent and update uncertainties encountered in the process of the management related situation assessment. First scenario represents the current status of the fishery management resulting in fishing mortality (F) higher than FMSY – the fishing mortality that corresponds to the Maximum Sustainable Yield (MSY). The second scenario demonstrates the assumed potential impact of economic incentives (e.g.zoning, individual transferable quotas (ITQs), territorial use rights etc.) on the reduction of excessive fishing capacity and bringing actual fishing mortality closer to FMSY.

The context of fisheries management, according to recent studies, tends to be complex, uncertain and controversial and it cannot be adequately addressed based on work done within the “linear” science-policy interface (SPI). It is believed that moving towards a more participatory and “collaborative” SPI model would favour implementation of more efficient economic incentives to reduce the fishing fleet capacity and bring the actual fishing mortality closer to the levels considered to be sustainable. Focusing the participatory research basically on fish biology and fish stock assessment seems to be too narrow while, on the other hand, it does not embrace many other important issues of contemporary fisheries management. It is suggested that further involvement of stakeholders into the full-scale bio-economic modelling based participatory research would contribute to better understanding of the dynamics of human natural interface concerned and, consequently, would improve the consensus between different interests.

The ICES Baltic Fisheries Assessment Working Group (WGBFAS) met 14-21 April 2015 (Chair: Mare Storr-Paulsen, Denmark), with 28 participants and 9 countries represented. The objective of WGBFAS was to assess the status of the following stocks:

1 ) Sole in Division IIIa, SDs 20-22

2 ) Cod in Kattegat, Cod in SD 22-24, Cod in SD 25-32

3 ) Herring in SD 25-27, 28.2, 29 and 32, Herring in SD 28.1 (Gulf of Riga), Herring in SD 30, Herring SD 31.

4 ) Sprat in SD 22-32

5 ) Plaice 21-23, Plaice 2425

6 ) Flounder 22-23; 24-25; 26+28 and 27+29-32, Brill 2232, Dab 2232, and Turbot 2232 (survey trends)

WGBFAS also identified the data needed, for next year’s data call with some suggestions for improvements in the data call as well as in InterCatch. The report contains an introduction with the summary of other WGs relevant for the WGBFAS, country specific fishery description, the methods used, and ecosystem considerations. The results of the analytical stock assessment or survey trends for the species listed above are then presented with all the stocks with the same species in the same sections. The report ends with references, list of Working Documents, recommendations and Stock Annexes. In first quarter 2015 the Baltic cod stocks and the plaice stocks were benchmarked. As a result the Baltic cod stocks now have to apply a splitting key in SD 24 were both stocks are present. This has changed the assessment from being an area based assessment to now being a stock based assessments and has implications for the advice. The principle analytical models used for the stock assessments were XSA and SAM. For most flatfishes, CPUE trends from bottom trawl surveys were presented (except plaice 2425 and her31 using relative SSB from SAM and XSA, respectively). Ecosystem changes have been analytically considered in the following stock assessments: Herring in SD 25-27, 28.2, 29 and 32, and Sprat in SD 22-32, in form of cod predation mortality. Last year a very large retrospective pattern in the Eastern Baltic cod stock caused that the WG rejected the analytic assessment. Several uncertainties in the data lead to this conclusion i.a age reading problems with large inconsistency between and within nations as well as a change in growth and natural mortality. However, even though a data compilation workshop and a benchmark have been conducted in the intermediate time it was not possible to solve the main issue on growth. The lack of knowledge on growth caused to that even the length based data required in the data call was very uncertain for the models and in the end the WG was not able to produce a better model than was presented last year which is based on survey trends. The Her-30 (Herring in the Botnian Sea) was by the working group down scaled from a category 1 stock to a category 3 stock due to the commercial tuning fleet used in the assessment having very uncertain estimates in the last couples of years. However, during the Baltic ADG an alternative assessment was suggested were the stock is still considered a category 1 stock but the last 8 years of the commercial tuning fleet was terminated. This assessment was conducted after the working group but has been included in the report.

The Inter-Benchmark Protocol for Herring in Subdivision 30 (IBP-Her30) worked by correspondence  between  February  28  and  March  28  2013. Verena Trenkel  (France) served as Chair with invited expert Noel Cadigan (Canada). There were six participants. The objectives of the groups were to review the work carried out in response to the benchmark working group WKPELA in 2012.

The ICES/HELCOM Working Group on Integrated Assessments of the Baltic Sea(WGIAB) was established in 2007 as a forum for developing and combining ecosystembasedmanagement efforts for the Baltic Sea. The group intends to serve as a scientificcounterpart and support for the ICES Baltic Fisheries Assessment Working Group(WGBFAS) as well as for efforts and projects related to Integrated Ecosystem Assessments(IEA) within ICES and HELCOM. The group works in cooperation with similargroups within the ACOM/SCICOM Steering Group on Integrated Ecosystem Assessments(SSGIEA).The 2015 WGIAB meeting was held in Cádiz, Spain, from 9–13 March, back-to-backwith the meeting of its counterpart in the Working Group on Ecosystem Assessmentof Western European Shelf Seas (WGEAWESS). The meetings had joint sessions as wellas WG specific work, and some participants effectively participated in both meetings.The WGIAB meeting was attended by 27 participants from nine countries. The meetingwas chaired by Christian Möllmann, Germany, Laura Uusitalo, Finland and Lena Bergström,Sweden.This was the last year of the ongoing three-year Terms of Reference (ToR) for WGIAB.The main working activities in 2015 were to i) conduct studies on Baltic Sea ecosystemfunctioning with the goal to publish case studies from different parts of the Baltic Seain peer-reviewed journals, ii) work on the demonstration exercise to develop ecosystem-based assessment and advice for Baltic fish stocks focusing on cod (DEMO) withmultiple approaches, iii) plan further how to integrate the social and economic aspectsmore tightly in the WGIAB work, and iv) discuss the future focus and format of theWGIAB work.The Baltic ecosystem functioning activity focused on identifying and exploring keytrends and linkages in the Baltic Sea foodweb. This was pursued by presentation andfurther discussion of ongoing intersessional work on foodweb modelling and integratedanalyses, and by exercises to develop conceptual models Baltic Sea foodwebsand the links to ecosystem function. Long-term monitoring datasets on the abiotic andbiotic parts of the Baltic Sea Proper ecosystem were updated for use in the continuedwork to develop environmental indicators for fisheries and marine management.The focus of the DEMO 3 (DEMOnstration exercise for Integrated Ecosystem Assessmentand Advice of Baltic Sea cod) was on finding a way to use the results from theDEMO1 and DEMO2 workshops in short and midterm projections/scenarios of Balticcod dynamics based on different types of modelling, as well as designing methodologyand modelling data for practical implementation of Integrated Advice for Baltic cod.The WGIAB was positively inclined towards including social and economic aspectsinto the integrated assessment. Openings to this path were provided by presentationon ongoing project work, and discussing their linkages to ecological aspects. It wasseen as crucial that experts on social and economic analysis should be included andtake an active part in the future work of the group.The group concluded that its upcoming work should focus more closely on functionaldiversity, which was identified as a recurring issue in the Baltic Sea. This approach wasalso identified as a useful connection point between scientific and management aspectsin order for the group to continue serving as a forum for developing ecosystem-basedmanagement efforts in the Baltic Sea. A focus on functional diversity was also seen as2 | ICES WGIAB REPORT 2015a potentially feasible way of bringing together management aspects for different sectors,by linking to ecosystem services concepts.The group proposed Saskia Otto, Germany and Martin Lindegren, Denmark as newincoming Chairs, together with Lena Bergström, Sweden and Laura Uusitalo, Finland.Having four Chairs is justified due to the wide scope of the group's work, as well asthe increased work load due to the planned new foci.

In many raptors and owls the male is the main provider of food in the early phase of the nestling period while the female incubates the eggs and broods the young. In the nestling period the female often helps the male to feed the young, but the factors affecting whether and when she leaves the brood to hunt have not been investigated in detail. We present a dynamic state variable model that analyses female behaviour and fat storage dynamics over the nestling period. The results show that in the first half of the nestling period the female faces a conflict between the need to brood the young and the need to hunt to provision them with food. This conflict arises because the energy needs of the young peak early in the nestling period, at a time when they still cannot thermoregulate and therefore need brooding from the female. The most critical period is the second nestling week, when both female and nestling fat reserves will decrease to low levels. Large female fat reserves in the early nestling period provide a solution to this conflict and are essential for successful breeding. Stochasticity in male provisioning is thus not needed to explain why females should be fat when the eggs hatch. Under normal circumstances, the female broods during the first two weeks and leaves the young only if hunting is absolutely necessary. After the second week the energy requirements are relaxed, and whether the female assists the male in hunting or not depends on factors such as male hunting success, environmental stochasticity, and energy requirements of the young. Our model provides a framework for empirical investigations on female behaviour during breeding in raptors, owls, and other birds with marked division of labour.

The eastern Baltic (EB) cod (Gadus morhua) stock was depleted and overexploited for decades until the mid-2000s, when fishing mortality rapidlydeclined and biomass started to increase, as shown by stock assessments. These positive developments were partly assigned to effective managementmeasures, and the EB cod was considered one of the most successful stock recoveries in recent times. In contrast to this optimistic view, theanalytical stock assessment failed in 2014, leaving the present stock status unclear. Deteriorated quality of some basic input data for stock assessmentin combination with changes in environmental and ecological conditions has led to an unusual situation for cod in the Baltic Sea, which posesnew challenges for stock assessment and management advice.Anumber of adverse developments such as low nutritional condition and disappearanceof larger individuals indicate that the stock is in distress. In this study, we (i) summarize the knowledge of recent changes in cod biology andecosystem conditions, (ii) describe the subsequent challenges for stock assessment, and (iii) highlight the key questions where answers are urgentlyneeded to understand the present stock status and provide scientifically solid support for cod management in the Baltic Sea.

The concept of maximum sustainable yield (MSY) rests on the notion that an intermediately sized stock exists at which the biomass production is at its maximum. the maximum is a result of density-dependent negative feedback on percapita production with increasing stock size. the harvesting is sustainable if it equals the stock productivity. the stock productivity depends on environmental variables, typically food-availability but also temperature and salinity may be important. Predation does not affect the productivity directly, but it is important to consider how the annual surplus production is shared with predators and hence affects the fishing yield. In the stock model we use for MSY analyses we differentiate between internal factors affecting density-dependence and external factors that either drive productivity or represent predation.

To estimate MSY reference points we modelled some stocks with Monte-Carlo simulations (see WD6 for details). The stochastic properties of the stock are estimated and implemented in the stochastic behaviour of the model in order to find the natural range of stock ariables under MSY management. This can be a basis for B trigger in ICES framework, the SSB at which the F MSY is re-evaluated or an adjusted F is adopted according to the harvest control rule. Under some conditions, stochastic simulations will have F MSY that are different from deterministic runs, i e yields are asymmetrical around their means.

The stock model has two variables, NAA (numbers at age) and WAA (weight at age). The functions and their parameters to update the variables annually were obtained from statistical analyses of the data and results from XSA runs provided by the WGBFAS stock coordinators. Four functions were estimated statistically:1) Number of recruits per SSB as a linear function of SSB2) Age-specific natural mortality as a linear function of predator abundance ( e. g. Cod SSB)3) Weight of recruits as a linear function of average parent weight4) Average weight increase as a linear function of WAA and year-specific growthThe MS errors of the regressions were used as variances of the normal distributions from which a random parameter was generated. If body growth is negatively correlated with weight it indicates a Von Bertalanffy type of growth. Also, if number of recruits per SSB is negatively correlated with SSB, there is a density-dependence required for a maximum production to exist.

There are also two external variables that affect the analyses: predator abundance and year-specific growth (named year-growth). The year-growth is a statistical variable tha t encapsulates all year-specific growth common for all age-classes. The year-growth is externally linked driver for variation in productivity of the stock. A Fourier function including the four longest wave-lengths was fitted to the predator abundance and the year-growth respectively. The Fourier functions were representing long term changes in the external conditions, and the residuals as uncorrelated inter-annual variation. The MS of the residuals was also used or generating randomscatter in the external variables between years.

The stock model was run for 10 500 (for sprat 40 500) years, under which the  first 500 years were not generating data and were only used to release the dependence on the initially set conditions. The time series of yields and SSB were collected. The runs were repeated for intervals of F to frame the maximum average yields with a step of 0.01. A weighted catchability (as estimated for each species) was applied so that the  F denoted the average for a given range of age-classes. We also present F MSY max and F MSY min which frames the range in which the average yields within 95% of the MSY (WKFRAME-2 2011). The B trigger is set as the lower 2.5 % percentile of the SSBs from the simulation.

The ecosystem of the baltic sea has undergone dramatic changes, perhaps a regime-shift, during the last four decades. The Baltic Sea Herring SSB has declined to a third and weight-at-age has halved due to plankton prey deficits. The management objective of the herring is currently in the transition to a cautionary maximum sustainable yield (MSY). We modelled the main basin Baltic Sea herring under the currernt regime, and analysed the effect of a ercovery of the cod stock and the food availability as they were in the early 1980s. We recommend a target F MSY of 0.16, but with a recovery of the cod, recommended target F MSY is is 0.10. A simultaneous increase in both cod and food availability is estimated to increase the yield with 40% at the target F MSY is 0.20. We present functions to calculate F MSY and estimate the expected yield depending on the abundance of cod and food-availability. A retrospective application of our functions indicates over-fishing in the 1990s  and early 2000s, and a net loss in yields with a landing value of about E440 Millions.

The Baltic Sea ecosystem has undergone dramatic changes, so-called ecosystem regime shifts, during the past four decades. Baltic Sea herring (Clupea harengus) spawning-stock biomass has declined to a third, and weight-at-age has halved as a result of food shortages and competition with sprat (Sprattus sprattus). The management objective for the herring stock is currently in transition from precautionary to maximum sustainable yield (MSY). The main basin Baltic Sea herring was modelled under the current ecosystem regime and the effect of a recovery of the cod (Gadus morhua) stock and the availability of planktonic food to levels found in the early 1980s analysed. A target of F-MSY = 0.16 for herring, which should decline to F-MSY = 0.10 with recovery of the cod stock, is proposed. An increase in the availability of planktonic food is estimated to more than double the yield at F-MSY = 0.27, overriding the negative effects of cod predation should there be a simultaneous increase in both cod and availability of planktonic food. The estimated net increase in yield is 40% at F-MSY = 0.20. Functions are presented to calculate FMSY and to estimate the expected yield depending on the abundance of cod and food availability. Retrospective application of the functions is indicative of overfishing of herring in the 1990s and early 2000s, and a net loss in yield, with a landing value of some E440 million.

Stochastic variability of biological processes and uncertainty of stock properties compel fisheries managers to look for tools to improve control over the stock. Inspired by animals exploiting hidden prey, we have taken a biomimetic approach combining catch and effort in a concept of Bayesian regulation (BR). The BR provides a real-time Bayesian stock estimate, and can operate without separate stock assessment. We compared the performance of BR with catch-only regulation (CR), alternatively operating with N-target (the stock size giving maximum sustainable yield, MSY) and F-target (the fishing mortality giving MSY) on a stock model of Baltic Sea herring. N-targeted BR gave 3% higher yields than F-targeted BR and CR, and 7% higher yields than N-targeted CR. The BRs reduced coefficient of variance (CV) in fishing mortality compared to CR by 99.6% (from 25.2 to 0.1) when operated with F-target, and by about 80% (from 158.4 to 68.4/70.1 depending on how the prior is set) in stock size when operated with N-target. Even though F-targeted fishery reduced CV in pre-harvest stock size by 19–22%, it increased the dominant period length of population fluctuations from 20 to 60–80 years. In contrast, N-targeted BR made the periodic variation more similar to white noise. We discuss the conditions when BRs can be suitable tools to achieve sustainable yields while minimizing undesirable fluctuations in stock size or fishing effort.

Sympatric speciation can arise as a result of disruptive selection with assortative mating as a pleiotropic by-product. Studies on host choice, employing artificial neural networks as models for the host recognition system in exploiters, illustrate how disruptive selection on host choice coupled with assortative mating can arise as a consequence of selection for specialization. Our studies demonstrate that a generalist exploiter population can evolve into a guild of specialists with an ‘ideal free’ frequency distribution across hosts. The ideal free distribution arises from variability in host suitability and density-dependent exploiter fitness on different host species. Specialists are less subject to inter-phenotypic competition than generalists and to harmful mutations that are common in generalists exploiting multiple hosts.

When host signals used as cues by exploiters coevolve with exploiter recognition systems, our studies show that evolutionary changes may be continuous and cyclic. Selection changes back and forth between specialization and generalization in the exploiters, and weak and strong mimicry in the hosts, where non-defended hosts use the host investing in defence as a model. Thus, host signals and exploiter responses are engaged in a red-queen mimicry process that is ultimately cyclic rather then directional. In one phase, evolving signals of exploitable hosts mimic those of hosts less suitable for exploitation (i.e. the model). Signals in the model hosts also evolve through selection to escape the mimic and its exploiters. Response saturation constraints in the model hosts lead to the mimic hosts finally perfecting its mimicry, after which specialization in the exploiter guild is lost. This loss of exploiter specialization provides an opportunity for the model hosts to escape their mimics. Therefore, this cycle then repeats.

We suggest that a species can readily evolve sympatrically when disruptive selection for specialization on hosts is the first step. In a sexual reproduction setting, partial reproductive isolation may first evolve by mate choice being confined to individuals on the same host. Secondly, this disruptive selection will favour assortative mate choice on genotype, thereby leading to increased reproductive isolation.

The Working Group on Multispecies Assessment Methods (WGSAM) met in Reykjavik, Iceland, 10–14 October 2016. In this tenth report of the pan-regional WGSAM, work focused on four (B, E, F, G) of the multi-annual ToRs.

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. Reporting on ToR A was scarce compared to previous years, partly because recent relevant work was reported against ToR E and G instead.

A Key Run (ToR B) of the Baltic Sea Ecopath with Ecosim (NS-EwE) model 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 the 4 experts in January. The Key Run is documented in a detail in Annex 3, with key outputs summarised in Section 3 and data files made available on the WGSAM webpage). WGSAM also conducted an informal review of the LeMans modelling framework for potential application in the Irish Sea, and recommended adjustments to the framework for further review. Because the LeMans framework is a within-model ensemble addressing parameter uncertainty, this review also related to ToR D.

Multispecies model skill assessment (ToR C) and multi-model ensemble methods (ToR D) were not emphasized this year. However, plans were made to coordinate future work for ToR C, and one ToR D presentation reviewed the utility of a dynamic multimodel ensemble for making inferences about the real world. This method can infer results for individual components of aggregate groups; the ensemble model uses correlations in other ecosystem models to determine what the models that group species would have predicted for individual species. A proof of concept for the North Sea was presented.

Ecosystem indicator analyses (ToR E) were presented from a wide range of ecosystems. A theoretical analysis comparing results from the Celtic and North Seas with 4 “idealized” fleets was presented to analyse the performance of selected indicators in a multispecies mixed fishery. Four indicators including the Large Fish Indicator (LFI) were examined, and shown to have mixed utility in measuring the impact of different fleet sectors, with the best indicator varying by ecosystem. A multivariate analysis of ecosystem responses to multiple drivers was conducted for four US ecosystems using gradient forest method to identify potential ecosystem thresholds. Other multivariate methods were reviewed that draw on the strengths of multiple indicators for the Northeast US shelf ecosystem. A food web based biodiversity indicator was presented with an application for the Baltic Sea. This could be extended to any ecosystem with an EwE or similar model. A community status indicator relating a species-area relationship to the LFI and mean trophic levels was presented for the Swedish west coast.

Impacts of apex predators on fisheries (ToR F) were examined with one presentation and a group discussion planning further work. A multipecies production model was parameterized to simulate interactions between three fish guilds, fisheries, and one marine mammal guild, concluding that fish reference points and trajectories change with marine mammal interactions. Fishery management was also important to reduce vessel interactions with and ensure prey supply to marine mammals. 4 | ICES WGSAM REPORT 2016 Exploration of practical advice for fisheries management incorporating multispecies, mixed fishery, and environmental factors (ToR G) was evident across regions. Two approaches for incorporating species, fleet, environmental, and other interactions are in progress in the Northeast US. One presentation outlined the New England approach, and another outlined the Mid-Atlantic approach. In New England, a management strategy evaluation is in progress to evaluate harvest control rules that consider herring's role as forage in the ecosystem. The modelling framework and stakeholder workshops were discussed. In the Baltic, a Nash Equilibrium optimisation approach incorporating environmental factors was presented for the cod-herring-sprat fishery to attempt to identify a solution that would give good yield for all species simultaneously. In the North Sea a theoretical analysis using 4 “idealized” fleets was presented to analyse the potential implications of "Pretty Good Yield" ranges around MSY. The model examined the likelihood of the fishery being precautionary for the different species given the uncertainties involved, and concluded that the upper ends of MSY ranges would not guarantee precautionarity.

Three meetings were held between 2013 and 2015 (Stockholm, London and Woods Hole), with progress being made on all the Terms of Reference. WGSAM has been making significant contributions required to enable ICES to develop its capability to give advice on the ecosystem impacts of fishing and climate change. This is a priority area identified in the ICES strategic plan and is consistent with scientific needs to support implementation of the Common Fisheries Policy and Marine Strategy Framework Directive. This final report summarises the key progress made against each ToR. A particularly important area established during this period has been the discussions that have led to guidelines on quality assurance of ecosystem models in-tended for advice giving. WGSAM prepared a specific briefing on this issue and con-tinues to work on issues related to model review processes, model validation and developing methods for generating advice from multi-model ensembles. These are all important areas of work in the evolution toward giving integrated, ecosystem-based advice to ICES clients. We recommend that ICES supports continuation of WGSAM new ToRs and considers more specifically how to support WGSAM in developing advice relevant products.

The current fisheries management goals set by the European Commission states that fish stocks should be harvested to deliver maximum sustainable yields (MSY) and simultaneously, management should take ecosystem considerations into account. This creates unsolved trade-offs for the management of the stocks. We suggest a definition of a multi-species-MSY (MS-MSY) where no alternative fishing mortality (F) can increase yield (long term) for any ecologically interacting stock, given that the other stocks are fished at constant efforts (Fs). Such a MS-MSY can be solved through the game theoretic concept of a Nash equilibrium and here we explore two solutions to this conflict in the Baltic Sea. We maximize the sustainable yield of each stock under two constraints: first, we harvest the other stocks at a fixed F (FNE); second, we keep the spawning stock biomasses of the other stocks fixed [biomass Nash equilibrium (BNE)]. As a case study, we have developed a multi-species interaction stochastic operative model (MSI-SOM), which contains a SOM for each of the three dominant species of the Baltic Sea, the predator cod (Gadus morhua), and its prey herring (Clupea harengus), and sprat (Sprattus sprattus). For our Baltic Sea case, MS-MSYs exist under both the FNE and the BNE, but there is no guarantee that point solutions exists. We found that the prey species’ spawning stock biomasses are additive in the cod growth function, which allowed for a point solution in BNE. In the FNE, the herring MSY was found to be relatively insensitive to the other species’ fishing mortalities (F), which facilitated a point solution. The MSY targets of the BNE and the FNE differ slightly where the BNE gives higher predator yields and lower prey yields.

Darwin’s Principle of Divergence explains sympatric speciation as gradual and directional. Contradicting evidence suggests that species’ traits evolve saltationally. Here, we model coevolution in exploiter-victim systems. Victims (resource population) have heritable, mutable cue phenotypes with different levels of defense. Exploiters have heritable, mutable perceptual phenotypes. Our simulations reveal coevolution of victim mimicry and exploiter spe-cialization in a saltational and reversible cycle. Evolution is gradual and directional only in the specialization phase of the cycle. Once linked to assortative mating, exploiter specialization provides conditions for speciation.

Our current understanding of sympatric speciation is that it occurs primarily through disruptive selection on ecological genes driven by competition, followed by reproductive isolation through reinforcement-like selection against inferior intermediates/heterozygotes. Our evolutionary model of selection on resource recognition and preference traits suggests a new mechanism for sympatric speciation. We find speciation can occur in three phases. First a polymorphism of functionally different phenotypes is established through evolution of specialization. On the gene level, regulatory functions have evolved in which some alleles are conditionally switched off (i.e. are silent). These alleles accumulate harmful mutations that potentially may be expressed in offspring through recombination. Second mating associated with resource preference invades because harmful mutations in parents are not expressed in the offspring when mating assortatively, thereby dividing the population into two pre-zygotically isolated resource-specialist lineages. Third, silent alleles that evolved in phase one now accumulate deleterious mutations over the following generations in a Bateson-Dobzhansky-Muller fashion, establishing a post-zygotic barrier to hybridization.

The ICES Baltic Fisheries Assessment Working Group (WGBFAS) met 19-26 April 2017(Chair: Tomas Gröhsler, Germany and Co-chair: Michele Casini, Sweden), with 41 participants and 9 countries represented. The objective of WGBFAS was to assess the status of the following stocks: Sole in Division 3.a, SDs 20–24; Cod in Kattegat, Cod in SDs 22–24, Cod in SDs 25–32; Herring in SDs 25–27, 28.2, 29 and 32, Herring in SD 28.1 (Gulf of Riga), Herring in SDs 30-31 (Gulf of Bothnia); Sprat in SDs 22–32; Plaice in SDs 21–23, Plaice in SDs 24–25; Flounder in SDs 22–23, in SDs 24–25, in SDs 26+28 and SDs 27+29–32; Brill in SDs 22–32, Dab in SDs 22–32 and Turbot in SDs 22–32. WGBFAS also identified the data needed for next year’s data call with some suggestions for improvements in the data call, and stock-specific research needs.The report contains an introduction with the summary of other WGs relevant for the WGBFAS, the methods used, and ecosystem considerations. The results of the analytical stock assessment or survey trends for the species listed above are then presented with all the stocks with the same species in the same sections. The report ends with references, recommendations, links to Stock Annexes and list of Working Documents.The principle analytical models used for the stock assessments were XSA and SAM. For most flatfishes and cod in SDs 25–32 (data limited stocks), CPUE trends from bottomtrawl surveys were used in the assessment (except plaice in SDs 24–25 for which relative SSB from SAM was used). For the data limited stocks, reference points based on length-based indicators were estimated (except cod in SDs 25-32 for which relative reference points were estimated using the SPiCT model). For cod in SDs 25–32, intersessional work was planned to hopefully allow returning to an analytical stock assessment in the near future. Ecosystem changes have been analytically considered in the following stock assessments: Herring in SD 25–27, 28.2, 29 and 32, and Sprat in SD 22–32, in form of cod predation mortality.